CDCR Audit - Needs to Better Ensure Against Conflicts of Interest and to Improve Its Inmate Population Projections, CA State Auditor, 2005
Download original document:
![Brief thumbnail](/media/cache/9f/fc/9ffc19d08353c2f4d8d40ab09df0547a.jpg)
Document text
![Brief thumbnail](/media/cache/9f/fc/9ffc19d08353c2f4d8d40ab09df0547a.jpg)
Document text
This text is machine-read, and may contain errors. Check the original document to verify accuracy.
S T I D U A E T A E A U O F S T It Needs to Better Ensure Against Conflicts of Interest and to Improve Its Inmate Population Projections R U B California State Auditor Department of Corrections: September 2005 2005-105 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. ���������� ����� ������� ������ �� ����� ������ �� ����������� ����� ������� ����� ������ ����� ������� September 13, 2005 2005-105 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 Bureau of State Audits presents its audit report concerning the Department of Corrections’ (department) contracting practices for private community correctional facilities (CCFs) and inmate population forecasting. This report concludes that in processing a no-bid CCF contract that was ultimately rescinded, certain contacts made by two of the contractor’s employees who formerly worked for the department may have violated conflict-of-interest laws, and the department does not ensure that its retired annuitants in designated positions file statements of economic interests. Nonetheless, information the department relied upon to determine the need for two no-bid CCF contracts appears accurate. Additionally, the department’s inmate population projections are useful for budgeting, but have limited value for longer-range planning, such as determining when to build additional facilities. Finally, because certain practices increase the subjectivity of the department’s projections and no documentation of the projection process exists, our statistical expert could not establish the validity of the projection process. Respectfully submitted, ELAINE M. HOWLE State Auditor ������ �� ����� ������ ��� ������� ����� ����� ���� ����������� ���������� ����� ���������� ����� �������� ���� ����� �������� ������������������ CONTENTS Summary 1 Introduction 5 Chapter 1 Although Consistent With State Requirements, the Department of Corrections’ Noncompetitive Contracting Policies Were Not Always Followed and Its Policies Could Better Ensure Against Conflicts of Interest 15 Recommendations 36 Chapter 2 The Department’s Inmate Projections Are Reasonably Accurate for Short-Term Purposes, but Their Value for Long-Range Planning Is Limited 37 Recommendations 50 Response to the Audit California Department of Corrections and Rehabilitation 53 California State Auditor’s Comments on the Response From the California Department of Corrections and Rehabilitation 57 SUMMARY RESULTS IN BRIEF Audit Highlights . . . Our review of the California Department of Corrections’ (department) processing of two no-bid community correctional facility (CCF) contracts and its projections of inmate populations revealed the following: þ Although one CCF contract was never executed, actions taken by two of the contractor’s employees who formerly worked for the department may have violated conflict-of-interest laws. þ The department does not ensure that retired annuitants in designated positions file statements of economic interests. þ The department, the facility owner, and the potential contractor all incurred costs before the department received approval to proceed with a no-bid contract. þ Information the department relied upon to determine the need for the no-bid contracts appears accurate. þ The department’s inmate population projections are useful for budgeting, but have limited value for longer-range planning, such as determining when to build additional facilities. T he California Department of Corrections (department), which operates California’s state prisons, housed more than 164,000 inmates on June 30, 2005. In 2004, to cope with a large increase in inmate population, the department remedied overcrowded conditions by using less secure areas in its facilities to house inmates. As a partial solution to the overcrowding, it also decided to reopen two closed community correctional facilities (CCFs) using one-year, noncompetitively bid contracts (no-bid contracts). CCFs are designed to house inmates who meet certain criteria, such as lower security classifications and no escape history. Although its policies and procedures for processing two no-bid contracts and identifying potential conflicts of interest are consistent with state requirements, the department did not ensure that these contracts were free of conflicts. For example, two employees involved in one contract had previously worked for the department and were still listed as retired annuitants. The contractor failed to disclose this information, and the department did not require the contractor to submit employee statements of economic interests. Moreover, contacts between these two employees and the department during the formulation of the contract possibly violated conflict-of-interest laws even though the contract was never executed. Further, the department lacks a system to ensure that retired annuitants file such statements. The department also forwarded one contract to the contractor for signing before receiving approval from the Department of General Services on its request to issue a nobid contract. Perhaps as a result, the department, the potential contractor, and the facility owner all incurred unnecessary expenses. The department later rescinded this request partially because of concerns about conflict of interest. Also, in justifying the two no-bid contracts for the CCFs, the department presented a misleading claim of cost savings because it used cost comparisons that did not include all comparable costs. Nonetheless, the information it used to determine the need for reopening the two CCFs appears reasonable. continued on next page . . . California State Auditor Report 2005-105 1 þ Because certain practices increase the subjectivity of the department’s projections and no documentation of the projection process exists, our statistical expert could not establish the validity of the projection process. To assist it in managing its resources, in the spring and again in the fall of each year, the department prepares a projection of the number of inmates it expects to house over the next six years. It uses these projections to determine its budget needs and assess its ability to house inmates. Although the projections are reasonably accurate for the first two years, they are significantly less accurate after the second year. As a result, the projections are useful for assessing budget needs—which rely on information relating to the first two years of the projection—but have limited usefulness for longer-range planning, such as determining when additional facilities should be built. Certain practices increase the subjectivity of the department’s projections. It uses staff experience to update information it uses for some of the variables that determine its inmate projection. Also, the unit that develops these projections did not always abide by its own policy that precludes using estimates that lack reasonable support. Because of these factors and others, including a lack of documentation, our statistical expert could not establish the validity of the process the department uses in making its inmate projections and believes it should consult with a group of statisticians to develop a more defensible projection process. RECOMMENDATIONS To strengthen controls over its processing of no-bid contracts, the department should do the following: • Wait until all proper authorities have approved the contract justification request before signing the contract or sending it to a contractor for signature. • Require key contractor staff to complete statements of economic interests. • Include all its costs when it decides to include cost comparisons in justification requests or state that the cost comparison is incomplete. To ensure that its staff is free from potential conflicts of interest, the department should require the retired annuitants it assigns to designated positions to submit statements of economic interests. 2 California State Auditor Report 2005-105 To increase the accuracy and reliability of its inmate projection, at a minimum the department should do the following: • Fully document its projection methodology and model. • Update its variable projections with actual information, such as the new security level data, whenever feasible to do so. Additionally, if the department intends to continue using the projections for long-term decision making, such as facility planning, it should ensure that it employs statistically valid forecasting methods. It should consider seeking the advice of experts in selecting and establishing the forecasting methods that will suit its needs. AGENCY COMMENTS The department generally concurs with 10 of our recommendations and has agreed to study the feasibility of implementing the remaining two. n California State Auditor Report 2005-105 3 Blank page inserted for reproduction purposes only. 4 California State Auditor Report 2005-105 INTRODUCTION BACKGROUND E stablished in 1944, the California Department of Corrections (department) operates California’s state prisons, oversees community correctional facilities (CCFs), and supervises parolees. As of June 30, 2005, it housed more than 164,000 inmates; with the completion of the Kern Valley State Prison facility in January 2006, it will operate prisons and other facilities with a maximum capacity of about 176,500. On July 1, 2005, the governor reorganized the Youth and Adult Correctional Agency into the Department of Corrections and Rehabilitation. The reorganized department operates California’s youth and adult correctional and rehabilitation operations within three main programs: adult operations, adult programs, and juvenile justice. The Division of Adult Institutions within adult operations is now responsible for the secure custody of adult inmates while providing effective programming to reduce recidivism, duties previously carried out by the Department of Corrections. In this report, “department” refers to the Department of Corrections prior to the reorganization. The department’s budget for fiscal year 2005–06 is $6.5 billion for its five programs: Institutions, Health Care Services, Inmate Education, Community Correctional, and Central Administration programs. The Institutions Program is responsible for operating 33 state prisons and 12 CCFs. The department contracts with both private and public independent operators of CCFs Housing for Different Security Levels to house just over 4,700 inmates. Level 1 Description Open dormitories without a secure perimeter 2 Open dormitories with secure perimeter fences and armed coverage 3 Individual cells, fenced perimeters, and armed coverage 4 Cells, fenced or walled perimeters, electronic security, more staff and armed officers both inside and outside the installation California State Auditor Report 2005-105 INMATE CLASSIFICATION The department assigns different security levels to inmates within its prisons, based on its assessment of behavior and other factors. It then houses inmates in facilities designed for their security levels, which range from level 1 to level 4, with level 1 being the minimum-security level and level 4 the maximumsecurity level. Additionally, the department has units designed to house the most violent and dangerous inmates. At reception centers, where 5 new or returning inmates are received after sentencing, a counselor evaluates each using a scoring sheet, which determines the level necessary to ensure the safety of inmates, correctional personnel, and the general public. In the past, the department has found it necessary to reclassify some inmates into higher security classifications because of factors or circumstances not considered in the initial screening. In October of 2002, it implemented a new scoring system that includes several measures intended to reduce the number of inmates being reclassified, including instituting minimum scores for offenses such as previous escape from custody, regardless of other factors. During a transition period inmates were reevaluated, and all incoming inmates are now evaluated with the new system. Inmates are sentenced to the department’s prisons for different reasons. As shown in Figure 1, based on prison census data, the majority of inmates at December 31, 2004, were admitted to prison after being sentenced for a newly committed crime. Another quarter of the inmate population committed a new crime while on parole for a previously committed crime. The remaining inmates violated the terms of their parole agreement and were returned to the department’s custody or were in the process of having their parole revoked due to violating the terms of their parole agreement. As of December 2004, 99 percent of the inmates were felons, and the majority of the other 1 percent were in prison for other offenses such as narcotic addiction. COMMUNITY CORRECTIONAL FACILITIES In 1965, to provide needed housing, supervision, counseling, and other correctional programs for inmates committed to its institutions, state law authorized the department to establish, operate, and contract for “community correctional centers,” commonly referred to as CCFs. The Community Correctional Facilities Administration, within the Institutions Division of the department, administers support functions, such as developing and processing CCF contracts. The department’s Office of Contract Services also participates in contract matters, such as the administration of the original contract and any amendments. 6 California State Auditor Report 2005-105 FIGURE 1 Reasons for Commitment to Prison As of December 31, 2004 ������ ���� ��������� �� ����� ��������� ����� ����� �� ������ ��� ������� ���������� �� ������ �� ����� ��������� ����� ��� Source: Department of Corrections’ prison census data, as of December 31, 2004. CCF Inmates must meet certain criteria including the following: • Cannot have more than 18 months left to serve on their sentence. • Not above a level 2 security classification. • Cannot have significant medical or behavior problems. • Cannot have an escape history. • Not in prison for certain sex crimes or certain types of assault. As shown in the text box, inmates the department houses in CCFs must meet certain criteria. The department houses these inmates along with inmates who meet the CCF criteria and who are registered sex offenders but are not considered sexually violent predators in modified CCFs, which have additional security features. Three of the 12 CCFs currently under contract with the department are modified. Though only certain inmates are eligible for transfer to a CCF, these facilities provide the same living situation for inmates as a prison. Inmates in CCFs are eligible for all services that inmates in other institutions receive, and the department treats inmates at the two types of facilities in the same way. Of the 12 CCFs in use as of July 2005, private companies operate six and local governments operate the other six. The department normally subjects the contracts for both private and public CCFs to a competitive bidding process. City and local governments run public CCFs, and the law mandates that persons providing California State Auditor Report 2005-105 7 security at the public facilities be peace officers. Private companies hire their own employees, but, because private citizens do not have peace officer powers, the department provides uniformed peace officers as an additional precaution in these six facilities. Private companies currently run all three of the modified CCFs. The 12 facilities provide capacity for 4,733 inmates. Figure 2 presents a map of CCF locations in California. McFARLAND AND MESA VERDE CCF CONTRACTS The department’s fiscal year 2003–04 budget did not include funds to continue the contracts for three private CCFs: McFarland, Mesa Verde, and Eagle Mountain. These three facilities closed on December 31, 2003. According to department records, the reason these contracts were not extended was because of a projected excess capacity of housing for this segment of inmate population—partly due to anticipated parole reforms, the effects of Proposition 36, and actual inmate population trends. Proposition 36 provides drug treatment instead of prison terms for certain nonviolent drug offenders. However, in 2004 the department experienced a large unexpected increase in inmate population because parole reform programs were not carried out and because new inmate admissions from counties increased. Since prior population projections had generally projected a stable population through 2009, the department did not expect this large increase. To respond to this situation, it put thousands of added beds into use, some located in “overcrowding” areas—temporary beds placed in areas that are more difficult to secure, such as gymnasiums and dayrooms. In summer 2004, the Youth and Adult Correctional Agency and the department decided to reactivate two of the closed CCFs, McFarland and Mesa Verde, using one-year, no-bid contracts, while initiating a competitive bidding process for a longer-term solution. The department did not consider reactivating the third closed CCF, Eagle Mountain, because it was in a remote location and the department had some concerns about the level of security provided by the facility. As of July 2005, the department was proceeding with an invitation for bid—a competitive bidding process—to add additional CCF beds or to replace those CCFs with expiring contracts. The invitation for bid requires the department to choose the lowest-priced responsible bidder. 8 California State Auditor Report 2005-105 FIGURE 2 Location of Community Correctional Facilities ������� ��������� ������������ ���������� � � � ��� ����� �������� ��� ������� ������� ���� ��� ���������� ��������� ������ ����� ������ ��������� ������������ ���������� � � � � � � ������� � ���������� ������� ���������� ��������� ������� ������� �������� ������� ������ �������� ������� ����� ���� ��������� �������� ������� �������� ��������� ������������ ���������� �� ������ ������ ��������� �� ������� ������� ��������� �� ������ ����� �������� ������ � ���������� ������� ��������� ������������ ���������� �� �������� ��������� ������������ ���������� ��� ��������� �� ��������� � � � ���� ������ ����������� ������� �������� ��� ����� ������ ������� �������� ������� ���� ��� ����� ����� ��������� ������ ������� �������� ��� ����� �������� ��� ��������� ������ � �� �� � � � � � ����������� �� � � � ����������� ��� ������� ����� � ��� ����� Source: Department of Corrections’ Web site. California State Auditor Report 2005-105 9 POPULATION PROJECTIONS UNIT The department’s Population Projections Unit (projections unit) generates population projections for time frames that span six fiscal years, monitors and reports on the quality of the projections, and explains inconsistencies between actual and projected populations. The annual population projections correspond with the State’s budget cycle and drive the department’s annual budget request. The department prepares its budget request using the fall population projection and submits this request to the Department of Finance (Finance) for use in preparing the Governor’s Budget. It revises its budget request based on the spring population projection and submits the revision to Finance for inclusion in the May revision of the Governor’s Budget. The department also uses these projections to assess the ability of its facilities to house the inmate population over a six-year timeline. The department uses a method referred to as microsimulation to develop its inmate projection. Conceptually, this method is designed to trace the path of individual offenders to project inmate populations on a month-to-month basis as they come into, move through, and move out of the correctional system. The projection process consists of the following four distinct steps: • Premises—the projections unit evaluates its most recent projection to assess the continuing validity of the assumptions that generated the projection. During this step it also updates its forecast of new inmate admissions coming from the State’s superior court system. Finally, the projections unit analyzes data and interviews key staff to develop assumptions that affect the projection. For example, it meets with key people within state government who set policies that can affect the amount of time an inmate spends in prison or on parole to identify specific policies that may change or new programs to be implemented, the effective dates of these actions, and dynamics of the programs. This step is considered complete once the premises have been presented and approved by the department’s executive staff. • Projections––the projections unit develops the population projection by translating the assumptions of the prior step into adjustments to the variables included in the microsimulation model. The unit bases the input variables on the prior projection and changes them as needed. 10 California State Auditor Report 2005-105 • Postscript—the projections unit refines its projection information to prepare a detailed final agency budget request. This process includes developing a number of specialized supporting projections needed for budgeting but not provided by the projection model. • Monitoring—the projections unit performs weekly and monthly comparisons of projected and actual populations, and prepares a monthly report that includes all the key measures for analyzing projection accuracy; the unit also assists the department’s management in formulating recommendations for upcoming projections. Monitoring is an ongoing process that runs through the duration of the projection cycle. According to the department, it developed its current microsimulation model in 1976 when laws allowed courts to sentence inmates to determinate prison terms, making it easier to determine inmates’ average length of incarceration. Since the department’s budget relies on the semiannual projections that it prepares, the projections unit considers it very important that its projections be accurate and objective. The unit makes an effort to include participation of key stakeholders, such as department staff from the institutions and parole divisions, when presenting issues that will affect the inmate and parole populations. The unit also receives occasional outside reviews. Although it is within the department, the projections unit considers itself independent of budget and policy-making decisions. SCOPE AND METHODOLOGY The Joint Legislative Audit Committee (audit committee) requested that the Bureau of State Audits evaluate the process the department used to negotiate and enter into two no-bid contracts for private prison facilities to determine whether its policies and procedures are consistent with and adhere to current laws and regulations, particularly in relation to conflictof-interest rules. In addition, the audit committee asked us to analyze information the department used in its decision to enter into the two no-bid contracts to determine whether such information was accurate and reliable, to analyze the reasonableness and consistency of its method of tracking and projecting inmate population, and to assess the validity of any cost savings it identified. Lastly, the audit committee asked us to obtain a legal opinion on whether the department entered California State Auditor Report 2005-105 11 into the two no-bid contracts in bad faith or in violation of state law and to identify any remedies open to the State and their potential impact. To determine whether the department’s policies and procedures for negotiating and awarding no-bid contracts are consistent with current state laws and regulations, we interviewed contracting staff and obtained an understanding of its policies and procedures relating to such contracts. We also documented state laws and regulations and the department’s written policies and procedures related to no-bid contracts. To determine if conflicts of interest were present, we reviewed the contracts’ provisions and contract files, identified all staff who participated in developing the two contracts, and evaluated any statements of economic interests. We then determined whether any staff had relationships or interests that would present a conflict of interest for these two contracts. We also reviewed the contract files and interviewed staff at the two companies to identify contractor staff involved in negotiating and completing the contracts, and assessed whether they had previously worked for the department. To assess whether the information the department used in its decision to reopen the two CCFs was accurate and reliable, we determined the information the department used to establish the need for the two no-bid CCF contracts. In addition, we obtained monthly inmate population counts for the last three fiscal years and compared them to the projected population estimates for that period to determine whether the inmate population rose unexpectedly. We also identified any cost savings the department claimed and assessed, to the extent possible, the validity of the claimed cost savings. To gauge the reasonableness of the contract amounts, we compared the annual average cost per bed for the two no-bid contracts to the annual average cost per bed of the department’s other CCF contracts. To determine whether the department’s method of tracking and projecting inmate populations is reasonable and consistent, we reviewed forecasts for the past three fiscal years to determine if the assumptions and variables included are reasonable and consistent, and further analyzed those that have a material effect on the population projection. We also consulted with a statistical expert to obtain an analysis of the department’s use of its microsimulation model. 12 California State Auditor Report 2005-105 In addition, we compared the actual inmate populations to the levels forecast by the department over the past 10 years to determine their accuracy and analyzed the sufficiency of the department’s method of projecting the number of needed inmate beds for facility planning purposes. Furthermore, we determined whether the department has a process to communicate with counties to identify policy and other changes that may affect the number of new inmates sentenced to state prison. In determining whether the department entered into the two no-bid CCF contracts in bad faith or in violation of the law, we did not find any situations that presented a conflict of interest in the processing of the one no-bid contract that was fully executed relating to the McFarland facility. For the other no-bid contract for Mesa Verde, the contract was rescinded before it was fully executed. Therefore, according to our legal counsel, the questions of whether the contract was entered into in bad faith or in violation of state law and what remedies are available to the State are moot. During this audit, it came to our attention that the department has not fully documented a computer application it uses to classify inmates and parolees. The State Administrative Manual requires agencies to fully document applications of information technology. Additionally, industry standards produced by the IT Governance Institute require organizations to document information technology applications for users, operators, and trainees. However, the department has not fully documented its classification application. Because this issue was not within the scope of our review, we have reported it in a separate letter to the secretary of the Department of Corrections and Rehabilitation. n California State Auditor Report 2005-105 13 Blank page inserted for reproduction purposes only. 14 California State Auditor Report 2005-105 CHAPTER 1 Although Consistent With State Requirements, the Department of Corrections’ Noncompetitive Contracting Policies Were Not Always Followed and Its Policies Could Better Ensure Against Conflicts of Interest CHAPTER SUMMARY O f two noncompetitively bid contracts (no-bid contracts) the Department of Corrections (department) developed for community correctional facilities (CCFs), one did not consistently comply with state requirements. Although department policies and procedures for processing no-bid contracts and identifying potential conflicts of interest are consistent with state requirements, in practice the department has not always ensured that such contracts are free of such conflicts. For example, the contractor for one of the no-bid contracts failed to disclose that two of its staff had previously worked for the department, thus creating a conflict of interest. The department ultimately rescinded this contract because the inmate population had declined and because of conflict-of-interest concerns. In the meantime, however, the department sent the contract to the contractor for signing before receiving final approval to award a no-bid contract from the Department of General Services (General Services). Moreover, the department, the potential contractor, and the facility owner began incurring costs to prepare the facility for operation. Because the contract was ultimately rescinded, the department, the potential contractor, and the facility owner all incurred unnecessary expenses. The information used by the department to determine its need to open the two CCFs appears accurate and reasonable. However, the cost comparisons it presented in its justifications for the two no-bid contracts did not include all comparable costs and, as a result, presented a misleading claim of cost savings. California State Auditor Report 2005-105 15 IN ONE NO-BID CONTRACT, THE DEPARTMENT DID NOT FULLY COMPLY WITH STATE REQUIREMENTS AND TWO CONTRACTOR EMPLOYEES MAY HAVE VIOLATED CONFLICT-OF-INTEREST LAWS As discussed in the Introduction, in 2004 the department decided to open two CCFs that had been closed the previous year because of a projected decline in inmate population. One of these contracts—for the McFarland facility (McFarland)—was appropriately approved and is currently in effect. Although its processing of the McFarland contract complied with state requirements, the department did not adhere to all requirements in processing the other contract—for the Mesa Verde facility (Mesa Verde). The department began to incur costs relating to the Mesa Verde contract before obtaining approval of its justification request for a no-bid contract from General Services and failed to identify potential conflicts of interest until after it had initially approved the contract. Although the Mesa Verde contractor did not identify any conflict of interest when signing the contract, two of its employees, who were involved in the contract process, had actually worked for the department in the 12 months prior to the contract being developed. The department processed the justification for the Mesa Verde contract and received all but the final approval from the director of General Services before rescinding it because of a decline in the population of inmates and concerns that one of the contractor’s employees had a conflict of interest, according to the former chief deputy for support services. The Department Began Incurring Costs Related to the Mesa Verde Contract Prior to Receiving Appropriate Approval The department, the facility owner, and the potential contractor all incurred costs before receiving approval from General Services to award the contract without competition. 16 Before awarding a contract without competition, the department must obtain the approval of General Services. Also, as part of the contract award process, after General Services’ approval of the request justifying an exemption from competitive bidding, the department operations manual requires contracts to be forwarded to the contractor for signature. This was the process the department used in executing the McFarland contract. However, it sent the Mesa Verde contract to the contractor for signature before obtaining General Services’ approval of its justification for exemption. The department later rescinded its request for exemption because of a decline in inmate population and because of conflict-of-interest concerns. It did notify the contractor by letter that the contract was not fully approved or in effect until General Services gave its final California State Auditor Report 2005-105 approval. Nevertheless, the department, the facility owner, and the potential contractor all incurred costs before receiving approval from General Services. As the department began the contract process for Mesa Verde, it developed a justification for requesting a no-bid contract for $5.8 million and submitted it to General Services for approval. As shown in Figure 3, on December 4, 2004, the director of General Services approved the initial request. However, later in that same month the department realized it had not included $1 million to equip and refurnish the facility so that it could be operated as a CCF. It submitted a revised no-bid justification request to General Services for $6.8 million on December 29, 2004. The revised request was being routed for review and approval by the appropriate staff at General Services when the department, on February 3, 2005, rescinded its request and recalled the justification for the no-bid contract from General Services because of the decline in inmate population and its concerns related to a conflict of interest with one of the contractor’s employees. FIGURE 3 Mesa Verde No-Bid Justification Request and Contract Time Line �������� �� ���� ������� �������� �������� �������� ������ ������������� �������� �������� �� ���� ���������� �������� �������� ������ ������������� �������� �������� ���� �������� ��� ���� ������� �������� ������� ������ ������������� ������� ��� ������� �� �� ������� ��������� �������� ���� �������� �� ���� ������� �������� �������� ������ ������������� ������� ��� ������� �� �� ������� ��������� ������� �� ���� ���������� ������� ������ �������� ���������� ������� ��� ���� ���������� ����� �������� ���������� ������� ���� �������� ��� ���� ���������� ����� �������� �� ���������� ��� ���������� �������� �� ���� ���������� �������� ������� �������� �� ������ ������� ������ ������������� �������� �������� ���� �������� �� ���� ������� �������� ������� ��� ������� ������ ������������� ������� �� ��� ����������� �������� ��� ���� ���������� �������� ��� ����� �� ������ ������� ������ ������������� �������� Source: Department of Corrections’ contract files. * Agency refers to the Youth and Adult Correctional Agency that, prior to July 1, 2005, had oversight responsibility for the Department of Corrections. California State Auditor Report 2005-105 17 According to the department’s associate director of the Office of Business Services, in some cases it is not feasible from an operations standpoint to wait until the no-bid justification has been approved by General Services to begin processing the contract for signature by the contractor and the department. He further stated that, because of the length of the time frames required by General Services to review and approve a no-bid justification request, the department typically must begin to process the contract while awaiting General Services’ approval in order to obtain the goods or services required to meet its business needs. According to the associate director, through its July 1, 2005 reorganization, the department has eliminated four approval levels in the internal processing of no-bid justifications. He believes this new process will help to ensure that whenever possible, no-bid justifications will be forwarded to General Services within their required time frames for review and approval. Even though it had not received approval from General Services for the revised justification to proceed with the no-bid Mesa Verde contract with its contractor (Civigenics), the department began to process the contract on December 31, 2004, and to incur preparation costs. 18 Even though it had not received approval from General Services for the revised justification to proceed with the no-bid Mesa Verde contract with its contractor (Civigenics), the department began to process the contract on December 31, 2004, by sending the agreement to the contractor for signature. Until it rescinded the contract justification, the department, Civigenics, and the owner of the Mesa Verde facility began incurring costs in preparation for operating the facility. According to the correctional captain who would have overseen the reactivation of Mesa Verde, the department began interviewing staff to fill the positions needed to operate the facility. Also, the captain said he ordered supplies such as clothing for the inmates who would be assigned to the facility. When we spoke to the owner of Mesa Verde in June 2005, he told us the inmate clothing had just recently been moved from the facility. The facility owner and Civigenics told us they spent $200,000 and $50,000, respectively, preparing to operate the facility on behalf of the department. The department could not estimate the cost of the administrative time it incurred in anticipation of operating the facility. According to the chief of its service contracts section, the reason the department proceeded before it obtained General Services’ approval was to expedite the contract process so that when General Services’ approved the justification, the department could quickly finalize it. California State Auditor Report 2005-105 Although the Department Has Controls in Place to Identify Conflicts of Interest, a Conflict May Have Existed With the Unexecuted Mesa Verde Contract Despite conflict-of-interest disclosure requirements in the contract, Civigenics did not disclose that two of its employees had worked for the department within the past year. As of July 2005, these same two Civigenics employees were also listed as current retired annuitants available to work at the department. According to Civigenics officials, the company hired one former high-ranking department employee to develop a strategic plan and the other to help with the reactivation of Mesa Verde. The employment of the two individuals by both the department and Civigenics created potential conflicts of interest that, had the contract been fully executed, could have rendered it void. Moreover, certain contacts between these two individuals and the department during the contract formation process raise the possibility that conflict-of-interest laws were violated even though the contract was never fully executed. Civigenics did not disclose that two of its employees, who would be involved with the contract, were former high-ranking employees who had retired from the department within the past year and were also current retired annuitants of the department. The two no-bid contracts we reviewed included standard terms regarding conflicts of interest, which require contractors to abide by state requirements related to conflicts of interest and to disclose in writing all interests and activities that create an actual or potential conflict of interest in performance of the agreement. The terms further provide that if a contractor violates those requirements, such action will render the agreement void. Additionally, the terms of both contracts provide that the department may require contractors and their employees to file statements of economic interests. These standard terms, if followed, appear to provide the department with the information needed to adequately assess the presence of conflicts on the part of contractors, subcontractors, and employees. Despite these requirements, Civigenics did not disclose that two of its employees, who would have involvement with the impending contract, were former high-ranking employees who had retired from the department within the past year and were also current retired annuitants of the department.1 The Political Reform Act of 1974 (act) imposes various restrictions on state officers and employees related to conflicts of interest, both while employed by state government and after they leave state service. These restrictions prohibit employees 1 We identified a third Civigenics employee who previously worked for the department; however, this person separated from state service on April 28, 2000, and was never hired as a retired annuitant. California State Auditor Report 2005-105 19 The act defines an attempt to influence a government decision as contacts, appearances, or other attempts to influence an officer or employee of a state agency, including contacts on behalf of a business entity. from participating in making a government decision, such as the decision to enter into a contract, if they have a material financial interest in the decision. The provisions of the act that apply during employment prohibit a public official from attempting to use his or her official position to influence a government decision, such as the decision to enter into a contract in which he or she has a financial interest. In addition, the so-called “revolving door” provisions of the act prohibit a public official who has left state employment from contacting his or her former employer within 12 months of leaving state employment in an attempt to influence the formation of a contract in which he or she has a financial interest. The act defines an attempt to influence a government decision as contacts, appearances, or other attempts to influence an officer or employee of a state agency, including contacts on behalf of a business entity. The act is designed to ensure that public officials perform their duties impartially, free from bias caused by their own financial interests or the financial interests of those who support them. State agencies have the discretion to employ retired individuals as retired annuitants. A state agency may reinstate a retired state employee, and that person may actively perform the duties of a state position. As shown in Figure 4, a deputy director in the Health Care Services Division—employee A in Figure 4—retired on January 31, 2003, and was reinstated as a retired annuitant effective the following day. The retired deputy director last worked for the department on June 3, 2004, but he remained on its payroll system as a currently employed retired annuitant as of July 2005. Similarly, the chief deputy director of field operations—employee B in Figure 4—retired from the department on May 30, 2003, and was reinstated as a retired annuitant effective two days later. The retired chief deputy director last worked for the department on May 27, 2004, but remained on its payroll system as a currently employed retired annuitant as of July 2005. Although these two individuals are not actively performing state service, they are currently employed as retired annuitants by the department, and it may still request them to perform active state service. The Fair Political Practices Commission (commission) is charged with interpreting and enforcing the provisions of the act. The commission considers an individual who is actively performing the duties of state service to be subject to the restrictions previously described that apply during employment, regardless of that person’s status as a full-time, intermittent, or retired annuitant employee. Consequently, these two individuals 20 California State Auditor Report 2005-105 FIGURE 4 Retired Annuitant Time Line ����� �� ����� ��������� � ��� � ��� ��� �������� ���������� �� �������� ��������� ��� ����� ������� ��� ���� �������� � �������� ��� ��� ���� �������� ��� ���� ���� ��� ��� ��� ����������� �������� �� ���� �������� � ��������� ������� ��������� ������ ��������� ���� �� ���� �������� ��� ���� ���� ��� ��� ��� ����������� ���� ���� ���� ����� �� ����� ��������� � ��� � ��� ��� �������� ���������� �� �������� ��������� ��� ����� ���� �� ���� �������� � ��������� ������� ��������� ����� ������ ��������� ������� �� ���� ���������� �������� ������ �� ������ ��� ��� ���� �������� � �������� ��� �� ���� �������� � �� ���������� �� � ������� ��������� ������������ ������������� ��������� Source: Department of Corrections’ personnel files. Note: Employee A is a former deputy director; employee B is a former chief deputy director. * The deadline for filing annual statements of economic interests is April 1. would be treated as state employees for purposes of the act both prior to their retirement and during the period they actively performed state service as retired annuitants immediately following their retirement. However, it is unclear whether a retired annuitant who is not currently performing active state duties would be treated as a current employee or as a former employee for purposes of the act. Nevertheless, any attempts to influence this contract during the period in question would violate the act. In October 2004, the department selected Civigenics, the current employer of these individuals, to operate Mesa Verde. It initially received approval from General Services to award this contract on a noncompetitive basis in December 2004 but rescinded its California State Auditor Report 2005-105 21 Several department staff told us the former chief deputy director contacted them during the time the contract was being processed to inquire on its progress. request in February 2005. Several department staff told us the former chief deputy director contacted them during the time the contract was being processed to inquire on its progress. The current department director and a chief deputy director of field operations told us the former chief deputy director visited the executive offices of the department, accompanied by representatives of Civigenics. In addition, department staff told us the former deputy director contacted the department at least once during this period in his capacity as a Civigenics employee. Because these individuals contacted the department during this period, we believe they may have violated the applicable provisions of the act. Despite the fact that the department rescinded its request for General Services’ approval and this contract was never executed, our legal counsel has advised that any “attempts to influence” the contract decision-making process may still constitute a violation of the act. The Public Contract Code imposes similar restrictions both during employment and post-employment. Section 10410 prohibits a current employee from receiving compensation for an activity funded by a state contract, and Section 10411 prohibits a former employee from entering into a contract with his or her former agency if the official held a policymaking position with the agency in the same general subject area as the proposed contract within the 12 months prior to leaving state service. Regardless of whether we treat these individuals as current employees by virtue of their continued status as currently employed retired annuitants or as former employees because they were not currently performing active state service, they would have violated the applicable provisions and voided the contract if it was fully executed. Identifying potential conflicts of interest early on in the contracting process helps avoid investing resources into a contract that may ultimately be void and require that the whole contracting process be repeated. One way the department could guard against situations such as this in the future would be to exercise its authority to require key contractor staff to complete statements of economic interests (statements) as part of the contract development process or to specifically ask whether any current or former department employees will be assigned to the contract. Moreover, had the department required its retired annuitants to file statements, it would have been aware the retired annuitants were also employees of Civigenics. 22 California State Auditor Report 2005-105 The department now apparently believes the involvement of these former employees in a contract with Civigenics would violate conflict-of-interest laws. In June 2005, Civigenics submitted a competitive bid for the Mesa Verde contract. However, the department disqualified this bid in a letter dated July 14, 2005, on the basis that the participation of a current employee created a conflict of interest. No Apparent Conflicts of Interest Exist in the McFarland Contract In February 2005 a newspaper article alleged a connection between the former director of the Department of Finance (Finance) and the GEO Group, Inc. (GEO), the company the department contracted with to operate the McFarland CCF. The former director of Finance was appointed as a trustee of a real estate trust that owns the McFarland facility. We reviewed this connection to determine whether the former director’s involvement constituted a conflict of interest. Even though the former Finance director serves on the board of trustees (board) of a real estate investment trust company that has an ownership interest in McFarland, we found no evidence that she influenced the department’s decision to reopen the facility. Even though the trust benefits from the lease on McFarland, the 10-year lease between the two companies began more than five years before the former Finance director joined the trust’s board. The former director left Finance in October 2004 and joined the board of Correctional Properties Trust (the trust), whose wholly owned subsidiary owns McFarland. The subsidiary leases McFarland to GEO, who operates the facility. According to its most recent annual report, neither the trust nor its wholly owned subsidiary has an ownership interest in GEO. Nonetheless, 90 percent of the trust’s rental income comes from leases with GEO. Even though the trust benefits from the lease on McFarland, the 10-year lease between the two companies began in April 1998, more than five years before the former Finance director joined the trust’s board. Moreover, GEO was the only company that could have operated McFarland on the department’s behalf because it had an existing noncancelable lease on the facility through April 2008. Because GEO’s lease payment is fixed during the 10-year lease period, allowing for annual inflation adjustments not to exceed 4 percent, the State’s contract with GEO to operate the facility does not appear to provide an additional benefit to the trust. Since the former Finance director is associated with the trust and not with GEO, it does not appear that she has a conflict of interest in regards to the McFarland contract. California State Auditor Report 2005-105 23 We also reviewed other GEO staff who were involved with or could have influenced the contracting process to determine if any had recently worked for the department and therefore presented a potential conflict of interest. However, none of the GEO staff we identified had worked for the department in the past five years. We identified no conflicts of interest relating to the department’s no-bid contract with GEO. THE DEPARTMENT CAN IMPROVE ITS COLLECTION AND REVIEW OF REQUIRED DISCLOSURE FORMS The two retired annuitants we reviewed in designated positions have not filed required statements of economic interests. State law requires agencies to adopt a conflict-of-interest code that designates employees in decision-making positions and requires them to file periodic statements. Accordingly, the department has adopted regulations that list the designated positions and spell out the disclosure requirements. Although most of the employees who are assigned to designated positions with a role in developing the CCF contracts completed the required statements, some did not. All 20 department staff we reviewed who had a role in developing the two no-bid facilities contracts filed statements covering all or part of 2004, but two retired annuitants associated with one of these contracts did not. According to the commission, retired annuitants in designated positions are subject to the same rules as other employees. However, one retired annuitant who, as of July 2005 is listed on department records as a retired annuitant chief deputy director, did not file the required statements for calendar years 2003 and 2004. Another retired annuitant initially assigned to a designated position required to file a statement and reassigned two months later to another designated position never filed a statement for the periods covering either position. A personnel analyst told us the department struggles to track retired annuitants when they are hired because they are not assigned to established positions, and it is sometimes not clear whether the duties included on their duty statements could lead to conflicts of interest. However, the duty statements for the retired annuitants we reviewed clearly indicated a potential for conflicts of interest to exist. For example, one of the retired annuitants is classified as a chief deputy director of field operations, which is a designated position. This retired annuitant’s position description includes broad responsibilities, such as to plan, organize, direct, and coordinate the field operations of the 24 California State Auditor Report 2005-105 department. Therefore, the department is not following the law when it fails to require retired annuitants assigned to designated positions to complete required disclosure forms. The department’s practice of continuing former employees as active retired annuitants when they are not actually working could create confusion about whether its retired annuitants are subject to revolving-door prohibitions or the conflict-of-interest provisions that apply to current employees. According to the department, one of the primary reasons it hires staff who retire at the deputy director level and above as retired annuitants is to provide expert testimony in pending litigation. Typically, the department appoints retired annuitants to one-year terms and will reappoint them in the subsequent year if their services are still needed. However, because of the state hiring freeze in effect during 2001, the former department director issued a memo directing each institution and the department’s headquarters personnel office to delete the expiration dates of all currently employed retired annuitants as of December 31, 2001, to eliminate the need to seek formal freeze exemptions approved by Finance each new calendar year. We believe the department’s practice of leaving the expiration dates blank when appointing retired annuitants contributed to potential conflicts with the two retired annuitants employed by Civigenics, and continues to pose the risk for conflicts involving the department’s other retired annuitants. To ensure a continuing need for the services of retired annuitants, the former department director initiated an internal procedure in the same memo requiring a form to be prepared justifying why each annuitant’s continued services are critical to the department’s mission. The forms were to be reviewed and approved by the appropriate chief deputy director before scheduling these individuals for work. For all appointments taking effect after December 31, 2001, the memo directed that the department must file a freeze exemption and obtain the approval of Finance before allowing the retired annuitant to begin work. The department continues to leave the expiration date blank when entering such appointments in the payroll system and does not use the internal procedure each year to ensure its retired annuitants’ services are still needed. We believe this practice contributed to potential conflicts of interest with the two retired annuitants employed by Civigenics, as previously discussed, and continues to pose the risk of conflicts involving the department’s other retired annuitants. This would be true even for employees who have not worked for the department in over a year and would otherwise no longer be subject to the revolving-door limitations. California State Auditor Report 2005-105 25 According to the chief of Personnel Services, although as of August 2005 the department is still abiding by its policy of not entering expiration dates on its appointments of retired annuitants, it plans to ask each division to annually advise Personnel Services’ staff which retired annuitants are no longer working. The department will then separate the identified retired annuitants from state service. However, until it implements this change, the department will continue to be at risk from potential conflicts of interest with its contractors and has no way of knowing if its retired annuitants are still needed. Further, the department did not adhere to its previous internal procedure, which required a form to be prepared justifying why each annuitant’s continued service is critical to the department’s mission. Therefore, we are not confident that the department will follow through with its proposed policy. Because the hiring freeze is no longer in effect, we believe a better control would be for the department to revert to its earlier process of appointing retired annuitants to one-year terms and reappointing them annually if there is still a need. If its only purpose in hiring these employees as retired annuitants is to provide expert testimony in pending litigation, the department could explore other potential options for securing these services rather than hiring former employees as retired annuitants. For example, state law exempts contracts for expert witness services from the conflict-of-interest restrictions. The department could contract with these retired employees rather than appointing them as retired annuitants. However, this option would limit these retired employees’ services to only providing expert testimony. Four of 20 employees’ statements we reviewed were filled out incorrectly. 26 Also, the department does not ensure the completeness of the statements employees do file. Four of the 20 employees whose statements we reviewed filled out their statements incorrectly. For example, one employee did not fill out the dates the statement covered, so it was unclear to what period the disclosures related. Another employee did not attach any supporting disclosure schedules and also did not indicate whether he had any reportable interests requiring supporting disclosure schedules. Since the statement was not complete, it was not possible to determine if the employee had a conflict of interest. Because the department does not review all the filed statements for accuracy or completeness, it cannot ensure that its employees in designated positions have met their respective disclosure requirements. California State Auditor Report 2005-105 Three of the four incorrect statements had errors or omissions on the cover page relating to information the filing officer is required to review. If the filing officer had properly reviewed these statements, she would have identified the errors in three of them. According to the conflict-of-interest filing officer, she randomly reviews for completeness 20 percent of the statements filed by department staff. Although state regulations require the filing officer to complete a more extensive review on at least 20 percent of statements that are filed on time, these regulations also require the filing officer to review every statement for certain information contained on the cover page. Three of the four incorrect statements had errors or omissions on the cover page relating to information the filing officer is required to review. If the filing officer had properly reviewed these statements, she would have identified the errors in three of them. THE DEPARTMENT’S POLICIES AND PROCEDURES FOR NO-BID CONTRACTS AND FOR IDENTIFYING CONFLICTS OF INTEREST ARE CONSISTENT WITH STATE REQUIREMENTS To protect the public from misuse of public funds and to provide all qualified bidders with a fair opportunity to enter the bidding process, the law requires the department to advertise and obtain competitive bids on all contracts that exceed $75,000 unless otherwise excluded from these requirements. State law exempts departments from competitive bidding requirements when, among other reasons, the director of General Services approves an exemption because the State’s best interests are better served by doing so. To promote sound business decisions and practices in securing necessary services, the State has developed the State Contracting Manual, which contains policies, procedures, and guidelines relating to no-bid contracts. The State Administrative Manual also provides policy direction for departments concerning contracts exempt from General Services’ approval and contracts exempt from advertising in the State Contracts Register. To ensure that public officials perform their duties in an impartial manner, free from bias caused by their own financial interests, state law also precludes public officials from making, participating in making, or using their official position to influence a government decision in which they have a financial interest. To accomplish this, the law requires agencies to adopt and publish a conflict-of-interest code. As part of this code, agencies are required to identify the positions in their organizations that are responsible for making, or that participate in making, decisions which may have a material effect on any California State Auditor Report 2005-105 27 financial interest. After identifying such positions, each agency must then determine the specific types of interests it will require staff occupying those positions to report. The department has established policies regarding the processing of contracts, as well as regulations and policies relating to conflicts of interest, that are consistent with state requirements for these areas. Consistent with state law, the department operations manual requires all contracts above a specified amount to be competitively bid, unless specifically approved for exemption by the director of General Services. The department also has established regulations in the California Code of Regulations and policies in its operations manual to guide its staff in identifying conflicts of interest that are consistent with state law. In its regulations, the department has established a conflictof-interest code, identified the positions designated to file statements, and specified the categories of information the employees occupying these positions are required to disclose. Also, it has established policies in its operations manual that prescribe when employees who are in designated positions must file statements and when employees should be disqualified from making, participating in, or influencing any decisions that have a foreseeable material effect on their financial interests. INFORMATION THE DEPARTMENT RELIED ON TO DETERMINE THE NEED FOR CCF CONTRACTS APPEARS ACCURATE The two main causes the department cited for the unexpected increase in inmate population were the failure of its parole reforms to reduce inmate numbers and the large increase in inmate admissions from counties. 28 In justifying the need to open two CCFs using no-bid contracts, the department cited an unexpected increase in inmate population as its reason for the sudden need for additional beds. The two main causes the department cited for this increase were the failure of its parole reforms to reduce inmate numbers and the large increase in inmate admissions from counties. It had reasonably relied on earlier population projections that reflected a stable or declining inmate population. Nonetheless, when justifying the no-bid contracts, the department included a cost comparison that did not include all applicable costs. The remaining information it relied on to establish the need for two no-bid CCF contracts appears reasonable. California State Auditor Report 2005-105 Parole Reforms Enacted by the Department Did Not Produce the Expected Results In January 2004, the department implemented a new parole model consisting of three alternatives to returning parolees to prison (see text box) who had committed certain types of parole violations. It expected these reforms to reduce by 24,000 the number of parolees who return to prison each year for parole violations. The Department Parole Reforms department included the expected effects of these • Electronic In-Home Detention parole reforms when it created one of its projections Program—1,000 electronic monitoring of future inmate populations. As we discuss in the devices used to monitor parole violators in this curfew or home detention program. Introduction, it uses these projections to assess the sufficiency of its facilities to house its inmate • Community Correctional Reentry Center or Halfway Back Center population six years into the future. According to Program—900 beds that provide housing former management staff, the department was not and other services for parole violators able to realize the estimated inmate reductions. outside of prisons. In fact, in April 2005, the agency secretary who • Substance Abuse Treatment Control Unit oversees the department discontinued the parole Program—1,316 beds to house and treat parolees for 30 days in a jail setting while reform programs because of implementation providing them with services, training, and problems. 90 days of aftercare in the community. The department’s parole reforms are presented in more detail in a separate report, Bureau of State Audits’ Report 2005-111, to be released in fall 2005, including a federal judge’s June 2005 decision requiring the department to reinstate certain elements of the parole reform programs. New Inmate Admissions From Counties Rose Unexpectedly The other main reason the department provided to explain the unexpected increase in the number of inmates was the higher-than-expected number of inmates local courts sentenced to prison in fiscal year 2003–04. Our review of new inmate admissions confirmed the department’s assertion. As shown in Figure 5 on the following page, new inmate admissions from counties increased by 4,818 during fiscal year 2003–04, a significant factor in the increase in population during this time. According to the department director, Los Angeles County had the largest increase in inmate admissions of all California counties. The department’s county admission data for new inmates indicated that Los Angeles County accounted for 36 percent of the overall increase in inmate admissions in fiscal California State Auditor Report 2005-105 29 FIGURE 5 Total New Inmate Admissions From Counties ��� ������ ���������� ��� ���������� �� �� �� �� �� ��������� ������� ������� ������� ������� ������ ���� Source: Department of Corrections’ Offender Information Services Branch. “Broken Window” Theory of Policing year 2003–04. The director attributed much of this increase to the city of Los Angeles’s new police chief’s “broken window” method of policing (see text box). This form of policing is based on the theory that if one window in a building is broken and is not repaired, all of the rest of the windows will soon be broken. As less serious offenses go unpunished, an atmosphere of disorderliness can permeate a community, leading to more serious offenses. The director told us that until recently, the department did not communicate regularly with counties to identify changes that would affect the number of inmates counties send to state institutions. Accordingly, the department did not anticipate the increase in new inmate admissions Aggressive policing of less serious crimes can beginning in fiscal year 2002–03. The director bring about a reduction in more serious future crimes. identified two efforts the department has recently undertaken to enhance communications with local Sources: Wilson and Kelling, “Broken Windows.” The governments. First, it began attending statewide city Atlantic Monthly, March 1982 and Worrall, “Does police chiefs’ meetings, which are held every other ‘Broken Windows’ Law Enforcement Reduce Serious Crime?” California Institute for County Government month. The director said that by attending these Research Brief, August 2002. meetings, the department hopes to identify the effect local decisions may have on the State’s inmate population by more effectively communicating with local agencies. According to the acting deputy director for the Parole and Community Services Division, a department representative began attending the police chiefs’ meetings in April 2004. Second, the department is in the process of reinstating a law enforcement consortium in which it teams with local law 30 California State Auditor Report 2005-105 enforcement agencies to exchange information to solve common problems. The acting deputy director stated he expects these meetings to begin by September 2005. The Unexpected Increase of Inmates Created Security Concerns Since the actual inmate population was increasing when the department had projected a stable or declining population, it had to accommodate the increase through placing thousands of new beds in temporary locations that are harder for staff to secure. Concern for public safety was one of the main reasons the department cited in its justification for the no-bid one-year CCF contracts. Since the actual inmate population was increasing when the department had projected a stable or declining population, it had to accommodate the increase through placing thousands of new beds in temporary locations that are harder for staff to secure. The department’s concern for public safety appears reasonable, and its reliance on this information to assess the overcrowded situation and to decide to open two CCFs also appears reasonable. As previously described, total inmate population rose unexpectedly, so the department’s short-term projections, which usually are fairly accurate, fell short of the actual numbers. Therefore, relying on the population projection it prepared in spring 2004 that included the estimated effects of the parole reforms, the department was not prepared for the surge in actual population it experienced later that year. As shown in Figure 6 on the following page, the population projection it prepared in spring 2004 anticipated a decline in future inmate population from 161,000 in June 2003 to 157,000 in June 2005. However, the actual inmate population increased by 3,317 from January 2004 to September 2004. This increase alone exceeds the inmate design capacity of some of the department’s institutions. This influx brought the total inmate population to 165,102 inmates, higher than the fiscal year-end population numbers for any of the previous seven years. Twice a year, the Program Support Unit within the department releases an institution activation schedule (activation schedule) to coincide with the population projections. This is a schedule of new beds to be added and existing beds to be removed from service for the next two fiscal years, based on the population projection. If the department needs more or fewer beds than originally planned, the assistant deputy director of the Institutions Division releases an emergency memo detailing where to place additional or remove excess beds. When the department cannot house inmates in the beds located in areas normally designated for that purpose, it places temporary beds California State Auditor Report 2005-105 31 FIGURE 6 Actual Versus Forecast Inmate Population June 2002 Through June 2005 ������ ���������� ������ ���� �������� ��� ��������� ���� ��� ������ �� ������� ��� ���������� ��� ������� ���� ��� ��� ��� ��� ��� ��� ��� ������� ������� �� �� �� �� �� �� � �� �� �� �� �� �� �� �� �� �� ��� ������� ������������ ���� Sources: Department of Corrections’ monthly report of population and spring 2004 population projection. in overcrowded areas such as dayrooms or gymnasiums through the use of these emergency memos. When this happens, safety concerns are heightened because these areas are harder for staff to secure. From April 2003 to October 2004, the department placed 7,087 new beds into service using the activation schedule or through emergency memos. Figure 7 shows the cumulative change in capacity from January 2003 to December 2004. Some of this 32 California State Auditor Report 2005-105 FIGURE 7 Cumulative Change in Inmate Capacity From January 2003 to December 2004 ����� ���������� ������ �� ������ �������� ����� ����� ����� ����� ����� � ������� �� �� � �� � �� �� � �� �� �� �� �� �� �� �� �� � � �� � � �� � � �� � �� �� �� �� � �� � �� �� � �� �� �� �� �� �� �� �� �� � � �� � � �� � � �� � �� �� ������� ���� ���� ���������� Sources: Department of Corrections’ monthly institution activation schedules and related emergency revision memos. increase in capacity came as a result of the department having to house inmates in dayrooms and gymnasiums, since the department did not open or reopen any institutions in 2003 and 2004. The Cost Comparisons the Department Used to Justify the No-Bid Contracts Were Incomplete Although the information on which the department based its decision to open two CCFs using no-bid contracts appears reasonable, its justification for these contracts included incomplete cost comparisons. The department stated in its justification that the two contracts represented a potential cost savings to the State because the per diem rates for the facilities are less than the daily jail rate of $59, the maximum the department can reimburse counties for detaining certain California State Auditor Report 2005-105 33 state parolees who have violated parole and therefore are being sent back to prison. However, the two costs are not comparable. Because the CCF contract amounts, unlike the daily jail rate, do not include all the costs of housing an inmate, the department’s claim of cost savings is misleading. Compared to other CCF contracts, however, the average annual per-bed cost of the two no-bid contracts appears to be within a reasonable range. The daily jail rate was established in law in 1993 and represents the department’s average cost of housing inmates. According to the department, it has adopted the same benchmark for comparison purposes when considering contracts to house inmates in CCFs. Therefore, in calculating the savings included in the justifications for the two no-bid contracts, the department compared the daily per-inmate costs from the CCF contracts to the daily jail rate. According to state law, if the department is unable to accept new inmates due to overcrowding, it can house the inmates waiting to be transferred to prison in local jails until it can identify space within one of its institutions. Therefore, this daily jail rate is the maximum daily amount the department can legally pay local governments for housing state inmates. Costs of CCF Inmates Paid Directly by the Department • Health services provided for anything other than first aid. • Certain transportation costs. • Discipline costs when inmates are transferred back to a state institution. • Department peace officer staff employed at private CCFs. • Correctional counselors. • Administration of the inmate appeal system. Although the per diem amounts included in the no-bid contracts—$45 per day for the McFarland contract and $54 per day for the Mesa Verde contract—are less than the $59 maximum daily jail rate, the amounts are not comparable because contract amounts do not include all of the costs of housing an inmate; the daily jail rate does include all such costs. The rate at which the department reimburses local entities is intended to include all the costs of detaining a state prisoner. However, it incurs the additional costs shown in the text box for state inmates housed in CCFs that are beyond those reflected in the CCF contract amounts. The department did not include these costs in its comparison to the daily jail rate. • Administrative services the department performs at the state level that also apply to CCF inmates. According to the acting chief of the Community Correctional Facilities Administration, to process the justification for the no-bid CCF contracts, the department had to indicate how each contract’s price was determined to be fair and reasonable. To do this, it elected to use the daily jail rate for comparative purposes, but she acknowledges the actual daily rates for the CCF contracts do not include all service and operation costs. 34 California State Auditor Report 2005-105 Nevertheless, the costs included in the two no-bid contracts still appear reasonable when compared to the other CCF contracts. As shown in the Table, we compared the average annual cost per bed for the two no-bid CCF contracts to the average annual cost per bed for the other CCF contracts existing in 2004. Although both no-bid contracts are more expensive than the average annual CCF contract cost of $15,278 per bed, they appear to be within a reasonable range of the other CCF contracts.2 TABLE Average Annual Cost Per Bed for No-Bid Contracts and Existing Community Correctional Facilities 2005 No-Bid CCF Contracts Facility Lassen CCF CCF Contracts in Place in 2004 $16,824 Mesa Verde CCF (Civigenics)* † McFarland CCF (The GEO Group, Inc.) $16,814 16,239 Leo Chesney Center CCF 16,238 Victor Valley MCCF‡ 16,165 Claremont Custody Center CCF 15,738 Baker CCF 15,562 Taft CCF 15,408 Shafter CCF 15,250 Delano CCF 14,705 Golden State MCCF‡ 14,523 Adelanto CCF 14,413 Central Valley MCCF‡ Desert View MCCF ‡ Average for 2004 CCF contracts 14,401 14,112 $15,278 CCF = Community Correctional Facility. * This amount does not include $1 million in start-up costs to furnish and equip the CCF that the department added to the contract. If the $1 million is considered, the average annual cost per bed for this contract would be $22,696, which is significantly above the range of current CCF contracts. † This contract was not executed. ‡ MCCF refers to a modified community correctional facility. 2 We computed the annual average cost for the Mesa Verde CCF contract excluding the $1 million start-up cost to furnish and equip the facility for operation. If the $1 million is included in computing the average, the amount, $22,696, is considerably above the range of current CCF contracts. California State Auditor Report 2005-105 35 RECOMMENDATIONS To strengthen controls over its processing of no-bid contracts, the department should do the following: • Wait until all proper authorities have approved the no-bid contract justification request before sending a contract to a contractor for signature or signing the contract itself. • Require key contractor staff to complete statements of economic interests. • Include all its costs when it decides to include cost comparisons in justification requests or state that the cost comparison is incomplete. To ensure that its staff is free from potential conflicts of interest, the department should do the following: • Ensure that its retired annuitants in designated positions submit required statements of economic interests. • Ensure that statements of economic interests submitted by staff are complete. • Consider contracting with retired staff to provide expert testimony in litigation instead of its current practice of hiring them as retired annuitants. • When appointing retired annuitants, limit such appointments to a one-year period and require annual reappointment. n 36 California State Auditor Report 2005-105 CHAPTER 2 The Department’s Inmate Projections Are Reasonably Accurate for Short-Term Purposes, but Their Value for Long-Range Planning Is Limited CHAPTER SUMMARY A lthough the inmate population projection of the Department of Corrections (department) has an average error rate of less than 5 percent for the first two years of its six-year projection, by the last year the average error rate climbs to almost 30 percent. As a result, the projection is useful for assessing the next two years’ budget needs but has limited usefulness for longer-range planning, such as the need to build new prisons. In addition, the department does not update inputs to its projection model with historical data using a statistical process but rather adjusts the variables by relying on staff experience. Also, because it does not update information related to prisoners’ security classifications using the most recent data, the department is projecting security needs based on an obsolete classification system that may differ significantly from the current inmate population. Furthermore, our statistical expert advised us the department’s process for making inmate population projections is not based on a statistically valid method for creating a forecast. Moreover, the department’s Population Projections Unit (projections unit) did not adhere to its policy precluding the use of estimates for which it does not have reasonable support, such as the estimated effect of proposed legislation. Such actions reduce the accuracy of a projection and make it difficult for the unit that produces the projections to defend its independence. Finally, until recently, the department did not effectively communicate with local governments, hindering its ability to anticipate factors that could have an effect on inmate populations. California State Auditor Report 2005-105 37 THE DEPARTMENT USES PROJECTIONS IN BUDGETING AND JUDGING WHETHER AVAILABLE BED SPACE IS SUFFICIENT Each fall, the projections unit projects the number of inmates it expects the department to house over the next six years and revises the projection the following spring.3 It uses a microsimulation model to trace the path of individual offenders to project inmate populations on a month-to-month basis. This movement of inmates through the projection model is depicted in Figure 8. FIGURE 8 Inmate Projection Model ������ ����� ������� ������ ������ �������� ������ ���� Source: Credible Projections Require More Than a Good Model, Jack Leonard and Scott Sweet, Population Projections Unit, Department of Corrections, presented at a prison population forecasting workshop, Washington, DC, December 10–11, 1997. The projections unit primarily uses two types of variables in developing its projection: one relating to the amount of time inmates stay in a particular stage of the correctional system and the other relating to where in the correctional system the inmates will move next. The first type of variable predicts, as a percentage, the probability that inmates will move to another stage of the correctional system in each of the months after entering the current stage. The other type of variable represents the percentage probability of different stages inmates can enter when they do move. These percentages are based on data from prior experience. When applied to the current inmate population and adding the projection for newly admitted inmates, the model simulates inmates’ projected movement between each stage of the corrections process over a six-year time span, up to when they exit the system, either through release from parole or death. By combining these movements between stages, the department creates a projection of expected population in each stage for each month for the following six years. 3 38 The department’s projection includes inmates, parolees, inpatients, and outpatients under its control. For the purposes of this report, we only discuss the projection as it relates to inmates. California State Auditor Report 2005-105 The projections serve two primary purposes. First, the department uses them in developing its budget for institutional and parole operations. Projecting the expected number of inmates is important in determining the budget for items such as correctional officers, food costs, and transportation. Second, the department uses the projection to assess the sufficiency of available bed space at its institutions to house the projected inmate population. Its Facilities Management Division (facilities division) compares the population projection to the number of inmate beds the department has or will have in the periods covered by the projection. This comparison allows the department to assess whether it needs additional beds or prisons in the future to meet its security requirements or whether it has excess capacity. WITH HIGH ERROR RATES, THE DEPARTMENT’S LONGER-TERM PROJECTIONS DO NOT ACCURATELY PREDICT ITS NEED FOR INMATE HOUSING While the average error rate of the department’s projections is less than 5 percent in the first two years, it increases rapidly beginning in the third year, reaching almost 30 percent by the end of the sixth year. The department’s process for projecting future inmate population is reasonably accurate for short-term purposes, such as for developing budgets, but is much less accurate for longer-term purposes, such as assessing the housing needs of future inmate populations. To determine the accuracy of the projections, we compared the actual inmate populations at June 30 for each year that was available as of June 2005 to the levels the department projected for each year of each projection between spring 1995 and spring 2005. We then determined the average error rate of each year for the projections it produced during this period. As shown in Figure 9 on the following page, although the projections are reasonably accurate for the first two-and-ahalf years of a projection period, they become increasingly less accurate beyond that point, quickly rising to average error rates that render them useless for their intended purpose.4 In developing its budgets, the department primarily relies on information from the first two years of a projection, which reflects the period for which the department is preparing a budget. As shown in Figure 9, the average error rate of the projection process in the first two years is less than 5 percent and therefore appears reasonable for this purpose. However, because of the time needed to build a new prison, the department also uses projections to assess the sufficiency 4 Because the time period of some projections had not yet lapsed as of June 2005, the number of projections used in computing the average error rates appearing in Figure 9 range from 11 for the first half-year to five for the sixth year of the projection. California State Auditor Report 2005-105 39 of its facilities to house future inmate populations. For this assessment the department uses all six years of the projection period. As Figure 9 shows, the average error rate increases rapidly beginning in the third year, reaching almost 30 percent by the end of the sixth year. Therefore, the department’s reliance on its projections in assessing the sufficiency of its facilities and planning future prison construction appears misplaced. FIGURE 9 Average Error Rate of Projections Completed Between Spring 1995 and Spring 2005 at Six-Month Intervals After Projection ��� �� �� �� ������� ����� ���� �� �� �� �� �� �� � � � � � ��� � � � � � ��� � ��� � ��� � ��� � ��� � ������ �� ����� ����� ���������� Source: Department of Corrections’ population projections. Note: Because the time period of some projections had not yet lapsed as of June 2005, the number of projections used in computing the average error rates range from 11 for the first half-year to five for the sixth year of the projection. According to the chief of the Offender Information Services Branch, the average error rate of the projections would be much smaller if projections completed prior to fall 1996 were excluded. The chief stated that the projections unit made a significant adjustment to the projection model in fall 1996 related to the State’s three strikes law. However, even if we exclude projections completed prior to fall 1996, our concerns remain. In fact, excluding these projections creates an average error rate that ranged from 1 percent after the first six months 40 California State Auditor Report 2005-105 to 25 percent by the projection’s sixth year. These results are not appreciably different than the original range of the average error rates shown in Figure 9. Also, the projections tend to overestimate future populations, which could result in the department building facilities that are not needed if it relies solely on the projections. The manager of the projections unit stated that she was not aware of the projections’ tendency to overstate inmate populations. She stated that over the last 10 years, the projections unit understated as much as it overstated the inmate population in the first two years of the projection cycle. However, the manager’s focus on the first two years of the projection cycle seems misplaced. The facilities division informed us that it takes at least three to five years to build a prison, so it is concerned with years three through six of the projection cycles. As shown in Figure 10 on the following page, most of the projections for periods beyond year three prepared between spring 1995 and spring 2000 projected populations that exceeded the actual number of inmates. For example, the department’s fall 1995 projection forecast that the inmate population would exceed 232,000 in 2001. However, the actual population in 2001 was 161,000, a difference of 71,000 inmates. The deputy director of the facilities division told us that it is department policy to request new or expanded prisons to house populations based on these projections. He further stated that the department requested funding for constructing new prisons in 1995 and again in 1996, but the Legislature did not approve these requests. Had the department actually built the prisons needed to house the number of inmates shown in its fall 1995 projection, it would have vastly overbuilt for its actual needs. In fact, our analysis shows that the department would be more accurate in its long-term planning if it simply used the actual inmate population at the time it created each projection and assumed the population would not change over the six-year projection period. As presented in Figure 11 on page 43, beyond the third year of a projection’s time period, the average error rate of assuming no change in the inmate population is less than that of the department’s projection. California State Auditor Report 2005-105 41 ���� ���� ���� Source: Department of Corrections’ population projections. � �� ��� ��� ��� ��� ���� ���� ���� ���� ���� ���� ���� ���� ������ ������ ���� ���� ���� ������ ���� ���� ���� ������ ���� ���� ���� ������ ���� ���� ���� ������ ���� ���� ���� ������ ���� ���������� Comparison of the Final Three Years of Each Projection to Actual for Spring 1995 Through Spring 2000 Projections FIGURE 10 ������ �� ������� ��� ���������� 42 California State Auditor Report 2005-105 FIGURE 11 Average Error Rate Over the Projection Period ����� ���� ����� ��������� ���������� ��� �� ����� ���� �������� �� ���������� ������ �� �� �� ������� ����� ���� �� �� �� �� �� �� � � � � � � � � � � � � � � � � � � � � ��� � ��� � ��� � ��� � ��� � ��� � ������ �� ����� ����� ���������� Source: Department of Corrections’ population projections. THE DEPARTMENT NEEDS TO REDUCE THE SUBJECTIVITY IN ITS PROJECTIONS The department includes over 50 variables in its model for developing projections of inmate populations, such as the amount of time left on an inmate’s sentence and the severity of the offense. Each variable has some effect on the projection. However, for over half of the variables, rather than using a statistical process based on historical data to update each variable, the department makes adjustments based on staff experience. Also, in developing some of its projections, the department included variables for which it did not have trend—or historical—information and therefore had no solid basis on which to determine their effect on the projection. As a result of these actions, the department has increased the level of subjectivity in the projection, making the result less reliable. California State Auditor Report 2005-105 43 Finally, until recently, the department did not effectively communicate with counties, hindering its ability to anticipate changes that may affect its prison population. The Department Does Not Appropriately Update Its Projection Data According to our statistical expert, the department cannot support its forecasts using its present methodology. The department’s projection model uses data from prior experiences to establish the likelihood of certain events occurring at steps along the projection process. For example, at a given point in the simulation model, an inmate hypothetically may have a 40 percent chance of being released on parole, a 50 percent chance of remaining in prison for at least another month, and a 10 percent chance of dying in prison. However, the department does not always appropriately update the frequencies—or relative percentages of the likelihood of different options occurring—using sufficient historical data. Rather than using a statistical process to develop the frequencies, the department takes the same frequencies used in its previous projection and then updates the numbers based on analysts’ experience and review of the actual data since the last projection. This method increases the possibility of bias entering into the projection. According to our statistical expert, the department cannot support its forecasts using its present methodology. Also, its practice of using amounts from a prior projection may have resulted in the department projecting inmate security classifications in error. It uses projected security levels of future inmate populations to ensure it has sufficient housing capabilities required for each security level. Although analysts have made adjustments based on their subjective judgment, the variables related to inmates’ security classifications have not been updated with actual numbers since October 2002. The manager of the projections unit stated that in 2002 the department implemented a new classification system and the projections unit has not yet received the information necessary to forecast the different security classification levels from the Classification Services Unit (classification unit). The manager further stated that the new security classification system has not yet generated enough data to determine a trend but suggests that there will be enough information by 2006. However, data we obtained from the classification unit shows that by relying on 2002 data, the projections unit may have underestimated the future need for security level 2 beds. 44 California State Auditor Report 2005-105 According to staff who maintain the new classification system, the projections unit has not requested, and therefore the classification unit has not created, the report showing the number of inmates by security level classification. The classification unit did provide us with information related to inmates’ security scores, which the department uses to determine the inmates’ security level classifications. However, the unit did not provide us with information regarding how many inmates are reclassified into a higher security classification than their initial score would indicate because of safety concerns or other special circumstances, which could potentially change the population in each classification. The acting deputy director of the Risk Management Division stated that it would be difficult for the department to generate the information concerning such reclassifications in time to use for our audit because of the amount of programming and data verification it would need to complete to develop this information. The data the department did provide to us on security classifications without the effect of inmate reclassifications is presented in Figure 12 on the following page, which shows that during the new classification system’s implementation period in 2003, the number of inmates with security scores of level 1 and level 2 changed significantly. Because it used outdated information, the projections unit may be forecasting significantly different security requirements for the inmate population than are truly needed. From June 2004 to June 2005, the respective populations settled into a more stable trend. Because it used outdated information from 2002, the department’s most recent projection forecasts a security level 2 population of 32,000 in June 2005, which is significantly less than the actual level 2 population of 46,000 reported by the new classification system, excluding the potential effect of any reclassifications. As a result, when the department uses the projection for infrastructure planning and staffing, it may be underestimating the number of security level 2 beds needed to house future inmates. By failing to use the information from the new classification system, the projections unit may be forecasting significantly different security requirements for the inmate population than are truly needed. We believe the department already has enough data to establish a reliable trend. According to our statistical expert, because the new classification system has caused a significant change in the populations of the various security classifications, the department should begin using actual data collected under the new system. Our expert also stated that the monthly information obtained during the last year and a half would be able to more accurately project the current security requirements California State Auditor Report 2005-105 45 FIGURE 12 Actual Prisoner Population by Security Classification Level ����� � ����� � ����� � ����� � ��� ��� ������ ������ �� ������� ��� ���������� �� �� �� �� �� �� � ������� ���� ���� ���� �������� ���� ���� ���� �������� ���� ���� ���� Source: Department of Corrections’ Classification Services Unit. Note: Data does not include the effect of any inmate reclassifications, which the department was unable to provide during our audit. of the population than the 2002 data based on the old classification system, providing the department enough data points to create a more accurate forecast. Contrary to Its Policy, the Projections Unit Used Speculative Estimates in Its Projections At the direction of the department and contrary to its own policy, the projections unit used estimates in its projections that are not based on past experience or that include information from programs whose effects could not be reasonably estimated in several instances. Specifically, in the 2004 spring and fall projections, the department’s former chief deputy director of support services directed the projections unit to include the estimated effects of various parole reforms. According to 46 California State Auditor Report 2005-105 the manager of the projections unit, these estimates were based on changing criteria, and the parole reforms in question had numerous issues that needed to be resolved before any reasonable expectation of population reductions could be estimated. From our review of department policy memos, we noted that criteria such as which inmates were eligible for these programs and the maximum amount of time inmates could be enrolled changed during the time period in which these projections were being made. Nonetheless, department management required the projections unit to include the estimates in its population projections, thus compromising the unit’s independence. Without being able to function independently of internal or external pressure to use certain data or arrive at certain conclusions, the credibility of the projections unit’s forecasts is diminished. According to the manager of the projections unit, including these estimates in the projection led to a major variance between actual and projected populations. In fact, our analysis confirmed that the spring 2004 projection was 2 percent off after six months and 4 percent off after a year and a half, resulting in one of the department’s least accurate projections in the last 10 years. Because these estimates are a part of the projection and are not identified as separate from a forecast based on trends in the data, our statistical expert concluded that the validity and independence of the entire projection is called into question and the value of the projection for decision-making purposes is reduced. The manager of the projections unit objected to including estimates in the projection, stating it compromises the integrity of the projections process. In fact, the manager of the projections unit, when required to include these estimates in the projection, made the following statements in a letter to her supervisor noting her concerns: I strongly object to this change because I believe it compromises the integrity of the projections process and the credibility of the projections produced by my unit. The projections have become a topic of political debate. We were told that the fall 2004 projections could not be released to the Governor’s Office because the numbers were too high and that we must include estimates of the [parole] reforms in the fall 2004 projections publication, even though they were not a part of the formal projections process. The projections unit stated that a projection process must isolate policymakers from the actual creation of the projection and that the projections unit must never include speculations about the California State Auditor Report 2005-105 47 effects of proposed legislation or policy changes. Furthermore, the unit has stated that no proposed legislative or policy change is included in the projections until the changes have been formalized and a date of implementation has been set. However, by insisting that the projections include the estimated effects of the parole reforms, the department caused the projections unit to violate these principles. The Department Failed to Obtain Information From Counties That Would Have Alerted It to Rising Admissions The department is developing ways to establish better communications with the counties. As discussed in Chapter 1, in addition to the unrealized effects of parole reforms, the spring 2004 population projection was also understated because of an unexpected rise in inmate admissions from counties. Because county superior courts sentence felons to state prison, changes in county policies on prosecuting criminals can affect inmate admissions at the state level. Los Angeles County was the primary source of the rising inmate admission rate during this period. According to the department’s director, the new chief of police of the city of Los Angeles changed the city’s approach to policing, increasing the number of people being sent to prison. However, until recently, the department did not have an effective process in place to communicate with local governments to identify such changes and their effect on the number of inmates being sentenced to prison. As further described in Chapter 1, the department is developing ways to establish better communications with the counties. The department may not be able to use the information it receives from these efforts in its projections because sufficient historical data may not be available to allow the projections unit to adequately assess the effects of the identified issues. Nonetheless, the department will be able to use the information in its decision-making process and therefore will be able to better anticipate future increases and decreases in inmate population that result from changes made at the local government level. LACK OF DOCUMENTATION CASTS DOUBT ON THE VALIDITY OF THE PROJECTION PROCESS To assess the statistical validity of its projection process, our statistical expert met with key department staff to review the documentation of the projection method. However, the department does not have documentation describing its complete projection model, so we were unable to assess its 48 California State Auditor Report 2005-105 validity. According to our statistical expert, documenting a projection process, including the computer program used, is important so others can evaluate the process and understand its limitations and capabilities. She added that, for staff within the department, such documentation is very valuable for the continuity of the forecasting process when current staff retire or leave. She concluded that data analysis is a constantly evolving process and appropriate documentation is crucial in all stages to continuously improve the analysis as more and more data become available. According to the chief of the branch that includes the projections unit, it is currently revising the projection model and plans to produce documentation for the revised version. Nonetheless, key department staff described the process to our expert to provide an understanding of the methodologies the projections unit uses. Based on these discussions, our statistical expert provided the following observations: Our statistical expert concluded that the department’s projection process, as described, is not based on a valid statistical method. Because there is no documentation to describe the projection process, we are unable to evaluate the model itself and are evaluating the process as described to us. The simulation process described to us by [department] staff is not based on a valid statistical method that is available in the statistical literature. The description of the simulation process also did not include appropriate theoretical justification. The underlying assumptions of the model, the accompanying theory, and the validity of the simulation process was not established either through published research papers or statistical journals. Both the State Administrative Manual (administrative manual) and information technology (IT) industry standards require the documentation of IT systems. The administrative manual specifically requires that documentation meet the needs of nontechnical users, technical users, outside auditors, and any agency measurement necessary. Additionally, the adequacy of documentation is a required factor when purchasing any IT components and a required element in any project plan. Without documentation, the department cannot ensure the proper use of the applications and technological solutions it has put in place. California State Auditor Report 2005-105 49 Therefore, because the department lacks documentation for its projection model, and because the model as described to our expert is not based on established statistical methods, its validity is uncertain. Nonetheless, our expert provided the following insights related to the department’s two primary uses of the projection: Although not the result of a standard statistical model, the results of the projection evaluated alone are reasonable in the short term, because the projected population does not increase by more than 10 percent per year and includes a significant amount of detail on subpopulations, which the department needs for budgeting purposes. However, as the model is used more, the accuracy of the projection should increase. The department should not provide a long-term forecast if it is unable to achieve a reasonable level of accuracy. To address the validity of the department’s projection model, our statistical expert stated the following: The department needs to bring a group of statisticians together to get advice on establishing a statistically valid forecasting methodology with the current data system. A time series specialist, a non-parametric statistician, and a finite population sampling theorist along with a statistical computing expert will be able to design a suitable forecasting tool for the department’s needs. There are national experts available in all these areas of statistics within California’s higher education system. Both the California State University and the UC campuses have researchers specializing in these areas right here in Northern California. One of the benefits for the department is that this group of statisticians will be able to provide the state-of-the-art statistical tools needed for the department’s short-term and long-term forecasts. RECOMMENDATIONS To increase the accuracy and reliability of its inmate projection, at a minimum the department should do the following: • Fully document its projection methodology and model. 50 California State Auditor Report 2005-105 • Update its variable projections with actual information, such as the new security level data, whenever feasible to do so. • Continue its recent efforts to enhance its communications with local government agencies to better identify changes that may materially affect prison populations. • Disclose when a projection includes estimates for which inadequate historical trend data exists, such as the estimated effects of a new policy, and the specific effect such estimates have on the projection. Additionally, if the department intends to continue using the projections for long-term decision making, such as facility planning, it should ensure that it employs statistically valid forecasting methods. It should consider seeking the advice of experts in selecting and establishing the forecasting methods that will suit its needs. 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 this report. Respectfully submitted, ELAINE M. HOWLE State Auditor Date: September 13, 2005 Staff: Doug Cordiner, CGFM, Audit Principal David E. Biggs, CPA Renee Davenport Jonnathon Kline Heather McIntier California State Auditor Report 2005-105 51 Blank page inserted for reproduction purposes only. 52 California State Auditor Report 2005-105 Agency’s comments provided as text only. Department of Corrections and Rehabilitation Risk Management Division P.O. Box 942883 Sacramento, CA 94283-0001 August 29, 2005 Elaine M. Howle* State Auditor Bureau of State Audits 555 Capitol Mall, Suite 300 Sacramento, CA 95814 Dear Ms. Howle: The Department of Corrections and Rehabilitation (CDCR) has reviewed your draft audit report entitled “California Department of Corrections: It Needs to Better Ensure Against Conflicts of Interest and to Improve Its Inmate Population Projections.” We appreciate the opportunity to respond to the draft report. Enclosed is the CDCR’s response to the report’s recommendations. We appreciate the attention to accuracy and detail that your staff put into the process. Your staff has been professional and at all times available to discuss the issues. Please extend our appreciation to those who participated in this review. As indicated in the enclosed response, the CDCR is committed to making further improvements by addressing the issues presented in the report. If you have any questions concerning the response, please contact me at 323-6001 Continued Success, (Signed by: Roderick Q. Hickman) RODERICK Q. HICKMAN Secretary California Department of Corrections and Rehabilitation Enclosure * California State Auditor’s comments appear on page 57. California State Auditor Report 2005-105 53 RESPONSE TO BUREAU OF STATE AUDITS (BSA) DRAFT REPORT California Department of Corrections: “It needs to Better Ensure Against Conflicts of Interest and to Improve Its Inmate Population Projections.” Chapter 1 BSA Recommendation #1: Wait until all proper authorities have approved the contract justification request (a.k.a. NCB) before signing the contract or sending it to a contractor for signature. 1 CDCR’s Response: CDCR’s general practice is to wait until all proper authorities have approved the contract justification request before signing the contract or sending it to a contractor for signature. However, when timing is critical, obtaining the contractor’s signature in advance helps to expedite the process but does not, in any way, execute the contract. CDCR’s Instructions to the Contractor clearly state that the Contract is of no force and effect until they receive a fully approved original signature copy of the Contract for their files. A contract is not valid or executable until approved by DGS’ Office of Legal Services (OLS). BSA Recommendation #2: Require key contract staff to complete statements of economic interests. CDCR takes under advisement BSA’s recommendation to have key contract staff complete the statement of economic interests and will immediately initiate a review of the recommendation for feasibility of implementation. It is anticipated the review will be completed within 60 days. BSA Recommendation #3: Include all its costs when it decides to include cost comparisons in justification requests or state that the cost comparison is incomplete. CDCR concurs with the recommendation. BSA Recommendation #4: Ensure that its retired annuitants in designated positions submit required statements of economic interests. CDCR’s Response: The CDCR agrees to adopt BSA’s recommendation to enter expiration dates on the appointments of retired annuitants. This will allow each retired annuitant (RA) appointment to be reexamined each year as to the applicability of conflict of interest (COI) compliance. The CDCR will also strive to educate and train the hiring authorities as well as the personnel analysts with the COI criteria. This will enable staff to question RA paperwork and forward questionable RA duty statements to the COI filing officer for review. 1 54 California State Auditor Report 2005-105 BSA Recommendation #5: Ensure that statements of economic interests submitted by staff are complete. CDCR’s Response: The CDCR filing officer will comply with the Regulation 18115 and ensure that each cover page is complete. In addition, the Department filing officer will review 20 percent of all timely statements (one half of which will be selected at random). The CDCR is in the process of adding COI information and filing site to the CDCR website. This site will assist staff and management in the filing of the COI forms and will provide listings of the classifications required to file as well as easy to follow instructions on correct completion of the cover page. Since this site is easily accessible to staff with internet or intranet access, it will make information regarding the COI form more convenient and may reduce the incomplete submission of the COI form. BSA Recommendation #6: Consider contracting with retired staff to provide expert testimony in litigation instead of its current practice of hiring them as retired annuitants. CDCR’s Response: The CDCR will review its current procedures for expert witnesses and evaluate the feasibility with legal staff regarding hiring retired staff as consultants rather than RAs. This may not be a feasible option in some cases, as the high level executives that are required to be expert witnesses possess the needed high level expertise to assist the CDCR in developing operational policies and procedures. If these employees are under contract, they may be precluded from working as RAs under Penal Code Section 10410 (Prohibitions as to state officers and employees). BSA Recommendation #7: When appointing retired annuitants, limit such appointments to a one-year period and require annual reappointment. CDCR’s Response, The CDCR will adopt BSA’s recommendation to start entering expiration dates on all RA hires. 2 California State Auditor Report 2005-105 55 Chapter 2 Responses BSA Recommendation #1: Fully document its projection methodology and model. CDRC’s Response: The Department agrees with this recommendation. The Department is in the process of revising the existing simulation model and writing documentation for the revised version. BSA Recommendation #2: Update its variable projections with actual information, such as the new security level data whenever feasible to do so. CDRC’s Response: The Department agrees with the recommendation to use the new security level data and intends to do so as soon as it is feasible. BSA Recommendation #3: Continue its recent efforts to enhance its communications with local government agencies to better identify changes that may materially affect prison population. CDRC’s Response: The Department concurs. BSA Recommendation #4: Disclose when a projection includes estimates for which inadequate historical trend data exists, such as the estimated effects of a new policy, and the specific effect such estimates have on the projection. CDRC’s Response: The Department agrees with this recommendation. BSA Recommendation #5: Additionally, if the Department intends to continue to use the projections for long-term decision-making, such as facility planning, it should ensure that it employs statistically valid forecasting methods. It should consider seeking the advice of experts in selecting and establishing the forecasting methods that will suit its needs. CDRC’s Response: The Department welcomes the advice of such statistical experts in an effort to improve its forecasting methods. The Department will be initiating an Interagency Agreement for an independent review of the forecasting methods. It is anticipated that the review can be completed within 12 months. 3 56 California State Auditor Report 2005-105 2 COMMENTS California State Auditor’s Comments on the Response From the California Department of Corrections and Rehabilitation T o provide clarity and perspective, we are commenting on the California Department of Corrections and Rehabilitation’s (department) response to our audit. The numbers below correspond to the numbers we have placed in its response. 1 We do not dispute what the department has stated. Nevertheless, as we cite on page 1 of our report, perhaps as a result of beginning to process the Mesa Verde contract before receiving approval from the Department of General Services, the department incurred needless costs relating to a contract it ultimately rescinded. 2 The department may have misunderstood our recommendation. As we state on page 45, the department already has sufficient actual security level data to establish a reliable trend. California State Auditor Report 2005-105 57 cc: 58 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 California State Auditor Report 2005-105