Addressing the State's Long-Term
We recommend that the Legislature authorize additional prison capacity in stages, beginning this year and continuing over the next four to five years. With regard to 1997-98, we recommend that the Legislature authorize two medium/ maximum level prisons this year and appropriate funds needed to design these institutions. (Construction funds will not be needed until 1998-99.)
|Total Inmate Population
Comparison of Recent CDC Projections
|Selected Yearsa||Date of CDC Projections|
|Spring 1996||Fall 1996||Spring 1997|
|a Population as of June 30 of each year shown.|
|b Not estimated.|
In our Analysis of the 1997-98 Budget Bill (page D-65), we discussed the reasons for the difference between the CDC's spring 1996 projections and the much lower fall 1996 projections. (As shown in the figure, the fall 1996 estimate for mid-2005 is almost 50,000 less than the prior estimate for that date.) We indicated that there were two main reasons for the reduced estimate: (1) a reduction, based on actual data, in the assumed average length of prison terms received by "second strikers" (convicted under the "Three Strikes and You're Out" law) and (2) a slower rate of growth in new admissions from court.
There is less change between the department's fall 1996 and spring 1997 forecasts. The latest projection of inmate population for mid-2006 (the ending date of the last two projections) is about 8,000 less than the prior projection for that date. The CDC indicates that the primary reasons for this difference are reductions in their assumptions regarding the (1) rate of new admissions from court and (2) rate at which parole violators are returned to prison with a new term imposed by the courts.
Even given the slower projected growth rates, these latest estimates still represent an increase of 92,000 inmates (or about 60 percent) by mid-2006--an average growth of about 10,000 inmates per year. As shown in Figure 2, the projections follow a growth trend that has continued since the mid-1980s with little interruption. Of course, projections that extend out nine years are subject to considerable uncertainty. Thus, the actual 2006 population could be considerably higher or lower than the forecast. The CDC staff involved with the forecasts believe, however, that given existing law and practices within the criminal justice system, the inmate population will continue to grow at a steady, rapid rate.
Within the state prisons, most inmates are housed in cells and dormitories. Most of the general inmate population is housed two per cell and with double-bunking in dormitories. Housing is less crowded in certain specialized areas, such as administrative segregation, security housing, and psychiatric treatment. In addition to the cells and dorms, the CDC has placed inmates in bunk beds in most prison gymnasiums and in many housing unit dayrooms.
In the fall 1997, construction will be completed on the last prison authorized by the
Legislature--the California State Prison at Corcoran II, which includes a substance abuse
treatment facility for 1,500 inmates. In addition, community correctional facilities to house 2,000
inmates will be completed by spring 1998. Figure 3 summarizes the capacity of the system
according to the housing arrangements described above after all funded housing is completed.
The figure shows that the total capacity of the system is about 171,000.
|Prison System Capacitya|
|Housing units (cells and dorms)||147,915|
|Subtotal for institutions||(160,820)|
|Community correctional facilities||8,186|
|a When all currently authorized beds are completed (spring 1998).|
In addition to this long-term capacity, the CDC will have additional space for about 7,200 inmates--mainly by adding third bunk beds in gymnasiums and dormitories and placing bunk beds on dayroom floors in celled housing units. The department believes, however, that these beds create a setting of undue risk of violence against staff and inmates, and thus they plan to use these beds only when necessary pending the availability of permanent housing.
Even without the "high risk" beds, the department will be operating a prison system with about 13,000 inmates in gymnasiums (10,500) and dayrooms (2,500). For many years the department has used gymnasiums to house inmates. In most prisons constructed since the early 1980s, restroom and shower facilities have been installed in the gymnasiums--in essence converting them to large dormitories. While it would be desirable to eventually replace these arrangements with permanent housing units, the state is facing significant inmate population growth. This means the existing prison system has to be used to the fullest extent possible, while still maintaining a safe environment for inmates and staff. Given this situation, we have included these housing arrangements in the estimated capacity of 171,000 for the funded prison system.
Figure 4 compares the prison system capacity with the estimated inmate population in mid-2006 for each security level. The figure shows that the largest housing needs will be in Levels I and III. We note that the capacities shown in Figure 4 include CDC's plan to convert about 8,500 cells for Level III inmates to instead house Level IV inmates. We assume that these conversions will remain in effect at least through 2006.
|California Department of Corrections
Prison Housing Gap by Security Level
|Capacity of Funded Prisons||June 2006 Inmate Population||Projected Housing Gap in 2006|
As noted above, the six prisons would accommodate about 30,000 inmates, assuming that inmates would be housed in the new prison gymnasiums. Figure 5 shows the impact of these new prisons on the 70,000 inmate housing gap by security level.
|Impact of Governor's Prison Proposal on Housing Gap|
|Projected Housing Gap in 2006||Capacity of Six Proposed Prisons||Housing Gap With Six New Prisons|
As shown in Figure 5, when compared to the expected population, the Governor's proposal would leave varying degrees of housing shortage in each housing level. The new prisons would increase the total capacity to about 200,000. This would accommodate the projected inmate population until about September 2002, or about one year after completion of all six prisons. We estimate that adding additional capacity to meet the housing needs for the projected inmate population in mid-2006 would require eight more prisons costing about $1.9 billion. Figure 6 depicts how the six proposed prisons would address the projected inmate population growth and the future housing gap that would remain to be addressed.
Two prior-year proposals by CDC to expand existing prisons that were not approved by the Legislature may merit reconsideration. One added four dormitories (800 beds) at San Quentin for an estimated cost of about $21,000 per bed (as compared to about $44,000 per bed for a new Level II institution). The other project modified the Northern California Women's Facility to (1) add two dormitories and (2) convert the facility to a men's institution for an estimated $8 million. This expansion and conversion would accommodate about 1,200 male inmates at a construction cost of about $7,000 per bed. Of course, this proposal should only be adopted if alternative approaches are implemented that would alleviate the need to house the 760 female inmates that occupy this institution.
Construct New State Prisons. Since 1982, the Legislature has authorized 21 new prisons at a cost of more than $4 billion. As discussed earlier, meeting the 70,000-bed housing gap in 2006 would require the construction of 14 prisons (5,000 beds each) at a cost of about $3.5 billion. From the time that funding is first authorized, it generally takes CDC at least three and one-half years to open a new prison. Thus, to complete all 14 prisons by June 2006, the Legislature would need to authorize a number of prisons annually from 1997-98 through 2002-03.
Leasing Prison Space. The state could address the inmate housing gap in part by leasing (1) existing excess prison or jail capacity of other jurisdictions or (2) additional prison space to be provided by the public or private sector.
If used for the development of large prisons, these contracting arrangements could be advantageous for California. The privately built prisons can probably be constructed in somewhat less time than a state prison due to the state's legal processes for awarding design and construction contracts. In addition, because the private firms would bid on a competitive basis, the operating costs of the prisons should be less than the costs for a state-owned institution. We do not imply that contracting with private prison operators should be the only solution to addressing the inmate housing gap, but it could be one cost-saving component of the Legislature's overall strategy.
Overcrowd Community Correctional Facilities. The state potentially could increase its systemwide capacity by placing additional inmates in existing community correctional facilities. (This could be done by using bunk beds instead of single beds.) This arrangement would, of course, be subject to reaching agreements with the respective contracting entities, and in the case of some publicly operated community correctional facilities, resolving ongoing legal disputes over their compensation by the state. We estimate that if bunk beds were added or other steps taken to create space, at least 2,000 additional minimum-security inmates could be accommodated.
Modify Classification System. There may be changes to the state's inmate classification system that could be made (see accompanying box on page 10). While such changes would not increase capacity of the system, it could significantly reduce the costs involved.
|Can the Costs of New State Prisons Be Reduced
By Changing Classification Outcomes?
|Prison construction costs vary considerably by the type of housing and associated security arrangements. For example, the combination Level III/IV prisons that the California Department of Corrections (CDC) is proposing to build are estimated to cost almost $40 million more than a comparably sized Level II prison. The need for the more expensive, higher security prisons is based on existing capacity and the CDC's projected number of Level III and IV inmates as determined by the department's inmate classification system. Classification of inmates into the various security levels is based on a scoring system that takes several factors into account.|
|The current classification system was developed in the mid-1980s. The department hired a consultant to assess whether this scoring system places inmates in the appropriate security level. The department's consultant recently concluded that, while the current system is basically sound, fine-tuning of the system is possible. The consultant identified the following factors as the best indicators of misbehavior in prison: age, term, commitment offense, prior criminal history, past behavior in prison, and other risk factors such as gang affiliation or psychiatric history. In addition, the consultant indicated that there is not enough history with the second- and third-striker population to determine how well they fit within the existing scoring system.|
|The CDC is reviewing the consultant's work to determine how and whether to proceed with any modifications to the existing system. Based solely on the above findings, we would not expect that any changes the department would implement would significantly alter inmate classification. Nevertheless, even a small reduction in classification outcomes could result in significant capital outlay and operational savings. The department should therefore begin evaluating the impact of changing its scoring system based on the main indicators identified by the consultant.|
|We believe that the department should also look at its classification system in terms of external security--involving protection of the general public from those incarcerated in a prison--and internal security--protection of staff and inmates from aggressive behavior on the part of other inmates. A highly secure perimeter, such as those provided at new CDC prisons, significantly reduces the potential for escape and protects the public regardless of whether inmates are housed in cells or dorms within the institution. The department should refine the classification system to maximize the number of inmates that can be housed in less costly dormitory settings while maintaining a safe environment.|
As shown in Figure 7, the options would reduce the inmate population by varying amounts, thereby resulting in differing amounts of savings in prison construction and operations. It is also important to note that the estimated savings shown in Figure 7 are based on the impact of each individual option. As several options are combined, the total impact on population would be somewhat less because various options would affect some of the same inmates.
Some of the options, such as expansion of substance abuse treatment programs, are likely to enhance public safety by reducing criminal recidivism. Other changes shown in Figure 7 are not without some risk to public safety. The Legislature and the Governor would want to weigh this risk in selecting options to reduce population growth.
Options for Reducing Inmate Population Growth
(Dollars in Millions)
(for June 2006)
|Reject short-term commitments. Inmates with less than a specified term would remain in county custody or be released.|
|Send directly to state parole. Inmates with less than a specified term would begin immediate supervision in the community on parole.|
|Increase good time/work time credits.||4,800||$100||$220|
|Punish offenders at the local level instead of state prison for the following ten specified nonviolent and nonserious offenses:||(30,500)||($670)||($1,360)|
|Upon release from prison, parole supervision would be provided only for felons convicted of:|
|Place inmates above age 60 in home or community facility detention.|
|Increase substance abuse treatment services for felons within the prison system.|
|Expand and enhance commitment of offenders to prison as civil addicts.|
|Improve inmate work and education programs.|
Given the impending housing shortage in the prison system and the time required to build new prisons, we believe that the Legislature and the Governor should take actions to address the long-term housing gap. There is no one clear-cut solution to the challenge of accommodating 70,000 additional inmates by mid-2006. In Figure 8, we offer a plan to address the 70,000-bed housing gap that is weighted almost evenly between adding new prison capacity and reducing the expected inmate population growth.
|An LAO Plan for Addressing Inmate Housing Gap|
|Reduction in Mid-2006 Population|
|Reduce population growth|
|Prison Industry Authority reform (assume reduces recidivism by 20 percent).||4,800|
|Expand substance abuse treatment for 10,000 inmates (phase in over seven years).||4,700|
|Expand and enhance the civil addict program by adding 3,000 beds over five years.||3,800|
|Place inmates over age 60 to home or community facility detention except violent and serious felons.||1,300|
|Direct placement to parole for inmates with six months or less to serve.||2,200|
|County supervision rather than parole for inmates convicted of nonviolent or nonserious offenses with no prior violent or serious offenses.||7,800|
|Punish offenders at the local level instead of state prison for:|
|Reduce total impact by 10 percent for overlapping effect||-3,880|
|Total population reduction||(34,920)|
|Increase System Capacity||36,080|
Plan Consistent With State Policy. Our plan is consistent with state policy established by the Legislature and the Governor. For example, Chapter 41, Statutes of 1994, First Extraordinary Session (AB 99x, Rainey), provides a framework under which criminals convicted of relatively low-level offenses could be diverted from state prison to alternative punishment options within local communities. The AB 99x policy recognizes that it makes sense to prioritize expensive state prison space for the most serious offenders--generally those with long terms convicted of violent or serious crimes and career criminals.
Assistance to Counties Is Warranted. Some of our recommendations would result in additional direct costs to counties, which would assume responsibility for those offenders no longer under state jurisdiction. (Other recommendations would have indirect impacts on counties--both savings and costs.) Given these impacts, it is important to consider the current fiscal condition of counties and the challenges they would face in assuming these responsibilities. For instance, as we indicated in our Analysis of the 1997-98 Budget Bill, almost all county jails are currently overcrowded. In 23 counties (representing over 70 percent of the state's total jail capacity), the federal courts have imposed limits on the number of people that can be held at any one time. These limits often require early release of inmates--an estimated 29,000 statewide each month.
Assuming adoption of our proposal, we believe that the state should provide funding to counties to address their additional criminal justice costs. There are a number of ways to accomplish this. Our proposal for replacing state parole supervision with county supervision of certain low-level offenders, for instance, assumes the state would provide a subsidy to county governments to help defray supervision costs. The state also could provide funding for additional jail construction and for alternative punishment programs--such as mandatory substance abuse treatment, day reporting centers, and electronic monitoring--as contemplated in AB 99x.
The Legislature should also consider increasing support for local efforts to reduce incidents of crime--particularly efforts that focus on juveniles. This is especially true at this time, given the coming surge in the state's juvenile population. Over the long run, effective programs to reduce crime by juveniles, as well as adults, will provide savings for all elements of the criminal justice system and will improve public safety for society in general. From a state fiscal perspective, such efforts could directly impact the pressure to expand the prison system.
State Savings With Plan. We estimate that adopting our plan would save the state about $1.6 billion in one-time capital outlay costs for new prisons and $700 million in annual operating costs for the CDC by 2005-06. (This estimate assumes state financial assistance for county supervision of state parolees. Annual savings would be offset further to the extent the state provides additional funding for local entities.) Average annual growth in operating costs for the CDC would be reduced from 7.4 percent (assuming no policy changes) to 6.1 percent--an amount that would still exceed overall General Fund revenue growth (assuming moderate economic growth in California over this time period).
One major benefit at the state level from implementing our plan would be to reduce the considerable impacts on the CDC's operations resulting from large inflows and outflows of inmates. In 1995-96, the total inmate population grew by about 10,000. Over the entire year, however, CDC took in almost 127,000 inmates and released 117,000 inmates. These huge flows through the system create a significant challenge for the department in performing such functions as classification, placement, transportation, and employment of inmates. To the extent that our recommendations reduce the number of offenders who are sent to prison for very short time periods, this strain on the department's operations would be less.
Impact on the Housing Gap by Security Level. Figure 9 shows the estimated impact on the projected housing gap for each security level if our proposals to reduce inmate population were adopted. It indicates that the proposals would have the most impact on lower-security male inmates and women, with little impact on higher-security inmates. In addition, the reduction in the numbers of what are currently short-term offenders going to state prison should reduce admissions and hence the need for reception center housing. Policy changes would more than address the housing gap for Level II inmates. This space could then be used to address a portion of the gap for Level I inmates.
|Impact of Proposed Policy Changes
On Housing Gap
|Projected Housing Gap in 2006||Impact of Policy Changes||Prison System Expansion Needs|
Therefore, coupled with our recommended policy changes to reduce inmate population, we recommend that the Legislature authorize two new Level III/Level IV prisons this year (accommodating about 10,000 inmates). The Legislature needs to develop a financing plan for these prisons. Financing options include the General Fund, general obligation bonds, lease-payment bonds, and federal crime bill grants. However, based on the time that will be required to design the prisons, the Legislature would only need to fund the design costs (approximately $20 million) at this time. Federal funds are available for this purpose. Funding for construction would not be needed until fiscal year 1998-99.
For expansion beyond the two prisons, we believe that a portion of the housing capacity needs could be met by means other than new state institutions. We therefore recommend that the Legislature direct the department to develop the following for legislative review for the 1998-99 budget.
The Legislature can then review this information next year and determine how to authorize and finance additional capacity.
Gerald Beavers and Craig Cornett.
Publications: (916) 445-2375
World Wide Web site: www.lao.ca.gov
Return to LAO Home Page