First, there have recently been fewer new admissions into the prison system than was projected in the spring 1994 estimate. The CDC now projects that total new admissions will still grow continuing a long-term trend but not as much as previously estimated. This affects the projections both for the base population (inmates and prison terms that would occur without the Three Strikes law) and for those inmates sent to prison under the Three Strikes law. The CDC assumes that there have been fewer Three Strikes admissions than previously anticipated in part because of the large backlog of Three Strikes cases awaiting adjudication.
Second, the CDC has lowered its projection of felons that, because of Three Strikes, would be sent to state prison instead of being sentenced to local jails or put on probation.
Third, the CDC incorporated the Three Strikes law into the computer model that is used for biannual projections for inmate population. The previous Three Strikes estimate was based on a simplified model that the CDC uses for assessing the impact of proposed legislation. (This third factor accounts for 19,000 of the 35,000 total reduction in the five-year estimate.)
In total, about 30,000 of the 35,000 reduction in estimated inmate population is due to the CDC's revised estimate of the impact of the Three Strikes law. The CDC is currently evaluating whether there are likely to be changes in the long-term impact of this law. (The department's spring 1994 long-term assessment was that Three Strikes would eventually increase inmate population by 275,000 in 2026-27.)
The Three Strikes law will have a far greater impact on the prison population than any prior single piece of legislation. Given the sweeping scope of this new law, population projections are subject to great uncertainty. Specifically, one area which is very difficult to estimate is any behavioral changes either on the part of criminals or the criminal justice system stemming from this law. In a companion Status Check, The Three Strikes and You're Out Law A Preliminary Assessment, we indicate that in response to Three Strikes, there are noticeable changes in the patterns of pleadings entered by defendants. Over time, such behavioral changes could result in wide variances from the CDC's current population estimate.
The actual inmate population is currently below the fall projections, which could be attributable to the behavioral changes mentioned above. As of early December, the population was unchanged over the previous four months and was about 2,600 less than the CDC's fall 1994 estimate. Based on discussions with the CDC staff and with local criminal justice officials, this is probably due to an increased backlog of cases awaiting trial. If the local criminal justice system takes steps to reduce the backlog, however, the inmate population might still increase to levels projected by the CDC.
Regardless of these short-term impacts of the Three Strikes law, planning for new prisons must focus on inmate population trends over several years. For new prison planning purposes, therefore, we believe that the CDC's projections provide at least a reasonable order of magnitude of future inmate population.
Although the department's projections have been revised downward, it still is projecting dramatic growth in the number of inmates. The CDC's current inmate population projection represents an increase of 86,000 inmates in the next five years from 125,000 to 211,000. This equals the increase that the state prison system incurred over the past ten years. The state's prisons, however, fall far short of having space to accommodate this projected growth in inmate population.
Figure 2 shows recent and projected overcrowding levels in the state prison system based on the scheduled completion of all authorized prisons. As shown in the figure, prison overcrowding increased from about 140 percent in June 1983 to around 180 percent in June 1990 and has remained fairly close to that level over the last four years as additional prisons have been opened. Based on the CDC's fall 1994 population projections, unless more prisons are built, overcrowding in the prisons will increase significantly over the next five years and will reach 256 percent in mid-1999 nearly three inmates for each space designed to house one inmate.
The CDC has established a goal of having sufficient prison capacity to house inmates at an average overcrowding level of 120 to 130 percent of design capacity. In order to attain this level by mid-1999, the state would have to construct 38 prisons at a cost of around $9.5 billion. (The state currently has 29 prisons.) This level of construction over a four-year time frame cannot be achieved.
Even a more moderate task (one that was included in the CDC's five-year plan) of building sufficient capacity to maintain the current 182 percent overcrowding level would be extremely difficult. This would require construction of prisons designed to house 33,000 inmates about 15 prisons at a cost of around $4.5 billion. (In comparison, over the last 13 years, the CDC has designed and constructed prison space for 43,000 inmates the largest prison construction program ever undertaken in the United States.)
The process of selecting a suitable and acceptable site, performing the environmental review, completing design and construction of the facilities, and then activating a single prison for occupancy can take four years. Accomplishing these tasks within four years for 15 prisons would be an unprecedented challenge. Furthermore, within the four-year time period, the department would need to hire and train over 15,000 new staff for these prisons.
In September 1994, the CDC reported that environmental impact reports were either completed or in process for four prison sites. The department also indicated that several other communities have expressed interest in having prisons. Based on the state's experience to date, however, it is unlikely that additional prisons, beyond those already authorized, can be completed before 1999.
Potential Sources of Prison Construction Funding. No new funding was authorized for new prisons or for emergency housing in 1994. In addition, no general obligation bond measure for prison construction or renovation was placed on the 1994 ballots. Currently, there is less than $10 million available from past prison bond measures for appropriation by the Legislature.
The state's new prisons have been almost entirely funded by either general obligation or lease-payment bonds. Barring a special election, the earliest that voters could authorize additional general obligation bonds for prisons would be at the March 1996 statewide election. Funding with lease-payment bonds does not require voter approval and thus could be authorized at any time with a majority vote of the Legislature and approval by the Governor. For several reasons, lease-payment bonds are more costly than general obligation bonds. We estimate that General Fund debt service for lease-payment bonds is about 15 to 20 percent higher over the life of the bonds.
Other than bonds, there are currently few practical options for financing prison construction. Direct appropriations, either from the General Fund or from the Special Account for Capital Outlay (SAFCO), have been minimal in recent years. In view of the near-term outlook for these funding sources, outlays of the magnitude needed to construct new prisons is highly unlikely.
The recently enacted federal crime bill, however, should provide funds to California to assist in part of this construction. As we note in our report, The Federal Crime Bill: What Will it Mean for California?, the state's potential share of federal prison construction grants over the next six years could be about $1.2 billion based on authorized funding levels specified in the federal legislation. The actual amount the state receives will depend on annual appropriations by the Congress. The initial grant will not be available before October 1995. In addition, the state would be required to provide 25 percent of the costs of projects funded with the federal grants. If the state receives the entire $1.2 billion (requiring $400 million in state funds), about five new state prisons (11,000 beds) could be constructed.
If the state were to authorize and construct 15 more prisons over the next five years, debt service costs would increase by about $400 million in 1999-00. (This includes debt service costs for five authorized-but-unfinished prisons and assumes the state receives $1.2 billion in prison construction grants from the federal government.) The annual costs to operate these additional prisons would be about $1.5 billion. Thus, General Fund costs for the CDC in 1999- 00 would be about $5 billion (in 1994-95 dollars) an increase of nearly 60 percent in five years.
In the current year, the General Fund cost of the CDC is about 7.6 percent of estimated General Fund revenues. For the CDC's expenditures to be an equal percentage in 1999-00, General Fund revenues would have to grow by an average of 10 percent per year. Given the state's current tax structure and economic outlook, revenue growth will probably not be this large. For example, in the 1980s when the state enjoyed relatively strong economic growth General Fund revenues increased by an average of 8 percent per year. It is therefore likely that operating the state's prison system will require an increasing portion of state revenues.