California Department
of Corrections and
Rehabilitation:
Inmates Sentenced Under the Three Strikes Law
and a Small Number of Inmates Receiving Specialty
Health Care Represent Significant Costs
May 2010 Report 2009-107.2

Independent NONPARTISAN
TRANSPARENT Accountability

The first five copies of each California State Auditor report are free. Additional copies are $3 each, payable by
check or money order. You can obtain reports by contacting the Bureau of State Audits at the following address:
California State Auditor
Bureau of State Audits
555 Capitol Mall, Suite 300
Sacramento, California 95814
916.445.0255 or TTY 916.445.0033
OR
This report is also available on the World Wide Web http://www.bsa.ca.gov
The California State Auditor is pleased to announce the availability of an on-line subscription service. For
information on how to subscribe, please contact the Information Technology Unit at 916.445.0255, ext. 456,
or visit our Web site at www.bsa.ca.gov.
Alternate format reports available upon request.
Permission is granted to reproduce reports.
For questions regarding the contents of this report,
please contact Margarita Fernández, Chief of Public Affairs, at 916.445.0255.

Elaine M. Howle
State Auditor

CALIFORNIA STATE AUDITOR

Doug Cordiner
Chief Deputy

Bureau of State Audits

555 Capitol Mall, Suite 300

S a c r a m e n t o, C A 9 5 8 1 4

May 18, 2010	

916.445.0255

916.327.0019 fax

w w w. b s a . c a . g o v

2009-107.2

The Governor of California
President pro Tempore of the Senate
Speaker of the Assembly
State Capitol
Sacramento, California 95814
Dear Governor and Legislative Leaders:
As requested by the Joint Legislative Audit Committee, the California State Auditor presents this
audit report concerning the effect of California Department of Corrections and Rehabilitation’s
(Corrections) operations on the state budget.
This report concludes that inmates sentenced under the three strikes law, and a small number of
inmates receiving specialty health care, represent significant costs. Specifically, about 25  percent
of the inmate population was incarcerated under the three strikes law, which requires longer terms
for individuals convicted of any felony if they were previously convicted of a serious or violent crime
as defined in state law. On average, we estimate that these individuals’ sentences are nine years longer
because of the requirements of the three strikes law and that these additional years of incarceration
represent a cost to the State of $19.2 billion. Furthermore, the current conviction for which many of
these individuals are incarcerated is not for a serious or violent crime, as defined in state law, and
many were convicted of multiple serious or violent crimes that occurred on the same day.
Our review also found that of the $529 million that California Prison Health Care Services incurred
for contracted specialty health care providers in fiscal year 2007–08, $469 million could be associated
with individual inmates. Among the inmates with specialty health care costs, 70 percent averaged
slightly more than $1,000 per inmate and cost $42 million in total, while the remaining 30 percent of
inmates amassed specialty health care costs totaling more than $427 million. Furthermore, specialty
health care costs for 1,175 inmates, or just one-half of 1 percent of the inmates incarcerated during the
year, totaled $185 million. In addition, specialty health care costs totaled $8.8 million for the 72 inmates
who died during the last quarter of the year, exceeding $1 million in the case of one inmate.
Finally, a significant amount of custody staff overtime is the result of a medical guarding and
transportation workload that does not have associated authorized positions. Overtime is also
necessary when custody staff positions are vacant, but is decreased by staff who do not use the
full amount of leave they earn. However, the unused leave of custody staff—increased by
the additional leave provided through the furlough program—represents a liability to the State
that we estimate is at least $546 million and could be more than $1 billion.
Respectfully submitted,

ELAINE M. HOWLE, CPA
State Auditor

California State Auditor Report 2009-107.2

May 2010

Contents
Summary	

1

Introduction	

7

Chapter 1
Longer Sentences Due to Three Strikes Represent a Significant Cost	

21

Recommendations	

34

Chapter 2
A Small Portion of the Inmate Population Accounts for Most 	
Contracted Specialty Health Care Costs	

35

Recommendations	

46

Chapter 3
Vacant Positions, Medical Guarding, and Leave Accruals Influence 	
Overtime Costs	

49

Recommendations	

63

Appendix
Serious or Violent Felonies as Defined by California State Law 	

65

Responses to the Audit
California Department of Corrections and Rehabilitation	

69

California State Auditor’s Comments on the Response From
the California Department of Corrections and Rehabilitation	
California Prison Health Care Services	
California State Auditor’s Comment on the Response 	
From California Prison Health Care Services	

71
73
75

vii

California State Auditor Report 2009-107.2

May 2010

Summary
Results in Brief

Audit Highlights . . .

The mission of the California Department of Corrections and
Rehabilitation (Corrections) is to ensure public safety through the
safe and secure incarceration of offenders, effective supervision of
parolees, and rehabilitation strategies to successfully reintegrate
offenders into communities. In a report we published in
September 2009 titled 2009-107.1: California Department of
Corrections and Rehabilitation: It Fails to Track and Use Data
That Would Allow It to More Effectively Monitor and Manage
Its Operations (prior report), we described how Corrections’
expenditures of $10 billion comprised 10 percent of the State’s
General Fund budget for fiscal year 2007–08. Although Corrections
tracks costs at the institutional level, it does not track costs in a
manner that would allow us to stratify them by specific inmate
characteristic. As a result, despite rising costs, Corrections does
not have sufficient information to identify how much specific
inmate or institution characteristics contribute to these costs and
how changes in its operations would influence expenditures. In
this report, we expand on certain issues presented in our prior
report and present additional information related to the inmates
incarcerated under the three strikes law (striker inmates). We
provide a breakdown of some health care costs by type of service,
the health care delivery method, and whether the inmates receiving
specialty health care were striker inmates. We also describe the
impact of vacancies on overtime, and we discuss California Prison
Health Care Services’ (Health Care Services) plans for containing
health care costs.

Our review of California’s increasing
prison cost as a proportion of the state
budget and California Department
of Corrections and Rehabilitation’s
(Corrections) operations revealed
the  following:

As of April 2009, 25 percent of the inmate population was
incarcerated under the three strikes law, which requires longer
sentences for individuals who are convicted of any felony and
have been convicted previously of crimes defined in state law as
serious or violent felonies, also known as strikes. As discussed in
our prior report, we estimated that on average, these individuals’
sentences are nine years longer because of the requirements of the
three strikes law. In further analyzing the nature of the crimes for
which striker inmates are incarcerated, we found that the current
conviction for which many are imprisoned is not a strike. However,
the longer sentences that striker inmates are receiving are based
on their previous strikes. We also found that significant portions
of the striker inmate population were convicted of committing
multiple serious or violent offenses on the same day, and that some
committed one or more serious or violent offenses as a juvenile.

»» Inmates incarcerated under the three
strikes law (striker inmates):
•  Make up 25 percent of the inmate
population as of April 2009.
•  Receive sentences that are, on average,
nine years longer–resulting in about
$19.2 billion in additional costs over the
duration of their incarceration.
•  Include many individuals currently
convicted for an offense that is not a
strike, were convicted of committing
multiple serious or violent offenses on
the same day, and some that committed
strikeable offenses as a juvenile.
»» Inmate health care costs are significant
to the cost of housing inmates. In fiscal
year 2007–08, $529 million was incurred
for contracted services by specialty health
care providers. Additionally:
•  30 percent of the inmates receiving
such care cost more than $427 million.
•  The costs for the remaining 70 percent
averaged just over $1,000 per inmate.
•  The costs for those inmates who died
during the last quarter ranged from
$150 for one inmate to more than
$1 million for another.
continued on next page . . .

1

2

California State Auditor Report 2009-107.2

May 2010

»» A significant portion of the increased
workload due to medical guarding
and transportation is covered
through overtime.
»» The large leave balances of custody
staff, to which the furlough program
has contributed a significant amount,
will eventually cost the State from
$546 million to more than $1 billion.

As we described in our prior report, the cost of housing striker
inmates for the additional years they were sentenced to under the
three strikes law represents a substantial liability to the State. We
estimated that the additional years imposed by the three strikes law
represent $19.2 billion in additional costs over the duration of the
incarceration of striker inmates identified as of April 2009. Of
this amount, $7.5 billion is associated with striker inmates whose
current convictions are for crimes that are not strikes. Although we
identified specific populations of striker inmates and the additional
years these inmates are sentenced to, there are some limitations in
Corrections’ data that limited our analysis.
A significant part of the overall cost to house inmates in
Corrections’ 33 institutions is inmate health care costs. Health care
provided to inmates can generally be classified as one of two types:
primary health care, which is provided by medical staff within
the institutions, and contracted specialty health care, which is
provided by contracted providers. We obtained the costs and nature
of contracted specialty health care provided to inmates from the
Contract Medical Database (the CMD), most of which we were able
to associate with individual inmates.
Of the $734 million in the CMD costs in fiscal year 2007–08,
$529 million was for contracted services by specialty health
care providers. The remaining $205 million was for contractors
that provide temporary staff to fill vacant health care positions to
provide medical care at the institutions, referred to as registry costs.
Of the $529 million in contracted specialty health care costs, we
were able to associate 89 percent, or $469 million, with valid inmate
identification numbers. We could not associate the remaining
$60 million with specific inmates. Our analysis of the specialty
health care costs associated with specific inmates revealed that
the majority of these costs for fiscal year 2007–08 were associated
with a relatively small population of inmates. Specifically, among
the inmates with specialty health care costs recorded in the CMD,
30 percent of the population cost more than $427 million, while
the costs for the remaining 70 percent averaged slightly more
than $1,000 per inmate. Further, just one-half of 1 percent of the
inmates incarcerated during the year, or 1,175 inmates, incurred
39 percent of such costs in fiscal year 2007–08. We also found that
of the nearly 15,800 inmates who incurred more than $5,000 in
specialty care costs during fiscal year 2007–08, 63 percent were
age 40 and older. In comparison, inmates age 40 or over represent
only 41 percent of all inmates. We also found that the 72 inmates
who died during the last quarter of fiscal year 2007–08 incurred,
on average, $122,300 for specialty health care services for that
fiscal year. Ranging from $150 for one inmate to more than
$1 million for another, these 72 inmates accounted for $8.8 million
in specialty health care costs during fiscal year 2007–08.

California State Auditor Report 2009-107.2

May 2010

According to Health Care Services—the organization overseen
by the federal receiver responsible for administering health care
provided to inmates—it is taking steps to contain health care costs.
These cost containment measures include implementing utilization
management, contracting with a third-party administrator
to process medical invoices, and expanding telemedicine.
Although Health Care Services has continued expanding its
use of telemedicine as part of its cost containment strategy, it
has not fully estimated the potential savings of using additional
telemedicine. Health Care Services did provide an estimate of
the medical guarding and transportation costs that are avoided
with each telemedicine consultation, which indicates that the
telemedicine consultations completed in fiscal year 2008–09
resulted in savings of $4.6 million to $9.2 million in guarding and
transportation costs. Estimates of the costs associated with projects
that will facilitate expanding telemedicine are significant. They
include $41 million for the Telemedicine Services Project, a broadly
defined project intended to provide technology and resources
to improve and expand telemedicine services by 400 percent to
80,000 consultations by fiscal year 2013–14. Using Health Care
Services’ most recent cost avoidance estimates, the $41 million in
costs would be offset by $46 million to $93 million in costs avoided
over the five-year period. Savings vary significantly by institution,
however, suggesting that some telemedicine consultations may not
be cost‑effective.
Finally, using payroll-related information from the State Controller’s
Office, we reviewed the relationship between Corrections’ vacant
custody staff positions and overtime. We determined that a
significant workload related to medical guarding and transportation
does not have associated authorized positions and is covered
through overtime. Additionally, we found that the formula
Corrections uses to determine how many custody staff should be
hired to adequately staff its institutions did not accurately account
for the factors in the documentation that supports Corrections’
calculation of custody staff that are unavailable to work, such as
vacation or sick leave. As a result, custody staff opportunities to
take the leave that they earn were decreased. These errors are offset
in some institutions by the way Corrections applies the formula to
the number of guarding assignments.
Whether due to fewer opportunities to use leave or other factors,
custody staff have accumulated significant leave balances. Although
the work provided when staff choose not to take the full amount
of leave they earn during the course of a year reduces the current
staffing costs of running the correctional system, the growing leave
balances represent a deferred liability. However, if custody staff used
the leave to which they are entitled, Corrections would incur higher
overtime costs. The furlough program has exacerbated this situation

3

4

California State Auditor Report 2009-107.2

May 2010

because of the need to maintain staffing levels in the institutions,
causing further increases in total leave balances. We estimate these
total leave balances will result in an eventual cost to the State of at
least $546 million and could be more than $1 billion, depending on
whether staff are paid for the leave when they quit or retire, or are
able to use the leave while they are employed.
Recommendations
Health Care Services should continue to explore methods of
reducing the costs of medical care to the State, including those
of inmates with high medical costs. These efforts could include
proposing a review of the program that allows for the early release
of terminally ill or medically incapacitated inmates, and other
possible means of altering the ways in which inmates are housed
without unduly increasing the risk to the public.
To determine whether the additional expansion of telemedicine is
cost-effective within the California correctional system, Health Care
Services should further analyze the cost‑effectiveness of
telemedicine through a more robust estimate of savings, including
consideration of factors such as the percentage of telemedicine
consultations that required subsequent in‑person visits because the
issue could not be addressed through telemedicine.
To ensure that the level of custody staffing meets institutional
needs, and to provide staff the opportunity to use the amount of
leave they earn in the future, Corrections should update its staffing
formulas to accurately represent each of the factors for which
custody staff are unavailable to work, such as vacation or sick
leave. In addition, Corrections should create a policy for regularly
scheduled reviews of the data used in the staffing formulas and
update the formulas as necessary.
To better communicate to policy makers the annual cost
of incarceration, and to provide a more accurate estimate of
expenditures associated with changes in the large leave
balances of custody staff—many of whom require relief
coverage when they are absent—Corrections should provide the
following as supplemental information to the relevant legislative
policy and fiscal committees:
•	 A calculation of the annual increase or decrease in its liability for
the leave balances of custody staff to better explain the cause of
changes in expenditures.
•	 An estimate of the annual cost of leave balances likely to be paid
for retiring custody staff.

California State Auditor Report 2009-107.2

May 2010

Agency Comments
Corrections did not disagree with our findings or
recommendations, but did not specifically address them in
its response.
Health Care Services states that it concurs with the audit
findings and recommendations, but also chose not to address
them specifically.

5

6

California State Auditor Report 2009-107.2

May 2010

Blank page inserted for reproduction purposes only.

California State Auditor Report 2009-107.2

May 2010

Introduction
Background
Established in 1944, the California Department of Corrections
and Rehabilitation (Corrections) operates California’s prisons,
oversees community correctional facilities, supervises parolees,
and operates the juvenile justice system. Corrections’ mission is
to enhance public safety through safe and secure incarceration of
offenders, effective parole supervision, and rehabilitative strategies
to successfully reintegrate offenders into communities. In a
July 2005 reorganization, Corrections added “Rehabilitation” to its
name to encompass its objective of addressing the rehabilitative
and reentry needs of incarcerated juvenile wards and adult
inmates. As of June 30, 2009, it was responsible for nearly
168,000 inmates, 111,000 parolees, and more than 1,600 juvenile
wards of the State.
Corrections’ Annual Expenditures
In a report we published in September 2009 titled 2009-107.1:
California Department of Corrections and Rehabilitation: It Fails
to Track and Use Data That Would Allow It to More Effectively
Monitor and Manage Its Operations (prior report), we stated that
Corrections’ expenditures comprised 10 percent of the State’s
General Fund budget in fiscal year 2007–08, having increased by
32 percent over the previous three years, to $10 billion. During the
same period, Corrections’ population of adult inmates decreased
by 1 percent while its population of adult parolees increased by
7 percent, juveniles in facilities decreased by 37 percent, and
juvenile parolees decreased by 29 percent. In fiscal year 2008–09
Corrections reported overall expenditures of $10.2 billion and the
percentages spent in the general areas of operation were similar to
those in fiscal year 2007–08. As shown in Table 1 on the following
page, most of Corrections’ expenditures are attributable to two
programs, Adult Corrections and Rehabilitation (adult corrections)
and Adult Health Care Services (adult health care). For fiscal year
2007–08, expenditures on adult corrections totaled $5.4 billion,
or more than half of Corrections’ expenditures for that year.
This amount includes expenditures for the 33 adult correctional
institutions, the conservation camps, and costs of inmates sent to
community correctional facilities and to out‑of‑state correctional
facilities.
Adult health care, the second most expensive cost area of
Corrections’ operations, cost $2.1 billion in fiscal year 2007–08,
or 22 percent of Corrections’ total expenditures. Some inmate
health care is offered internally at all adult institutions, but

7

8

California State Auditor Report 2009-107.2

May 2010

significant amounts of the adult health care expenditures
are related to specialty health care services from contracted
providers. Specialty health care includes care provided in areas
such as surgery, cardiology, and dermatology. These costs totaled
$529 million in fiscal year 2007–08, or 25 percent of all health care
related costs. According to the 2009–10 Governor’s Budget, the
objective of adult health care is to provide medical, dental, and
mental health care to the inmate population statewide, consistent
with adopted standards for the quality and scope of services
within a custodial environment. The adult health care program
also operates three licensed hospitals and a skilled nursing facility
for female inmates, 16 correctional treatment centers, a hospice
care wing at California Medical Facility, and eight HIV units at
various institutions.
Table 1
California Department of Corrections and Rehabilitation’s Program
Expenditures for Fiscal Year 2007–08
PROGRAM

TOTALS

PERCENT

Adults
Adult corrections and rehabilitation
Correctional inmate health care
Education, vocational, and other rehabilitation
Community partnerships
Subtotals

$5,407,735,654

54%

2,144,083,416

22

428,347,373

4

8,999,779

0

$7,989,166,222

80%

$733,818,494

7%

Parole
Adult parole operations
Board of parole hearings
Subtotals

85,202,618

1

$819,021,112

8%

$213,396,454

2%

Juveniles
Facilities operations
Programs
Health care services
Parole operations
Subtotals

162,545,866

2

93,871,269

1

30,124,029

0

$499,937,618

5%

$291,410,083

3%

Other
Administration
Corrections standards authority

220,823,787

2

Capital outlay

149,649,936

2

Subtotal
Total

$661,883,806

7%

$9,970,008,758

Source:  Bureau of State Audits’ analysis of California Department of Corrections and Rehabilitation’s
accounting records for fiscal year 2007–08.

California State Auditor Report 2009-107.2

May 2010

The Three Strikes Law
California enacted the three strikes law in 1994,1 with the intent of
ensuring longer prison sentences and greater punishment for those
who commit a felony and were convicted previously of a serious
or violent felony. As Figure 1 shows, when a person is convicted
of a felony and has one prior conviction for a serious or violent
felony, the three strikes law provides a minimum sentence that is
double the sentence a court would have imposed on an individual
with no prior serious or violent felony convictions; this type of
offender is often referred to as a second striker. The three strikes
law also generally provides a minimum sentence of 25 years to life
imprisonment for someone who is convicted of a felony and who
has two or more prior convictions for a serious or violent felony;
this type of offender is often referred to as a third striker.
Figure 1
General Application of the Three Strikes Law
Offender Convicted of ANY Felony

Number of offender’s prior
violent or serious felony convictions*
NONE

Sentence is term required for
current felony conviction per
applicable law.

Three times the term selected
by the court for each current
felony conviction.

ONE

Sentence is two times the
term required for current
felony conviction.

25 years

TWO or MORE

Sentence is life imprisonment
with the minimum term being
the greater of the following:

Term selected by the court
under California Penal Code,
Section 1170, plus enhancements;
or sections 190 or 3046.

Source:  California Penal Code, sections 667 (b)–(i) and 1170.12.
*	 Per California Penal Code, sections 667 (f)(2) and 1170.12 (d)(2), upon recommendation by
the district attorney, the court may dismiss or strike a defendant’s prior felony conviction
allegation(s). Moreover, per People v. Superior Court (Romero), a judge may also dismiss a prior
felony conviction.

1	

The Legislature enacted the three strikes law in March 1994, and in November 1994 voters
approved Proposition 184 to enact a virtually identical version of the law.

9

10

California State Auditor Report 2009-107.2

May 2010

The three strikes law specifies the offenses for which a prior felony
conviction counts as a strike. A prior conviction for an offense
that California law defines as a violent felony or a serious felony
counts as a strike. In addition, the three strikes law specifies that a
prior conviction for an offense committed in another jurisdiction
counts as a strike if the offense would have been punishable by
imprisonment in state prison in California and if the offense has all
the elements of a serious or violent felony as defined by California
law. The three strikes law further specifies that offenses committed
by juveniles may count as strikes in certain circumstances.
Although these juvenile offenses are not expressly defined as serious
or violent felonies, we use the term “serious or violent felony” to
describe any offense that could count as a strike. The Appendix
provides a list of all serious or violent felonies.
Under the three strikes law, only a defendant’s prior convictions
for serious or violent felonies constitute strikes. A defendant
subsequently convicted of any felony, even a felony that is not a
serious or violent felony, may be sentenced under the three strikes
law. For example, courts have sentenced defendants who were
convicted of felony petty theft for shoplifting to 25 years to life
imprisonment under the three strikes law because they had at
least two prior convictions for serious or violent felonies. In
People v. Romero,2 the defendant had prior convictions dating to the
1980s for burglary, hit and run, battery on a peace officer, and lewd
conduct with a child under age 14. After Romero was convicted of
felony petty theft for stealing a magazine in 1999, he was sentenced
to 25 years to life imprisonment under the three strikes law because
of his prior convictions.
Although the three strikes law may result in circumstances like
Romero, the law allows prior strikes to be dismissed in certain
circumstances. Even if a prior felony conviction otherwise would
count as a strike, a prosecutor may ask a court not to count it as
a strike in the furtherance of justice or if insufficient evidence
exists to prove the prior felony conviction. Additionally, in 1996
the California Supreme Court held that a court may also decide,
without a request from a prosecutor, not to count a prior felony
conviction as a strike in the furtherance of justice.
In addition to the minimum prison sentences required, the
three strikes law has other provisions that may affect the size of
California’s inmate population. For example, a second or third
striker may not be granted probation and may not be committed
to any facility other than a state prison. The law also limits the
amount of credit that inmates incarcerated under the three strikes
2	

People v. Romero (2002) 99 Cal. App. 4th 1418.

California State Auditor Report 2009-107.2

May 2010

law (striker inmates) can earn against their prison sentence for
participation in work training or education programs while
in prison. Generally, striker inmates may earn a maximum of
20 percent credit against their sentences, while other inmates may
be paroled after serving less than 40 percent of their sentences.
Moreover, a striker defendant convicted of multiple current
offenses generally must serve consecutive sentences rather than
concurrent sentences.
Many other states have strike laws; however, differences exist
among states regarding what constitutes a strike and the lengths of
sentences for second- and third-strike offenses. For instance, the
three strikes law in California includes the sale of certain drugs to
minors as a strike offense, while New Jersey and Arkansas do not
include drug-related crimes as strike offenses. Additionally, an
individual who commits any felony can be sentenced under the
three strikes law in California if prior felony convictions were strike
offenses. In contrast, Arkansas, New Jersey, and Washington
require that, for an inmate to be sentenced as a striker, the current
conviction must be a strike. Further, some states’ striker laws
provide a mandatory sentence of life in prison without the
possibility of parole for inmates with multiple strikes, while the
three strikes law in California sentences such individuals to a term
of 25 years to life. Finally, Washington provides a sentence of life
without parole for a second conviction for certain sex offenses.
The Federal Court‑Appointed Receiver’s Role
In 2006 a federal court‑appointed a receiver
(receiver) to provide leadership and executive
management over the California prison medical
health care system. In 2006 the receiver established
the California Prison Health Care Receivership
Corporation to assist with his responsibilities for
developing, implementing, and sustaining a health
care system that provides constitutionally adequate
medical care to all inmates. In June 2008 the court
approved the receiver’s Turnaround Plan of Action,
which was designed to address the constitutional
deficiencies in California’s prison health care
system. The plan focuses on six goals, as described
in the text box. The receiver uses the name
California Prison Health Care Services (Health Care
Services) to describe the organization he oversees.

Federal Court-Appointed Receiver’s Goals
1.	 Ensure timely access to health care services.
2.	 Establish a prison medical program that addresses a full
continuum of health care services.
3.	 Recruit, train, and retain a professional‑quality medical
care workforce.
4.	 Implement a quality assurance and continuous
improvement program.
5.	 Establish a medical support infrastructure.
6.	 Provide for necessary clinical, administrative and
housing facilities.
Source:  Federal court-appointed receiver’s Turnaround Plan of
Action, June 6, 2008.

11

12

California State Auditor Report 2009-107.2

May 2010

As we discussed in our prior report, one way in which Health
Care Services plans to improve the provision of specialty care is
through the expanded use of telemedicine. Basically, telemedicine
involves the delivery of medical services through the use of
multimedia technology such as voice, video, and data. This allows
the inmate and a medical specialist or psychiatrist to meet through
video conferencing. According to Health Care Services, it has
expanded the specialty services offered through its telemedicine
program from 14 medical specialties in April 2009 to 22 medical
specialties as of March 2010.
The Federal Court-Ordered Plan to Reduce the Prison Population
To address the constitutionally inadequate mental and medical
health care available to inmates in the California prison system
and the threat to the health and safety of inmates posed by
overcrowding, in August 2009 a three‑judge federal court ordered
Corrections to provide the court with a plan that would reduce
the population of Corrections’ adult institutions over the next
two years. According to Corrections, this court order would require
a reduction of more than 40,000 inmates. In response to the
court order, the secretary of Corrections stated that Corrections
believes the federal courts are exceeding their authority under
the Prison Litigation Reform Act, and Corrections will continue
to fight against a population cap or court‑ordered early release.
In January 2010 the three-judge panel stayed its order pending
Corrections’ appeal of the matter to the U.S. Supreme Court.
Scope and Methodology
The Joint Legislative Audit Committee requested that the Bureau
of State Audits evaluate the effect of California’s rapidly increasing
prison population on the state budget. We were asked to focus on
specific areas of Corrections’ operations to provide the Legislature
and the public with information necessary to make informed
decisions. Specifically, we were asked to do the following:
•	 Review the current cost to house inmates; stratify the costs by
their security level, age, gender, or any other relevant category
tracked by Corrections; and determine the reasons for any
significant cost variations among such levels and categories.
•	 Determine the number of inmates Corrections has sent to
other states and calculate the State’s cost and impact on
Corrections’ budget.

California State Auditor Report 2009-107.2

May 2010

•	 Analyze Corrections’ budget to determine the amounts allocated
to vocational training, rehabilitation, and education programs.
•	 For a sample of institutions offering vocational training,
rehabilitation, and education programs, review Corrections’
system for determining the number of instructors and custody
staff needed for inmates to participate in these programs. If
such staffing is inadequate, determine if any inmates have been
denied access to these programs.
•	 To the extent possible, determine the costs for incarceration
under the three strikes law. At a minimum, determine the
incarceration cost for each of the following three scenarios:
-	 The third strike was not a serious and violent felony.
-	 One or more of the strikes was committed as a juvenile.
-	 Multiple strikes were committed during one criminal offense.
•	 Calculate annual overtime pay since 2002 for Corrections’
employees, including correctional officers and custody staff, and
investigate the reasons for significant fluctuations.
•	 Review the number of vacant positions during the last five years
and determine whether they affect the annual overtime costs and
whether filling vacancies would save Corrections money.
•	 Determine the extent to which Corrections currently uses
and plans to use telemedicine. Further, determine if by using
telemedicine Corrections is reducing inmate medical and
custody costs and the cost to transport and guard inmates
outside the prison environment.
We addressed many of these objectives in our prior report.
Specifically, we reviewed the current cost to house inmates,
summarized the information available by program and cost area,
and ultimately determined the average cost per inmate. We also
analyzed the cost variances among institutions and institution
missions and identified potential causes of these variances based
on information obtained from interviews with Corrections’ staff.
Additionally, we determined the number of inmates Corrections
has sent to other states and the cost of transporting and housing
those inmates out of state. We also compiled the amounts allocated
to vocational training, rehabilitation, and education programs
and reviewed Corrections’ process for determining the number of
instructors and custody staff necessary for inmates to participate
in these types of programs. We estimated the cost of incarcerating
striker inmates and stratified the striker and non-striker inmate

13

14

California State Auditor Report 2009-107.2

May 2010

population by age. We also calculated the annual overtime pay
since fiscal year 2003–04 for Corrections’ adult operations
employees, including correctional officers and other custody staff.
In addition, we identified and investigated significant fluctuations
in annual overtime pay, including an examination of the effect of
vacant positions on overtime levels. Finally, we reviewed Health
Care Services’ use of telemedicine and its plans for expanding
telemedicine, and identified limitations in Health Care Services’
data regarding its estimate of potential cost savings related to the
use of telemedicine.
For this report, we determined the number of striker inmates
whose current offense was not a serious and violent felony, striker
inmates who committed one or more serious or violent offenses as
a juvenile, and striker inmates who committed multiple serious or
violent offenses on the same day. We also estimated the potential
cost of the additional years of incarceration imposed by the
three strikes law for each of these groups. Further, we reviewed
additional information regarding vacant positions and leave usage
and examined state laws, policies, and procedures relevant to these
subjects. In addition, to expand on the information presented in
our prior report regarding the stratification of incarceration costs
by inmate characteristics, we analyzed cost data for contracted
specialty health care and reviewed certain characteristics of inmates
receiving specialty care. We also reviewed Health Care Services’
plans for containing health care costs, including its plan and
associated costs for increasing the use of telemedicine.
To expand on the information we reported previously on the cost
to house inmates stratified by relevant categories, we examined
the cost of adult health care, which is the second most expensive
area of Corrections’ operations. Using Health Care Services’
Contract Medical Database (the CMD) for fiscal year 2007–08, we
identified and analyzed the costs of contracted specialty health care
services for inmates receiving such care. We reviewed the type of
service, the health care delivery method, and whether the inmates
receiving specialty health care were striker inmates. We also
reviewed the inmate characteristics and type of service provided
to inmates who received more than $5,000 in specialty health care
in fiscal year 2007–08. Finally, we identified inmates who died
during the last quarter of fiscal year 2007–08 and the amount
of contracted specialty health care costs associated with those
inmates during the entire fiscal year. As part of this work, we also
reviewed Health Care Services’ plans for containing the costs of
specialty health care in the future. To do so, we interviewed Health
Care Services’ staff to gain an understanding of the various cost
containment measures that Health Care Services is implementing.

California State Auditor Report 2009-107.2

May 2010

To complete our work regarding the audit objectives related to
the incarceration costs for specific portions of the striker inmate
population, we used data obtained from Corrections’ Offender
Based Information System as of April 2009. To identify striker
inmates incarcerated for a current offense that is a nonserious
and nonviolent felony, we identified the population of striker
inmates housed by Corrections and then excluded strikers who are
incarcerated for a crime that is violent or serious. Corrections’ data
identify offenses that are classified as violent, enabling us to identify
and exclude striker inmates who are incarcerated for such crimes.
To identify striker inmates who are incarcerated for a serious
felony, our legal counsel reviewed the elements of the remaining
crimes committed by striker inmates as of the date the crime was
committed to determine which crimes were serious as defined by
the relevant state law.
To identify striker inmates who committed multiple strikes during
one criminal offense, we identified all offenses in the data that
were committed by individuals incarcerated as strikers. We then
identified all convictions for offenses committed on the same date,
using information in Corrections’ database to identify offenses
classified as violent, and our legal counsel’s review of the elements
of the remaining offenses to determine which ones were serious as
defined by the relevant state law. In calculating this figure, we were
attempting to identify instances in which a person was convicted of
more than one crime stemming from the same criminal offense, or
set of circumstances. For example, an individual could be charged
with robbery, intimidating a witness, and assault with a deadly
weapon due to actions they took in the course of a single event.
However, this method could include crimes occurring on the same
day but related to different offenses or sets of circumstances. If a
person is guilty of a criminal act or omission that is punishable in
different ways under different laws, state law prohibits punishing
the person under more than one law, requires that punishment be
imposed under the law that provides the longest potential prison
sentence, and requires a court to stay the sentence for the law that
provides the shorter sentence. For this reason, we also identified
the number of striker inmates who would not have had convictions
for multiple serious or violent offenses on the same day if the
convictions for offenses whose sentences were stayed had been
excluded. This was intended to provide another estimate of the
number of striker inmates who have multiple strikes on record
stemming from a single set of circumstances.
To identify individuals with one or more strikes committed as
juveniles, we obtained the Offender Based Information Tracking
System data from Corrections’ Division of Juvenile Justice. We then
identified the primary offense for each juvenile referral that could
be linked to a currently incarcerated striker. Our legal counsel

15

16

California State Auditor Report 2009-107.2

May 2010

reviewed the elements of each crime as of the date the crime
was committed to determine if it met the definition of a serious
or violent crime as established in state law, or a juvenile offense
counting as a strike, as defined in state law. Corrections’ juvenile
data contained all juvenile records since January 1988 and some
records from as far back as 1972. Because these data do not contain
all juvenile records prior to January 1988, the juvenile history of
some older inmates currently incarcerated under the three strikes
law may have been excluded, and as a result, our analysis likely
understates the number of striker inmates who received one or
more strikes as a juvenile.
Although the Offender Based Information System data that we
obtained does not indicate which prior crimes were considered
when sentencing a criminal as a striker, we attempted to make
reasonable inferences from the information available to us.
However, we identified certain limitations in the data that may
affect the results of our analysis of the striker inmate population.
For example, the three strikes law specifies that a prior conviction
for an offense committed in another jurisdiction counts as a strike
if the offense would have been punishable by imprisonment in
state prison in California and if the offense has all the elements of
a serious or violent felony as defined in California law. However,
Corrections’ data does not include convictions from other
jurisdictions, so we were not able to incorporate them into our
analysis. Without an electronic record of convictions occurring
outside of California, it is difficult to determine whether previous
crimes committed within California or convictions from another
jurisdiction were the underlying cause of an inmate being sentenced
as a striker.
Finally, our analysis uses the versions of the laws defining serious
and violent crimes that currently apply under the three strikes law.3
By using the currently applicable definitions of serious and violent
felonies when determining the number of striker inmates with
one or more strikes committed as a juvenile, we may have counted
strikers whose prior juvenile offense was not considered serious or
violent at the time they were sentenced for their current conviction
because voters and the Legislature have amended the laws defining
which offenses are considered serious and violent to include
additional felonies. Further, by using the currently applicable
definitions of serious and violent crimes when identifying striker
inmates who committed multiple strikes during one criminal
offense, we may have included strikers with one or more prior
convictions that were not considered serious or violent when they
3	

State law specifies that the statutes referenced in the three strikes law, including the statutes
defining serious and violent felonies, are the versions of those statutes that existed on
September 20, 2006.

California State Auditor Report 2009-107.2

May 2010

were sentenced for their current conviction. However, given the size
of the striker inmate population, the number of crimes committed,
and the nature of the data available to us, we determined that this
was the most efficient method of accomplishing our audit objective.
To determine the extent to which Health Care Services
currently uses or plans to use telemedicine, we interviewed key
management at Health Care Services. We reviewed Health Care
Services’ documentation of its planning and budgeting for the
expansion of the telemedicine program. Also, we asked Health
Care Services to estimate the number of visits that they expect
the expanded telemedicine program to replace. To determine
whether Health Care Services’ expansion of telemedicine is
reducing costs, we obtained the budgets of the projects that Health
Care Services indicated are being implemented to facilitate the
expansion of telemedicine and attempted to compare the expected
costs of the telemedicine expansion project with the expected
savings of increased telemedicine usage.
To determine the number of vacant positions for the past five years
at Corrections, we obtained the State Controller’s Office position
roster file for fiscal years 2004–05 through 2008–09. By comparing
the total number of authorized positions and the number of those
positions that were paid, we calculated the number of vacancies for
custody staff positions, including correctional officers, sergeants,
and lieutenants. We also interviewed staff at three of Corrections’
institutions and key management at headquarters regarding
vacant positions.
To determine whether the number of vacant positions affects
annual overtime costs, we reviewed Corrections’ policies and
procedures on staffing and interviewed key management. We
obtained information regarding Corrections’ staffing formulas and
interviewed staff at three institutions to obtain their perspective
on how the formula is implemented. We also obtained records
from the State Controller’s Office payroll system regarding the
number of hours of overtime worked by custody staff during
fiscal years 2004–05 through 2008–09. We compared the
number of actual hours of overtime worked with the number
of hours positions were vacant during this period. Additionally,
we asked Corrections to identify the key policy changes that
affected overtime, leave, and vacancies during this period and
analyzed the overtime, leave, and vacancy data in an attempt to
determine the causes for overtime.
We relied upon various electronic data in performing this audit.
The U.S. Government Accountability Office, whose standards we
follow, requires us to assess the sufficiency and appropriateness of
computer processed data. To identify the striker inmate populations

17

18

California State Auditor Report 2009-107.2

May 2010

for the various scenarios described earlier and to determine the
additional cost of inmates incarcerated under the three strikes law,
we used information from the Offender Based Information System.
We assessed the reliability of the Offender Based Information
System by performing electronic testing of key data elements and
by testing the accuracy of the data. To test the accuracy of the
data, we selected a random sample of inmates and traced key data
elements to source documents. During the testing, we identified
errors in the inmate identification information that we used for
associating striker inmates with crimes they committed in the
past. Therefore, we concluded that these data were not sufficiently
reliable for the purposes of this audit related to the analysis of prior
crimes. Further, we did not conduct completeness testing because
the source documents required for this testing are stored at the
33 institutions located throughout the State. Nevertheless, we
present the results of our analysis of these data as they represent the
best available source of information.
To identify the striker inmates who received a strike as a juvenile,
we also used information from the Offender Based Information
Tracking System. We assessed the reliability of the Offender Based
Information Tracking System by performing electronic testing of
key data elements and by testing the accuracy of the data. To test
the data’s accuracy, we selected a random sample of inmates and
traced key data elements to source documents. During the testing
we identified errors in the inmate identification information that
we used for associating striker inmates with crimes they committed
in the past. Therefore, we concluded that these data were not
sufficiently reliable for the purposes of identifying striker inmates
who received a strike as a juvenile. In addition, we did not conduct
completeness testing because the source documents required for
this testing are stored at multiple juvenile facilities throughout the
State. Nevertheless, we present the results of our analysis of these
data as they represent the best available source of information.
We determined that the data we obtained from the State
Controller’s Office payroll system were sufficiently reliable for
the purposes of presenting data on overtime. We assessed the
reliability of the payroll data by performing electronic testing of
key data elements. In addition, we reviewed testing of the payroll
system’s major control features performed as a part of the State’s
financial audit.
We determined that the data we obtained from the State
Controller’s Office California Leave Accounting System were
sufficiently reliable for the purposes of identifying the amount
of leave used and accrued by custody staff. We assessed the
reliability of the leave data by performing electronic testing of

California State Auditor Report 2009-107.2

May 2010

key data elements. In addition, we reviewed testing of the leave
system’s major control features performed as a part of the State’s
financial audit.
We determined that the data we obtained from the State
Controller’s Office position roster file were sufficiently reliable
for the purposes of identifying the number of custodial positions
authorized by Corrections. We assessed the reliability of the roster
file by performing electronic testing of key data elements and by
testing the accuracy and completeness of the data. To test the
accuracy of the data, we selected a random sample of custodial
staff and traced key data elements to source documents. To test
the completeness of the data, we pulled a haphazard sample of
source documents and verified that corresponding entries existed in
the data.
To identify the costs associated with adult health care we used
information from the CMD. We assessed the reliability of the
database by performing electronic testing of key data elements and
by testing the accuracy and completeness of the data. To test the
accuracy of the data, we selected a random sample of inmate health
care costs and traced data elements to source documents. To test
completeness we pulled a haphazard sample of source documents
and verified that corresponding entries existed in the data. We
concluded that these data were sufficiently reliable for the purpose
of this audit.

19

20

California State Auditor Report 2009-107.2

May 2010

Blank page inserted for reproduction purposes only.

California State Auditor Report 2009-107.2

May 2010

Chapter 1
LONGER SENTENCES DUE TO THREE STRIKES REPRESENT
A SIGNIFICANT COST
Chapter Summary
The California Department of Corrections and Rehabilitation
(Corrections) housed 171,500 inmates as of April 2009. At that time,
inmates sentenced under the three strikes law totaled 43,500, or
approximately 25 percent of the total inmate population in California
institutions. As discussed in the Introduction, the three strikes law
provides a minimum sentence that is double the sentence a court
would have imposed on a person who is convicted of a felony and
who also has one prior conviction for a serious or violent felony;
this type of offender is often referred to as a second striker. The
three strikes law also generally provides a minimum sentence of
25 years to life imprisonment for someone who is convicted of a
felony and who has two or more prior convictions for a serious or
violent felony; this type of offender is often referred to as a third
striker. Courts commonly refer to a person’s prior convictions for
serious or violent felonies as strikes. In our prior report4 we found
that inmates sentenced under the three strikes law, known as striker
inmates, were sentenced on average to an additional nine years of
incarceration due to the three strikes law. Further, we found that
these additional years represent $19.2 billion in additional costs over
the duration of the sentences of current striker inmates.
In preparing this report, we determined that many striker
inmates’ current convictions are for felonies that are not defined
as serious or violent, and thus are not strikes. We also estimated
that $7.5 billion of the $19.2 billion in additional costs associated
with striker inmates is attributable to strikers whose current
convictions are for felonies that are not strikes. Although we
identified specific populations of striker inmates and the additional
years these inmates are sentenced to, some limitations exist in
Corrections’ data that limited our analysis and may affect various
analyses performed by Corrections.
Inmates Sentenced Under the Three Strikes Law Are a Significant
Portion of the Inmate Population
In April 2009 Corrections housed more than 43,500 inmates
incarcerated under the three strikes law, representing 25 percent
of the 171,500 inmates under Corrections’ responsibility.
4	

2009-107.1: California Department of Corrections and Rehabilitation: It Fails to Track and Use Data
That Would Allow It to More Effectively Monitor and Manage Its Operations, September 2009.

21

22

California State Auditor Report 2009-107.2

May 2010

Figure 2 shows that strikers have made up approximately the
same percentage of the population since 2001. As discussed in
the Introduction, California enacted the three strikes law in 1994,5
with the intent of ensuring longer prison sentences and greater
punishment for those who commit a felony and were convicted
previously of a serious or violent felony. The list of offenses that
are considered strikes has expanded since California enacted the
three strikes law in 1994. As Figure 2 shows, in 2000 voters approved
Proposition 21, which expanded the list of offenses that constitute
strikes. In 2006 the Legislature again expanded this list of offenses.
Figure 2
Striker Inmates as a Percentage of the Inmate Population and Time Line of the Three Strikes Law
Three strikes law enacted.

Percentage of Inmate Population

25%

Proposition 21 approved by voters,
expands the list of serious and
violent felonies.

Legislature further expands the list of
serious and violent felonies.

People v. Superior Court
(Romero) authorized a court to
dismiss a prior strike without a
prosecutor’s request.

3rd strikers
2nd strikers

20
15
10
5
0
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
Year*

Sources:  California Department of Corrections and Rehabilitation’s Quarterly Second and Third Strike Inmate Population reports, Historical Trends,
1987–2007 report, Prison Census Data reports 2008 and 2009, sections of the California Penal Code, and a court decision.
*	 Population as of December of each year.

Many Striker Inmates Received Longer Sentences for Felonies That
Are Not Strikes
The three strikes law imposes longer sentences for inmates who
commit any felony and have previously been convicted of a violent
or serious felony. The felony for which the offender is currently
convicted need not be serious or violent for an individual to be

5	

The Legislature enacted the three strikes law in March 1994, and in November 1994 voters
approved Proposition 184 to enact a virtually identical version of the law.

California State Auditor Report 2009-107.2

May 2010

sentenced under the three strikes law. In fact, our review of the
controlling offenses of Corrections’ striker population found that
the current convictions of more than half of the striker inmates
are not strikes. As shown in Table 2, 23,099 striker inmates fit this
description, representing 53 percent of all strikers as of April 2009.
In addition, the audit request asked us to determine the number
of striker inmates who received multiple strikes committed during
one criminal offense or who committed one or more strikes as a
juvenile. The table shows the number of inmates we identified in
each category. We noted some overlap among the three categories.
For example, of the 23,099 inmates whose current convictions are
not strikes, 4,477 also committed more than one serious or violent
offense on the same day, and 793 committed one or more serious or
violent offenses as a juvenile. Further, as the table shows, 233 striker
inmates meet all three of these criteria.
Table 2
Striker Inmates by Category
As of April 2009

Number of
Striker Inmates

CONVICTED OF COMMITTING WHO COMMITTED ONE
SERVING FOR A
OTHER
OR MORE SERIOUS OR
MULTIPLE SERIOUS OR
CURRENT OFFENSE
THAT IS NONSERIOUS VIOLENT OFFENSES ON THE VIOLENT OFFENSES AS STRIKER
INMATES*
A JUVENILE
SAME DAY
AND NONVIOLENT

Total Individuals
Meeting This Criteria

17,596
793†
4,477†
233†
23,099
53%‡

8,056

671

11,057

793†
631†
4,477†

631†

233†

233†

13,397
31%‡

2,328
5%‡

11,057
25%

Source:  Bureau of State Audits’ analysis of data obtained from the California Department of
Corrections and Rehabilitation’s Offender Based Information System and Juvenile Offender Based
Information Tracking System.
*	 Included in the category of Other Striker Inmates are striker inmates who were excluded from the
analysis of the three categories of striker inmates shown in this table. The convictions associated
with these inmates were for crimes whose descriptions in law are sufficiently broad that some
instances of the offenses identified could be considered a strike and others not.
†	 The total striker inmates as of April 2009 was 43,514. Because inmates with convictions that
met the criteria of more than one of the three categories of striker inmates are included in
each applicable column there is overlap and therefore the column amounts do not agree to
the total number of striker inmates.
‡	 Percentage of total inmate strikers within category (includes overlap with other categories).

Many offenses for which striker inmates currently are serving
sentences are not on the list of crimes defined as serious or
violent and thus do not count as strikes, but they are not trivial

23

24

California State Auditor Report 2009-107.2

May 2010

offenses. They include crimes such as certain types of battery,
assault, and various sex crimes. According to the general
categories in which Corrections’ data classifies crimes, of the
23,099 strikers whose current offenses are nonserious and
nonviolent, 4,180 were categorized as crimes against persons.
An additional 7,213 were classified as property crimes, including
burglary in the second degree, and 8,065 crimes were categorized
as drug crimes, such as possession of certain controlled substances.
Finally, 3,328 were classified as other, and 313 crimes in the data did
not include a classification.
Corrections’ Data Lack Details Necessary to Accurately Identify
Certain Striker Inmate Characteristics
Although we present information in Table 2 on the number of
striker inmates who met certain criteria, the data have several
limitations, as we discussed in the Scope and Methodology. In
addition, as we discuss later in this chapter, we found instances
in which data on sentencing guidelines were outdated or incorrect,
which complicated our analysis of striker inmates. However,
according to a correctional case records administrator (correctional
administrator), the Offender Based Information System is the
only system that Corrections has available to identify all inmates
sentenced under the three strikes law. Because of the limitations
of the data, such as the fact that Corrections’ data do not indicate
which prior convictions were considered when sentencing inmates,
we were not able to ensure that the convictions we identified were
in fact the convictions leading to the striker sentence.
There are some limitations in
Corrections’ data. For example,
although the three strikes law
states that prior convictions for
offenses committed in another
jurisdiction may count as a strike,
such information is not included in
the data.

For example, although the three strikes law states that prior
convictions for offenses committed in another jurisdiction
count as a strike if the offense would have been punishable by
imprisonment in state prison in California and if the offense
has all the elements of a serious or violent felony as described
by California  law, the correctional administrator told us that
Corrections’ data do not include information on offenses
committed in other states. Although knowledge of which of an
inmate’s prior convictions were considered in their sentencing is
not crucial information for Corrections’ day-to-day operations, it is
important to realize that we cannot determine the extent to which
the convictions we identified were the convictions that led to an
inmate’s current sentence. Some inmates may have been sentenced
as strikers because of crimes other than those we identified or may
have prior convictions for crimes that are strikes in addition to
those we identified.

California State Auditor Report 2009-107.2

May 2010

Additionally, although we attempted to identify all convictions
related to each individual, it is possible that some convictions were
not appropriately related. Corrections assigns inmates a new
identification number if they have been discharged and then
returned to prison for committing a new crime. Because of this
practice, inmates who return to prison have multiple identification
numbers, which results in multiple records for some individuals.
Although Corrections performs a manual process to identify other
identification numbers for an inmate, we have concerns regarding
the accuracy of the data used in this process. We are unsure
whether Corrections accurately identified all relevant records,
but it would not be cost-effective for the purposes of this audit to
review the paper records of all inmates. Therefore, we relied on
Corrections’ analysis to identify the relevant crimes associated
with an individual. As a result, it is possible that some crimes we
identified are not in fact related to currently incarcerated strikers,
or that there are additional crimes related to currently incarcerated
strikers that we did not review.
Corrections’ data show that many striker inmates committed
multiple serious or violent offenses on the same day. However, as
described in the Scope and Methodology section, some of these
offenses may not have arisen from the same set of circumstances
but may merely have occurred on the same day. As a result, the
13,397 striker inmates we identified who were convicted of multiple
offenses committed on the same day could include some inmates
whose multiple convictions were related to multiple sets of
circumstances. For this reason, we performed additional analysis
to identify inmates who would not have been counted had stayed
sentences been excluded. When the court stays a sentence, it can
be an indication that multiple convictions arose from the same set
of circumstances. Specifically, if a person is guilty of a criminal act
or omission that is punishable in different ways under different
laws, state law prohibits punishing the person under more than
one law, requires that punishment be imposed under the law that
provides the longest potential prison sentence, and requires a court
to stay the sentence for the law that provides the shorter sentence.
Thus, the number of these inmates with stayed sentences may be a
more accurate measure of inmates whose convictions arose from
the same set of circumstances. However, it is possible that courts
may have chosen to stay the sentences associated with some of
these convictions for other reasons, or that some sentences were
not stayed due to an exception to this requirement described in the
following paragraph.
Of the 13,397 striker inmates convicted of committing multiple
serious or violent offenses on the same day, 1,753 would not
have been counted if stayed sentences had been excluded.

Because Corrections assigns
inmates a new identification
number if they have been
discharged and then returned
to prison for committing a
new crime, some inmates have
multiple records.

25

26

California State Auditor Report 2009-107.2

May 2010

Sentences stayed for the reason described earlier must meet
certain conditions. If a sentence is to be stayed due to this section
of law, the circumstances must be indivisible; that is, the individual’s
actions are reviewed to determine whether they arose from a single
underlying intent to take an action in violation of the law. However,
if the actions are crimes of violence against different victims, there
is an exception to the requirement to stay the sentence. Because the
number of relevant inmates depends on the precise definition of
the words single offense, we calculated both figures.

The ability to identify juvenile
history of some older inmates
incarcerated under the three strikes
law is limited because Corrections’
juvenile data does not contain all
records prior to January 1988.

Compared with the number of striker inmates committing multiple
offenses on the same day, the data indicated that relatively few
striker inmates committed serious or violent offenses as a juvenile.
However, as also described in the Scope and Methodology section,
Corrections’ juvenile data does not contain all juvenile records
prior to January 1988, which limits our ability to identify the
juvenile history of some older inmates incarcerated under
the three strikes law as of April 2009. Because the three strikes
law allows prosecutors and judges to take into account serious
and violent crimes that were committed before the three strikes
law came into existence, our analysis likely excludes some current
striker inmates’ serious or violent offenses committed prior to 1988
when they were juveniles.
Finally, it is important to note that we excluded the convictions
of some striker inmates from our analysis because we were
unable to determine whether or not their convictions constituted
strikes. Specifically, we excluded convictions of 760 inmates from
the strikers whose current offense was not serious or violent,
346 inmates from the strikers who committed multiple serious
or violent offenses on the same day, and 235 inmates from the
strikers who committed one or more serious or violent felonies as a
juvenile. The convictions of these individuals were for crimes whose
descriptions in law are sufficiently broad that in some instances
the offenses could be considered serious, and thus a strike, and in
other instances they would not be considered serious. For example,
conspiracy to commit a serious felony is a strike. Therefore, if
an inmate was convicted of conspiring to commit murder, that
conspiracy would be considered a serious felony under state law.
However, if an inmate was convicted of conspiring to commit a
crime that is not a serious felony, the conspiracy offense would not
be a strike. According to a correctional case records administrator,
the only way to verify whether such offenses were considered
serious would have been to review the legal documents associated
with the inmates’ individual cases, and we determined that such a
review would not be cost-effective for the purposes of this audit.

California State Auditor Report 2009-107.2

May 2010

The Additional Years of Incarceration Imposed Due to the
Three Strikes Law Represent a Significant Cost
As we discussed in our prior report, we estimate that striker
inmates received sentences that are, on average, nine years longer
than if they had not been sentenced under the three strikes law.
These additional years represent a significant cost to the State. In
our previous report we calculated a total cost of $19.2 billion for the
additional years imposed on these inmates’ sentences. To calculate
this cost, we used the average cost of incarceration for fiscal
year 2007–08—a figure that includes all costs associated with adult
correctional operations, including administration and overhead.
However, the actual number of additional years of incarceration due
to three strikes could be even greater due to the fact that strikers
cannot earn as much credit toward an early release.
As described in our prior report, we determined the additional
cost of striker inmates by identifying the sentence for the
controlling offense, if it was related to a three strikes case, or
the longest sentence related to a three strikes case, for striker
inmates currently housed in Corrections’ adult institutions. We
then compared the estimated lengths of these sentences with an
estimate of the prison terms they might have received had they not
been sentenced under the three strikes law. To estimate the term
an inmate would have received in the absence of the three strikes
law, we identified the midterm of the possible prison terms
prescribed by state law for the crime for which the striker inmate
was incarcerated. If there were no midterm, we used the maximum
prison term specified in state law. We then added any applicable
enhancements—additional prison terms prescribed by state law
under certain circumstances. For example, an inmate currently
convicted of petty theft who had two prior convictions for serious
or violent felonies receives a minimum sentence of 25 years to life
under the three strikes law. Had he been sentenced to the midterm
of the normally applicable sentence, he would have received
two years. In this case, we estimated that the inmate would be in
prison for 23 years longer under the three strikes law. Additionally,
as discussed in the Introduction, a striker defendant convicted
of multiple current offenses generally must serve consecutive
sentences rather than concurrent sentences. Finally, we excluded
inmates whose sentences were likely not to have been increased by
the three strikes law.
Our analysis does not take into account certain factors that could
affect this additional cost calculation. For instance, the law limits
the amount of credit that striker inmates can earn against their
prison sentences to a maximum of 20 percent for participating in
work, training, or education programs, while other inmates can
reduce their prison sentences by more than 50 percent. However,

The actual number of additional
years of incarceration due to
three strikes may be greater since
striker inmates cannot earn as
much credit toward an early release.

27

28

California State Auditor Report 2009-107.2

May 2010

we did not take this factor into consideration because poor behavior
can reduce inmates’ credits and alter their credit‑earning status.
Additionally, to provide a more realistic estimate, we considered
inmates’ ages when calculating the estimated additional years of
incarceration they actually would serve if they reached the full
life expectancy for Americans, estimated at 77.7 years by the
Centers for Disease Control and Prevention.6 Finally, according to
Corrections, on average 54 percent of all paroled felons released
from prison for the first time in 2005 returned to prison within
two years of being released. This statistic—known as the recidivism
rate—would affect the results, but we did not consider it in our
analysis because it too is dependent upon inmate behavior.

The number of second‑striker
inmates with nonserious or
nonviolent convictions is more
than four times the number of
third‑striker inmates—yet, the
third strikers were sentenced
to nearly twice the number of
additional years.

For this report we performed additional analyses of the striker
inmate population. Specifically, in an analysis of the additional
years imposed by the three strikes law for inmates who meet
one or more of the three criteria specified in the audit request, we
found that the number of second-striker inmates who met one or
more of these criteria was significantly higher than the number
of third‑striker inmates. For example, as shown in Figure 3, the
number of second‑striker inmates whose current conviction is for
an offense that is not serious or violent was 19,045, or more than
four times the number of third-striker inmates meeting the same
criterion. Despite this fact, these third strikers were sentenced to
96,589 additional years, or nearly twice the 55,327 additional years
to which the second strikers were sentenced. This is because under
the three strikes law, third strikers generally receive a minimum
sentence of 25 years to life, but second strikers receive double
the normally imposed sentence. Because there is a significant
difference in the number of additional years to which second
and third strikers are sentenced, of the striker inmates who met
one or more of the criteria we reviewed, the cost of the longer
sentences for third strikers represents the majority of the additional
cost of the three strikes law.
In addition, we found that inmates who were convicted of
multiple serious or violent offenses occurring on the same day
were responsible for the largest portion of the estimated cost of
the additional years resulting from sentencing under the three
strikes law. The additional years imposed by the three strikes law
for these inmates represent $9.2 billion over the duration of their
incarceration, using the average cost of incarceration of $49,300
per year. Further, as shown in Figure 3, we estimate that the costs of
the additional years for striker inmates currently incarcerated for a
crime that is not serious or violent represents $7.5 billion over the
duration of their incarceration.

6	

Subsequent to our analysis, the Centers for Disease Control and Prevention reported that life
expectancy in the United States has risen to 77.9 years of age.

California State Auditor Report 2009-107.2

May 2010

Figure 3
Additional Years and Estimated Cost of Striker Inmates’ Sentences
Convicted of committing
multiple serious or violent
offenses on the
same day (13,397)

$6.1 billion
5,234 inmates

Serving for current offense
that is nonserious and
nonviolent (23,099)

$4.8 billion
4,054 inmates

$3.1 billion
8,163 inmates

Third striker
Second striker
Remaining inmates

$2.7 billion
19,045 inmates

$668 million
521 inmates

Committed one or more
serious or violent offenses
as a juvenile (2,328)

$508 million
1,807 inmates
$4.8 billion
1,902 third-striker inmates and
9,155 second-striker inmates

Other striker inmates (11,057)

0

20,000

40,000

60,000

80,000 100,000 120,000 140,000 160,000 180,000 200,000
Total Additional Years

Sources:  Bureau of State Audits’ estimate of costs based on data from California Department of Corrections and Rehabilitation’s Offender Based
Information System and juvenile Offender Based Information Tracking System.
Note:  Due to the fact that some striker inmates fall into more than one category, as illustrated in Table 2 on page 23, the number of strikers
represented in the figure above add up to more than the total number incarcerated. Because strikers whose characteristics meet the requirements of
more than one of the categories, the total years and dollars would need to be reduced by 68,180 years, or $3.4 billion, to correctly represent the total
additional years and cost.

To estimate the cost of the additional years of incarceration for
individuals sentenced under the three strikes law, we used the
average cost of incarceration for fiscal year 2007–08 as calculated
in our prior report. Although Corrections calculates a marginal cost
figure—an estimate of the cost to house one additional inmate in a
Corrections’ institution under overcrowded conditions, also known
as the overcrowding rate—we chose not to use this figure for
several reasons. Corrections provided a summary showing that, as
of March 2010, it is housing roughly 10,300 inmates in overcrowded
conditions, meaning that inmates are housed in nontraditional
beds in dayrooms or gymnasiums. However, as of April 2009, more
than 43,500 inmates were incarcerated under the requirements
of the three strikes law, many for serious or violent crimes, which
would reduce the likelihood that Corrections could place them in a
nontraditional housing area.
Further, as described in the Introduction, in August 2009 a federal
three-judge panel ordered Corrections to provide the court with
a plan that would reduce the population of Corrections’ adult
institutions over the next two years. The court order states that

29

30

California State Auditor Report 2009-107.2

May 2010

Corrections’ plan should reduce the population of its institutions
to 137.5 percent of their combined design capacity. According to
Corrections, the court order would require a reduction of more
than 40,000 inmates. Reducing the number of striker inmates
would be difficult, however, due to the requirements of the
three strikes law. Specifically, the additional years of incarceration
imposed by the three strikes law requires Corrections to house
these inmates for a longer period of time. In addition, striker
inmates cannot earn as much credit to reduce their sentences
as non-striker inmates, meaning Corrections cannot reduce the
overcrowding by paroling these inmates more rapidly.
Further, even if Corrections were to address this issue through
a reduction in the inmate population, the reduction would not
necessarily result in a proportional decrease in costs. Although
the custody costs associated with overcrowding would be reduced,
institutions operating at 137.5 percent of design capacity would still
be incarcerating more inmates than they were designed to house,
suggesting that administrative and overhead costs may be similar.
Because of the longer sentences and the reduced credit earning
opportunities imposed by the three strikes law, striker inmates
will likely continue to represent a significant portion of the inmate
population. Therefore, the average cost of incarceration appears to
be a more appropriate figure for determining the additional cost of
the three strikes law.

Strikers must serve at least
80 percent of their sentence before
they are eligible for release while
non-striker inmates can be paroled
after serving less than 40 percent of
their sentences.

In our prior report we estimated that striker inmates were
sentenced to an additional 389,000 years of incarceration due
to the requirements of the three strikes law. However, the actual
difference in time served between strikers and non-strikers may
be even greater than the difference in sentences. According to
Corrections’ Spring 2009 Adult Population Projections report for
fiscal year 2007–08, newly admitted inmates and parole violators
who were returned with new terms were sentenced to an average of
4.1 years. In its 2007 and 2008 demographic reports on prisoners
and parolees, Corrections reported that, on average, inmates were
first paroled after serving two years in 2007 and 2.1 years in 2008.7
This is because inmates may earn credit toward an early release by
participating in certain programs. However, strikers must serve
at least 80 percent of their sentence before they are eligible for
release. In contrast, due to a change in state law that occurred in
January 2010, non-striker inmates can earn credit that allows them
to be paroled after serving less than 40 percent of their sentences.
Although the amount of credit earned also depends on inmate
behavior, the number of additional years of incarceration due to

7	

The amount does not include the time served by some parole violators, because these individuals
returned to prison for a parole violation and were not sentenced to a new term.

California State Auditor Report 2009-107.2

May 2010

three strikes could be even greater than indicated in Figure 3 on
page 29 because strikers cannot earn as much credit toward an
early release. Because the credit earned by individuals depends on
their behavior, we did not attempt to more precisely calculate the
additional years served by striker inmates.
Outdated and Erroneous Information Reduces the Usefulness of
Corrections’ Data
During our analysis of data related to striker inmates, we identified
some concerns with the Offender Based Information System data
we obtained from Corrections. We found some inmate convictions
that were associated with outdated sentencing data and some
sentencing information that indicated specific versions of crimes
were active for one day, indicating errors in the data. Although
Corrections indicated that it intends to prepare its new data system
with accurate information, its plan to correct existing data before
they are put into the new system does not appear to address all
the areas with which we have concerns. These errors may require
Corrections’ staff to perform additional analysis to determine if an
individual’s sentence is accurate and may reduce the accuracy in
projecting the terms of inmate incarceration when estimating the
inmate population for future years.
In our analysis of Corrections’ data related to inmates, we identified
some convictions that had been assigned outdated sentencing
information—such as the possible term lengths and requirements
for notifying local law enforcement upon the inmate’s release. This
not only complicated our analysis, but also may reduce the accuracy
of some analyses performed by Corrections, such as evaluating the
accuracy of sentences imposed by the judicial system, or estimating
the average daily population in future years for the purpose of
estimating budgetary costs or savings. Specifically, when comparing
inmates’ offense dates to the sentencing information associated
with the offense, we identified approximately 85,000 convictions
that appeared to be associated with outdated information. When
we discussed our concerns with Corrections, a correctional case
records administrator (administrator) stated that case records
updates its sentencing information to reflect changes in the law
once a year and sometimes only once every two years. However,
the administrator told us that after the new laws go into effect, case
records’ staff do not subsequently review convictions associated
with the sentencing information that has been updated. Further,
the administrator told us that there are situations when case
records’ staff will correct sentencing information, but some inmate
convictions associated with the incorrect sentencing information
may go undetected.

We identified approximately
85,000 convictions in Corrections’
data that appeared to be
associated with outdated
sentencing information.

31

32

California State Auditor Report 2009-107.2

May 2010

Corrections’ record of one inmate’s
conviction is associated with
information that had not been
accurate for approximately
22 years prior to the date the
offense occurred.

In one case, Corrections’ record of an inmate’s conviction for
injuring a spouse or cohabitant indicates that the offense occurred
in March 2003. However, this conviction is associated with
sentencing information that was effective between July 1977 and
December 1980. As a result, although state law was revised several
times subsequent to 1980, this conviction is associated with
information that had not been accurate for approximately 22 years
prior to the date that this offense occurred. Further, changes to
state law in 1980 imposed a longer sentence for this offense than
the previous version of this law. Therefore, the conviction for this
offense is associated with sentencing information that does not
agree with state law at the time the offense occurred.
According to the administrator, inmate sentences imposed
by the judicial system are based on legal documents and are
tracked separately from the table that contains sentencing data in
Corrections’ data system, so errors in the sentencing information
do not affect the actual sentences that inmates serve. However,
she stated that convictions associated with incorrect sentencing
information may require Corrections’ staff to perform additional
analysis to determine if an individual’s actual sentence was
inappropriate. According to the chief of Corrections’ offender
information services branch, incorrect sentencing information
could lead to inaccurate estimates of the average daily inmate
population in future years, which are used to estimate budgetary
costs or savings. These errors also can increase Corrections’
workload in other ways.
For example, state law requires Corrections to notify local law
enforcement and the district attorney 45 days before the release
of an inmate in their jurisdiction who was convicted of a violent
felony. Our analysis of data indicated that a high number of
convictions were associated with outdated sentencing information,
and we were concerned that inmates associated with incorrect
sentencing information in the database were not always identified
in a way that would trigger Corrections’ staff to comply with the
notification requirements.
When we discussed our concern with Corrections, the
administrator stated that analysts perform audits of the files of all
inmates who are scheduled to be paroled or released within 60 days
to 75 days. As part of this audit, analysts determine whether an
inmate is violent and requires notification of any parties before
release. We reviewed a random sample of inmates requiring
notification of law enforcement and district attorneys before release
and found that for the records we reviewed, Corrections’ staff
sent proper notifications regarding the inmate’s release to parole.
Although Corrections’ manual process appears to be effective for
meeting the notification requirements, there may be an opportunity

California State Auditor Report 2009-107.2

May 2010

for Corrections to increase efficiency by ensuring that the
information in the database is accurate, thus reducing the need for
manual review by case records analysts.
Additionally, we found that some sentence information is
inaccurate and that some inmate convictions are associated with
this inaccurate information. For example, 42,000 of the 2.8 million
convictions we considered in our analysis were associated with
sentencing information related to 53 offenses that—according to
Corrections’ records—were effective for only one day, indicating
errors in the data. For instance, several convictions for grand
theft of an automobile or firearm were associated with sentencing
information that was effective for only one day. Some of these
convictions are for offenses that occurred during 2001; however,
the sentencing information associated with these convictions
was effective only on January 1, 1997. When we asked about some
of these records, the administrator told us that the ending date
for these records reflects the same day or day after the beginning
date. She stated that these records indicate errors in Corrections’
sentencing data that could not be deleted. We found thousands of
convictions associated with this erroneous information, which may
reduce the accuracy of some of Corrections’ analyses and increase
its workload.
Corrections is developing a new data system, the Strategic Offender
Management System. According to the project’s director, the new
system is intended to replace approximately 40 separate systems
and databases, provide consistency among juvenile and adult
data, track each inmate from intake through parole, and provide
inmate history upon return to prison. As part of the process of
implementing this new system, the project director stated that the
project’s data conversion unit would clean up the data that will be
transferred into the new system. Specifically, she stated that the
outdated sentencing information and other offenses that were out
of the ordinary, such as a crime being effective for only one day,
should be identified and cleaned up during the second release
of phase one of the project. The project director indicated that
this release is expected in October 2011. Although the project
director indicated that cleanup of the existing data will be part
of implementing the new system, Corrections’ data processing
manager—who is overseeing the data conversion—told us that
the current plan is to ensure that the edit tables, which include
sentencing information, are set up correctly in the new system,
but the plan does not address in detail how historical data will be
reviewed or corrected. Although the new data system may result
in more accurate information going forward, it is important that
Corrections ensure that its historical data is as accurate as possible
before they are converted into the new system. If Corrections is

33

34

California State Auditor Report 2009-107.2

May 2010

unable to ensure that its historical data are accurate, the results of
some analyses of inmates’ prior convictions and sentences may be
less reliable.
Recommendations
To address the erroneous sentencing information and
inappropriately assigned convictions in its data system,
Corrections should:
•	 Complete its cleanup of data that will be transferred into the
new system, ensuring that this review includes a detailed
evaluation of convictions that have been assigned outdated
sentencing information as well as deleting erroneous sentencing
information, before it begins using its new data system.
•	 Create a schedule for regular checks of the accuracy of existing
sentencing information, as well as the accuracy with which
sentencing information has been assigned to convictions.

California State Auditor Report 2009-107.2

May 2010

Chapter 2
A SMALL PORTION OF THE INMATE POPULATION
ACCOUNTS FOR MOST CONTRACTED SPECIALTY
HEALTH CARE COSTS
Chapter Summary
A significant contributor to the cost of housing inmates is their
health care. As we describe in our prior report,8 the California
Department of Corrections and Rehabilitation (Corrections)
tracks costs at the institutional level but does not track costs in
a manner that would allow their stratification by specific inmate
characteristics. As a result, we were unable to associate costs
for medical care provided by institution medical staff, including
primary care, with specific inmates. However, using the Contract
Medical Database (the CMD) maintained by California Prison
Health Care Services (Health Care Services), we were able to
associate some contracted medical specialty health care costs with
specific inmates.
In reviewing this information, we found that a large proportion
of the contracted specialty health care costs were for inpatient
hospital care, including acute medical and surgical care, and that
most specialty health care costs are associated with a small number
of inmates. Although our analysis also found that the health care
of older inmates was generally more costly, the average costs
associated with inmates that died were significantly greater than
the average costs for any specific age group. In November 2009 the
federal court-appointed receiver (receiver) submitted a required
report to the Legislature discussing several cost containment
measures to address rising health care costs. As part of these
efforts, Health Care Services is continuing to expand its use of
telemedicine and estimates significant costs for several projects
that it believes are necessary for this expansion, but it has not fully
estimated how much the telemedicine program expansion will save.
The Largest Portion of Contracted Specialty Health Care Costs Were
Classified as Inpatient Hospital Stays for Medical and Surgical Care
As shown in Table 1 on page 8 in the Introduction, inmate health
care costs totaled $2.1 billion during fiscal year 2007–08. Much
of this cost is related to primary care provided by physicians

8	

2009-107.1: California Department of Corrections and Rehabilitation: It Fails to Track and Use Data
That Would Allow It to More Effectively Monitor and Manage Its Operations, September 2009.

35

36

California State Auditor Report 2009-107.2

May 2010

For fiscal year 2007–08, $529 million
of the $734 million in contracted
health care costs was related to
specialty health care.

and nurses within Corrections’ institutions. As we described
in our prior report, Corrections tracks costs at the institutional
level but does not track them in a manner that would allow for
their stratification by specific inmate characteristics. As a result,
we were unable to associate costs for medical care provided by
institution medical staff, including primary care, with specific
inmates. However, using Health Care Services’ CMD, we were able
to associate a significant portion of contracted medical specialty
health care costs with specific inmates. According to the data in
the CMD for fiscal year 2007–08, Health Care Services incurred
more than $734 million in contracted health care costs. Of this
amount, $205 million was related to the salaries of temporary
contracted staff hired to fill vacant health care positions within
the institutions—referred to as registry costs. The remaining
$529 million was related to specialty health care services provided
to inmates through contracted providers, of which we were
able to associate more than $469 million, or 89 percent, with valid
inmate identification numbers. We were unable to associate the
remaining $60 million in contracted specialty health care costs with
specific inmates.
Of the $529 million in contracted specialty health care costs, Health
Care Services’ CMD classified more than $291 million as inpatient
hospital care and nearly $122 million as outpatient hospital care.
Each of the remaining health care delivery methods—such as costs
for laboratory and X-ray procedures, dialysis, and telemedicine—did
not represent a significant amount individually. For example, care
provided via telemedicine totaled $786,000. In total, these other
delivery methods cost $116 million.
Our review indicated that a few types of contracted specialty
health care represented a significant portion of the costs. By far, the
majority of contracted specialty care costs were for inpatient acute
medical and surgical care. As shown in Figure 4, $292 million, or
55 percent of the total specialty health care costs of $529 million,
was for inpatient acute medical and surgical care. Further, an
additional $53 million was for outpatient surgical care. We
combined the remaining service types that were not part of another
category as shown in Figure 4 and each individually represented
less than 2 percent of total specialty health care costs into the
category labeled “Other.” These costs included numerous types
such as cardiology, dermatology, and dialysis care not provided by
Corrections’ staff.
It is important to point out that the variance in the amounts
spent on different types of service does not necessarily relate to
an increased number of inmates served, because some types of
specialty health care are significantly more costly per inmate than
other specialty types. For example, inpatient acute medical and

California State Auditor Report 2009-107.2

May 2010

surgical care totaled $292 million, averaging $29,584 per inmate
for the 9,870 inmates receiving this type of care, while emergency
services totaled $36 million and averaged only $2,787 per inmate for
the 12,940 inmates served.
Figure 4
Contracted Specialty Care Costs by Type of Service
Fiscal Year 2007–08
$300

Dollars in millions

250
200
150
100
50
0

l
s
c
ort
ce
sti
are
ica
sp
rvi
no tory
ed care
n
al c
e
g
a
c
s
m
i
a
r
i ra
y
rg
te al
al t
nc
y/d abo
su
cu gic
dic
rge
nt
t a sur
log nd l
e
e
e
o
n
i
i
M
t
e d
d a
Em
ati an
tpa
Ra
Inp
Ou

r*

he

Ot

Type of Service
Sources:  Bureau of State Audits’ analysis of cost data from California Prison Health Care Services’
Contract Medical Database for fiscal year 2007–08.
*	 This category includes other types of care such as cardiology, dermatology, and dialysis.

Most Specialty Health Care Costs Were Associated With a Small
Population of Inmates, and Older Inmates Were Generally More Costly
Our analysis of the information in the CMD for fiscal year 2007–08
revealed that most specialty health care costs were associated with
a relatively small population of inmates. Specifically, 70 percent of
the inmate population with specialty health care costs averaged just
more than $1,000 per inmate and cost $42 million in total, while the
remaining 30 percent of inmates amassed specialty health care costs
totaling more than $427 million.
Of the specialty health care costs we could associate with specific
inmates, we determined that 58,726 inmates incurred costs for
specialty health care services in fiscal year 2007–08. However,
as shown in Table 3 on the following page, 1,175 of these inmates,

37

38

California State Auditor Report 2009-107.2

May 2010

or 2 percent, represented 39 percent of the total costs. Specialty
health care costs for these 1,175 inmates totaled $185 million, for
an average cost per inmate of nearly $158,000. In contrast, a large
majority of the population of inmates incarcerated during 2007
and 2008 did not have any specialty health care costs. Specifically,
although Corrections reported that more than 171,000 inmates were
incarcerated as of December 31, 2007, and 2008, it also reports that
it paroled, discharged, or otherwise released from custody more than
139,000 inmates and admitted nearly 143,000 inmates during 2008.
As a result, the 1,175 inmates incurring 39 percent of specialty health
care costs in fiscal year 2007–08 represent approximately one-half of
1 percent of the inmates incarcerated during the year.
Table 3
Specialty Health Care Costs for the Highest-Cost to Lowest-Cost Inmates
Fiscal Year 2007–08
PERCENTAGE OF THE
POPULATION WITH
SPECIALTY HEALTH
CARE COSTS

TOTAL COST OF
SPECIALTY HEALTH
CARE FOR INMATES
WITHIN THE RANGE

PERCENTAGE OF
TOTAL SPECIALTY
HEALTH
CARE COSTS

AVERAGE COST
OF SPECIALTY
HEALTH CARE
PER INMATE

1,175 (1 to 1,175)

2%

$185,305,361

39%

$157,707

2,350 (1,176 to 3,525)

4

96,811,134

21

41,196

4,700 (3,526 to 8,225)

8

80,035,638

17

17,029

9,400 (8,226 to 17,625)

16

65,419,783

14

6,960

41,101 (17,626 to 58,726)

70

41,673,892

9

1,014

NUMBER OF INMATES WITHIN
RANGE (FROM HIGHEST COST TO
LOWEST COST)

58,726

$469,245,808

Source:  Bureau of State Audits’ analysis of data from California Department of Corrections and
Rehabilitation’s Offender Based Information System and California Prison Health Care Services’
Contract Medical Database for fiscal year 2007–08.

In our prior report, we noted that older inmates generally require
more health care. Our current analysis of the cost of inmates
receiving specialty medical services by age is consistent with
that conclusion; specifically, for the inmates that we were able to
associate with contracted specialty health care costs who incurred
more than $5,000 in such costs, the average cost generally increased
with the age of the inmate. While reviewing data associated with
the nearly 15,800 inmates that incurred more than $5,000 in
specialty health care costs during fiscal year 2007–08, we noted
that 63 percent were age 40 and older. In comparison, inmates age
40 or over represent only 41 percent of all inmates. As shown in
Figure 5, during fiscal year 2007–08 average specialty health care
costs generally increased with age, and the oldest inmates were the
most costly on average. Specifically, among all inmates with more
than $5,000 in specialty health care costs, specialty health care for
inmates age 60 and older averaged $42,000 per inmate.

California State Auditor Report 2009-107.2

May 2010

We also compared specialty health care costs of inmates
incarcerated under the three strikes law (striker inmates) to those
of non-striker inmates. We found that the average specialty health
care costs in fiscal year 2007–08 were $8,700 for striker inmates, or
13 percent more than the average of $7,700 for non‑striker inmates.
The higher amount for striker inmates is not surprising given the
disparity in age between striker inmates and non-striker inmates.
As described in our prior report, the largest age group of striker
inmates ranges between 40 and 44, while the range for the largest
age group of non-strikers is 25 to 29.
Figure 5
Average Contracted Specialty Health Care Costs by Age for Each Inmate Who
Incurred More Than $5,000 in Specialty Care
Fiscal Year 2007–08
$45
40

Dollars in thousands

35
30
25
20
15
10
5
0
Under 18-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59
18
Age categories

60
and
over

Sources:  Bureau of State Audits’ analysis of data from California Department of Corrections and
Rehabilitation’s Offender Based Information System and California Prison Health Care Services’
Contract Medical Database for fiscal year 2007–08.

Additionally, although we noted that average contracted specialty
health care costs generally increased with the age of inmates, the
cost of specialty health care associated with inmates that died
during the last quarter of fiscal year 2007–08 were significantly
greater than those of any specific age group. Each of the 72 inmates
who died during the last quarter of fiscal year 2007–08 incurred,
on average, $122,300 for specialty health care services for that fiscal
year. Ranging from $150 for one inmate to more than $1 million for
another, these 72 inmates accounted for $8.8 million in specialty
health care costs during fiscal year 2007–08. A more detailed
analysis of the types of care these inmates received reveals that

39

40

California State Auditor Report 2009-107.2

May 2010

92 percent of the costs were for inpatient acute medical and
surgical care. In contrast to the average of $122,300 associated
with inmates that died, as shown in Figure 5, the average was
$42,000 for the oldest age group of inmates with specialty health
care costs exceeding $5,000.
State law establishes a program that allows Corrections to
recommend to a court that it consider the early release of certain
inmates who would not threaten public safety and who either
are terminally ill with six months or less to live or are medically
incapacitated as assessed by a health care provider. According to the
receiver, after both Health Care Services and Corrections agree that
an inmate meets the criteria for the program, the information is
forwarded to a sentencing judge who makes the final determination.
The receiver stated that this process occurs very rarely and that few
inmates reach the point of having their cases presented to a judge
or being released from prison.
Health Care Services Is Taking Steps to Contain Health Care Costs

According to the receiver, past
efforts to reduce the amount of
medically unnecessary off-site
referrals were ineffective and
not only affected costs but also
hindered the ability to support
patient care and movement.

In November 2009 the receiver submitted a cost containment
report to the Legislature in compliance with state law, which
requires that Health Care Services provide detailed written reports
on actions taken and planned to reduce and better manage medical
service contract costs in fiscal year 2009–10 and for future fiscal
years. The report discussed several cost containment measures the
receiver has implemented or is in the process of implementing.
Additionally, when we interviewed the receiver about his plans
for containing medical costs, he explained that he and his staff
had reviewed expenditures and found that there was a two- to
three-year period of large increases in medical costs as more
inmates were provided with access to care and began receiving
necessary specialty medical care. He explained that he and his
staff are focusing on efforts to contain costs, such as utilization
management—defined as assessing the efficiency of the health
care process and the appropriateness of decision making related
to the site, frequency, and duration of care; contracting with a
third‑party administrator to process medical invoices electronically;
and expanding telemedicine. The receiver’s November 2009 cost
containment report states that Health Care Services is working
to achieve a zero growth rate in contract medical expenditures
for fiscal year 2009–10, instead of the projected growth rate of
17 percent.
In Health Care Services’ cost containment report and in his
discussion with us, the receiver explained that there were concerns
that past utilization management efforts were ineffective in
reducing the amount of medically unnecessary off-site referrals.

California State Auditor Report 2009-107.2

May 2010

The report states that these unnecessary referrals not only affect the
costs but also overuse scheduling, nursing, and custody staff,
hindering their ability to support patient care and movement.
As part of its effort to address the concern about unnecessary
referrals, Health Care Services has implemented the use of
specialty referral guidelines, referred to by the receiver as
InterQual. According to the cost containment report, InterQual is
a licensed software product containing a library of evidence-based
decision‑support criteria, is supported by a clinical advisory team
of academic experts, and is updated annually. The report indicates
that the use of InterQual promotes the provision of care that meets
the criteria of medical necessity. In its cost containment report,
Health Care Services reports that since the implementation of
InterQual in the fall of 2008, the number of requests for services
decreased by 31 percent between April 2009 and September 2009.
Additionally, in its April 2010 cost containment report, the
receiver reported additional decreases in requests for services
through January 2010 for a total decrease of 41 percent between
April 2009 and January 2010. Despite this decrease in referrals for
specialty care overall, the receiver stated that Health Care Services
has not calculated the cost savings associated with the reduction
in referrals. However, he indicated that he believes the number
of unnecessary referrals has decreased significantly. According
to Health Care Services’ chief medical officer for utilization
management, the data captured by the utilization management
databases do not interface with any of Health Care Services’
contract or claims databases. Because utilization management data
do not interface with the contract or claims databases, Health Care
Services is unable to associate specialty health care utilization with
the cost of providing the care. As a result, Health Care Services has
not calculated a savings associated with the decrease in the number
of referrals.
The receiver also discussed with us, and reported in the cost
containment report, his concerns about the length of hospital
stays by inmates in acute care hospitals in the community. These
are inmate stays for emergencies or urgent care that is beyond
the scope of Corrections’ staff and facilities. According to the
receiver, inmates sometimes must remain in the hospital longer
than necessary because they cannot be placed back into the general
inmate population and the institutions do not have recovery beds
available when the hospital is ready to release them. The receiver
told us that Health Care Services recently started monitoring 14 or
15 high-volume contracted hospitals that house 80 percent of the
hospitalized inmates. The cost containment report also states that
Health Care Services implemented a Web-based electronic database
in October 2009 that allows institutions and headquarters to
monitor the use of infirmary beds at institutions and at community

Although the receiver indicates that
the number of unnecessary referrals
has decreased significantly, Health
Care Services has not calculated
a savings associated with the
decrease because its utilization
data do not interface with
certain databases.

41

42

California State Auditor Report 2009-107.2

May 2010

hospitals. Under this new monitoring program, the database
is updated daily and shows the inmates that were admitted,
discharged, and still in the hospital, according to the chief medical
officer for utilization management. The information collected
in the database is distributed to physicians at the institutions to
better manage infirmary beds available at the institutions and
to avoid keeping inmates in community hospitals when it is not
medically necessary.
According to the receiver, as a result of lawsuits and the
appointment of the receiver, inmates were provided additional
access to medical services, resulting in major increases in the
amount of health care provided. The receiver reported in the cost
containment report that this increased access to care has meant
that Health Care Services experienced a significant increase in
the volume of invoices received for processing from contracted
providers. To address the additional volume of invoices, improve
the accuracy and timeliness of invoices that were processed
manually, reduce duplicate payments, and avoid the penalties
incurred due to late payments, the receiver contracted with
a third‑party administrator to automate invoice processing.
According to the director of Health Care Services’ administrative
support services (director), 80 percent of the invoices will be
processed through the third-party administrator once the transition
to automated invoicing is fully implemented. The director also told
us that contracting with the third-party administrator reduced
errors associated with processing health care invoices and the cost
of penalties for late payments. Further, the director indicated that
Health Care Services contracted with a company to conduct a
two‑year post audit of health care invoices and estimates that this
audit will identify $10 million in overpayments.

Health Care Services
continues to expand the use of
telemedicine as part of its cost
containment strategies.

Finally, as we discuss in more detail in the next section, Health Care
Services continues to expand the use of telemedicine as part of its
cost containment strategies. According to the cost containment
report, Health Care Services is taking steps to increase the use
of telemedicine to deliver medical care, including expanding the
number of medical specialties available through telemedicine and
introducing an initiative to increase telemedicine use at selected
institutions. Based on the initiative experience, it will develop
methods and procedures to improve and expand the use of
telemedicine at all institutions. In the solution vision statement of
the Telemedicine Services Project summary, Health Care Services
indicates that the project will result in significantly improved access
to care and quality of care while significantly reducing costs and
increasing effectiveness of health care delivery at institutions. The
statewide program director of the office of telemedicine services
stated that using telemedicine saves costs by reducing the need for
guarding and transporting inmates outside of the institutions to

California State Auditor Report 2009-107.2

May 2010

see a medical specialist and improves safety by keeping the inmate
within an institution’s walls. As we discussed in our prior report,
medical visits outside the institutions have increased significantly
since the establishment of the receiver in 2005, as part of an effort
to improve the quality of health care provided to inmates. Taking
into account increases in custody staff salary, medical guarding
and transportation overtime increased by 449 percent from fiscal
years 2003–04 to 2007–08, or more than $111 million. For the most
recent year of our review period, fiscal year 2007–08, overtime for
medical guarding and transportation totaled $136 million.
Although Health Care Services is implementing several cost
containment measures that might result in savings, its director
of administrative support services stated that it is unable to
determine the extent to which each cost containment measure
separately contributed to reducing contract medical costs. Further,
the director stated that Health Care Services did not perform
cost-benefit analyses on the various cost containment measures
described in this section because they either were mandated
through the court’s approval of the receiver’s turnaround plan or are
integral parts of a modern managed health care organization. The
director also told us that although the benefit analysis was not done
in the traditional way, the implementation of the cost containment
measures are projected to save approximately $250 million in
fiscal year 2009–10. This estimated cost savings figure—according
to the receiver’s April 2010 cost containment report—is based
on projected costs of $741 million for fiscal year 2009–10 and an
assumption that absent the cost containment measures, contract
medical service expenditures would have continued to increase to
more than $992 million, a growth of 17 percent over the prior year.
Health Care Services Has Not Fully Estimated the Cost Benefit of
Expanding Telemedicine Use
Although Health Care Services has continued expanding its use
of telemedicine as part of its cost containment strategy, it has
not fully estimated the potential cost savings of using additional
telemedicine. Health Care Services has significantly expanded
the use of telemedicine within Corrections’ institutions over the
past two years and reported that it has an ongoing initiative at
selected institutions designed to continue increasing the use
of telemedicine. According to Health Care Services, in fiscal
year 2008–09 it conducted 15,900 medical specialty consultations
using telemedicine, a 51 percent increase over fiscal year 2007–08
and a 144 percent increase over the fiscal year 2006–07 amount. In
addition, Health Care Services told us it has expanded the specialty
services offered through its telemedicine program from 14 medical
specialties in April 2009 to 22 medical specialties as of March 2010.

From fiscal years 2003–04 to 2007–08
the cost of overtime for medical
guarding and transportation
increased by 449 percent to
$136 million.

43

44

California State Auditor Report 2009-107.2

May 2010

When we asked Health Care Services to provide us with an
estimate of the number of medical specialty visits that could be
replaced with telemedicine, the statewide program director for
the Office of Telemedicine Services (program director) stated that
the data systems needed to generate data that Health Care Services
could use to estimate the percentage or number of medical specialty
visits that could potentially be provided using telemedicine are not
available. The program director stated that the specific data needed
for an estimate of this type would include a breakdown of specific
procedure codes and information regarding the nature and purpose
of visits. The program director stated that it might be possible to
develop an estimate by using this data.

In estimating the medical guarding
and transportation costs that are
avoided with each telemedicine
consultation, Health Care Services
needs to include other factors.

Health Care Services did provide an estimate of the medical
guarding and transportation costs that are avoided with each
telemedicine consultation. This cost avoidance estimate of $290 or
$580 per telemedicine consultation has been updated from the
estimate we discussed in the prior report and is significantly lower
than the $800 previously indicated. The $580 represents the cost
avoidance for transporting and guarding one inmate per visit and
the $290 represents a transport of two inmates per visit. The new
estimate provided uses the per visit cost avoidance of $290 and
$580 to indicate that the telemedicine consultations completed
in fiscal year 2008–09 resulted in $4.6 million to $9.2 million in
avoided guarding and transportation costs. However, according
to Health Care Services’ staff, the cost avoidance for most visits
would be $290 per visit. Although the updated methodology
used to create this estimate addresses one of the concerns described
in our prior report, other concerns remain. For example, the
calculation continues to exclude consideration of other factors
that might affect costs, such as whether a subsequent in-person
visit must be performed because the issue could not be treated
through telemedicine.
Estimated Costs for Improvements to the Outdated Information
Technology Infrastructure Necessary to Facilitate the Expansion of
Telemedicine Are Significant
Although Health Care Services has not estimated the expected
total savings of telemedicine, estimates of the costs associated with
projects that will facilitate expanding telemedicine are significant.
According to Health Care Services’ chief information officer, many
of the projects under development are not specific to telemedicine
but are necessary to develop a modern information technology
infrastructure to take the place of outdated or nonexistent
information technology. The chief information officer stated that,

California State Auditor Report 2009-107.2

May 2010

in addition to bringing Corrections closer to current information
technology standards, implementing these technology projects
would facilitate the provision of additional telemedicine care.
Health Care Services identified seven projects, shown in Table 4,
as important and necessary to facilitate the expansion of the
telemedicine program. The chief information officer stated that it is
important to note that these projects will facilitate a more efficient
functioning of the health care program as a whole. He also stated
that many of the technological deficiencies that will be addressed
by these projects have constrained Health Care Services for many
years and said that some of the projects will bring technological
resources online that are considered fundamental in most medical
facilities. For example, according to the project description for
the Health Care Network project, it will establish a high-speed
data network that will allow all 33 institutions and headquarters
to communicate electronically. The program director told us
that the current network inhibits the growth of telemedicine
because telemedicine equipment cannot be moved, and that a new
computer network would allow telemedicine equipment to be used
at many locations in the institutions.
Table 4
Projects Identified by California Prison Health Care Services That Will
Facilitate the Expansion of Telemedicine

PROJECT NAME

ESTIMATED PROJECT
EXPENDITURES,
FISCAL YEARS 2009–10
THROUGH 2013–14

Telemedicine Services

$41,383,065

Health Care Network

181,501,008

Health Care Scheduling System

21,841,574

Clinical Imaging Services

37,207,667

Clinical Data Repository

118,476,434

Utilization Management

15,118,191

Health Care Data Center
Total

107,539,314
$523,067,253

Source:  Bureau of State Audits’ summary of costs from California Prison Health Care Services’
Economic Analysis Workbook forms.

Health Care Services also is projecting to spend more than
$41 million on the Telemedicine Services Project, which is a broadly
defined project intended to provide technology and resources to
improve and expand telemedicine services. More specifically, the

45

46

California State Auditor Report 2009-107.2

May 2010

project’s budget estimates one‑time costs for purchasing hardware,
software, and additional telemedicine equipment, as well as ongoing
funding for staff and maintenance costs.
According to Health Care Services the Telemedicine Services
Project—which is estimated to cost more than $41 million during
fiscal years 2009–10 through 2013–14—will increase telemedicine
services by 400 percent, to 80,000 consultations per year by fiscal
year 2013–14. Using Health Care Services’ most recent per visit cost
avoidance estimates of $290 and $580 and the increased estimated
number of consultations each year due to the expansion project, we
estimated that the project would result in $46 million to $93 million
in cost avoidance over the five-year period. However, as previously
described, the telemedicine service project director indicates that
Health Care Services is unsure of how many of the current specialty
health care consultations can be provided through telemedicine.
Further, even without considering the degree to which telemedicine
consultations are unsuccessful because the issue must be treated
through an in-person consultation, the underlying information used
in Health Care Services’ cost avoidance figures varies between $94
and $1,233 per visit, suggesting that telemedicine consultations may
not be cost-effective at some institutions. As a result we believe
that additional analysis could allow Health Care Services to better
identify visits that would result in a larger cost avoidance if they
were replaced with telemedicine. In addition, obtaining a better
understanding of the extent to which telemedicine consultations
can replace traditional consultations seems prudent, given the
infrastructure and staffing costs associated with expanding the
program.
Recommendations
Health Care Services should continue to explore methods of
reducing the costs of medical care to the State, including those
of inmates with high medical costs. These efforts could include
proposing a review of the program that allows for the early release
of terminally ill or medically incapacitated inmates, and other
possible means of altering the ways in which inmates are housed
without unduly increasing the risk to the public.
To improve its ability to analyze and demonstrate the effectiveness
of current and future utilization management efforts in containing
health care costs, Health Care Services should identify a method to
associate cost information with utilization management data.
To determine whether the additional expansion of telemedicine is
cost-effective within the California correctional system, Health Care
Services should do the following:

California State Auditor Report 2009-107.2

May 2010

•	 Identify and collect the data it needs to estimate the savings
of additional telemedicine through an analysis of the cost of
specialty care visits currently provided outside of the institution
that could be replaced with telemedicine.
•	 Further analyze the cost-effectiveness of telemedicine through
a more robust estimate of savings, including considering factors
such as the percent of telemedicine consultations that required
subsequent in-person visits because the issue could not be
addressed through telemedicine.

47

48

California State Auditor Report 2009-107.2

May 2010

Blank page inserted for reproduction purposes only.

California State Auditor Report 2009-107.2

May 2010

Chapter 3
VACANT POSITIONS, MEDICAL GUARDING, AND LEAVE
ACCRUALS INFLUENCE OVERTIME COSTS
Chapter Summary
For our prior report, we identified and reviewed a number of
factors that influenced custody staff overtime costs.9 Specifically,
we found that vacant positions, increased medical guarding,
and increased salaries and benefits have led to increased
overtime costs. Custody staff vacancies and salary increases were
specifically identified as having a significant impact on overtime,
due to the need to cover shifts regardless of staffing levels and the
increased cost of doing so. Although vacant positions increase the
use of custody overtime, they decreased over the last two years
we reviewed.
In reviewing more detailed information regarding leave, vacancies,
and overtime use by the California Department of Corrections and
Rehabilitation (Corrections) for this report, we identified issues
that limited our ability to determine the relationships among these
factors. One such issue is that a significant workload exists that is
covered through overtime in addition to the workload associated
with authorized positions. For example, staff at San Quentin State
Prison (San Quentin) indicated that there are roughly 100 guarding
assignments, on average, necessary each day to guard inmates in
community hospitals. Although these inmates must be guarded
each day, 58 of these guarding assignments, on average, are covered
through overtime as there are not enough positions authorized for
this purpose.
We also determined that Corrections’ staffing formulas contain
several errors that may cause institutions to schedule more or less
leave than they should, depending on the number of vacant positions.
Further, we found that custody staff have accumulated leave balances
that represent a future liability to the State that we estimate is at
least $546 million and could be more than $1 billion, depending on
whether staff are paid for the leave when they quit or retire, or are
able to use the leave while they are employed. We also estimate that
upon its termination the furlough program will have contributed
more than 8.7 million hours—or $332 million to $518 million—to
the liability for the leave balances of correctional officers alone, not
including other custody staff such as sergeants or lieutenants.

9	

2009-107.1: California Department of Corrections and Rehabilitation: It Fails to Track and Use Data
That Would Allow It to More Effectively Monitor and Manage Its Operations, September 2009.

49

50

California State Auditor Report 2009-107.2

May 2010

Vacancy Levels Have Fluctuated Over the Last Five Years and Can
Increase the Use of Overtime
Using position data from the State Controller’s Office for the
five‑year period ending June 2009, we calculated that Corrections
had more than 2,200 vacant custody staff positions as of June 2009.
Although this number is significantly lower than the average of
2,700 vacancies that we calculated for fiscal years 2004–05 through
2008–09, it still represents a 7 percent vacancy rate. As shown in
Figure 6, the number of vacant custody staff positions fluctuated,
ranging from a peak of 3,500 positions in February 2006 to a low of
1,800 in May 2008.
Figure 6
Custody Staff Vacancies
Fiscal Years 2004–05 Through 2008–09
Lieutenant
Sergeant
Officer

4,000

Number of vacant positions

3,500
3,000
2,500
2,000
1,500
1,000
500
0

July

July
2004–05

July
2005–06

July
2006–07

July
2007–08

2008–09

Fiscal Year
Source:  Bureau of State Audits’ analysis of data from the State Controller’s Office position roster file for fiscal years 2004–05 through 2008–09.

Although the number of custody staff employed by Corrections
increased overall during this same period, vacancies persisted
as Corrections added new authorized custody staff positions.
According to the State Controller’s Office data, Corrections’
authorized custody staff positions increased from 25,400 in
July 2004 to nearly 30,200 in June 2009—an increase of 4,800.

California State Auditor Report 2009-107.2

May 2010

Two key factors contributed to the increase in authorized positions
for custody staff over the period we reviewed. First, changes in
the inmate population directly affect the number of custody staff
needed. The inmate population increased by 5 percent during
fiscal years 2004–05 through 2006–07. Consistent with this
increase in population, Corrections’ authorized positions for
custody staff increased as well. The second factor was an increase
in inmate access to medical care brought about by the actions
of the federal court-appointed receiver (receiver). Although the
inmate population decreased during the remaining two years of
our review period, fiscal years 2007–08 through 2008–09, the
receiver identified a need for more custody staff to be assigned to
the medical delivery system in the institutions, due to the increase
in access to medical care. According to Corrections, the receiver
requested 1,750 additional custody staff positions, and Corrections
filled these positions during fiscal years 2007–08 and 2008–09.
To determine the number of vacant custody positions, we identified
the number of authorized positions, using the State Controller’s
Office position roster file, and subtracted from it the number
of paid positions. However, when we compared this result with
a summary of data collected by Corrections’ program support
unit from each institution in November 2008, we found that
Corrections’ summary indicated that it had filled about 1,070 more
correctional officer positions than we calculated using the State
Controller’s Office position roster file. When we discussed the
differences with Corrections’ staff, the chief of the program support
unit (chief ) stated that there are a number of factors that could
cause the differences. Some of the reasons he provided include a
variance in the methodology used by each institution to determine
the number of filled positions, significant time lag between when
positions are filled at institutions and when the institutions submit
the paperwork and it is processed by the State Controller’s Office,
and institutions counting staff that are in temporary positions—
referred to as blanket positions—as filled positions. The chief
told us that institutions have many staff in blanket positions. In
our calculation of filled positions, we did not include individuals
in blanket positions because they are not filling staffing needs
associated with authorized positions.
We also discussed the differences with the deputy director of fiscal
services (deputy director), who told us that the State Controller’s
Office position roster file is not an accurate account of actual
authorized and paid positions on a month-to-month basis. He
stated that due to Corrections regularly activating and deactivating
inmate beds in order to match changes in the inmate population,
the number of authorized positions changes every month, and
over the course of a year may vary by 5 percent or more. The
deputy director also stated that as a result of the activations and

According to Corrections, the
receiver requested 1,750 additional
custody staff positions, and
Corrections filled these positions
during fiscal years 2007–08
and 2008–09.

51

52

California State Auditor Report 2009-107.2

May 2010

deactivations, forms to add and remove positions are sent to
the State Controller’s Office and there is a time lag between the
positions being opened up at the institutions and the processing
of the forms to change the positions. He stated that, as a result,
the position roster file might not accurately reflect a particular
point in time. Further, the deputy director stated that there are an
unknown number of positions that may not have been removed in
the State Controller’s Office records after a deactivation or staffing
change, or may not have been removed in the position roster file
in a timely manner. He stated that although Corrections reconciles
its budgeted authority to its scheduling system to find and correct
differences in the number of custody staff positions, there is no
reconciliation between the scheduling system or the budgeted
authority data and the State Controller’s Office position roster file.
Although the deputy director stated that the information from
Corrections’ scheduling system would have the most accurate
representation of actual staffing because shifts and assignments in
the system determine actual daily shift coverage and assignments,
Corrections was unable to provide us with data from its scheduling
system for the five-year period we were asked to review.

Although reducing overtime by
filling vacancies would not reduce
costs, there are concerns regarding
the safety of staff working a high
number of overtime hours.

Although Corrections increased the number of custody staff
employed during the period of our review, vacancies continued to
exist because Corrections also added positions. The existence of
vacant positions can increase the need for overtime. For example,
if the vacancies exist in positions that are designated to cover
custody staff on leave, Corrections must have an existing custody
staff person work overtime to cover the absent person’s guarding
assignment. However, as we discussed in our prior report, when
we compared the cost of hiring a new correctional officer to the
cost of overtime for correctional officers, we determined that
the hourly cost of hiring a new correctional officer during the
first year of work at an institution is slightly higher than the cost of
paying an officer to work overtime. Therefore, reducing overtime
by filling vacancies would not reduce costs. However, there are
concerns regarding the safety of staff working a high number of
overtime hours.
A Significant Number of Medical Guarding Assignments Are Covered
With Overtime
Another factor influencing overtime is that positions have not been
created to address a significant number of guarding assignments.
This additional guarding work is for custody-related assignments
such as medical guarding and transportation, and although much
of this work must be covered through overtime, the overtime
does not appear to be related to an established workload. Staff we
interviewed at three institutions told us they either did not have

California State Auditor Report 2009-107.2

May 2010

authorized positions for medical guarding and transportation or
the authorized positions were insufficient. For example, an associate
warden at San Quentin told us that it guards inmates receiving
inpatient care at Bay Area hospitals. She stated that the number of
inmates in community hospitals varies from 10 to 35 per day, but
averages 19. She stated that guarding these inmates requires about
100 guarding assignments for 24 hours. This is because guarding
an inmate out of the institution typically requires two correctional
officers per inmate for each of the three shifts in a day. Further,
she stated that on average 58 of these guarding assignments are
not associated with authorized positions and are covered through
overtime. This information is consistent with the receiver’s
February 2010 Monthly Health Care Access Quality Report
(Monthly Health Care Report). The report indicates that, as of
February 2010, the monthly average for fiscal year 2009–10, based
on information reported by the institutions, is 1,900 personnel
years related to medical guarding and transportation that are being
covered by overtime, redirected, and part-time staff.
The 1,900 personnel years described in the Monthly Health Care
Report include a year-to-date monthly average of 243,500 hours
of overtime for medical guarding and transportation. Using
the overtime rate of $52 per hour for a correctional officer with
six years of experience, as described in our prior report, the
243,500 hours of overtime per month would cost $12.7 million, or
$152 million a year. Overtime accounts for 78 percent of the medical
guarding and transportation hours, with the remaining 22 percent
of hours worked by staff that are redirected from other duties or
by part-time employees. Although the actual cost of this overtime
may be slightly lower because some correctional officers are paid
for overtime at a lower rate, this number of hours represents a
significant amount of total custody staff overtime.
The receiver’s 11th Tri-Annual Report to the federal court, issued in
June 2009, states that Health Care Services had planned to assess
the hospital guard positions needed and to implement them at
each institution. However, the report further states that Health
Care Services stopped these efforts and intends to assess and create
strategically placed locked guarding units at community hospitals
throughout the State to reduce the custody resources required
and to gain operational efficiencies. According to the director of
administrative support services, Health Care Services decided not
to request additional custody staff positions because it believes that
referrals for outside specialty services will decrease in the future
as a result of cost containment measures, such as the utilization
management and telemedicine projects we discuss in Chapter 2.
In addition, according to the receiver, Health Care Services is
considering a plan to place inmates with higher specialty care needs
in institutions that can provide some of those specialties, reducing

The receiver’s report states that on
average there are 243,500 hours of
overtime for medical guarding and
transportation each month.

53

54

California State Auditor Report 2009-107.2

May 2010

the number of inmates receiving care outside the institution.
Finally, the director of administrative services stated that, because
emergency transportation cannot be predicted, it would be
inefficient to staff for this item through established positions. Using
overtime for a workload that is expected to decrease in the future
seems practical, and it may be more efficient to cover some of these
hours through overtime if they vary significantly from day to day.
However, given the amount of medical guarding and transportation
work covered through overtime, care must be taken to ensure that
the total amount of overtime worked by custody staff does not
impact the safety of operations.
Corrections’ staffing formula accounts for factors such as vacation
and sick leave, so overtime should be necessary only when positions
are vacant, emergencies occur, or staffing needs differ from the
percentages described in the formula. As shown in Figure 7, we
compared our estimate of the number of hours that guarding
assignments were vacant to the number of hours of overtime for
fiscal years 2004–05 through 2008–09, and found that the number
of hours of overtime actually worked exceeded the estimated need
for overtime to cover for authorized vacant positions for nearly all
of the period.

Medical guarding and
transportation costs were
responsible for nearly 32 percent of
overtime costs—$136 million—in
fiscal year 2007–08.

As we discussed in our prior report, institutional and miscellaneous
security issues, such as emergencies and firefighting, also contribute
to overtime costs. We found that these factors accounted for
21 percent of custody staff overtime costs in fiscal year 2007–08.
Medical guarding and transportation costs were responsible for
nearly 32 percent of overtime costs, or $136 million, in fiscal year
2007–08. The difference between the amount of overtime and
the estimated hours for which positions were vacant increased
in the months after the establishment of the receiver in 2005,
and generally decreased in fiscal year 2008–09, a time when
Corrections indicated that it filled 1,265 of the 1,750 custody staff
positions added to provide increased inmate access to health care in
fiscal years 2007–08 and 2008–09.
Additionally, during our analysis of overtime, the deputy director
explained to us that even though the staffing formulas were
established to ensure that there is adequate staff to cover guarding
assignments, if more than the average number of staff call in
sick on a particular day, additional staff may be needed to work
uncovered guarding assignments, and institutions may need to
cover the guarding assignments by using overtime. We were unable
to determine how often this situation arises. Our analysis of the
reasons for overtime, which covered a five-year period, was based
on reviewing monthly overtime and leave data we obtained from
the State Controller’s Office. Although reviewing daily overtime
data would most likely have allowed us to identify more detailed

California State Auditor Report 2009-107.2

May 2010

causes for overtime, Corrections did not have complete overtime
and leave data for the period we reviewed. As a result of this and
other factors that cause undetermined fluctuations in overtime,
we were not able to measure precisely the relationship between
changes in overtime and the number of vacancies.
Figure 7
Correctional Officer Overtime Worked Compared With Estimated Vacant Guarding Hours at Adult Institutions From
Fiscal Year 2004–05 Through 2008–09
485 custody staff
positions filled for
medical guarding
and transportation

Federal court
establishes health
care receivership

1,265 custody staff
positions filled for
medical guarding
and transportation

800
700
Actual correctional officer overtime hours

Hours in thousands

600
500
400
300
200
Estimated vacant guarding hours

100
0

JUN

JUN
200405

JUN
200506

JUN
200607

JUN
200708

200809

Fiscal Year
Sources:  Bureau of State Audits’ analysis of data from the State Controller’s Office position roster and payroll records for fiscal years 2004–05
through 2008–09; an estimate of monthly guarding hours for vacancies based on Department of Corrections and Rehabilitation’s (Corrections)
correctional officer staffing formula; and number of medical guarding positions filled in fiscal years 2007–08 and 2008–09 provided by the chief of
Corrections’ Program Support Unit.
Note:  Correctional officers earn overtime based on thirteen 28-day work periods each year. However, they are paid for these overtime hours as part
of the normal monthly pay process. As a result, some months include the overtime payments for two work periods. In these months we divided the
number of hours of overtime worked by two and added a second data point with that amount of overtime.

Some Aspects of Corrections’ Staffing Formulas Are Outdated and
Others Appear to Be Flawed
To ensure that it hires sufficient staff to handle the guarding
assignments that exist, Corrections uses staffing formulas to ensure
that, when the regularly scheduled custody staff are unavailable,
additional staff can work the assignment. These staffing formulas

55

56

California State Auditor Report 2009-107.2

May 2010

also are used to determine how many individuals can take vacation
at any time. However, Corrections’ current staffing formulas contain
errors and outdated information. Because these formulas are used
for staffing, such errors have an effect on Corrections’ ability to
ensure that custody staff are able to use the leave they earn.
Staffing Formulas Identify the Number of Custody Staff Necessary for
Guarding Assignments and Dictate Leave Use
Because Corrections is responsible for operating the institutions
that house inmates 24 hours per day, 365 days per year, it must
schedule many custody staff to work guarding assignments every
day. However, staff may not work their regularly scheduled guarding
assignment because of vacation, training, and other reasons, so
Corrections must hire a sufficient number of individuals to ensure
that it has custody officers available when needed. According to
the chief of the program support unit, Corrections determines the
custody staffing needs of institutions based on a number of factors,
including the inmate population and the staffing needs of various
program areas—such as education—at each institution.
Corrections determines the number of custody staff needed for
its guarding assignments using staffing formulas. The staffing
formulas generally include one position to regularly work each
assignment plus an allocation for relief coverage—an employee who
will work an assignment if the regularly scheduled employee is
absent. For example, Corrections’ current staffing formula for a
five‑day‑per‑week correctional officer guarding assignment includes
an allocation of 8 percent of another position for working the
guarding assignment during the primary officer’s vacation leave.
This and the other relief values, such as sick leave and training,
are added to the one full-time position to determine the total
personnel necessary to ensure that someone is available to work in
the guarding assignment every day necessary, resulting in a formula
value of 1.24 correctional officers for this guarding assignment.
Corrections has additional staffing formulas for sergeants and
lieutenants, as well as formulas for administrative assignments.
Another common guarding assignment—the seven-day-per-week
assignment—requires a correctional officer to be on duty every day,
resulting in a formula of 1.74 correctional officers needed to cover
an eight-hour shift every day of the week. In addition to identifying
the total number of staff needed to ensure that custody staff are
available to work all of the guarding assignments, staff at the
institutions use the staffing formulas to identify how many staff can
sign up to take vacation or holiday leave at the same time.

California State Auditor Report 2009-107.2

May 2010

Figure 8 illustrates one of the staffing formulas, showing that
Corrections budgets about five full-time employees for
four guarding assignments that require coverage five days per
week. In our example, each of four employees is assigned to an
individual guarding assignment and is expected to work that
assignment each day unless they attend training, use leave, or are
absent for other reasons. As the figure shows, the fifth employee
provides relief coverage for the other employees when they
are absent.
Figure 8
Staffing Formula Hours for Four Correctional Officers Working Five-Day Per‑Week Guarding Assignments
Working
Vacation
Holiday
Sick
Training
Other
Available for
other assignments

Guarding
assignment 1

Employees

Guarding
assignment 2
Guarding
assignment 3
Guarding
assignment 4
Relief coverage

Post 1

0

261

Post 2

522

Post 3

783

1,044

Post 4

1,305

1,566

1,827

2,088

Hours Per Year
Source:  Bureau of State Audits’ analysis of Department of Corrections and Rehabilitation’s (Corrections) correctional officer staffing formula, including
adjustments by Corrections’ Program Support Unit.

Errors in the Staffing Formula Affect the Amount of Leave Custody Staff
Can Sign Up to Take
We reviewed the documentation Corrections provided to support
the specific calculations it used when updating the correctional
officer staffing formula. Although we found the factors that make
up the formula agreed in total, some factors do not match the
documentation provided as support for the calculations used to
update the formula, which occurred about six years ago. According
to a captain in the program support unit, this occurred because
of limitations of the computer system into which the formula is
entered. Specifically, the system cannot account properly for certain
relief factors.
The staffing system cannot accommodate the percentage values for
some factors such as holiday leave, so the number of opportunities
to use this leave—which must be requested in advance—are lower

57

58

California State Auditor Report 2009-107.2

May 2010

than needed to allow correctional officers to use all the leave they
earn each year. In contrast, some factors are overstated, such
as vacation, bereavement leave, or leave under the provisions
of the Family Medical Leave Act. Correctional officers can use
two of these three leave types only in certain circumstances. As
a result, the staffing system does not properly reflect the staffing
levels Corrections calculated it needs to allow correctional
officers to use various types of leave. When correctional officers
are not able to use all the leave they earn, their leave balances
increase. For example, based on Corrections’ documentation for
a five‑day‑per‑week correctional officer guarding assignment, and
the number of authorized positions that were paid in June 2009, the
errors associated with the vacation and holiday leave factors would
have caused Corrections to provide 19,400 fewer opportunities
to sign up for days of vacation and holiday leave than is earned by
correctional officers in a year.
The captain told us that Corrections is in the process of revising its
staffing formula methodology. He also stated that the system used
to schedule custody staff guarding assignments is being replaced
with a new system and will account for the revised methodology.
The captain also stated that updating the current scheduling system
would entail months to conduct a change order to accept any
revised formula changes and is not cost-efficient because the system
is being replaced this year.
Some of the staffing formulas
Corrections uses to determine the
number of custody staff needed for
guarding assignments have errors
and Corrections has not modified
the formulas.

Corrections recently made changes to its staffing, but it did not
use the opportunity to correct for errors in the formula. State law
changed in February 2009 to eliminate two paid holidays, reducing
the number of holidays earned from 14 to 12 per year. Instead of
changing the staffing formula by reducing the overstated categories
described earlier, according to a November 2009 memo from the
acting director of the division of adult institutions, the institutions
were directed to reduce their holiday relief positions to account for
the two holidays. As a result of this change, opportunities for staff
to use holiday credit remain below the level they earn.
The errors previously described regarding the amount of hours
placed in leave categories for which custody staff can schedule
time off are reduced by errors in the way in which Corrections
calculates the amount of vacation leave that it allows custody staff
to take. Specifically, the number of staff who can take vacation and
holiday leave at an institution is based on the number of authorized
guarding assignments and does not change based on the number
of custody staff positions actually filled. However, when there are
vacant positions, less vacation coverage is needed because there
are fewer employees. In addition, individuals working overtime
in place of staff who would otherwise fill vacant positions do not
earn additional leave. As a result, at institutions that have vacant

California State Auditor Report 2009-107.2

May 2010

positions, the staffing formula allows for more holiday or vacation
leave than the formula indicates custody staff earn. For example, if
an institution has 1,000 correctional officer positions, 100 of which
are vacant, the staffing formula would create opportunities for
the 900 correctional officers employed to sign up for the amount
of vacation and holiday leave available to 1,000 staff. If these staff
choose to take more leave than they earn in a year by using accrued
leave from prior years, additional overtime would be necessary to
cover their positions.
Using Corrections’ staffing formula and associated documentation
for a five-day-per-week correctional officer guarding assignment
and the number of authorized positions that were vacant in
June 2009, we estimate that because of the way in which the
staffing formula is applied to the number of authorized guarding
positions, institutions would have allowed correctional officers the
opportunity to sign up for nearly 46,800 more days worth of leave
in a year than it should. However, because this issue is specific to
institutions with vacant positions, the net effect of this error may
be offset by the error previously discussed regarding the amount
of hours placed in leave categories for which custody staff can
sign up to take leave. As a result, depending on the number of
vacant positions at a specific institution, correctional officers may
be provided too many or too few opportunities to use the leave
they earn.
Growing Leave Balances, Due in Part to Vacancies and Errors in
the Staffing Formulas, Reduce Current Costs but Represent a
Future Liability
Various factors have caused Corrections’ custody staff to accrue
large leave balances. When this occurs, current staffing costs are
reduced but the State incurs the cost in the future when staff take
the leave or are paid for the balances when they quit or retire.
Currently, the state furlough program is the most significant cause
of the accumulation of leave balances for Corrections’ custody staff.
Although many custody staff have chosen to use the furlough hours
they have earned, the amount of leave they can take is limited, as
defined in the staffing formulas. As a result, their use of furlough
hours means they use less of other types of leave, causing balances
in those categories to increase. The liability these balances represent
must be paid out at employees’ retirement, if it is not addressed
before then. Moreover, these costs will be higher in future years
as custody staff salaries increase. In contrast, because accrued
leave postpones the compensation of staff for additional work, it
decreases the cost of the correctional system in the current year.

Currently, the state furlough
program is the most significant
cause of the accumulation of leave
balances because some staff use
furlough days when taking time off
rather than other types of leave.

59

60

California State Auditor Report 2009-107.2

May 2010

Furloughs Have Created New Liabilities

Corrections cannot allow many
custody staff to take additional
time off and as a result, under the
furlough program custody staff
receive reduced pay and earn leave
credit for future use.

In February 2009, pursuant to an executive order issued by
the governor concerning the State’s fiscal crisis, Corrections
implemented a mandated furlough program, requiring custody
staff to take two days of unpaid leave each month, resulting in
the reduction of custody staff pay by 9 percent. In July 2009 this
program was expanded to three days per month. As applied to
many other state workers, the furlough program generally reduced
pay and the amount of time worked, resulting in many state offices
and operations closing three days per month. However, due to the
unique nature of its operations, Corrections cannot allow many
custody staff to take additional time off without hiring individuals
to take their place or requiring overtime. As a result, under the
furlough program Corrections’ custody staff receive reduced pay
for the same amount of work and receive furlough leave credit for
future use. A lawsuit was filed on behalf of correctional officers
to pay employees full wages due or at least the minimum wage
for all hours worked, and in December 2009 a judge ruled in the
correctional officers’ favor. However, as of April 2010 the ruling had
been stayed while an appellate court considers the case.
Although the total amount of leave that custody staff are earning
has increased, the amount of time off they can sign up to take
is determined by the staffing formulas, which do not include
allocations for the furlough program. As a result, custody staff
have earned significantly more leave than they are able to use,
and regardless of whether staff use furlough hours and their other
leave balances increase or staff accrue their furlough hours, total
leave balances have increased due to the furlough program. It is
difficult to predict exactly how the increased leave balances earned
as a result of the furlough program will be paid out or used, but
unless the staffing formula is altered to provide additional staff
for the purpose of allowing higher levels of leave use, custody
staff will not generally be able to reduce their leave balances. As a
result, leave balances will remain high until they are paid out when
individuals quit or retire. Thus, the furlough program postpones
the compensation of custody staff for work they perform until the
time they are able to use the additional leave they accrued. Because
the amount of leave custody staff can sign up to use is defined
by the staffing formula, in order for individuals to use both their
regular leave as well as the additional leave accrued as a result of
the furlough program, other staff must accrue their leave instead
of using it, or staff must wait until they retire or quit to receive
compensation for their leave balance.
By multiplying the number of hours of furlough credit provided to
an individual each month since the program began by the number
of filled correctional officer positions in those months, and using

California State Auditor Report 2009-107.2

May 2010

the June 2009 staffing level to project the number of correctional
officers employed in future periods, we estimate that more than
8.7 million hours of furlough credit will have been earned by
correctional officers under this program by the time it is scheduled
to end in June 2010. At the typical pay rate of $35 per hour for an
officer with six years of experience, this represents a liability of
$304 million. However, if the increase in leave balances due to staff
using furlough instead of their regular types of leave is not paid
out until individuals retire, this liability will likely increase over
time as individuals receive pay increases or are promoted. Further,
if Corrections were to increase staffing or allow more overtime
so correctional officers are able to use the additional leave they
accrued, this liability would represent $455 million in staffing costs.
Finally, because staff continue to earn leave while absent, or when
leave balances are paid out when they quit or retire, our estimate of
the total number of furlough hours provided to correctional officers
will result in more than 1.2 million additional hours of vacation,
sick, and holiday leave. Consequently, we estimate that the total cost
of the furlough program for correctional officers—not including
sergeants and lieutenants—will be $28 million to $63 million higher
because of this factor, for a total of $332 million or $563 million,
depending on whether correctional officers are paid for their leave
balances when they quit or retire or use them.
Leave Accruals Reduce Current Costs
As shown in Figure 7 on page 55, Corrections has had a significant
number of vacant custody positions for several years. According
to the chief of the program support unit, vacant positions increase
overtime because they reduce the number of staff available to
cover leave. However, if staff are not able to, or choose not to use
leave, there is not a need for their work to be performed by someone
working overtime. In fact, our analysis of the State Controller’s Office
leave data indicates that custody staff earned nearly 8.4 million hours
more leave than they used between July 2004 and January 2010.
Although 6.8 million of these hours10 were the result of the furlough
program, we estimate it would have resulted in almost $435 million
in additional costs if Corrections had been required to increase
staffing or use overtime to cover the use of this leave. The furlough
program—which is intended to reduce current costs—represents
more than $354 million of this amount. This increase contributed
to a total leave balance, including sick leave, that represents a

10	

The January 2010 balance for the furlough leave of custody staff in adult institutions was
2.4 million hours. However, because total leave use is limited by the staffing formulas, use of
furlough leave decreases the use of other types of leave. For this reason we have attributed
6.8 million hours of the increase in the total leave balance—the amount of furlough leave earned
up to that point in time—to the furlough program.

We estimate more than 8.7 million
hours of furlough credit—or a
liability of $304 million—will have
been earned by correctional officers
under this program by the time it is
scheduled to end.

61

62

California State Auditor Report 2009-107.2

May 2010

liability we estimate to be approximately $940 million if Corrections
were to increase staffing or overtime to allow custody staff to take
their accrued leave. Alternatively, if paid out as lump sums when
individuals retire or quit, the leave balance—minus sick leave
that can be credited toward the amount of time an individual is
considered to have worked when they retire but is not paid out as
other types of leave are—represents a liability to the State that we
estimate at approximately $500 million. The actual liability may be
higher, as some of these hours are associated with sergeants and
lieutenants who receive a higher salary. Although guarding needs
are reduced by individuals who choose not to take their accrued
leave before they retire, or are unable to sign up for enough leave to
deplete their balance, the leave balance still represents an eventual
cost to the State when it is paid out in a lump sum upon retirement.
Further, the liability previously described is even higher than
the current number of hours in custody staff leave balances. If
Corrections increased staffing or overtime to allow staff to use
this leave, they would continue to earn leave while absent, which
we estimate would increase the liability to Corrections by roughly
$130 million. Even if Corrections pays the leave in lump sums when
employees quit or retire, state law requires that the payment include
the amount of leave staff would earn had they used their leave
balances, increasing the lump-sum payments by more than $46
million to a total of more than $546 million.
When vacant positions exist, custody staff who do not use the
amount of leave they earn reduce the need for overtime to work
the guarding assignments of those vacant positions. Although this
reduces staffing costs in the near term, it contributes to the growth
of staff leave balances and essentially defers the costs to later years.
Because the existence of vacant positions can result in additional
overtime when leave use increases, custody staff decisions to use,
accrue, or take a payout at retirement for accrued leave can cause
Corrections’ staffing costs to fluctuate. In addition, according to the
deputy director, custody staff commonly reach retirement with
hundreds of hours of accrued leave. Moreover, the deputy director
also informed us that Corrections does not budget for leave payouts
upon retirement, so these costs represent an unbudgeted expense
each year.
Providing leave coverage and paying out lump-sum payments
represent significant costs. Some custody staff may choose not to
take the full amount of leave that they earn each year. Others may
use accrued leave from past periods, so in any single period some
of these costs may be offset. However, in years that the cumulative
leave balance grows, Corrections is deferring staffing costs to future
years. For example, this might occur when the number of custody
staff increases and new staff are willing to forgo using their leave

California State Auditor Report 2009-107.2

May 2010

while other staff with large balances either take leave or receive
lump‑sum payments that do not exceed the leave amounts earned
by new staff. If the number of custody staff does not change for a
significant period of time or begins to decrease, there may be fewer
individuals willing to forgo using their leave, and the percentage of
leave costs deferred would decrease.
The large leave balances previously described indicate that some
portion of the cost of the correctional system was deferred in past
years, that is, correctional system costs were lower than they could
have been because staff did not use the full amount of leave they
earned each year. Essentially, custody staff who did not take the full
amount of leave they earned were providing more work each year
than expected based on the staffing formula. However, the furlough
program causes custody staff ’s leave balances to accumulate, which
will likely result in larger leave payouts for retiring staff. In addition,
if custody staff choose to take more of the leave the staffing
formula allows them to use in a year because of the additional
leave provided through the furlough program, fewer costs will be
deferred in the current and future fiscal years.
Recommendations
To ensure that the State Controller’s Office has accurate
information on the number of authorized and filled positions,
Corrections should determine why the number of positions the
State Controllers’ Office indicates are vacant is higher than the
number of vacant positions it is aware of, and submit information
to the State Controller’s Office to correct this situation as necessary.
To ensure that the total amount of overtime worked by custody
staff does not unduly reduce their effectiveness and result in
unsafe operations, Health Care Services should monitor overtime
closely. If its efforts to reduce the number of referrals of inmates to
outside specialty services do not reduce the amount of overtime
worked by custody staff for the purpose of medical guarding and
transportation, Health Care Services should explore other methods
of reducing the total amount of overtime worked by custody staff.
To ensure that custody staffing meets institutional needs, and to
provide staff the opportunity to use the amount of leave that they
earn in the future, Corrections should update its staffing formulas
to accurately represent each of the factors for which custody staff
are unavailable to work, such as vacation or sick leave. Corrections
should attend to this project before implementing its new business
information system to ensure the updated formulas can be used as

63

64

California State Auditor Report 2009-107.2

May 2010

soon as practical. In addition, Corrections should create a policy for
regularly scheduled reviews of the data used in the staffing formulas
and update the formulas as necessary.
To better communicate to policy makers the annual cost
of incarceration, and to provide a more accurate estimate of
expenditures associated with changes in the large leave balances
of custody staff—many of whom require relief coverage
when they are absent—Corrections should provide the
following as supplemental information to the relevant legislative
policy and fiscal committees:
•	 A calculation of the annual increase or decrease in its liability for
the leave balances of custody staff to better explain the cause of
changes in expenditures.
•	 An estimate of the annual cost of leave balances likely to be paid
for retiring custody staff.
We conducted this review under the authority vested in the California State Auditor by Section 8543
et seq. of the California Government Code and according to generally accepted government auditing
standards. We limited our review to those areas specified in the audit scope section of the report.
Respectfully submitted,

ELAINE M. HOWLE
State Auditor
Date:	

May 18, 2010

Staff:	
	
	
	
	
	
	

Tammy Lozano, CPA, CGFM, Project Manager
Jonnathon D. Kline
Daniel P. Andersen
Michelle J. Baur, CISA
Richard W. Fry, MPA
Jennifer D. Loos, CPA
Laura Peth

Legal Counsel: 	Scott A. Baxter, JD
For questions regarding the contents of this report, please contact
Margarita Fernández, Chief of Public Affairs, at (916) 445-0255.

California State Auditor Report 2009-107.2

May 2010

Appendix
SERIOUS OR VIOLENT FELONIES AS DEFINED BY
CALIFORNIA STATE LAW
The three strikes law is intended to ensure longer prison sentences
and greater punishment for offenders who commit a felony and
were convicted previously of a serious or violent felony. The felony
for which the offender currently is convicted need not be serious
or violent. As shown in Table A, state law describes which felonies
are considered serious or violent, and some crimes are considered
both serious and violent. For example, murder is considered both
a serious and a violent felony. Courts commonly refer to a person’s
prior convictions for serious and violent felonies as strikes. The list
of offenses that may count as strikes has expanded since California
enacted the three strikes law in 1994. Specifically, in March 2000,
through the passage of Proposition 21, voters expanded the list
of offenses that constitute strikes. In 2006 the Legislature again
expanded the list of offenses.
Table A
Felonies Considered Serious or Violent
OFFENSE

SERIOUS (PC 1192.7 (C))

VIOLENT (PC 667.5 (C))

Murder





Voluntary manslaughter





Mayhem





Rape



*

Sodomy

†

*

Oral copulation

†

*

Lewd or lascivious act on a child under 14 years of age

†

*

Any felony in which the defendant inflicts great bodily injury on any person
other than an accomplice

†

*

Any felony in which the defendant personally uses a firearm



*

Robbery

†



Arson



*

Sexual penetration

†

*

Attempted murder





Exploding or igniting, or attempting to explode or ignite, any destructive
device with the intent to commit murder



Willfully and maliciously exploding or igniting any destructive device, which
causes mayhem, death, bodily injury, or great bodily injury of any person



Exploding a destructive device or any explosive with the intent to injure or
murder or that causes bodily injury, great bodily injury, or mayhem



Kidnapping



Assault with the intent to commit rape, sodomy, or other crimes


*
continued on next page . . .

65

66

California State Auditor Report 2009-107.2

May 2010

OFFENSE

SERIOUS (PC 1192.7 (C))

VIOLENT (PC 667.5 (C))

Continuous sexual abuse of a child





Carjacking





Extortion and threats to victims or witnesses committed as part or in
association with a criminal street gang



Burglary of the first degree



*

Using, discharging, and injuring someone with a firearm while committing
certain felonies such as murder, rape, or kidnapping



 

Using a weapon of mass destruction

†

 *

Assault with the intent to commit rape or robbery



Assault with a deadly weapon or instrument on a peace officer



 

Assault by a life prisoner on a non-inmate



 

Assault with a deadly weapon by an inmate



 

Holding of a hostage by a person confined in state prison



Any felony in which the defendant personally used a dangerous or
deadly weapon



Selling, furnishing, administering, giving, or offering to sell, furnish,
administer, or give to a minor any heroin, cocaine, phencyclidine
(PCP), or any methamphetamine-related drug, or any precursors of
methamphetamines



Grand theft involving a firearm



Certain felonies committed as part of or in association with a criminal
street gang



Assault with the intent to commit mayhem, rape, sodomy, or oral copulation



 

Throwing acid or flammable substances with the intent to injure or disfigure
another person



 

Assault with a deadly weapon, firearm, machine gun, assault weapon, or
semiautomatic firearm, on a person, peace officer, or firefighter



Assault with a deadly weapon against a public transit employee, custodial
officer, or school employee



Discharge of a firearm in an inhabited dwelling, vehicle, or aircraft



Rape or sexual penetration in concert with another person



Shooting from a vehicle



Intimidation of victims or witnesses



Criminal threats



 

Any attempt to commit a serious felony listed in California Penal Code,
Section 1192.7 (c) other than assault



 

Any conspiracy to commit an offense listed in California Penal Code,
Section 1192.7 (c)



 

Any felony punishable by death or imprisonment in the state prison for life





Attempt to commit a felony punishable by death or imprisonment in the
state prison for life



 



California State Auditor Report 2009-107.2

May 2010

OFFENSE

JUVENILE (WIC 707 (b))

Assault with a firearm or destructive device

‡

Assault by force likely to produce great bodily injury

‡

Discharge of a firearm at an inhabited or occupied dwelling

‡

Any felony offense in which the minor personally used a certain weapon
such as a dagger, metal knuckles, or short-barreled shotgun

‡

Bribing or inducing witnesses

‡

Manufacturing, compounding, or selling a controlled substance

‡

Escape by a juvenile where great bodily injury is intentionally inflicted on an
employee of the juvenile facility

‡

Torture

‡

Aggravated mayhem

‡

Committing or attempting certain felonies against a senior citizen or
disabled person

‡

Kidnapping for ransom, robbery, with bodily harm, while carjacking, or with
the intent to commit specified sex crimes

‡

Carjacking while armed with a dangerous or deadly weapon

‡

Sources:  California Penal Code, sections 667.5 (c) and 1192.7 (c) and California Welfare and Institutions Code, Section 707(b).
*	 Certain types of this crime are violent felonies.
†	 Certain types of this crime are serious felonies.
‡	 While juvenile offenses are not expressly defined as serious or violent felonies, we use the term “serious or violent felony” to
describe any offense that could count as a strike.

67

68

California State Auditor Report 2009-107.2

May 2010

Blank page inserted for reproduction purposes only.

California State Auditor Report 2009-107.2

69

May 2010

(Agency response provided as text only.)
Department of Corrections and Rehabilitation
Office of the Secretary
P.O. Box 942883
Sacramento, CA 94283-0001

April 30, 2010
Ms. Elaine Howle*
State Auditor
Bureau of State Audits
555 Capitol Mall, Suite 300
Sacramento, CA 95814
Dear Ms. Howle:
This letter represents the California Department of Corrections and Rehabilitation’s (CDCR) response to
the Bureau of State Audits (BSA) report entitled, California Department of Corrections and Rehabilitation:
Inmates Sentenced Under the Three Strikes Law and a Small Number of Inmates Receiving Specialty Health
Care Represent Significant Costs. This audit, requested by the Joint Legislative Audit Committee, was
intended to determine the effects of the three strikes law on incarceration costs and review custody staffing
formulas to determine its effects on annual overtime and leave balances.

1

We agree with your assessment that the issues addressed in this audit are extremely complex and concur
with your recommendations to ensure accurate sentencing guideline data will be transferred into our new
data system. The review of this data is in fact part of our implementation process for the new system. Our
analysts review all sentencing documents when performing their audits of case files, which includes a review
for any notification or registration requirement.
In implementing the Business Information System (BIS), the Department will have an improved ability to
track and account for budgeted and filled positions. However, the data the State Controller’s Office tracks
will likely continue to vary from the Department’s data. We will continue to monitor these discrepancies in
an effort to reduce the scope of these differences.
Your report states that we should update our staffing formulas to represent factors such as vacation and
sick leave. We are reviewing staffing formulas for accuracy, but changes in the formulas will need to go
through the budget process. In the interim, we are working with the vendor for BIS to ensure this project can
accommodate any necessary changes to the relief formulas.
Your report indicates the Department should provide as supplemental information to the Legislature a
calculation of the increase or decrease in leave liability, as well as estimate of costs of leave balances likely to
be paid out for retiring custody staff. There are a number of factors that influence retirement decisions that

*  California State Auditor’s comments appear on page 71.

2

70

California State Auditor Report 2009-107.2

May 2010

Ms. Elaine Howle
Page 2
lead to wide variances in the likely expenditures, which make it very difficult to estimate these costs. The
Department is committed to ensuring that its budget reflects the cost of operation, and will work closely
with the Administration to ensure the budget accurately reflects those costs.
We would like to thank BSA for their work on this report and will address BSA’s specific recommendations
in a corrective action plan at 60-day, 6-month, and one-year intervals. If you should have any questions or
concerns, please contact Lee Seale, Deputy Chief of Staff, at (916) 323 6001.
Sincerely,
(Signed by: Brett H. Morgan)
BRETT H. MORGAN
Chief of Staff

California State Auditor Report 2009-107.2

May 2010

Comments
CALIFORNIA STATE AUDITOR’S COMMENTS ON THE
RESPONSE FROM CALIFORNIA DEPARTMENT OF
CORRECTIONS AND REHABILITATION
To provide clarity and perspective, we are commenting on the
response to our audit report from the California Department
of Corrections and Rehabilitation (Corrections). The numbers
below correspond to the numbers we have placed in the margin of
Corrections’ response.
The imprecision of statements regarding the nature of the audit,
our assessments, and the six recommendations we directed to
Corrections is disappointing. The reiteration of several points
presented in the report regarding Corrections’ current actions to
address the items we discussed does not provide any additional
information, and the indefinite comments related to changes in
operations do not appear to intimate any alteration of policies or
procedures. We look forward to reviewing the more specific actions
Corrections plans to take related to each of the recommendations
in the report when we receive its corrective action plans at 60 days,
six months, and one year from the date of the publication of
the report.

1

Although we agree that various factors make it difficult to precisely
estimate the costs of leave balances paid out to retiring custody
staff, given the number of correctional officers, it seems unlikely
that there would be a year in which no custody staff retired. As
such, it seems that even a conservative estimate of the cost of leave
balance payouts—using, for example, historical information on
such payouts during prior years—would provide a more accurate
estimate of the costs of the correctional system than no estimate
at all.

2

71

72

California State Auditor Report 2009-107.2

May 2010

Blank page inserted for reproduction purposes only.

California State Auditor Report 2009-107.2

73

May 2010

(Agency response provided as text only.)
California Prison Health Care Services
P.O. Box 4038
Sacramento, CA 95812-4038
April 30, 2010

Ms. Elaine M. Howle*
State Auditor
Bureau of State Audits
555 Capitol Mall, Suite 300
Sacramento, CA 95814
Re:	

Response to BSA Draft Report – California Department of Corrections and Rehabilitation: Inmates
Sentenced Under the Three Strikes Law and a Small Number of Inmates Receiving Specialty Health
Care Represent Significant Costs

Dear Ms. Howle:
We have reviewed the above draft report from the Bureau of State Audits. We are committed to reform
the California prison medical health care utilizing best practices with the most cost effective manner.
While we welcome and concur with the audit findings and recommendations, tremendous efforts
and ongoing improvements have been undertaken as addressed in the Receiver’s Turnaround Plan of
Action Tri‑Annual and monthly reports. You can find the Tri-Annual and monthly reports on our website at
http://www.cphcs.ca.gov/receiver.aspx.
Again, we would like to thank you and your staff for the valuable review and recommendations.
Sincerely,
(Signed by: J. Clark Kelso)
J. Clark Kelso
Receiver

*  California State Auditor’s comment appears on page 75.

1

74

California State Auditor Report 2009-107.2

May 2010

Blank page inserted for reproduction purposes only.

California State Auditor Report 2009-107.2

May 2010

Comment
CALIFORNIA STATE AUDITOR’S COMMENT ON THE
RESPONSE FROM CALIFORNIA PRISON HEALTH
CARE SERVICES
To provide clarity and perspective, we are commenting on the
response to our audit report from California Prison Health
Care Services (Health Care Services). The number below
corresponds to the number we placed in the margin of Health Care
Services’ response.
Although Health Care Services concurs with our findings
and recommendations, we are disappointed that the federal
court‑appointed receiver’s response does not specify how Health
Care Services intends to implement the recommendations. We look
forward to reviewing the more specific actions the receiver plans
to take related to each of the recommendations when Health Care
Services submits its corrective action plans at 60 days, six months,
and one year from the date of the publication of this audit report.

1

75

76

California State Auditor Report 2009-107.2

May 2010

cc:	

Members of the Legislature
Office of the Lieutenant Governor
Milton Marks Commission on California State
Government Organization and Economy
Department of Finance
Attorney General
State Controller
State Treasurer
Legislative Analyst
Senate Office of Research
California Research Bureau
Capitol Press