Special Examination Report No. 18-12 December 2018 Why we did this review The House Appropriations Committee requested this special examination. Based on the committee's request, this review was designed to determine what GDC's strategic plan is as it relates to demand for prison facilities and bed space. In addition, the review sought to determine what role private prisons will play and whether they offer a more effective reform and cost alternative for the state. About GDC GDC is charged with managing the state prison population. In fiscal year 2018, it managed an average of 50,012 offenders per day. Its expenses totaled $1.2 billion. GDC currently operates 34 state prisons and contracts operations for 4 private prisons and 22 county correctional institutions (CCIs). Georgia began making reforms to the judicial system in 2011 that have impacted the GDC population. Prior to the reforms, GDC facilities held approximately 52,800, and the corrections system had a backlog of more than 4,000 offenders. It expected the population to increase by over 7,000 offenders in the next five years. Instead, the increase was 800 offenders, and the backlog was reduced to 1,000. The reforms focused on reducing the sentences of nonviolent offenders; as a result, the remaining population is increasingly made up of violent offenders. Georgia Department of Corrections Requested information on bed demand and agency strategy What we found Although the prison offender population growth was slowed as a result of state's recent criminal justice reforms, we estimate the population will increase by approximately 1,277 (2.5%) over the next 5 years. The increase will most likely result in more than 97.5% of the Georgia Department of Correction's (GDC) standard (operational) beds being filled in most facilities. For some offender populations, the utilization rate may come close to, or exceed, 100%. Because current utilization rates and bed capacity vary by facility type, the increase in offenders will likely result in slightly different bed demand throughout the state's prison system. GDC will need to develop strategic plans based on the facility types that house distinct offender populations. Absent such planning, GDC risks overestimating its ability to accommodate increases. GDC has long relied on a combination of state prisons, private prisons, and county correctional institutions (CCIs) to manage the state's prison population. In fiscal year 2018, GDC had an average of 50,012 offenders per day housed in prisons. Of these, an average of 37,509 (75%) were in state prisons, 7,818 (15.6%) were in private prisons and 4,684 offenders (9.4%) were in CCIs. However, GDC does not assign all types of offenders to private prisons and CCIs but restricts assignments for these facilities to healthier, lowerrisk male offenders. GDC officials indicated that any permanent significant increase in bed space to accommodate population growth during the next five years will likely be achieved by expanding capacity within state prison facilities. Officials provided a proposed facility expansion plan that identified six prison facilities (four are operating 270 Washington Street, SW, Suite 1-156 Atlanta, Georgia 30334 Phone: (404)656-2180 www.audits.ga.gov currently) that could accommodate the expected population increase. Cost estimates provided for the four facilities currently in operation ranged from $1.0 to $15.0 million. Estimates for the two that are not currently operational ranged from $3.5 to $15.0 million. It should be noted that GDC staff indicated it has submitted a budget request for additional funding for one of these two. In addition, contracts with two private prisons currently include expansion clauses that allow an increase of 1,000 beds each. Private prison officials estimated that approximately 400 additional standard beds are currently available for contract that would not require construction build out. When determining how to house the increase, cost is a consideration. The GDC daily cost to manage offenders during fiscal year 2017 in private prisons and state prisons was comparable when controlling for offender sex, facility size, and facility risk classification. However, the GDC cost to manage offenders in CCIs was significantly less than either state prisons or private prisons. GDC pays CCIs a daily rate of $20 per offender; however, this rate has not been adjusted since 1999. State prisons averaged a per offender/per day cost of $65.58 across all facilities. To compare costs to private prisons, we identified a comparable set of state prisons (same offender sex, facility size, and risk classification) and found costs averaged $44.56 per offender/per day. Private prisons averaged a per offender/per day cost of $49.07. CCIs averaged a per offender/per day cost of $21.63. These costs include allocated indirect costs. What we recommend This report is intended to answer questions posed by the House Appropriations Committee and to help inform policy decisions. In addition to the information provided, we recommend GDC consider differentiating facility utilization rates by bed type (standard or special needs) and evaluating offenders by distinct sub-populations such as gender, special need, and offender risk to improve future bed demand planning. See Appendix A for a detailed listing of recommendations. Summary of Response: In its response to the report, GDC indicated its agreement with the information presented. It also noted that it will "incorporate bed space utilization (by bed type) and sub-population analysis by distinct sub-populations (gender, special needs, and offender risk classification) into the population planning process." Georgia Department of Corrections i Table of Contents Purpose of the Special Examination 1 Background National Justice Reinvestment Reforms 1 Georgia Criminal Justice Reforms 2 Prison Facility Profile 4 Offender Management 6 Bed Capacity Planning 6 GDC Financials 8 Requested Information Question 1: Does GDC have a strategic plan that considers necessary factors? 9 Question 2: What is the projected population increase over the next five years and does GDC have sufficient plans to address it? 12 Question 3: Are private prison beds a more effective reform and cost alternative for the state? 18 Matters for Consideration Further review is needed to determine whether the type and size of the population served by private prisons could be changed. 21 Before GDC decides to increase its reliance on contracted bed space, it should assess the potential risk thoroughly and have considered contingencies. 22 Consideration should be given to reviewing the reimbursement rate for County Correctional Institutions (CCIs). 22 Appendices Appendix A: Table of Recommendations 24 Appendix B: Objectives, Scope, and Methodology 25 Appendix C: Inventory of State Prisons, Private Prisons, and County Correctional Institutions 30 Appendix D: Population Projection Based on FY 2018 Average Daily Population 32 Appendix E: Population Projection Based on FY 2018 Highest Offender Count Day 33 Georgia Department of Corrections ii Georgia Department of Corrections 1 Purpose of the Special Examination This review of the Georgia Department of Corrections (GDC) was conducted at the request of the House Appropriations Committee. Based on the committee's request, this review was designed to answer the following questions: 1. What is GDC's strategic plan as it relates to demand for prison facilities and bed space and what role will private prisons play? 2. Are private prison beds a more effective reform and cost alternative for the state? A description of the objectives, scope, and methodology used in this review is included in Appendix B. A draft of the report was provided to GDC for its review, and pertinent responses were incorporated into the report. Background National Justice Reinvestment Reforms As shown in Exhibit 1, 27 states adopted "justice reinvestment reforms" between 2007 and 2016. Generally, the goals of these reforms have been to reduce correction costs and decrease offender recidivism while maintaining public safety. Actions have focused on reducing sentences for nonviolent offenders and replacing time in prison with community supervision programs aimed at rehabilitation. States have reported reduction in prison commitments and costs as a result of these reforms. Exhibit 1 27 States Implemented Major Justice Reforms (2007-2016) AK ME VT NH WA MT ND MN WI MI NY MA RI ID WY SD IA IL IN OH PA NJ CT OR NV CO NE MO KY WV MD DE CA UT KS AR MS TN NC VA DC AZ NM OK LA AL GA SC HI TX FL Source: National Conference of State Legislatures States with Justice Reinvestment Reforms Georgia Department of Corrections 2 Georgia Criminal Justice Reforms In May 2011, the Georgia Assembly passed HB 265 in recognition of an increasing number of GDC inmates and the corresponding increases in incarceration costs to taxpayers. HB 265 created the temporary Special Council on Criminal Justice Reform for Georgians and charged it with researching potential reforms to Georgia's criminal justice system. In the years that followed, the council offered policy recommendations, and several major legislative reforms were passed (see Exhibit 2) . In May 2012, at the recommendation of the Special Council, the Georgia Assembly passed HB 1176. It included reforms to reduce the imprisonment of nonviolent offenders, increase referral of offenders to accountability courts, and focus on investing the potential savings of these reforms into community supervision programs. In March 2013, the Governor signed HB 349 to recreate the Special Council (renamed the Georgia Council on Criminal Justice Reform) and extend its term through June 2018. In subsequent years other significant justice reforms were put forth by the Council and passed into law, including HB 310 (which codified the Office of Transition, Support, & Reentry created by executive order in 2013) and HB 328 (which created the Council of Accountability Court Judges) in 2015. Exhibit 2 Timeline of Major State Legislative and Executive Reforms Year 2011 2012 2013 2014 Bill / Executive Order HB 265 HB 1176 Executive Order 05.24.12.03 HB 242 HB 349 Executive Order 6.28.13.04 Executive Order 6.24.13.01 SB 365 Purpose Created the Special Council on Criminal Justice Reform for Georgians and charged it with researching potential reforms to improve Georgia's criminal justice system Initiated justice reinvestment reforms Created an Accountability Court Funding Committee to assist with allocating funds to court divisions Reformed the juvenile justice system Recreated Special Council and renamed as Georgia Council on Criminal Justice Reform Appointed Council members and grants them 5-year terms Created the Office of Transition, Support, & Reentry Implemented post-incarceration reforms; developed a program to assist with re-entry 2015 HB 310 HB 328 Created the Department of Community Supervision to oversee parole and established the Governor's Office of Transition, Support, & Reentry in statute Created the Council of Accountability Court Judges to set standards and practices for all accountability court divisions based on national practices 2016 2017 SB 367 SB 174 SB 406 2018 SB 407 Created reforms to help offenders reenter society and expand accountability courts Improved the probation system for the goal of reducing recidivism rates Enacted the Georgia Long-term Care Background Check Program Codified the Governor's Criminal Justice E-filing Project, reformed the misdemeanor bail process, provides judges with opportunities to convert monetary fees/fines into community service, and imposed tougher penalties on firearm-related offenses Source: Official Code of Georgia Annotated Georgia Department of Corrections 3 Accountability Courts One of the major reforms to Georgia's criminal justice system has been an increased investment in and expansion of accountability courts. Accountability courts are statefunded alternatives to incarceration for qualified offenders based upon their offense and characteristics such as drug addiction, mental health concerns, or history as a military veteran. Accountability courts provide supervision, community service options, and treatment to these offenders. As of June 2018, there are 149 accountability courts throughout the state, with at least one type of accountability court in each of the ten judicial districts. Reform Results Collectively, the Georgia criminal justice reinvestment reforms have significantly affected the projected offender population between 2011-2018. In 2011 (before the passage of any reform bills), GDC facilities held approximately 52,800, and the corrections system had a backlog of more than 4,000 state inmate commitments held at county prisons awaiting transfer. GDC projected an increase of approximately 7,000 additional offenders during the periods 2012-2016. In fact, GDC commitments during the period 2011-2017 increased only 800, while reducing the backlog of offenders awaiting assignment to less than 1,000. Exhibit 3 shows the GDC offender population trends during the period 2011-2018, with projections through 2023. Exhibit 3: GDC Offender Population Declined Significantly After Reforms 70,000 66,000 62,000 58,000 54,000 52,804 59,684 53,750 55,451 50,000 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 Source: Georgia Department of Corrections Because of the justice reforms, the inmate population is changing. Reforms have focused on reducing the sentences of nonviolent offenders; as a result, the remaining population is increasingly made up of violent offenders (e.g., those who committed a major violent or sexual offense) serving longer sentences and requiring greater security. GDC reports that the percentage of violent offenders increased from 62.8% Georgia Department of Corrections 4 in fiscal year 2011 to 67% in fiscal year 2016.1 GDC measures offender risk using a 0.002.00 scale. Prison Facility Profile GDC relies on a combination of state prisons, private prisons, and county correctional institutions to manage the prison offender population.2 In fiscal year 2018, the agency operated 34 state prisons and contracted operations for 4 private prisons and 22 county correctional institutions. GDC classifies facilities into sub-groups using a combination of factors such as the offender's sex, the general population's risk classification, and whether the facility serves "special needs" (e.g., specialized medical conditions). Facilities can technically house most offenders; however, the risk classification denotes the classification of the majority of the facility population. As a management strategy, GDC indicated it tries to house lower-risk offenders in minimum or medium level facilities and strives to keep the percentage of close (or high risk) offenders in these facilities at or below 5%. The majority of offenders in close risk facilities are high-risk offenders; approximately 10-40% of the offenders in these same facilities are classified as medium or low risk. Exhibit 4 presents a summary profile of facilities in GDC's prison system by owner/management. Appendix C provides a detailed list of the facilities. Exhibit 4 GDC Prison Facility Owner/Management Groups Owner/Manager State Prisons Private Prisons County Correctional Institutions Total # of Facilities 331 4 22 # of Special Needs Facilities 8 0 0 Offender Sex Male and Female Male Male Type of Facility Risk Classification Minimum, Medium and Close2 Minimum and Medium Minimum and Medium % of Total Offenders 75% 15.6% 9.4% 1 At the end of fiscal year 2018, the Metro State Prison opened. Because it was not open during the period reviewed, it was not included in our analysis. 2 Close facilities are the highest general risk classification for offenders/facilities. Source: GDC Data/DOAA Analysis State Prisons GDC owns and operates 34 state prisons. Of these, 29 house only male offenders; 4 house only female offenders. One medical facility (Helms) houses male and female offenders. Eighteen state prisons are classified as medium risk; seven are classified as high (or close) risk. Nine state prisons are classified as special mission facilities which serve specific populations or offender needs3 . For instance, Baldwin State Prison holds offenders with severe mental health issues, and Burruss Correctional Training Center houses juveniles. 1 From March 2014 to July 2017, the average male offender's score increased from 0.33 to 0.45; the average female offender's score increased from 0.22 to 0.23. 2 Transitional centers, detention centers, parole revocation centers, and residential substance abuse treatment centers were excluded. 3 Special mission facilities also have a security level, but this is secondary to its special mission status for the purposes of our examination. Georgia Department of Corrections 5 State prisons house a large majority of the GDC prison population. In fiscal year 2018, state prisons held 75% of offenders. The daily average offender population in state prisons was 37,509 in fiscal year 2018. Private Prisons Georgia has a long-standing history of working with private contractors for the construction of prisons and offender management. Currently, GDC has long-term contracts with two companies for the operation of four private medium security level prisons. Contracts are renewed annually. Exhibit 5 presents a summary of private prisons in the GDC system. According to staff, GDC currently restricts the offender sub-group it assigns to private prisons to healthy, male offenders with a lower risk classification. Offenders in private prisons typically do not have significant special needs. According to the Bureau of Justice Statistics, at least 27 states, including Georgia, use private prisons. Among these, eight states have a total prison population comparable to Georgia's in terms of number of prisoners. The percentage of total population housed in private prisons averages 13% and ranges from 0.1% (North Carolina) to 26.6% (Oklahoma) . In fiscal year 2018, the daily average offender population in private prisons was 7,818. During the same year, private prisons held 15.6% of GDC prison offenders. Exhibit 5 Two Companies Operate Four Private Prisons, Fiscal Year 2018 Facility Name Coffee Correctional Facility Owned and Operated By Core Civic Contract Period1 (Begin/End) 1997 / 2034 Jenkins Correctional Facility Core Civic 2010 / 2035 Riverbend Correctional Facility Geo Group 2010 / 2051 Wheeler Correctional Facility Core Civic 1997 / 2034 1 Includes years GDC has option to renew, beyond the initial contract period. 2 A per diem rate shall apply for beds at Wheeler Correctional Facility above 2,628. Source: GDC records Contracted Bed Capacity 2,628 1,150 1,500 2,6282 County Correctional Institutions (CCIs) Georgia has had a network of county correctional institutions for many decades. As with private prisons, GDC currently restricts the offender sub-group it assigns to CCIs. GDC officials indicated that CCIs are only assigned male offenders with a lower-risk classification without special needs. In addition, (because CCI offenders are often utilized as labor for county governments) offenders must be able to work and be supervised outside of the facility. GDC renews CCI contracts annually. In fiscal year 2018, CCIs held 9.4% of GDC prison offenders. The daily average offender population in CCIs was 4,684 in fiscal year 2018. Georgia Department of Corrections 6 Offender Management After sentencing, offenders typically are held in the jail of the county where they were sentenced. Once GDC receives the sentencing information from the court, the agency confirms the information, assigns the offender an ID number, and schedules an offender transfer. It should be noted that offenders may remain in the county jail until space is available in an appropriate GDC prison facility. GDC reports a nine-day average offender wait time in fiscal year 2018. Initial Processing and Assignment Male offenders are transported to Georgia Diagnostic and Classification State Prison or Coastal State Prison for screening and intake. Arrendale State Prison serves the same purpose for female offenders, while Burruss Correctional Training Center processes juveniles. The intake process is conducted by GDC staff and includes assessing the offender's medical and mental health status, as well as identifying programming needs (e.g., educational counseling, substance abuse, etc.) . The intake process also includes assessing affiliations or relationships with staff or other inmates that may pose a security risk (e.g., gang affiliations). Staff input this information into its information system, SCRIBE, which has an embedded automated assessment instrument. This instrument generates a security risk classification score based on the offender's past and current sentences, nature of the crime, history of violence, and needs identified in the initial review. Staff then match the offender's risk classification and programming needs to appropriate facilities. Once they confirm space is available, staff arrange for the offender to be transported to the selected facility. Subsequent Transfers/Reassignment(s) Although offenders are initially assigned to a facility, they may be reassigned or relocated for many reasons, including: a significant change in special need occurs (e.g., serious illness/injury) , a significant change in risk security classification (i.e., offender is reclassified as a greater or lower risk) , or a more appropriate facility type is preferable (e.g., the offender's sentence is nearing an end and the offender is moved to a transition center). While GDC leadership reports the annual number of transfers and reassignments is significant, its data systems are not able to quantify these moves. Bed Capacity Planning GDC plans and manages its bed capacity using utilization reports generated from SCRIBE, as well as other reports that detail offender population trends affecting bed utilization (e.g. admission/releases, concentration of close security offenders throughout the system, offender completion of programming, etc.). Calculating utilization rates by facility requires several variables, including bed type, bed capacity, and offender counts. These variables, and how we applied them, are discussed in detail below. Georgia Department of Corrections 7 Bed Type GDC's prisons have different types of bed space to house offenders. The two bed types are: standard beds and special use beds. They are defined as follows: Standard beds are those beds available within a facility for general offender assignment. Special use beds are used for offenders based on an identified need. Examples include infirmary beds, isolation or segregation beds and mental health observation beds. These assignments are not permanent. Taken together, standard beds and special beds are called physical beds. Bed Capacity Bed capacity is the sum of the beds (by bed type) available for offender assignment within a facility at any given time. Facilities, therefore, have a standard bed capacity as well as a special use bed capacity. Taken together, they make up the physical bed capacity. Bed capacity is not fixed and can increase or decrease temporarily (due to facility repairs or maintenance) or permanently (due to facility opening, expansion, or closing). Offender Counts GDC maintains a daily count of offenders in all facilities. Offenders are counted according to the type of bed they occupy. For example, a daily count for Arrendale State Prison for July 1, 2017 was 1,289 offenders in standard beds and 112 offenders in special beds. Utilization Rates For purposes of this report, we used the definitions above in calculating utilization rates within and among facilities. Utilization, which is calculated by bed type, is the count of offenders in a particular type of bed divided by the total number of those beds. For example, Macon State Prison has 1,426 standard beds. On June 30, 2018, it had 1,402 offenders in standard beds. Therefore, the standard bed utilization rate was 98.3%. Utilization rates are calculated the same way for special and physical beds. Physical Capacity versus Contracted Capacity GDC contracts with private prisons and county correctional institutions to house offenders. According to GDC staff, the "contracted capacity" is the number of beds the state has entered into a contractual agreement to occupy. However, the agency may not contract for the entire supply of available offender beds within a facility. For example, according to a company representative, Coffee Correctional Facility has a physical bed capacity of 2,992. However, the GDC fiscal year 2018 contracted capacity is 2,628. During any given month, GDC may or may not fully utilize the contracted beds but makes decisions month-to-month based on the availability of funds. Georgia Department of Corrections 8 GDC Financials Annually, GDC expends over $1.2 billion to manage the state's offender population. As shown in Exhibit 6, over 50% of GDC's expenditures have been for operation of state prisons over the past two years. Another 20% has been for health related expenses. The next largest annual expenditure has been for private prison operations. GDC is budgeted to expend $1.2 billion in fiscal year 2019. Exhibit 6 Expenses Have Been Consistent over the Period Fund Source State Federal Other Total Actual 2017 2018 $1,161,720,344 $1,182,282,786 2,674,260 3,323,557 67,076,829 63,454,605 $1,231,471,433 $1,249,060,948 Budgeted 2019 $1,188,970,280 1,016,171 16,617,667 $1,206,604,118 Expenses (by Program) Administration $ 38,632,217 $ 39,058,952 Food and Farm Operations 28,056,029 28,220,264 Health Offender Management1 249,283,006 44,297,386 239,706,029 43,618,339 Jail Subsidy 0 60,690 Private Prisons 135,787,976 135,395,608 Detention Centers 40,809,024 41,919,553 State Prisons 662,494,941 688,502,897 Transition Centers 32,110,854 32,578,617 Total $1,231,471,433 $1,249,060,948 1 Includes contract payments for County Correctional Institutions. Source: Budget Comparison Reports $ 37,633,793 27,608,741 239,372,467 43,646,572 5,000 139,784,108 50,450,237 635,604,221 32,498,979 $1,206,604,118 GDC currently contracts with two companies for the four private prisons. The companies are responsible for the operations of the prisons, including staffing. GDC pays a cost per offender per day rate. The contracts contain a tiered structure whereby the cost per offender per day rate is determined by the number of offenders assigned to the facility. For example, one rate may apply for up to 1,000 offenders; a lower rate may apply to the next 500 offenders. Typically the per offender per day rate decreases after a specified threshold is met. In addition to the contracted costs, there are also indirect costs associated with the private prisons and CCIs. GDC provided a cost allocation for administration and offender management that it applied to the private prisons and CCIs when calculating overall cost per offender per day. Health care costs of offenders in the private prisons are also assigned for cost calculation. Georgia Department of Corrections 9 Requested Information Strategic Planning as it Relates to Bed Space To answer this question, we considered whether GDC's strategic plan considered necessary factors in assessing future facility and bed space demands. As noted in the following sections, overall GDC has a substantial planning process that considers future needs; however, it can be refined. By analyzing information at a facility subpopulation level, GDC can assess potential facility-level impacts allowing for improved population management. In assessing the significance of such an approach, we also sought to determine how large and what type of population increase GDC faces. The total projected increase over the upcoming five-year period is 1,277 (2.5%). Without differentiating facility and bed types, the agency risks overestimating its ability to accommodate the increase. By differentiating facility and bed types, the agency will be better able to identify future capacity needs. As context for the following sections, GDC's average daily prison population in fiscal year 2018 was 50,012 ranging from 49,174 to 50,520. The average daily physical bed capacity was 53,255. When broken down further, the average daily standard bed capacity was 47,802 and the average daily special bed capacity was 5,453. Utilization rates for standard beds were typically higher than 95% across facility sub-groups, while special bed utilization was approximately 82%. Question 1: Does GDC have a strategic plan that considers necessary factors? GDC plans at both a strategic and tactical level. At a strategic level, GDC participates in multi-agency planning to consider policy changes that may affect the prison population and tracks population projections for increases over time. GDC also submitted a strategic plan covering fiscal years 2019-2022 to the Governor's Office of Planning and Budget (OPB) as part of the fiscal year 2019 annual budget process. The plan includes goals related to facility security and safety, offender re-entry, workforce quality, and technological improvements. At a tactical level, GDC uses weekly reports on offender counts and bed space to identify when facilities are approaching maximum capacity in standard or special use beds. However, we found that the planning does not include specific goals or objectives related to bed space utilization nor detail on how GDC intends to accommodate specific subpopulations of offenders within the projected increase. Additionally, the strategic plan does not project whether or how state prison facility condition or contracts with outsourced management might affect GDC's ability to accommodate the prison population. Below is a description of the factors that GDC should add to its planning process. Offender Population Projections The state has a long-standing contract with an Atlanta-based consulting firm to aggregate offender data and develop five-year prison population projections. The projections are presented annually, and offender population trends that influence bed space availability are reviewed by a group composed of specified members of the Governor's executive staff, OPB, the Governor's Public Safety Council, the Georgia Board of Pardons and Parole, the Criminal Justice Coordinating Council, the Department of Community Services, and GDC. Approximately 20 representatives from these same organizations make up the Offender Population Committee (OPC), which meets monthly to discuss the implications of these population projections. Georgia Department of Corrections 10 GDC population projections have closely approximated actual population counts, with error rates under 1% over the last two annual projection cycles. However, these projections are not broken down into sub-populations. By refining this projection, GDC could more accurately identify bed space demand by facility types such as by sex, by risk classification, and special mission. Facilities Facility condition can affect bed capacity as facilities, or portions of facilities, may be closed for maintenance or renovation, affecting bed space availability. As a result, facility condition should be a consideration when planning for population growth. Based on a 2007 comprehensive state prison facility condition assessment (the last one completed), there are facilities that require maintenance and/or renovation. The facilities were rated on building structure, utilities, and electronic and physical security. The average score was 62.5%, with the highest facility score being 79.2% and the lowest being 45.9%. GDC indicated no substantial improvements or renovations to facilities have occurred since this 2007 study; staff indicated that funding availability has driven facility maintenance and capital improvements. According to staff, no major facility closures are scheduled for fiscal years 2019-2023. In addition to maintenance or renovations, according to staff, GDC also initiated efforts to improve facility infrastructure weakened by a higher concentration of violent offenders and to prevent manipulation of locking mechanisms that was occurring. GDC refers to this process as "hardening" of the facilities. In 2013, GDC identified seven prison facilities that required hardening; it then prioritized and completed the projects over the next two years. Staff indicated that requests for facility hardening projects originate from the Facilities Department, executive leadership, or managers of individual facilities. In April 2017, GDC purchased a computerized maintenance management system (CMMS), which is used to track these projects. According to staff, work plans are generated using the data, and these plans are discussed weekly with multiple divisions and the Facilities division leadership staff. CMMS should help make information on facility alterations more readily available for consideration during planning. Contracted Offender Management Contracting for bed space provides GDC flexibility. However, there are risks associated with contracting that should be considered during strategic planning. GDC currently contracts with two private prison companies for the operation of four private prisons and with 22 counties for space in CCIs. In fiscal year 2018, private prisons supervised a daily average offender count of 7,818 while CCIs supervised a daily average offender count of 4,684. Together, these two facility types supervised approximately 25% of all offenders in the GDC prisons system in fiscal year 2018. GDC expects to continue relying on this capacity as part of any strategic planning to manage the offender prison population. However, it is worth noting that changes in GDC's overall prison bed capacity can be (and have been) affected when a contracting entity decides to cease operations. (See Matters for Consideration.) Georgia Department of Corrections 11 Definitions Applied to the Analysis of Bed Space Bed Types Standard (Operational) beds are those beds available within a facility for general offender assignment. Special use beds are for offenders with an identified need. Examples include infirmary beds, isolation or segregation beds and mental health observation beds. These assignments are not permanent. Generally, offenders are not assigned to special use beds upon initial entry into the prison system. Physical beds are the two, standard and special use, bed types combined. Bed Capacity GDC manages capacity differently depending on whether it is managing the facility directly or contracting for the bed space. For purposes of our analysis, we defined capacity as the number of beds available to GDC within a short-time frame (e.g., a few months) . We defined capacity across the different types of prisons as follows. State prison capacity is the total number of beds, standard and special, that are available for use within prisons operated by GDC. Private prison capacity is the number of beds currently under contract and the ones that, contractually, could be activated without additional construction. This activation would result in additional contract costs. CCI capacity is the number of beds currently under contract. Utilization Rates Utilization is the count of offenders in a particular type of bed divided by the total number of those beds. Utilization rates are used as an indication that the population has reached a capacity that will require additional planning to avoid overcrowding. Facility Sub-Groups GDC serves different sub-groups of the prison population based on sex, risk classification and type of services needed in its facilities . For purposes of this report, we selected the following categories as subgroups of facilities in which to analyze the impact of the population increase: GDC Male Facilities 29 facilities that house only male offenders Female Facilities 4 facilities that house only female offenders GDC Medium Facilities 18 facilities, identified by GDC, in which the population is predominately classified as minimum or medium risk level based on a risk classification score Close Facilities - 7 facilities, identified by GDC, in which the population is predominately classified as high risk level based on a risk classification score Special Mission Facilities - 9 facilities classified by GDC as such based on their operations (e.g., medical, mental health, diagnostic) State Prisons 34 prisons owned by the state and operated by GDC Private Prisons 4 prisons owned and operated by private companies, with whom GDC contracts for services County Correctional Institutions (CCIs) 22 facilities owned and operated by county governments, with whom GDC contracts for services Source: GDC records and interviews/written correspondence with GDC and private facility leadership Georgia Department of Corrections 12 Question 2: What is the projected population increase over the next five years and does GDC have sufficient plans to address it? Although the prison offender population growth was slowed because of the state's recent criminal justice reforms, we estimate the prison offender population will increase by approximately 1,277 (2.5%) during fiscal years 20192023.4 As a result of the increase, bed utilization rates for standard (operational) beds in many prison facility types will most likely exceed 97.5%. During periods of above average facility population, utilization rates will likely be higher. For some facility types, the utilization rate may come close to or exceed 100%. Current Bed Capacity and Utilization Rates During fiscal year 2018, the GDC prison system had an available daily average standard bed capacity of 47,802 across 32 state prisons5, 4 private prisons, and 22 county correctional institutions (CCIs). Standard bed capacity in male-only facilities totaled 44,387, while standard bed capacity in female-only facilities totaled 3,415. The average count of available standard beds (vacancies) in male-only facilities was 2,089, while available standard beds in female-only facilities was 148. Across the entire prison system, the utilization rate for available standard beds was 95.3%. However, standard bed capacity and utilization rates varied among facility owners and offender sub-groups. For example, among only state prisons, the utilization rate was 96.5%, with a standard bed capacity of 34,587 and 1,223 bed vacancies. Among these, certain facility types had higher utilization rates. For example, in male-only state facilities for medium risk offenders, the utilization rate was 98.0%, with a total bed capacity of 14,952 and 298 bed vacancies. In the sections that follow, we present estimates to show areas in the prison system where offender population increases will likely create very high bed demand and utilization rates. We applied utilization rate thresholds of 95.0% and 97.5% for standard beds in specific facility sub-groups, such as gender, risk classification, and special mission.6 These utilization rates can be used by management to signify when additional planning to increase bed capacity may be required. We applied the projected offender increase to both the fiscal year 2018 daily average and the day with the highest offender count.7 4 GDC estimates the offender population will grow in total by 1,428. We estimate 151 offenders will be assigned to facilities not included in our analysis, such as transition and probation detention centers based on the current distribution of offenders. 5 Two prisons were excluded from our analysis. Metro state prison opened at the end of fiscal year 2018 and therefore did not have complete data. Helms state prison is a mixed-gender facility and the population could not be separated for our analysis. Helms houses a small population; therefore, it did not impact the results of our analysis. 6 We excluded transition and detention centers from this analysis of state prisons. 7 These projections assume offenders are assigned to facilities throughout the state prison system according to current ratios. If GDC changes its distribution of offenders, the utilization rates would change. In addition, our analysis is based on a fixed inventory of beds because staff indicated GDC does not plan to close any facilities due to maintenance or renovation over the next five years. Finally, the analysis presumes CCI contracted beds will remain at current levels. Should these assumptions change, the utilization rates would change, potentially significantly. Georgia Department of Corrections 13 Projected Bed Capacity and Utilization Rates: Average Offender Population As noted earlier, we estimate the prison offender population will increase by 1,277 (2.5%) during fiscal years 20192023. The increase will most likely result in bed utilization rates for standard beds among many prison facility types to exceed 97.5%. (Appendix D provides a detailed breakdown of projected bed capacity, vacancies, and utilization rates by facility sub-groups using the fiscal year 2018 daily offender average.) As shown in Exhibit 7, using the fiscal year 2018 daily averages as the base for projections, we estimate the state prisons operated by GDC (excludes private prisons and CCIs) will have an overall standard bed utilization rate of 98.9%, with 371 vacancies by 2023. However, utilization rates and corresponding available beds are not identical across facility sub-groups. For example, facilities designated for the largest overall offender population (male medium-risk offenders) will have demand that exceeds current capacity by 76 beds. Female standard facilities will likely have utilization rates near 99.0%, with a very low available bed count of 20. Exhibit 7 After Applying Projected Increases, Most State Prison Types Have Standard Bed Utilization Rates over 97.5% (Based on Daily Average Count) Types of State Prisons STATE PRISONS Male Close Male Medium Female Medium Fem ale Special Mission Male Special Mission Utilization Rates FFYY232M02a3leprSoMjected FFYY182M01a8le SM 75% 80% 85% 90% 95% 97.5% 100% 105% 110% Source: DOAA Analysis Standard Bed Capacity and Vacancies Based on Daily Average Available Beds Based on Capacity Projections (Fiscal Year 2018) (Fiscal Year 2023) State Prisons 34,587 3711 Male Close Risk 7,572 131 Male Medium Risk 14,952 -76 Female Medium Risk 2,012 20 Female Special Mission 1,403 44 Male Special Mission 8,648 251 1 Due to rounding, subgroups add to 370. Source: DOAA Analysis Georgia Department of Corrections 14 Exhibit 8 presents projections for male medium-risk facilities in the state's prisons system and includes beds and offenders in state prisons, private prisons, and CCIs.8 The offender population for these facilities is by far the largest sub-group in the state prison system and accounts for over half (56.3%) of all offenders. The projections indicate that utilization rates in state prisons will go beyond current standard bed capacity (100.5%), while utilization in CCIs will likely reach 99.0%, with approximately 39 available beds. With respect to private prisons, we included all standard bed capacity in this analysis, not just beds currently under contract by GDC. Providers report that approximately 400 standard beds are currently available but not under contract. As a result, the current utilization rate is lower (89.8%) than in state prisons or CCIs, as will be the future utilization rates (92.1%). If we applied a count of only beds currently under contract, utilization rates would be similar: fiscal year 2018 was 98.0% and fiscal year 2023 will likely be 100.5%. Exhibit 8 After Applying Projected Increases, State Male, Medium Risk Prisons Have Standard Bed Utilization Rates over 100%1 (Based on Daily Average Counts) Types of Medium-Risk Prisons MALE MEDIUM-RISK PRISONS State Private County Utilization Rates FYF2Y32C0o2u3nptryojected FFYY182C0o1u8nty 75% 80% 85% 90% 95% 97.5% 100% 105% 110% Source: DOAA Analysis 1 This analysis presents standard beds available in private prisons, not just those currently under contract. Providers report an additional 400 beds are available, but not currently contracted. Using only current contracted bed capacity for private prisons, fiscal year 2018 utilization rates was 98.0% and fiscal year 2023 rates are likely to reach 100.5%. 8 As noted earlier, this analysis assumes the same distribution as the current population; any changes GDC made to that distribution (e.g., sending more offenders to the private prisons before housing them in state prisons) would change the projections. There is an incremental cost associated with adding offenders to the private prison population. The same incremental cost is not associated with adding offenders to the state facilities. Georgia Department of Corrections 15 Standard Bed Capacity and Vacancies Based on Daily Averages Available Beds Based on Capacity Projections (Fiscal Year 2018) (Fiscal Year 2023) Male Medium Risk Prisons 28,167 6271 State 14,952 -76 Private 8,372 663 County 4,843 39 1 Due to rounding, subgroups total 626. Source: DOAA Analysis Projected Bed Capacity and Utilization Rates: Highest Number of Offenders In fiscal year 2018, the average daily offender population was 50,012. The population ranged from 49,174 to 50,520. To illustrate potential projected bed capacity and utilization rates on days with high offender counts, we applied population increase projections to the 50,520 offender population. It should be noted that by applying only a one-day population as a baseline, projections will not reflect the typical supply and demand for standard beds throughout the state prison system. However, the prison bed supply must be able to accommodate this variation of overall and specific subpopulation variation. Appendix E provides a detailed breakdown of projected bed capacity, vacancies, and utilization rates by facility sub-groups using the fiscal year 2018 high offender population. Exhibit 9 shows actual fiscal year 2018 utilization rates and capacity as well as projected utilization rates and vacancies for fiscal year 2023 in state prisons by facility sub-group. Using the fiscal year 2018 high offender count day as the base for projections, we estimate the state prisons operated by GDC (excludes private prisons and CCIs) will have an overall standard bed utilization rate of 99.6%, with 150 vacancies. However, utilization rates and corresponding available beds are not identical across facility subgroups. For example, facilities designated for the largest overall offender population, male medium-risk offenders, would have demand beyond current capacity by approximately 48 beds. Female special mission facilities would be near capacity at 99.3%, with a very low available bed count of 11. Georgia Department of Corrections 16 Exhibit 9 Projections Based on Highest Day Highlight Potential Variations in Capacity Issues Types of State Prisons STATE PRISONS Male Close Male Medium Female Medium Female Special Mission Male Special Mission Utilization Rates FFY2Y32M02a3leprSoMjected FFYY1820M1a8le SM 85% 90% 95% 97.5% 100% 105% 110% Source: DOAA Analysis Standard Bed Capacity and Vacancies Based on Highest Day Available Beds Based on Capacity Projections (Fiscal Year 2018) (Fiscal Year 2023) State Prisons 34,879 150 Male Close 7,611 76 Male Medium 15,096 -48 Male Special Mission 8,720 73 Female Medium 2,012 38 Female Special Mission 1,440 11 Source: DOAA Analysis GDC Plan During the course of this review, GDC officials indicated that any permanent significant increase in bed space capacity in the next five years would likely come from expanding state prison capacity. Expansion could include adding beds in current facilities, as well as renovating current facilities or reopening currently closed facilities. In September 2018, GDC officials submitted a plan to renovate Wayne State Prison (which would increase bed capacity by 200 beds) to the Governor's Office of Planning and Budget. Officials indicated there were currently no other expansion plans. We shared our analysis with GDC officials. In response, they provided the information in Exhibit 10 as the potential facility expansions that could occur if deemed necessary and funding were available. Georgia Department of Corrections 17 Exhibit 10 Potential GDC Facility Expansion Plan Facility Name Rogers State Prison Metro Reentry Prison Wayne State Prison1 Western Probation Detention Center Hays State Prison Valdosta State Prison # Beds 48 480 192 180 Sex Male Male Male Female 400 Male 200 Male Security Level Medium Medium Medium Medium Close Medium 1 A design budget was submitted to OPB in September 2018. Source: GDC Current Status Operating Operating Non-Operating Non-Operating Projected Cost to Increase Not provided $13 million $15 million $3.5 million Operating Operating Not provided $1 million GDC officials indicate that GDC expects to continue utilizing private prisons and CCIs as it has previously, for a specific type of lower-risk male offender. However, the agency has no specific plans to address any increases in offenders by expanding the bed capacity in private prisons. Officials noted that usage of private prisons and county correctional institutions is limited by the availability of appropriated funds for contract reimbursement payments; however, if additional funds were available, GDC would consider increasing usage. RECOMMENDATION: 1. GDC should consider differentiating facility utilization rates by bed type (standard or special use) and evaluating offenders by distinct sub-populations such as gender, special need, and offender risk to improve future bed demand planning. Georgia Department of Corrections 18 Question 3: Are private prison beds a more effective reform and cost alternative for the state? We sought to determine whether private prisons offer a more effective option by comparing the costs of operation for the various types of facilities in which GDC manages its offenders. In the process, we identified county correctional institutions as an additional option. While this analysis focused on comparable costs to assess effectiveness, it did not assess output or outcome measures.9 A comprehensive performance assessment would require a detailed comparison of output and outcome measures in addition to cost to supervise offenders. GDC currently has three different prison types in which it manages state offenders: state prisons, private prisons, and county correctional institutions (CCIs). The average cost to manage offenders in private prisons is less than the average cost in state facilities; however, when controlled for offender sex, facility size, and risk classification, the state prison costs are comparable (see Exhibit 11). CCIs are substantially less expensive than either state or private facilities. According to GDC, it assigns lower risk male offenders to private prisons and CCIs. The four private prisons house between 1,100 to 2,650 offenders each. In fiscal year 2017, the 23 CCIs housed between 100 and 530 offenders each. To compare costs with private prisons, we identified a group of five comparably sized state prisons that includes male-only, medium risk facilities. As shown in Exhibit 11, in fiscal year 2017, the overall cost to manage per offender/per day was $65.58 for state prisons. However, when compared to comparable facilities, the daily cost was $44.56 per offender, which is similar to the private prisons' daily cost of $49.07. While state and private prisons were similar, during the same time, the CCI's cost per offender per day was substantially lower at $21.63. Exhibit 11 State and Private Prisons' Costs per Day are Comparable, CCIs are Significantly Less Fiscal Year 2017 # of Avg. Daily GDC Cost Offender Offender Facility Type Facilities State Prisons all1 33 State Prisons comparables2 5 Offender Count 38,146 8,252 (millions) $913.1 $134.2 Cost Per Year Cost Per Day $23,937 $65.58 $16,266 $44.56 Private Prisons CCIs3 4 7,897 $141.4 $17,909 $49.07 23 4,975 $39.3 $7,894 $21.63 1 During fiscal year 2017, there were 33 facilities. At the end of fiscal year 2018, the Metro State Prison opened. 2 Includes male-only, medium risk facilities. 3 At the end of fiscal year 2017, two CCIs closed. A new one opened in fiscal year 2018, bringing the current number to 22. Source: DOAA analysis of GDC data The offender management costs are presented by facility type and discussed in more detail in the following sections. 9 The analysis did not assess outputs such as programming, health care, facility condition or outcomes, such as offender recidivism. Georgia Department of Corrections 19 State Prisons GDC estimated a total cost of $913.1 million to manage an average daily offender count of 38,146 in 33 state prisons during fiscal year 2017. The total cost was composed of $617.6 million in direct costs, which included operating costs of the facilities (e.g., personal services, regular operating expenses), and $295.5 of indirect and overhead costs, which included administrative, offender management and health related costs allocated to the facilities by GDC based on offender population. The per offender per day cost among state prisons ranged widely (from a low of $38.34 to a high of $490.60). The cost varied greatly depending on facility classification. As shown in Exhibit 12, facilities classified as "special mission" incurred a higher cost per offender per day than those classified as close risk (the highest security classification) or medium risk. Of the 18 facilities classified as medium risk, we identified five comparable to private prisons because of similar offender types and population size. Exhibit 12 GDC Expenses Vary By Facility Classification, Fiscal Year 2017 Facility Classification Special Mission Close Risk Medium Risk1 Totals2 Averages 1 Excludes Metro Reentry Facility. 2 Totals may be off due to rounding. Source: DOAA analysis of GDC data # Facilities 8 7 18 33 -- Avg. Daily Offender Count 10,581 9,390 18,175 38,146 -- GDC Cost (millions) $363.9 $210.3 $338.8 $913.1 -- Cost Per Offender/ Per Year $34,396 $22,397 $18,643 -$23,937 Cost Per Offender/ Per Day $94.23 $61.36 $51.08 -$65.58 Private Prisons GDC estimated a total cost of $141.4 million to house an average daily population of 7,897 in four private prisons in fiscal year 2017. The total cost included $135.8 million in direct costs, which are specified in the contract, and $4.5 million in indirect and overhead costs, which GDC assigns to the facilities based on offender population. As shown in Exhibit 13, the cost per day per offender does not vary among private prisons as much as state prisons. Because direct costs are derived using a per offender per day reimbursement rate, these costs remain fairly static.10 Further, the population within private prisons is more homogenous than those supervised in state prisons. GDC currently restricts offender assignments of only male offenders with a lower risk classification (minimum/medium) and fewer health risks. In fiscal year 2017, it ranged from a high of $54.17 to a low of $46.57. 10 Contracts may contain a "tiered" payment structure, whereby the per offender per day rate is determined by the number of offenders assigned to the facility. For example, one per offender per day rate may apply to each offender up to an agreed-upon population (e.g., 1,000 offenders), then a different per offender per day rate may apply thereafter (e.g., for the next 500 offenders). Typically, the per offender per day rate decreases after the threshold number is assigned. Georgia Department of Corrections 20 Exhibit 13 Private Prison Offender Management Costs GDC Approximately $50 Per Offender Per Day (Fiscal Year 2017) Facility 1 2 3 4 Totals Averages Avg. Daily Offender Count 2,655 2,616 1,485 1,141 7,897 -- GDC Cost (millions) $45.1 $44.6 $29.1 $22.6 $141.4 -- Cost Per Offender/ Per Year $17,000 $17,050 $19,617 $19,771 -$17,909 Cost Per Offender/ Per Day $46.57 $46.71 $53.75 $54.17 -$49.07 Source: DOAA analysis of GDC data It is worth noting that when GDC has contracted with private vendors for offender management, the vendor has constructed and owned the prison facility. We attempted to isolate costs for only the supervision of the offenders within private facilities. However, this data was unavailable. County Correctional Institutions In fiscal year 2017, GDC used 23 CCIs to house an average daily population of 4,975. It incurred costs of approximately $39.3 million, of which $38 million were direct costs and $1.3 million were in allocated indirect costs. The direct costs are based on contract rates and terms. As with private prisons, GDC currently restricts offender assignments of only male offenders with a lower risk classification (minimum/medium) and fewer health risks. Further, because offenders are often used as labor for county governments, offenders with job skills (e.g., heavy equipment operating skills) and the ability to conduct physical labor are typically assigned to CCIs. GDC officials indicated that the number of offenders that fit these criteria is limited. Despite this constraint, it is important to note that GDC relied on the capacity of the existing CCI facilities to supervise approximately 10.1% of the fiscal year 2018 daily average male prison population. Because contract reimbursement rates are identical among all CCIs, the per offender per day costs do not vary between units. GDC contracts with counties with a reimbursement rate of $20 per offender per day, less than half the average per offender per day reimbursement rates contracted with private prisons. GDC officials indicate that this rate has remained unchanged since 1999. We interviewed two CCI wardens to discuss operations. Both asserted that the reimbursement rate does not cover the cost to house and manage an offender. According to one warden, the reimbursement rate covers less than half of the average offender management cost. The withdrawal of two counties from the CCI program last year may be further evidence of the financial strain related to the CCI reimbursement rate. We address the reimbursement rate of CCIs in the following section entitled Matters for Consideration. Georgia Department of Corrections 21 Matters for Consideration Further review is needed to determine whether the type and size of the population served by private prisons could be changed. If GDC chose to, it could change the population it currently assigns to private prisons. However, it would have to first ensure the risks associated with increasing the type or number of offenders housed in private facilities is fully evaluated. These two points are discussed below. Offender Type: GDC has elected to restrict the offender sub-group it assigns to private prisons. Officials indicated that private prisons are only assigned male offenders with a lower-risk classification (minimum/medium) who do not have significant special needs. This decision to limit the criteria could be revisited. Private prison representatives indicated that they could provide services to other types of offenders beyond those currently housed. According to these officials, private prisons are capable of supervising women and/or higher-risk offenders. GDC would need to identify and assess the risks of changing the offender type. For example, it would need to assess the risk of housing higher risk offenders or those who are not healthy. It should also consider available research on this topic. Number of Offenders: In addition to the type of offender, the size of the population served by private prisons has been determined by GDC. In fiscal year 2018, the average daily offender population supervised by private prisons was 7,818, which is 15.6% of all GDC prison offenders. Based on our analysis, this group is a subset of the total offender population that meets the criteria that would make them eligible for private placement. GDC's offender management system does not identify offenders as "private eligible"; however, GDC currently has approximately 18,000 offenders housed in state operated male, medium risk level facilities; a percentage of these offenders would likely meet current criteria for private prison placement. Additional funds would be required to increase the number of contracted beds. Current contracts provide for additional expansion above the number of beds currently under contract. Contract clauses also allow for physical expansion of the facilities. Additionally, the state could decide to contract for additional private prison facilities. A brief review of eight other states11 found that some outsourced up to 27% of their populations to private prisons. 11 These eight states were identified as comparable because they contract for private bed space and have a prison population similar in size to Georgia's. Georgia Department of Corrections 22 Before GDC decides to increase its reliance on contracted bed space, it should assess the potential risk thoroughly and have considered contingencies. Before deciding to increase its use of contract beds, GDC should consider bed space capacity risks associated with outsourcing. Several episodes have resulted in material reduction in overall GDC prison bed capacity in the past. A sudden reduction in bed space capacity that occurs from external parties choosing to discontinue partnership with GDC may result in additional strain on the remaining bed space capacity throughout the prison system. In fiscal year 2018, private prisons (15.6%) and county correctional institutions (9.4%) supervised approximately 25.0% of the daily average prison population housed in standard beds. According to staff, GDC expects to contract for bed space as part of any strategic planning to manage the offender prison population. At the end of fiscal year 2017, Troup and Thomas Counties declined to renew their fiscal year 2018 contract with GDC for CCI operations. Together, these CCIs had provided a total of 459 beds. This change resulted in an immediate reduction of approximately 1% of total bed space capacity in the prison system and approximately 9% of the bed space capacity in the CCIs. In 2010, GDC was required to relocate offenders assigned to the D. Ray James private prison after the company discontinued its contract with the state in order to pursue contracting with the federal government. The facility housed approximately 1,800 offenders. In response to the closure, GDC assigned a portion of the offenders to two existing private prisons, which required a temporary surge in bed capacity. GDC then issued a request for proposal to construct a new private prison facility. It should be noted that two of the contracts (covering three facilities) allow GDC the option to purchase the facility, should the contract be terminated, according to terms established in the contracts. Consideration should be given to reviewing the reimbursement rate for County Correctional Institutions (CCIs) . The reimbursement rate for CCIs is currently $20 per offender per day and has not increased since 1999. This rate is less than half the daily rate paid to private prisons and $10 less than the rate paid to county jails to hold offenders after sentencing.12 Unlike the county jail subsidies, which are appropriated as a distinct program in the GDC appropriation, CCI reimbursement funds are appropriated within GDC's general offender management program along with state prisons. During our review, we identified two CCIs that recently ceased contracting with the state. GDC staff indicated the low reimbursement rate was a factor in their decisions. The number of CCIs has apparently declined over time as well. According to a longserving CCI warden, there were 57 CCIs operating in 1969; that number declined to 31 in 1992, and there are currently 22 operating. 12 According to a 2005 special examination, the county jail subsidy rate of $20 per day per offender was substantially less than the cost to manage offenders. The estimated county jail costs, by facility, were estimated to be between $34.98 and $41.51 per day. In July 2014, the county jail subsidy rate was increased to $30 per offender per day. Georgia Department of Corrections 23 Georgia has relied on a network of CCIs to house offenders for decades. O.C.G.A. 425-53 establishes the CCIs and permits GDC to contract directly with county governments for CCI operations. These contracts are renewed annually and establish all terms, including the maximum number of daily offenders the CCI will oversee and the daily compensation for offender supervision. According to staff, offenders assigned to CCIs are male offenders with a lower risk classification (minimum/medium) with no special needs. These offenders often provide labor for county governments and therefore must be able to work and be supervised outside of the facility. GDC currently contracts with 22 CCIs (from 22 counties) . As noted above, during fiscal year 2017, Troup and Thomas counties opted not to renew their contracts. In fiscal year 2018, the CCIs supervised an average daily offender count of 4,684, or 9.4% of the total GDC offender prison population. Housing offenders in CCIs remains a cost-effective option over state and private prisons. When the two CCIs ceased operations, approximately 425 offenders had to be relocated. Had all of them been relocated to a comparable state facility, it would have cost $44.56 per offender per day, which is over twice the cost of housing them in the CCIs. It should be noted that, because of the criteria listed above, and the need to maintain balance in the private and state prisons, the actual number of offenders who could move to CCIs is not currently known. GDC would need to determine the actual number of offenders who could be assigned to CCIs. Georgia Department of Corrections 24 Appendix A: Table of Recommendations Question 1: Does GDC have a strategic plan that considers all necessary factors? None Question 2: What is the projected population increase over the next five years and does GDC have sufficient plans for how to address it? 1. GDC should consider differentiating facility utilization rates by bed type (standard or special use) and evaluating offenders by distinct sub-populations such as gender, special need, and offender risk to improve future bed demand planning. Question 3: Are private prison beds are a more effective reform and cost alternative for the state? None Georgia Department of Corrections 25 Appendix B: Objectives, Scope, and Methodology Objectives This report examines the Georgia Department of Corrections (GDC). Specifically, it examines strategic planning related to prison bed space supply and demand, the role of private prisons in addressing identified demand, and considerations of private prisons as a cost alternative to public prisons. Our examination set out to answer the following questions: 1. What is GDC's strategic plan as it relates to demand for prison facilities and bed space and what role will private prisons play? 2. Are private prison beds a more effective reform and cost alternative for the state? Scope This special examination generally covered strategic planning activity related to bed space supply and demand and the role of private prisons in meeting that demand for GDC. Program activity for fiscal years 2016, 2017, and 2018 was examined with consideration of earlier or later periods when relevant. Information used in this report was obtained through: Interviews conducted with GDC leadership as well as staff within the state, private, and county correctional prisons funded by GDC, and statistical consultants advising GDC. Site visits at state and private prison facilities. Analysis of bed utilization data for fiscal years 2016, 2017, and 2018 from GDC's offender management data system, SCRIBE. Review of federal and state laws, GDC's policies and procedures, and program documentation. Review of literature, reports, and research discussing corrections bed utilization planning and measurement. The examination reviewed data from GDC's SCRIBE system, a database used to house information on offenders sentenced to prison and probation. Key data used to support the findings include prison facility bed space capacity and vacancy data used to calculate utilization rates sub-groups for both standard and special use beds. Basic data reliability testing was completed to assess accuracy and completeness of the data, and we concluded the data was sufficiently reliable for our purposes. Methodology To determine what GDC's strategic plan is as it relates to demand for prison facilities and bed space and what role private prisons will play, we interviewed GDC staff about their current strategic planning procedures and reviewed documents, projected demand for bed space, and plans to meet this demand including anticipated role of private prisons. We also interviewed representatives from the consulting firm that routinely calculates 5-year offender population projections used by GDC and other corrections agencies comprising the Offender Population Committee to understand: the methodology used to develop population projections; the accuracy of the projections; and, the sub-populations examined. We communicated with representatives of corporations operating private prisons to quantify additional bed Georgia Department of Corrections 26 capacity that existed beyond capacity currently contracted with GDC. We also interviewed two wardens of county correctional institutions to gain an understanding of how these facilities are used. Site visits were also conducted at one private prison facility (Riverbend Correctional Facility) and one close security state prison (Hancock State Prison) to better understand bed prison space utilization in practice (e.g. various configurations of cells and cell blocks, difference between standard and special use beds, etc.). We visited Georgia Diagnostic & Classification Prison, the initial facility encountered by male offenders entering the prison system, to better understand the complexities of prisoner diagnostics and classification for the purpose of placement. In order to familiarize ourselves with GDC's operations, facilities, and strategic plan methodology, various agency documents were reviewed including: GDC's strategic plan for 2013-2016 and a more recent 2019-2022 strategic plan submitted to OPB. Annual population projections produced by the contractor and corresponding testament of accuracy provided by the company. Private prison and county correctional institution contracts. PowerPoint presentations from Offender Population Committee meetings occurring August 2017-September 2018. Reports used by Offender Administration leadership for operational planning related to bed space utilization. Inventories and descriptions of prison facilities. GDC's most recent facilities master plan completed in 2007. A report of bed utilization terminology/definitions generated from the agency's data system. In order to determine current and future demand for prison space, we requested data from GDC's SCRIBE system, a database that houses information on offenders sentenced to prison and probation. We collected and examined daily offender population data for fiscal years 2016, 2017, and 2018 for state prisons, private prisons, and county correctional institutions. The data contained an aggregate offender count for each day of the year by facility broken down by bed type (standard and special use) with associated utilization rates for each bed type that was hard coded (i.e. not a calculated field). Basic data reliability testing for accuracy testing was completed to assess accuracy and completeness of the data, and we concluded the data was sufficiently reliable for our purposes. We recalculated hard coded utilization rates with existing data fields to confirm accuracy and then estimated the proportion of bed space in facility subgroups (males/female, medium/close security, special mission). Combinations of subgroups were also considered (e.g. male/medium/special mission). Facility subgroup estimates were calculated as follows: Gender: Given that all but one small facility (Helms Medical Facility--71 offender bed capacity) are gender specific (only house males or females), all beds in female facilities were considered "female" beds and all beds in male facilities were considered "male" beds. Helms facility was excluded from the analysis due to inability to differentiate proportion of male/female offenders within the facility with the existing facility-level data set. This allowed us to Georgia Department of Corrections 27 approximate an exact proportion of male/female offender beds for this subgroup. Security Level: Determining the precise proportion of offenders with a particular security level (i.e. minimum/medium or close) was not possible with our facility level data set, nevertheless estimates were generated using facility security level as a proxy for offender security level. Factors limiting the precision of this subgroup estimate are indicated below. Risk classification dispersion throughout the prison system The risk classification of a facility denotes the classification of the majority of the population within, however, facilities can technically house most offenders. GDC strives to house lower risk offenders in minimum or medium level facilitates, yet approximately 10-40% of offenders in closed facilities were noted to be minimum or medium risk offenders according to an offender count obtained for June 30, 2018, which undermined our ability to estimate security level of offenders within with precision. Likewise, the percentage of close (or high risk) offenders in lower security facilities can approximate 5%. We estimate the number of offender beds at risk for miscategorization due to our methods to be approximately 9% of all beds represented in the data set. Facility Security Classification of "Special Mission" Further complicating matters was the fact that increased flexibility exists in placement of offenders into special mission facilities (e.g., the need for the programming provided at a particular facility can be the primary driver). This makes facility security level an even less accurate predictor of offender security level for special mission facilities. For instance, Phillips State Prison is a close security special mission facility, however, 73.1% of offenders placed within are minimum or medium security level. As a result, special mission facilities were excluded from the analysis when considering bed space availability by security level classification. Special Mission: Facilities that serve specific populations or offender needs are categorized as "special mission" prisons by GDC. For instance, Baldwin hold offenders with severe mental health issues and Burruss houses juveniles. All beds within special mission facilities comprised this sub-population for our analysis, providing an exact count (not an estimate). The proportion of offender beds in each sub-population above (and combination of sub-populations, e.g. male min/medium risk/non-special mission) were then applied to the aggregate 5-year population projection developed by the contractor in order to project what the offender bed count would be in 2023 if prison facility capacity remained constant. A comparison of projected offender count to GDC prison capacity for each of the various sub-groups resulted in a projected utilization rate in 2023. We applied a 95% threshold to projected utilization rates calculated as the threshold GDC should not exceed for safe and efficient operation, based on standards indicated in GDC's most recent 10-year facility master plan document. Preliminary results from the analysis of current and future bed space availability were shared with agency leadership, with a request for indication of how the agency intends Georgia Department of Corrections 28 to absorb the projected incoming population (according to projections). GDC leadership provided a proposed list of four currently operational facilities it intended to renovate in order to bring additional beds online, and two non-operational facilities it intended to renovate and re-open, with a corresponding count of beds to be added and associated cost indicated by facility. To determine whether private prison beds are a more effective reform and cost alternative for the state we reviewed and analyzed information submitted by GDC regarding costs. While this analysis focused on comparable costs to assess effectiveness, it did not assess output or outcome measures.13 A comprehensive performance assessment would require a detailed comparison of output and outcome measures in addition to cost to supervise offenders. Academic efforts have been undertaken for years to compare cost and performance between private and public prisons, but efforts to complete comprehensive, cost-effective analyses have proven difficult to conduct accurately. GDC submitted a detailed cost allocation analysis for state prisons, private prisons, and county correctional institutions for fiscal year 2017. Fiscal year 2017 was the most recent year for which the analysis was available. The cost allocation included direct and indirect costs. We reviewed the content for accuracy and completeness and deemed it reliable for use. Direct costs for state prisons included operating expenses directly attributable to prisons. Indirect costs included agency administrative and offender management unit expenses, offender health costs, and the food and farm program. Direct costs for private prisons and county correctional institutions consisted of offender management reimbursement payments established in contracts. Indirect costs allocated to private prisons and county correctional institutions included items such as agency administrative and offender management unit expenses. Costs were allocated proportionally according to the offender population assigned to these facility types. In addition, directly attributable offender health costs were assigned to private prisons and county correctional institutions. Costs were analyzed to establish comparison between the sectors by establishing a per unit metric and by limiting comparison based on facility attributes. GDC established a per offender per day (and per year) metric which we relied on to compare costs between the facilities. In addition, we limited direct cost comparison of state prisons and private prisons to those with similar offender sex (male only), risk classification (minimum/medium), and mission (non-special mission). It is worth noting that representatives from the private prisons noted that facilities must gain/maintain accreditation from the American Correctional Association (ACA) and comply with contract requirements. In addition, one representative submitted a list of additional quality/compliance work that is conducted within facilities. These include inspections, annual audits, and internal facility audits. 13 The analysis did not assess outputs such as programming, health care, facility condition or outcomes, such as offender recidivism. Georgia Department of Corrections 29 This special examination was not conducted in accordance with generally accepted government auditing standards (GAGAS) given the timeframe in which the report was needed. However, it was conducted in accordance with Performance Audit Division policies and procedures for non-GAGAS engagements. These policies and procedures require that we plan and perform the engagement to obtain sufficient, appropriate evidence to provide a reasonable basis for the information reported and that data limitations be identified for the reader. Georgia Department of Corrections 30 Appendix C: Inventory of State Prisons, Private Prisons, and County Correctional Institutions Operator Facility Gender Risk Level FY18 Average Standard Bed Count FY18 Average Special Use Bed Count FY18 Average Physical Bed Count Arrendale State Prison Female Special Mission 1,325 95 1,421 Emanuel Women's Facility Female Medium 393 3 396 Pulaski State Prison Female Medium 1,123 41 1,164 Whitworth Women's Facility Female Medium 426 5 431 Helms Medical Facility Mix Medium 31 0 31 Hancock State Prison Male Close 813 255 1,068 Hays State Prison Male Close 699 343 1,042 Macon State Prison Male Close 1,389 289 1,679 Smith State Prison Male Close 1,200 288 1,488 Telfair State Prison Male Close 1,106 242 1,348 Valdosta State Prison Male Close 778 269 1,047 Ware State Prison Male Close 1,269 211 1,480 Autry State Prison Male Medium 1,559 139 1,698 Calhoun State Prison Male Medium 1,438 170 1,608 Central State Prison Male Medium 1,074 24 1,098 Dodge State Prison Male Medium 1,175 45 1,219 State Dooly State Prison Male Medium 1,590 56 1,645 Johnson State Prison Male Medium 1,352 159 1,511 Lee State Prison Male Medium 717 11 727 Long Unit Male Medium 203 2 205 Montgomery State Prison Male Medium 387 16 402 Rogers State Prison Male Medium 1,286 86 1,372 Rutledge State Prison Male Medium 589 25 614 Walker State Prison Male Medium 403 5 408 Washington State Prison Male Medium 1,223 123 1,346 Wilcox State Prison Male Medium 1,659 124 1,783 Augusta State Medical Prison Male Special Mission 1,031 117 1,147 Baldwin State Prison Male Special Mission 854 95 949 Burruss Correctional Training Center Male Special Mission 711 11 722 Coastal State Prison Male Special Mission 1,698 70 1,768 GA Diagnostic & Classification Prison Male Special Mission 2,103 319 2,422 GA State Prison Male Special Mission 1,019 419 1,437 Metro Reentry Facility14 Male Special Mission 114 0 115 Phillips State Prison Male Special Mission 773 89 862 14 Metro Reentry Facility was not included in our bed utilization analysis because it recently re-opened. Its inclusion would have skewed the analysis due to its partial data set for FY18. Georgia Department of Corrections Appendix C: Continued Private County ADOC Bulloch CCI Carroll CCI Clark CCI Clayton CCI Colquitt CCI Coweta CCI Decatur CCI Effingham CCI Floyd CCI Gwinnett CCI Hall CCI Harris CCI Jackson CCI Jefferson CCI Mitchell CCI Muscogee CCI Richmond CCI Screven CCI Spalding CCI Sumter CCI Terrell CCI Coffee Correctional Facility Jenkins Correctional Facility Riverbend Correctional Facility Wheeler Correctional Facility Source: GDC records Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Male Medium Medium Medium Medium Medium Medium Medium Medium Medium Medium Medium Medium Medium Medium Medium Medium Medium Medium Medium Medium Medium Medium Medium Medium Medium Medium 11 156 240 123 241 187 214 191 186 425 211 189 148 146 140 136 520 227 144 372 344 137 2,499 1,079 1,423 2,516 31 0 11 0 156 0 240 0 123 0 241 0 187 0 214 0 191 0 186 0 425 0 211 0 189 0 148 0 146 0 140 0 136 0 520 0 227 0 144 0 372 0 344 0 137 106 2,605 56 1,135 58 1,481 81 2,597 Georgia Department of Corrections Appendix D: Population Projection Based on FY 2018 Average Daily Population 2018 5 yr 2023 Facility Types All Male Female Capacity Vacancies Utilization % Projected Growth Vacancies Utilization % O S P O S P O S P O S P O S P O S P 47,802 5,453 53,255 2,237 1,007 3,243 95.3% 81.6% 93.9% 1,163 114 1,277 1,073 44,387 5,292 49,678 2,089 990 3,079 95.3% 81.3% 93.8% 1,080 110 1,190 1,009 3,415 161 3,576 148 17 164 95.7% 89.6% 95.4% 83 4 87 64 893 1,966 880 1,889 13 77 97.8% 83.6% 96.3% 97.7% 83.4% 96.2% 98.1% 91.9% 97.8% Special Mission Male Female 10,051 1,472 11,523 538 257 795 94.6% 82.6% 93.1% 243 31 274 295 226 521 97.1% 84.7% 95.5% 8,648 1,367 10,015 460 247 707 94.7% 81.9% 92.9% 209 29 238 251 218 470 97.1% 84.0% 95.3% 1,403 105 1,508 78 10 88 94.4% 90.7% 94.2% 34 2 36 44 7 51 96.9% 93.0% 96.6% Male - Close Risk 7,572 2,187 9,758 316 290 606 95.8% 86.7% 93.8% 185 48 234 131 241 373 98.3% 89.0% 96.2% Male - Min/Med Risk 28,167 1,738 29,905 1,313 453 1,765 95.3% 74.0% 94.1% 686 33 719 627 420 1,047 97.8% 75.8% 96.5% State Prisons 14,952 1,198 16,150 298 214 513 98.0% 82.2% 96.8% 374 25 399 -76 189 113 100.5% 84.2% 99.3% 1 Private Prisons 8,372 540 8,912 855 239 1,094 89.8% 55.8% 87.7% 192 8 200 663 231 894 92.1% 57.3% 90.0% 2 CCIs 4,843 0 4,843 159 0 159 96.7% N/A 96.7% 120 0 120 39 0 39 99.2% N/A 99.2% Female - Med Risk 2,012 Legend O - Operational Beds S - Special Use Beds P - Physical Beds 56 2,068 70 7 77 96.5% 87.6% 96.3% 50 1 51 20 6 26 99.0% 89.8% 98.8% Notes 1 - Capacity reflects actual physical capacity, not current contracted capacity. Acces sing additional capacity (over contracted capacity) would not require construction; however, it would incur additional cos ts . 2 - Capacity reflects contracted capacity. It does not reflect any additional capacity that may be available, as this level of detail was not readily available. Source: DOAA Analysis of GDC Data 32 Georgia Department of Corrections Appendix E: Population Projection Based on FY 2018 Highest Offender Count Day Date of Occurrence: 5/24/18 Facility Types All Male Female 2018 5 yr 2023 Capacity Vacancies Utilization % Projected Growth Vacancies O S P O S P O S P O S P O S P 48,112 5,394 53,506 2,018 968 2,986 95.8% 82.1% 94.4% 1,165 112 1,277 853 856 1,709 44,660 5,240 49,900 1,885 966 2,851 95.8% 81.6% 94.3% 1,081 108 1,189 804 858 1,662 3,452 154 3,606 133 2 135 96.1% 98.7% 96.3% 84 4 88 49 -2 47 Utilization % O S P 98.2% 84.1% 96.8% 98.2% 83.6% 96.7% 98.6% 101.2% 98.7% Special Mission Male Female 10,160 1,467 11,627 332 275 607 96.7% 81.3% 94.8% 256 23 279 76 252 328 8,720 1,369 10,089 286 275 561 96.7% 79.9% 94.4% 213 28 241 73 247 320 1,440 98 1,538 46 0 46 96.8% 100.0% 97.0% 35 2 38 11 -2 8 96.2% 82.9% 95.0% 99.2% 81.9% 96.8% 99.3% 102.5% 99.5% Male - Close Risk 7,611 2,140 9,751 262 333 595 96.6% 84.4% 93.9% 186 46 231 76 287 364 99.0% 86.6% 96.3% Male - Min/Med Risk State Prisons 1 Private Prisons CCIs2 28,329 1,731 30,060 1,337 358 1,695 95.3% 79.3% 94.4% 682 35 717 655 323 978 97.7% 81.3% 96.7% 15,096 1,191 16,287 325 148 473 97.8% 87.6% 97.1% 373 26 400 -48 122 73 100.3% 89.8% 99.6% 8,372 540 8,912 968 210 1,178 88.4% 61.1% 86.8% 187 8 195 781 202 983 90.7% 62.7% 89.0% 4,861 0 4,861 44 0 44 99.1% N/A 99.1% 122 0 122 -78 0 -78 101.6% N/A 101.6% Female - Med Risk Legend O - Operational Beds S - Special Use Beds P - Physical Beds 2,012 56 2,068 87 2 89 95.7% 96.4% 95.7% 49 1 50 38 1 39 98.1% 98.9% 98.1% Notes 1 - Capacity reflects actual physical capacity, not current contracted capacity. Acces s ing additional capacity (over contracted capacity) would not require cons truction; however, it would incur additional cos ts . 2 - Capacity reflects contracted capacity. It does not reflect any additional capacity that may be available, as this level of detail was not readily available. Source: DOAA Analys is of GDC Data 33 The Performance Audit Division was established in 1971 to conduct in-depth reviews of state-funded programs. Our reviews determine if programs are meeting goals and objectives; measure program results and effectiveness; identify alternate methods to meet goals; evaluate efficiency of resource allocation; assess compliance with laws and regulations; and provide credible management information to decision makers. For more information, contact us at (404)656-2180 or visit our website at www.audits.ga.gov.