Analysis of the 2008-09 Budget Bill: Criminal Justice

Capital Outlay (5225)

Implementing AB 900’s Infill Bed Plan: Progress and Concerns

The California Department of Corrections and Rehabilitation has made some progress, but encountered some obstacles, in its early efforts to implement a $7.7 billion package of prison construction projects. In this analysis, we provide an update on the implementation of the 16,000–bed “infill” component of the plan to address overcrowding in prison facilities. We highlight the significant changes that are already being made in this component of the prison construction package, and discuss their implications for meeting legislative goals for expanding inmate rehabilitation programs, complying with court requirements for improved health facilities, and dramatically increasing state costs.

On May 3, 2007, after extensive negotiations with the Legislature, the Governor signed into law Chapter 7, Statutes of 2007 (AB 900, Solorio), now generally referred to as “AB 900.” Among other changes, this legislation authorized the construction of up to 16,000 so–called infill beds to replace temporary housing for state prison inmates in gyms, classrooms, and other public spaces. In this analysis, we provide background on the key components of AB 900, update the Legislature on the implementation of AB 900’s infill bed plan, and discuss some key issues relating to AB 900 that we believe the Legislature may wish to address in the future.


Key Components of AB 900. Chapter 7 (hereafter referred to as AB 900) contained a broad package of prison construction and rehabilitation initiatives which were intended to relieve the significant overcrowding problems facing state prisons.

Some parts of the measure, which are not examined in detail in this analysis, relate to the expansion of drug treatment, academic education, and other rehabilitative programs for inmates and parolees. However, the measure also contained a number of significant provisions to finance the construction of both state prison facilities and county jail space primarily using lease–revenue bonds as well as a General Fund appropriation of $300 million. The key components of the $7.7 billion construction package are summarized in Figure 1 and include the following:


Figure 1

AB 900 Spending Plan

(Dollars in Millions)


Phase I


Phase II














































The lease–revenue bonds are available to fund two phases of projects. For example, 12,000 infill beds are provided in Phase I and another 4,000 beds in Phase II. Under the terms of AB 900, the second phase of funding is to be made available only after certain specified goals related mainly to the rehabilitation of offenders—such as the expansion of prison drug treatment services—have been met.

Governor’s Capital Outlay Budget Proposals. The 2008–09 budget plan reflects the continued implementation of AB 900 using the funds appropriated in the measure. However, the budget plan itself provides little detail as to the specific facilities to be built with these funds, and displays only general categories of expenditures relating to AB 900 in the spending plan for CDCR. As shown in Figure 2, the budget plan indicates that about $86 million will be spent to move forward with AB 900 projects (primarily infrastructure) in 2007–08 and that almost $600 million would be spent on infill bed, reentry, health, and infrastructure projects in 2008–09.


Figure 2

AB 900 Projects in the
Governor's Budget Plan

(In Millions)




Infill (Phase I)



Reentry (Phase I)



Medical/mental/dental (Phase I)



AB 900 infrastructure







More detailed information about what specific projects will actually be built, and in what time frames, has been presented by the administration in briefings and documents. Still further details are expected to be presented in a master plan for CDCR facilities. The master plan had not yet been released at the time this analysis was prepared.

The Governor’s budget plan proposes to shift $2.2 billion originally designated in AB 900 for Phase II of the infill and reentry projects to the Receiver’s various capital outlay projects. This funding for the Receiver would be in addition to the $857 million previously allocated in Phase I and the $286 million to be allocated in Phase II of AB 900 for health facilities, bringing the total funds available for these purposes to $3.4 billion. As a result of the shift of $2.2 billion to the Receiver, the $5 billion previously available for the construction of infill beds and reentry facilities would be reduced to $2.8 billion. We assess this proposal in more detail in the “Adult Correctional Health” analysis of CDCR.

Recent Developments Could Affect Size of Inmate Population. Two developments since the enactment of AB 900 could significantly affect the size of the state inmate population and, thus, state prison construction plans.

First, the Governor’s January 10 budget plan proposes the early release of certain nonviolent and nonserious offenders 20 months early and the granting of summary parole to other nonserious offenders, a policy change that would also significantly affect the number of parolees returned administratively to the prisons for parole violations. The administration estimates that the combined effect of the early release and summary parole proposals would be a reduction in the average daily prison population of 35,000 inmates by 2009–10.

Second, in July 2007, federal courts in the Plata and Coleman cases ruled together on motions filed on behalf of inmate plaintiffs to create a three–judge panel to decide if overcrowding in the prison system is delaying efforts to improve inmate medical services and mental health care. If the panel finds that to be the case, it could order the state to release prisoners, cap the inmate population, or both, significantly lowering the population—perhaps by tens of thousands of inmates. The trial in the case had been scheduled to start in February 2008, but has been postponed indefinitely while the state tries to work out a legal settlement.

How the AB 900 Projects Have Progressed

Our analysis indicates that CDCR has made some progress, but encountered some obstacles, in its early efforts to implement AB 900.

Hiring of Key Management Staff and Staff Reorganization. After the enactment of AB 900, the Governor convened a facilities construction “strike team” to assess the best way to implement the construction components of the new law. Leadership for the AB 900 planning and construction projects was given to the chair of the strike team. A number of key construction management staff have been hired and a major reorganization of the department’s capital outlay unit has occurred. This reorganization and hiring of key management staff was completed in fall 2007. At the time this analysis was prepared, CDCR had filled about 32 of 88 new positions created in the capital outlay unit in the 2007–08 budget to move forward with AB 900 implementation.

Development of New Infill Bed Plan. Following the enactment of AB 900, the administration began efforts to significantly revise its previous infill bed component, including both the proposed location of beds and the types of beds to be constructed. The revised plans are being integrated into an overall master plan for CDCR capital outlay which, we have been advised, is nearing completion.

The design of the first four infill projects is now in development, with the completion of scope, budget, and schedule packages expected early in 2008. The first four projects at the time of this analysis include 1,000 dormitory beds at Kern Valley State Prison (KVSP) near Delano; 950 celled beds at Northern Kern State Prison (NKSP) near Delano; 1,900 celled beds at Wasco State Prison (WSP); and 2,200 celled beds at California Correctional Institution (CCI) near Tehachapi. Three of these four prison projects (NKSP, WSP, and CCI) also include beds in reception centers—facilities, we discuss later in this analysis, designed for newly arriving inmates.

Lack of Access to Planning Funding Has Slowed Projects. Last year, following the enactment of AB 900, the Legislature enacted clean–up legislation to ensure that the Legislature had opportunities to review infill, infrastructure, and other AB 900 projects at specific stages of the development process. The measure, Chapter 175, Statutes of 2007 (SB 81, Committee on Budget and Fiscal Review), specifically required that the administration submit to the Legislature capital outlay planning packages for these projects, documents which include their intended design and the resulting scope, budget, and schedule of the project based on these designs.

Assembly Bill 900 funded studies, preliminary plans, working drawings, and construction costs for its projects. However, the measure did not fund capital outlay planning packages or any expenditures that are made prior to approval of these projects by the State Public Works Board. As a result, the delivery of capital outlay planning packages has been delayed and overall efforts to deliver projects have been slowed. The CDCR has notified the Legislature of its efforts to address this problem for infill bed projects through the temporary redirection of other funds—in particular, the monies appropriated in the AB 900 package for infrastructure. Capital outlay planning packages are an important source of information for meaningful legislative oversight of these projects. However, under the provisions of AB 900, it is not clear that it is appropriate for CDCR to use these funds to prepare capital outlay planning packages for infill bed projects. Notably, the Governor’s proposed trailer bill language makes capital outlay planning packages reimbursable from AB 900 funding.

Infill Bed Plan Has Changed Significantly

The major components of the infill bed plan have changed considerably since the original infill bed plan was prepared by CDCR almost a year before the passage of AB 900. While the master plan outlining final construction plans has not been released, preliminary documents and briefings provided by CDCR to legislative staff make it clear that major changes to the infill bed plan are in the works. The major changes we have identified are discussed below.

More Cells, Fewer Dormitories. Overall, the revised infill bed plan appears to be shifting toward building more cells and fewer dormitory beds, which can only be used to house lower–security inmates. The original infill bed plan proposed building 62 percent of all infill beds as dormitory beds. The infill bed plan presented to legislative staff in the fall had only 11 percent of all currently planned infill beds as dormitories. All the remaining beds in this version of the department’s plans would be cells, which can be used to house inmates at any classification level. The CDCR has indicated that this mix of dormitories and cells could change further as its plans continue to evolve. Notably, all of the currently planned dormitory beds are included at the very first infill bed project, KVSP.

Fewer Sites With More Beds Outside of Inmate Areas. The original infill bed plan proposed locating infill bed projects at 25 different prison facilities. The size of individual infill bed projects would have ranged from 150 beds at California Substance Abuse Treatment Facility and State Prison, Corcoran to 1,940 beds at NKSP near Delano. The majority of infill beds were to be constructed inside of inmate–occupied areas.

The new infill bed plan will locate infill beds at only ten sites, each generally averaging more beds than under the previous plan. The infill projects at NKSP and six sites (yet to be determined) would have 950 beds, while CCI would have 2,200, the most beds of any infill bed site. Most of the construction of the new infill bed plan will be outside of inmate occupied areas—for example, vacant land on the outskirts of prison grounds.

Changes in Reception Center Facilities. When a prisoner enters prison, he initially is sent to a particular type of prison facility known as a “reception center.” Reception centers provide security, health care, and educational assessments of prisoners so that they can be assigned to the appropriate housing security classification. The length of time that new prisoners are in a reception center is generally brief, ranging from a few weeks to a few months, before they are sent from there to other general population institutions to serve the duration of their sentences.

The new infill bed plan would include three reception center facilities among the first four construction sites, representing approximately 24 percent of currently planned infill bed construction. (The number of reception center beds in the original infill bed plan was not specified, but our analysis suggests it would have resulted in roughly the same share of beds for this purpose.) As is the case for existing reception centers, the new reception centers would have more space for intake and inmate assessments than regular prisons, because of their unique role. But they would have no programming or academic classroom space because of the short amount of time prisoners are expected to stay at a reception center facility before being moved to other types of prison facilities.

Significant Increase in Space for Health Care and Academic Education. The new infill bed plans would greatly increase the space that would be available to deliver medical, dental, and mental health care. The square footage devoted in the new projects to health care may be more than seven times the amount of space constructed at the last state prison to open in 2005, KVSP near Delano.

The revised infill bed plan will provide some additional space for vocational programs, while space for academic education would triple. The size of academic classrooms would increase from 20 students, as constructed at KVSP, to accommodate classes of up to 27 students. Second, the total number of classrooms would be double those built at the existing KVSP. The original infill bed plan devoted much less space for these programs.

Implications of the Revised Infill Bed Plan

The changes to the infill bed plan discussed above have some significant ramifications. In some respects, they will provide increased flexibility to the state prison system to change with changing circumstances, such as the potential for significant reductions in the inmate population provided under the Governor’s budget plan or possible court orders. One element of the revised bed plan appears to be in line with the direction provided by the Legislature in AB 900 to provide more space for inmate programs, and appears likely to be consistent with the requirements of the federal courts in several cases for improved health facilities. All of these changes, however, could come at significant additional costs compared to those initially contemplated under AB 900. We discuss our specific findings below.

Overall Plan Provides More Flexibility… Our analysis suggests that the construction of more celled housing and less dormitory space in the new infill bed plan is generally a move in the right direction. This overall approach now being proposed would give the state more flexibility to respond to a sudden, major reduction of the inmate population, either due to the adoption of the Governor’s budget plan or due to a court–ordered population cap on the number of inmates. If a large reduction in the inmate population occurs, it is likely to disproportionately affect lower–security inmates, making the building of additional dormitory housing unnecessary at this time. If a large reduction does not occur, celled housing can be used for low– or high–security inmates with the proper staffing and correctional procedures, making it the most flexible housing option. While a celled housing facility costs somewhat more than a dormitory facility because it is significantly larger, it makes little sense to spend money on dormitory space that does not meet the prison system’s future needs.

…But May Not Be Flexible Enough. One important question is whether the revised infill bed plan goes far enough in the direction of building cells instead of dormitories. As noted earlier, the revised bed plan still contemplates construction of all proposed dormitory housing at the first of four projects now being scheduled. If the budget plan or court orders result in sharp reductions in the number of lower–security inmates held in the state prisons, the state could find itself having built the wrong kind of beds (1,000 such beds at the first prison project) at just the wrong time.

Similarly, the proposal in the revised infill bed plan to build additional reception center beds may also prove to be a mistake under these circumstances. Under the revised plan, all of the reception center beds would be built as part of the first four infill bed projects. However, the Governor’s summary parole proposal would, among other provisions, significantly reduce the number of parole violators returned to state prison each year. That means the state may need to build and staff fewer reception center beds than it now has, rather than expand these types of beds. Moreover, because the reception centers would be built without programming space, they could not easily be converted to other uses for the general inmate population.

The dormitory and reception center beds contained in the revised infill bed plan may ultimately be needed. Future court actions and policy changes affecting inmate population levels are by their nature not known. But these uncertainties mean it may not make sense for the state to build these types of infill bed facilities first. It may make more sense to build celled space rather than dormitory or reception center space in the initial four infill bed projects.

Increase in Academic Space Appears Warranted, Health Facility Need Less Clear. As noted above, the revised infill bed plans significantly increase the square footage allotted to education purposes. Our analysis suggests this approach is warranted and consistent with the provisions of AB 900. This added academic space will increase facility cost. However, based on the data we have reviewed, the cost of academic space would be relatively low, on a square footage basis, compared to the costs of other types of space in infill bed plan facilities. More importantly, our analysis suggests this increase in academic space is consistent with AB 900’s programming and rehabilitation goals.

We would acknowledge that, given ongoing litigation over medical care, mental health care, and dental care, steps should be taken to ensure that adequate facilities for these purposes are provided in the newly built prison facilities. However, the potential sevenfold increase in space for health facilities in the new infill bed projects, when compared with the space provided for such purposes in past prison construction projects, has not yet been justified, in our view.

Consolidation of Projects Would Reduce State Costs. This proposal to consolidate infill beds in fewer but larger projects outside of inmate–occupied areas is a reasonable one that is likely to result in major cost savings for these projects. Fewer sites result in fewer infrastructure projects. Also, fewer construction management teams would need to be hired to coordinate the infill projects, resulting again in significant cost savings.

The move of projects outside of inmate–occupied areas could reduce construction costs by more than 15 percent. Construction within such areas reduces productivity, as construction workers must spend as much as one hour at the beginning and one hour at the end of each eight–hour day going through security processes, adding greatly to the labor costs that ordinarily make up 60 percent of the cost of construction.

Construction Cost Estimates Increasing Dramatically. At the time this analysis was prepared, CDCR had not finalized its estimates of the costs of the first four infill bed projects it intends to pursue. However, our review of cost models and other working documents that the department and its consultants are using suggests that the estimated costs of these new facilities have increased dramatically.

Based on CDCR data we have reviewed, the department is projecting that the cost of constructing beds for 950 inmates and one group of support buildings will be $210 million. In contrast, the cost of constructing an entire prison of 4,600 beds—including more costly types of beds—at KVSP a few years ago was a total of $377 million. The cost per bed at KVSP was $82,000, while the cost per bed in the new infill bed plan is $222,000. Our analysis suggests that these higher estimates for the infill beds cannot be explained by the increases in labor and material costs that have occurred since KVSP was built.

So–called “soft” costs and contingencies appear to be major additional factors driving up the department’s cost estimates for the projects. Soft costs is the term often used to refer to nonconstruction costs of projects, such as architectural and engineering fees, project management and construction management fees, and inspection fees. Contingencies are funds earmarked to cover uncertainties, such as cost increases that may occur over time or project cost estimates that are too low. The CDCR estimates of both soft costs and contingencies are linked by ratios to “hard” construction costs. Double the amount of hard costs, and soft costs and contingencies double too. These ratio–driven factors appear to be contributing significantly to the higher costs for infill beds that are now being estimated for the CDCR projects.

The escalation of these costs is prompting the CDCR to scale down its plans on the number of infill beds it intends to build. If these estimates prove correct, the state would ultimately build several thousand fewer infill beds with AB 900 funding than the 16,000 that were contemplated in the legislation. In the alternative, the administration might build more dormitory beds instead of cells to deliver more beds, but would then run the risk that these could prove to be the wrong kind of beds to build.

Analyst’s Recommendations

In light of these significant concerns related to the implementation of AB 900 infill bed plans, we recommend that the Legislature take the following actions:

Obtain Independent Construction Cost Estimates. In order to address concerns relating to rising cost estimates for AB 900 infill bed projects, the Legislature should direct CDCR to obtain during the spring an independent estimate of these construction costs from a private sector firm that has no involvement in these projects. We are advised that the Department of General Services (DGS), which oversees the construction of many large state construction projects, routinely obtains two private sector estimates for its more costly projects. The cost of getting a second opinion on these construction costs could amount to $175,000. However, these costs could be paid from existing lease–revenue bond funding. Moreover, we believe this is a worthwhile investment compared to the possibility of overspending hundreds of millions of dollars on the AB 900 infill projects.

Additionally, the Legislature should establish staff positions within the CDCR’s Facility Planning, Construction and Management program, who would be permanently assigned to provide effective and continuous monitoring and validation of all capital outlay cost estimates associated with the new infill bed plan. The DGS has a similar construction estimating group to verify the accuracy of the project estimates it reviews. The department currently lacks the ability to independently verify the accuracy of these estimates, even though they often involve hundreds of millions of dollars in state costs for a single project. We recommend two positions be established to provide these estimates. These senior estimating positions would cost about $300,000 annually, including salaries, benefits, and related operating expenses. These costs also would be reimbursable from AB 900 funds.

Revise the Infill Bed Plan Using New Cost Estimates. The Legislature should direct CDCR to prepare an alternative infill bed plan using the new cost estimates by May 1, 2008 that takes into account the possibility that the Governor’s plan for early release and summary parole of offenders, or some alternative approach, is adopted by the Legislature. If the average daily population of inmates in the prison system is reduced, as the Governor has proposed, by as many as 35,000 inmates, a new approach to implementation of AB 900 will be warranted.

Any revised plan should specifically justify the number of beds that would be built and address how the type of housing constructed—cells, dormitories, and reception centers—would be changed to reflect the new situation. Additionally, this revised plan should justify the increase in health care services space.

Recommend Approval of Language on Planning Packages. The Legislature should adopt the Governor’s proposed trailer bill language that permits capital outlay planning packages for infill, reentry, and medical bed projects to be funded using funding from AB 900. This change would reduce potential delays in the development of these projects that could eventually increase their cost, and ensure that the Legislature has complete information available for its review of these projects at key points in the facility planning process.

Condemned Inmate Complex at San Quentin Raises Key Questions

The Governor’s budget requests an additional $136 million in lease–revenue bond financing to complete the construction of a new Death Row facility, known as the Condemned Inmate Complex (CIC), at the San Quentin state prison. We recommend that the Legislature not approve additional funding to complete CIC unless the department can resolve questions about inmate population restrictions at the prison and the tripling of housing unit construction costs for the project.

Proposal. The budget requests an additional $136 million in lease–revenue bond financing to complete the construction of a new Death Row facility, known as the Condemned Inmate Complex (CIC), at California State Prison, San Quentin. These new costs are due primarily to significant increases in construction cost estimates that have occurred since the inception of the project. The CDCR now estimates that, when finally completed, the project would cost $356 million.

According to the Department of Finance, CDCR has spent about $20 million to date on the project from the $220 million in lease–revenue bond financing that was approved in 2003–04. Construction of the project was temporarily put on hold in 2006 when escalating costs caused the project to exceed available appropriations. A request for additional funding to complete the project was presented last year as part of CDCR’s budget plan, but was not approved by the Legislature and then later was temporarily withdrawn by the administration. Last year, the Legislature passed AB 1743 (Huffman), a measure that would have prohibited any money being spent on the CIC until the Bureau of State Audits evaluated alternative prison sites for the CIC. The Governor vetoed this legislation. The 2008–09 budget request is $20 million more than the department had requested last year to complete the CIC. The CIC would have 768 cells, providing capacity for 1,152 male inmates on Death Row.

Increasing Construction Costs Not Justified. As we noted last year, the department’s estimated costs for the project have escalated dramatically over time even as the scope of the project has been reduced from an original plan for 1,408 beds to the present plan for 1,152 beds for condemned inmates. The Legislature previously authorized the CIC project at the $220 million funding level in 2003–04 based on an original project budget prepared in late 2002.

The latest cost estimate for CIC of $356 million appears to be out of line with the costs incurred by the state the last time it built housing units with the same design at KVSP near Delano, a facility that was activated in 2005. Some additional costs for a project of this type since KVSP was built would be expected due to increases in labor and materials costs. However, the costs now being assumed for housing units at the San Quentin project far exceed what would be expected. Our analysis considered the original cost of constructing housing units at KVSP which are generally similar to those at San Quentin. We then adjusted those costs based on a construction index that measures both the inflation and market–driven components of these costs on an ongoing basis. Our analysis indicates that the so–called hard construction costs directly related to construction, including labor and materials such as concrete and steel, have together increased construction costs by 26 percent since KVSP was built. However, the department’s estimates for San Quentin are triple those for similar housing units at KVSP.

Part of this escalation in costs might be explained by the unique way that housing units would be configured at San Quentin. Because of space constraints, three two–story housing units are to be constructed at San Quentin instead of the customary configuration of six single–story housing units. This approach would add some costs to the project because the concrete walls would have to be poured in place rather than using precast concrete. However, based on construction cost data provided by CDCR, this unique configuration appears to cost only 9 percent more than the standard, single–story housing units. It does not explain why the costs of such units would now be triple the cost of constructing comparable beds at KVSP. (As discussed earlier, this is also an issue for AB 900 infill beds planned at other prisons.)

Questions Remain About Inmate Population Restrictions. Last year, during the course of our review of the 2007–08 budget proposal to complete the construction of the new Death Row, questions arose as to the appropriateness of the condemned housing project in light of constraints imposed on the overall San Quentin inmate population. The department initially indicated to us that, as part of the environmental impact review process for the CIC, the state had agreed to a restriction of 6,558 on the total number of inmates that can be housed at San Quentin. Environmental documents prepared for the CIC indicate that the project, together with the existing housing, would provide a maximum potential capacity of about 7,100 inmates at San Quentin—well in excess of the 6,558 limit. Subsequent data provided to us by CDCR suggest that the maximum potential population at San Quentin would be well under the population limit previously cited by the department. We have asked CDCR to clarify this conflicting information, but the department was unable to resolve the matter at the time this analysis was prepared.

The issue of a population cap at San Quentin has important ramifications for the state’s overcrowded state prison system. If the additional capacity added through the CIC project is added to the existing cells at San Quentin, including those now housing Death Row inmates, it is possible that San Quentin someday might not be able to use its full bed capacity and the state would thus have to let hundreds of beds at the prison remain empty to comply with an agreed–upon population limit. It would not make sense, in our view, to invest such a large sum of state funds for new prison capacity only to have to let hundreds of other prison beds sit idle someday. On the other hand, if the actual capacity of the prison with the added CIC beds is below the agreed upon population cap, this should no longer be a concern.

Analyst’s Recommendation. Due to the concerns outlined above, we withhold recommendation at this time on the administration’s request for additional lease–revenue bond financing authority to complete the San Quentin CIC. Questions about the cost of the new Death Row complex, and the possible effect of inmate population limits at the prison, should be resolved before the Legislature considers the project further. There are two key steps that could be taken to resolve these issues.

Earlier in the analysis of the implementation of AB 900, we recommend that CDCR be directed to retain an independent outside expert to assess the department’s costly estimates for the construction of four new infill bed projects. We further recommend that the same approach be applied to San Quentin. The Legislature needs to know why construction cost estimates of housing units proposed to be built at San Quentin have tripled in five years when increases in labor and material costs or the unique configuration of the beds do not justify this increase.

Meanwhile, CDCR should provide a written report before budget hearings that specifies: (1) the maximum capacity of San Quentin now, including potential overcrowding of the facility; (2) the maximum potential capacity of San Quentin, including potential overcrowding of the facility, if CIC is completed; (3) any specific limits on the inmate population at San Quentin to which the state has agreed as a result of the environmental review process for the CIC; and (4) the department’s rationale for building the CIC at San Quentin if in fact that means other existing space at the prison could not be used to hold inmates in the future. This information would enable the Legislature to determine whether this large construction project at San Quentin is warranted, or whether, as we recommended last year, that the facility be moved to another site without such constraints.  

Return to Judicial and Criminal Justice Table of Contents, 2008-09 Budget Analysis
Return to Full Table of Contents, 2008-09 Budget Analysis