1Sonia Tetlow conducted this study while a student at Rollins School of Public Health, Emory University. She is currently an Oak Ridge Institute for Science and Education (ORISE) Research Fellow at the Centers for Disease Control and Prevention, Atlanta, USA
Received Date: June 02, 2016; Accepted Date: July 07, 2016; Published Date: July 14, 2016
Citation: Tetlow S, Phillips V (2016) The Cost of Incarceration: A Descriptive Analysis of Low-income Mentally Ill Misdemeanor Offenders in DeKalb County, Georgia (2012). J Community Med Health Educ 6:450. doi:10.4172/2161-0711.1000450
Copyright: © 2016 Tetlow S, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Visit for more related articles at Journal of Community Medicine & Health Education
Objectives: We collected the jail housing unit histories and calculated the costs of incarcerating mentally ill misdemeanor offenders who were arrested in DeKalb County, Georgia in 2012 and represented by the Public Defender’s Office.
Methods: We obtained records from the Sheriff’s and Public Defender’s Offices for the 2012 cohort. Days of detainment in and movement among jail housing and forensic hospital units were determined for those with moderate and severe illness. The costs of incarcerating offenders with mental illness, relative to those without, were calculated.
Results: Offenders with severe illness spent an average of 64 days (SD=129) in jail compared to 39 days (SD=65) for those with moderate illness (p<0.05). DeKalb County spent 26 times more per offender for those with severe illness compared to those with none, 89% of which was spent on specialized mental health units.
Conclusions: Mentally ill offenders spend a disproportionate amount of time in jail relative to those with no mental illness. Duration and cost vary with degree of illness. Diverting these offenders, particularly severe cases, into treatment could improve health outcomes and reduce criminal justice costs.
Jail; Incarceration; Mentally ill; Cost analysis
SPMI: Severe and Persistently Mentally Ill; MMI: Moderately Mentally Ill; JMS: Jail Management System
The link between incarceration and mental illness is well established. People with mental health conditions are arrested relatively frequently and housed in jails for significant periods of time in relation to minor, non-violent offenses, compared to those with no mental illness [1-5]. As a result, they impose substantial costs on the judicial system .
Numerous factors contribute to the relatively high rates of arrest among the mentally ill. Mental illness can manifest in behaviors that themselves can be perceived as criminal offenses, for example disorderly conduct [4,7], and result in relatively lower socio-economic status, including higher rates of unemployment and homelessness [8,9]. Despite higher rates of arrest, the majority of mentally ill offenders in the jail commit minor crimes, with only an estimated 27% having committed violent offenses [3,8].
Studies have shown an association between mental health treatment and reduced rates of arrest and recidivism [10-18]. However, many Americans with mental health conditions who are at a higher risk of arrest, due to low income for example, go untreated . Barriers to treatment include: cost, lack of insurance, inadequate availability of appropriate services, and symptomology that can impair a person’s ability to navigate complex health care systems [20-22].
Prior research has examined the relationships between mental illness, treatment, and costs to the mental health and criminal justice systems. Evidence on the cost-effectiveness of interventions aimed at mentally ill offenders has been mixed. Several studies have shown that effective treatment is associated with lower legal system costs [3,23]. Studies have also shown that jail diversion programs which send mentally ill offenders into treatment resulted in reductions in the durations of incarceration and number of arrests [24,25] and were cost-saving or cost neutral . One study found that use of police crisis intervention teams as a pre-booking diversion strategy resulted in annual net savings . Other analyses, however, have found that diversion programs increase overall costs, primarily due to the added costs of mental health treatment .
For the majority of these studies, researchers used a single estimate for the cost/per day of jail housing and did not take into account the higher cost of jail mental health units. As such, they may underestimate the cost of incarcerating mentally ill offenders and the costs and benefits of diversion programs. There were two exceptions to this rule. Domino et al., included a higher estimate for the cost of jail psychiatric units for detainees that were assessed positively for a mental illness, although this was not adjusted for severity . Cowell, et al.  used a weighted average for the per day cost of jail housing that included the higher per day cost of jail mental health units. However, neither study captured the actual number of days mentally ill detainees were held in specialized mental health units in order to quantify the associated costs.
This study provides an in-depth analysis of the rates of arrests, charges, and the durations of incarceration stratified by jail housing unit types, including crisis stabilization, mental health, and general population, and forensic hospital units, among a cohort of misdemeanor offenders who were arrested in DeKalb County, Georgia in 2012 and referred to the social workers in the Public Defender’s Office Mental Health Division. It contributes to the literature by distinguishing the severity of illness among the mentally ill cohort and calculating the costs associated with incarceration by the number of days spent in specific housing units within the jail. These data allow for more accurate assessments of the costs to the criminal justice system for incarcerating this population and the subsequent benefits of reducing arrest rates among mentally ill misdemeanor offenders. They can also be used to inform a targeted intervention for those most frequently arrested for minor, non-violent offenses.
Over 700,000 residents, or 7% of Georgia’s population, reside in DeKalb County. The DeKalb County Jail averages 40,000 admissions per year, 17%-24% of whom are estimated to have mental health conditions . The jail contains multiple types of housing units, including general population, crisis stabilization, mental health, and medical units. The DeKalb County Public Defenders Office represents an average of 16,000 clients annually . In order to qualify for indigent defence, offenders must earn below 100% of the Federal Poverty Level. The DeKalb County Public Defender’s Office includes a Mental Health Division consisting of two attorneys and four licensed social workers.
We cross-referenced a report provided by the Public Defender’s Office Mental Health Division of offenders referred to the social workers for exhibiting behavioral health problems (n=478) with Jail Management System (JMS) records to identify our cohort of lowincome mentally ill misdemeanor offenders arrested in 2012 (n=172). The Mental Health Division social workers assessed this cohort using the View Point Health Assertive Community Treatment Eligibility Requirements  to determine n=67 offenders met criteria for a severe and persistently mentally ill diagnosis (SPMI), while n=105 were classified as mild or moderately mentally ill (MMI). We collected demographic information, charges, housing within the jail, history of DeKalb County arrests, and permanent numbers for the cohort from JMS records.
We used SAS analytics software to calculate descriptive statistics for population demographics and incarceration characteristics and to conduct chi-square, t-tests, and Wilcoxon-Mann-Whitney tests to compare the SPMI and MMI populations.
We used the cost of jail housing ($57 per day) from an intergovernmental agreement between the DeKalb County Sheriff’s Office and Fulton County to estimate the cost of general population housing . Daily costs of housing in the DeKalb County Jail’s crisis stabilization and mental health units were not available, so we used the estimate from the 2004 report by the Council of State Governments for the cost of housing a mentally ill inmate updated to 2012 dollars ($167 per day) for both housing unit types . We used the estimate for holding an inmate on the forensic unit of Georgia Regional Hospital ($330 per day) to calculate the cost for days when an inmate was temporarily released from the jail and transferred to a forensic hospital unit for competency restoration .
We combined the number of days spent in crisis stabilization and mental health units into one overall estimate of the total number of days offenders were held in specialized mental health housing units. We calculated the total number of days inmates were temporarily released from jail as an estimate of the number of days inmates transferred to Georgia Regional Hospital for competency restoration spent in forensic hospital units. We then subtracted the total number of days in mental health housing units and forensic hospital units from the overall number of days of incarceration to derive an estimate for the number of days spent in general population housing in the jail. We then multiplied the estimates for the total number of days in each housing category by its cost estimate and calculated summary housing cost estimates.
The 67 SPMI offenders accounted for 109 arrests with 19 (28%) arrested multiple times in 2012. The 105 MMI offenders accounted for 174 arrests with 41 (39%) arrested multiple times in 2012 (Table 1). Both groups were predominantly male and African American. The mean ages for female SPMI and MMI offenders were 40.6 and 35.5 years respectively, and the difference was statistically significant. The mean age for both groups of male offenders was approximately 41 years. On average SPMI offenders had been arrested in DeKalb County 16 times and MMI offenders had been arrested 11 times prior to January 2012.
|Number of 2012 of arrests (n)||109||174|
|Number of offenders||67||105|
|People with multiple arrests during 2012||19 (28%)||41 (39%)|
|2 arrests||6 (32%)||25 (61%)|
|3 or more arrests||13 (68%)||16 (39%)|
|Female||25 (23%)||35 (20%)|
|Male||84 (77%)||139 (80%)|
|Mean age in years (SD)|
|Female||40.6 (7.8)*||35.5 (10.8)|
|Male||40.6 (12.4)||41.2 (14)|
|Non-Hispanic black||96 (88%)||148 (85%)|
|Non-Hispanic white||9 (8%)||25 (14%)|
|Other||4 (4%)||1 (1%)|
There were statistically significant differences in the amount of time that SPMI and MMI offenders were held in specialized jail housing units. SPMI offenders spent an average of 64 days in jail, including 8 days in the jail’s crisis stabilization unit (Table 2). Mental health sections are segregated by gender. Males spent an average of 39 days in the mental health units for men, and females spent an average of 28 days in the mental health units for women. Thus, SPMI offenders were held in specialized crisis stabilization and mental health units for more than half of their incarceration. MMI offenders spent an average of 39 days in jail and 3 days in crisis stabilization. Male and female MMI offenders averaged 13 and 21 days respectively in mental health units. SPMI offenders were temporarily released and transferred to forensic hospital units for competency restoration for an average of 2 days. MMI offenders were transferred for an average of less than one day.
|Variable||SPMI (n=109)||MMI (n=174)|
|Days in Jail|
|95% CI||(39, 88.2)||(28.8, 48.4)|
|Days in Crisis Stabilization Unit|
|95% CI||(4.6, 11.4)||(1.6, 4.1)|
|Days in Mental Health Unit (male)|
|95% CI||(15.7, 62.2)||(5.8,21.1)|
|Days in Mental Health Unit (female)|
|Days Temporary Transfer to Hospital|
|95% CI||(-.9, 5.2)||(.02, .26)|
|Arrests between Jan 2013 - May 2014|
|95% CI||(3, 4.2)||(2.4, 3.5)|
|Days Between Release and Re-arrest|
|95% CI||(59.3, 105.8)||(68.2, 109.6)|
SPMI offenders were re-arrested in DeKalb County an average of nearly 4 times in the 18 months between January 2013 and the beginning of data collection in May 2014. MMI offenders were rearrested an average of 3 times. The average length of time between release and re-arrest was less than three months for both SPMI and MMI offenders.
Incarcerating the 67 SPMI offenders cost the county approximately $1 million dollars, 89% of which was spent on specialized housing units. Incarcerating the 105 MMI offenders cost the county over $700,000, 72% of which was spent on specialized housing units. Thus, the total estimated cost of incarcerating 172 misdemeanor offenders in 2012 was over $1.7 million, 82% of which was due to the higher costs of holding offenders in crisis stabilization, mental health, and forensic hospital units (Table 3). This translated to an average of $14,800 for each SPMI misdemeanor offender and $6,943 for each MMI misdemeanor offender. In comparison, a misdemeanor offender without mental health conditions spent an average of 7-10 days in jail  for an estimated cost of $399-$570. Thus, DeKalb County spent approximately 26 times more for incarcerating SPMI offenders and 12 times more for MMI offenders than for misdemeanor offenders with no behavioral health conditions.
|Variable||SPMI (n=67)||MMI (n=105)|
|Total number of days in Jail||6,932||6,716|
|General Population ($57/ day)|
|Number of days||1,855||3,604|
|% Total Cost||11%||28%|
|Mental Health & Crisis Stabilization ($167/ day)|
|Number of days||4,844||3,088|
|% Total Cost||82%||71%|
|Forensic Hospital Units ($330/ day)|
|Number of days||233||24|
|% Total Cost||8%||1%|
|Total Housing Unit Costs||$991,573||$729,044|
|Average Cost Per Offender||$14,800||$6,943|
An analysis of the misdemeanor offenses the cohort was charged with showed that the most common charge for both SPMI and MMI individuals was Criminal Trespassing (25% and 26% respectively). The subsequent top four most common charges for SPMI individuals were Probation Violation, Theft by Shoplifting, Obstruction of Officers, and Disorderly Conduct. For MMI individuals, they were Battery, Probation Violation, Failure to Appear, and Obstruction of Officers.
Since it was the most common offense, we conducted further analysis specifically on the population of severely mentally ill offenders who were charged with Criminal Trespassing. Non-parametric tests on the history of arrest and jail housing among this population did not find statistically significant differences between the SPMI offenders charged with Criminal Trespassing versus other offenses, however the sample of Criminal Trespass offenders was small (Table 4). In addition, though the results were not statistically significant, they reveal a subset of the SPMI population with a long history of frequent arrests in DeKalb County (an average of 22 prior and nearly 5 subsequent) who were held for 91% (for women) and 77% (for men) of their incarceration in specialized crisis stabilization and mental health units.
|Variable||CT (n=27)||no CT (n=82)|
|DeKalb County Arrests before Jan 2012|
|95% CI||(15, 29.8)||(10.9, 17.5)|
|Arrests between Jan 2013 - May 2014|
|95% CI||(3.3, 6.1)||(2.6, 3.9)|
|Average number of days between arrests|
|95% CI||(22.5,74.8)||(64.2, 123.2)|
|Days in Jail|
|95% CI||(12.4,75.6)||(38.9, 101.2)|
|Days in Crisis Stabilization|
|95% CI||(.7,17.4)||(3.9, 11.4)|
|Days in Mental Health Unit (male)|
|95% CI||(-1.7,51)||(13.6, 74.4)|
|Days in Mental Health Unit (female)|
|Days Temporary Transfer to Hospital|
|95% CI||0||(-1.2, 6.9)|
Table 4: SPMI Offenders: Criminal Trespass Subset Comparison.
We described the detailed incarceration histories of one small cohort of low-income mentally ill misdemeanor offenders from the time of their 2012 arrests through their releases. We found that they spent a disproportionate amount of time in incarceration relative to their charges, and that the amount of time increased relative to the severity of their mental health conditions. We also found that these offenders were moved frequently between housing unit types while in jail and held in specialized mental health units for the majority of their incarcerations. This finding supports the literature’s assessment that the criminal justice system has become a de facto mental health services system. Jail crisis stabilization and mental health units require specific protocols, such as specialized staffing and the use of single versus double cells for the purposes of seclusion, which incur greater costs than general population units. Subsequently, the costs incurred for incarcerating misdemeanor offenders with mental illness was exponentially higher than the average cost for those without. Since we also found that our cohort had been arrested frequently and had long histories of re-arrest, these costs to the criminal justice system would compound over time and appear to result in no return on investment.
Our study faced several limitations. We used a report of the clients referred to the Public Defender’s Office social workers between January 1, 2012 and January 31, 2013, to identify all 2012 arrests. If there was a greater time lag between arrest and referral, some clients with late 2012 arrests might not have been referred to the social workers before January 31st and would not have been included in the study. In addition, some misdemeanor offenders with mental health conditions arrested during 2012 may not have been referred to the social workers by Public Defender attorneys and thus would not have been included in the report.
Since this was a retrospective study, information could have been lost between when clients were seen and when the social workers assessed the study population for SPMI criteria determinations. In addition, since there were three social workers in the Mental Health Division in 2012, there might have been variation in how each assessed and applied the SPMI criteria to their individual clients. One social worker also had left the office after 2012 and then was rehired in 2014. This limited her access to notes on clients that were stored in email exchanges no longer available to her. All of these limitations could have affected the consistency and completeness of the SPMI dataset.
The numbers of prior and subsequent arrests were limited by the study’s parameters and potentially compounding factors such as age and geography. For example, the study examined records from only one of the ten counties in the greater Atlanta region, so rates of arrest could have been even higher for this population if adjacent county jail data had also been considered.
The study focuses only on the costs of jail housing and does not capture additional costs to the criminal justice system such as the costs associated with per arrest processing, transport to and from Georgia Regional, or court ordered evaluations. In addition, neither cost nor protocol information was available for the jail crisis stabilization or the mental health units. Thus, we used the same cost estimate for both, which was likely an underestimate, since crisis stabilization units may require more security, medication, and medical and psychiatric supervision than mental health units. The cost estimate for Georgia Regional forensic unit housing was applied to all days that an inmate was temporarily released from Jail under the assumption that a mentally ill inmate would only be temporarily released and transferred to the hospital for competency restoration. If inmates had been temporarily released to other facilities, we could have overestimated the cost of housing for forensic hospital units. Additionally, if an offender’s housing history listed him or her in two different units on the same day due to the timing of the transfer between units, that day could have been counted twice in the study, once under each unit. This would overestimate the number of days listed by housing unit category.
The dataset did not include a population of offenders with no behavioral health conditions. Thus, we compared the cost estimates associated with specific estimates of jail housing unit duration that were calculated through statistical analysis for the study cohort to general estimates provided by the DeKalb Public Defender’s Office for misdemeanor offenders with no behavioral health conditions. Because the latter estimates were not generated by the same study, this could overestimate the differences between the two populations. Further analyses should be conducted on a cohort of offenders with mental illness and a comparable control group of offenders with no mental illness.
In order to act as good stewards of public funds, policymakers must take into account the effectiveness of policies and the returns received for resources invested. Though the criminal justice system has adapted to the increasing population of mentally ill offenders churning through its courts and jails by incorporating mental health services, because these services focus on stabilization (for the Jail) and release (for the Public Defender’s Office), they can offer only temporary aid. If this aid ends upon release, any benefits from the services received will be short-lived, as illustrated by the high rates of recidivism among this cohort. Community mental health services offer those with mental health conditions a therapeutic setting and the possibility of treatment and disease management. As such, they can help to mitigate the severity of mental health symptomology and reduce the rates of arrest and recidivism among the mentally ill.
Research has shown that programs that divert this population from the criminal justice system into a community behavioral health system providing adequate mental health treatment, housing, and case management can help to stop repeated cycles of arrest, improve health outcomes among the mentally ill, and also reduce costs to the criminal justice system . To maximize impact, such programs could initially focus on the subset of non-violent misdemeanor offenders who are frequently re-arrested for minor charges such as Criminal Trespassing. Without such a diversion into community mental health services, mental health costs to the criminal justice system could continue to grow over time with no return on investment. The results of this study, thus, could be used to support the fiscal case for shifting from a criminal justice approach focused on the repeated arrest and incarceration of mentally ill offenders to strategies supporting a more coordinated approach to connect this population with mental health treatment and social services, which would better serve both mentally ill individuals and the public good.
The authors would like to thank the DeKalb County Public Defender’s Office and the DeKalb County Sheriff’s Office for their support of this study. We would like to especially recognize the contributions made by the social workers in the Mental Health Division of the Public Defender’s Office in identifying the study’s cohort and SPMI dataset.