Geriatric Inmates: Policy and Practice

Segregated Units

As Brie Williams expressed, “What do we, as a society, want to do with a prisoner who is so demented he doesn’t remember his name, or who has had a stroke and is completely paralyzed?” (Tokar, 2006, para. 10). One possible solution is having segregated units with elder-specific programming and treatments. In 2008, at least 13 states had units dedicated to older inmates, 6 had dedicated prisons, 9 had dedicated secure medical facilities, and 8 had dedicated hospice facilities (Chiu, 2010). Such units would provide safety against younger inmates and concentrate appropriate treatments and programming. Virginia and Pennsylvania are just two states that have built geriatric prison facilities as a “mini-hospital” model to confine and care for older inmates (Stal, 2013, p. 71). Some argue that segregating aging inmates is a cost-effective way to provide medical services that are age appropriate to meet their physical needs (Rosefield, 1993) and is more efficient because medical services are centralized (Aday, 2003; Yates & Gillespie, 2000). There has been very little research on whether segregated housing and centralized medical services are more effective in delivery as compared to other available medical services. In a study by Thivierge-Rikard and Thompson (2007), researchers sought to test the association of available care in segregated (geriatric) units and care in nonsegregated units of prisons. Results indicated no relationship between housing models for elderly inmates and availability of physical health services. However, they did find that housing models for aging inmates were associated with mental health services. They found the number of mental health services is higher in consolidated (geriatric) housing models as compared to institutions with no geriatric services (Thivierge-Rikard & Thompson, 2007). Elderly inmates coping with long sentences, mental health, or illness-related issues (e.g., Alzheimer’s or dementia) having a greater need for services could explain the benefits of this consolidated model. Nonetheless, several states have decided to create consolidated units to accommodate their prisoners’ needs based on age and illness. Additional research is required to validate this study. Further research is also needed to weigh the costs and benefits of segregated housing models. Research should also evaluate inmate feedback on quality of services in segregated facilities as compared to the general population to gauge its necessity and usefulness. Table 1 highlights units addressing elderly care.

Structured Programming for Elderly Inmates

The True Grit Program housed in the Northern Nevada Correctional Center offers programming for those serving their “‘golden years’ behind bars” (Harrison, 2006, p. 47). Established in 2003, this program aims to enhance physical, mental, and spiritual health by providing physical and recreational therapy, group therapy, and individual therapy. Activities include visits from the Delta Society, which brings therapy dogs to prison monthly. All materials utilized by the program were provided by outside community members and organizations, indicating a high level of community engagement with the program. This program currently has over 170 participants (Linn, 2017).

There has been very little empirical research on this program, its benefits, effects on inmates, or successful reentry and recidivism. Since the program’s inception, Kopera-Frye and colleagues (2013) noted that of the 265 inmates who have passed through the program, 38 died in prison, 91 were paroled or finished their sentences, and the remainder (135) either voluntarily terminated involvement or were dismissed for violations of rules. Kopera-Frye et al. (2013) conducted a descriptive study of this program and its benefits for veterans compared to nonveterans to determine whether there were any differences in program benefits based on veteran status. Results indicated that regardless of group membership, both veteran and nonveterans reported relatively low levels of psychological symptoms and moderate levels of psychological symptoms related to adjusting to prison life. Both groups reported benefits from treatment, with nonveterans reporting slightly better, but nonsignificant results. More rigorous studies employing a control group, perhaps comparing those in the general population, would give a better indication of the program’s effects.

Geriatric Parole

Geriatric parole or medical parole is a viable option in managing the elderly inmate population. According to a content analysis (Maschi et al., 2015), 46 states, Washington, DC, and the federal government have compassionate or geriatric release laws. Statutes vary by state and can include medical or compassionate release, inmate furloughs, or discretionary parole. Release mechanisms are typically categorized by age, physical or mental health, postrelease support, and personal or criminal history (Maschi et al., 2015). Mechanisms for release can be narrow and can conflict with standard parole practices, leaving inmates less likely to exercise this option. In Virginia, those who apply for geriatric release forfeit that year’s automatic parole hearing as the parole board will not hear cases on both grounds in the same year (Chiu, 2010). There is little incentive to apply under the geriatric release mechanism.

Colorado released three prisoners under this policy from 2001 to 2008; Oregon has released two prisoners per year since 2009; and from 2001 to 2007, Virginia released four inmates (Anno et al., 2004; Chiu, 2010; Williams et al., 2012; Williams et al., 2011). This may be explained by elderly inmates who are still considered “high risk”; those who were high risk in the past but have few social contacts outside, thereby inflating their risk level while not taking age into account; and not agreeing on the concept of inmates “aging out” of crime (Kempker, 2003; Psick et al., 2017).

Ethridge and White (2015) investigated how and why Texas implemented the Medically Recommended Intensive Supervision (MRIS) program. In 1991, the Texas Comptroller of Public Accounts, John Sharp, made recommendations regarding “special needs inmates” that included transfers to other state programs. The aim was to control the costs of medical care for the elderly and the mentally fragile inmates (Ethridge & White, 2015). In 1997, this statue was revised to MRIS. Displeased with the low release numbers, there were efforts to expand the statute, including expanding eligibility for release and reducing the age requirement from 65 to 60 for elderly inmates. The revised legislation expanded the definition of terminally ill (increasing the length of time remaining from 6 months to 1 year) and revisions to parole eligibility, but this new bill was never voted on. By 2013, the political sensibilities had shifted and there was a push to limit who would be eligible for release. Legislators wanted to eliminate the elderly provision all together, stating that medical condition, not age, should be the basis for release consideration (Ethridge & White, 2015). Legislators did not vote on this revision to the bill and the 1997 bill remains.

Etheridge and White (2015) found that between years Fiscal Year (FY) 2007 and FY2012, the numbers of referrals to MRIS increased from 941 to 1,857 inmates per year. During the same period, the total count of inmates presented to the parole board for MRIS deliberation increased from 290 to 491. In FY2007, 34.8% (n = 101) of those presented for parole were granted release. By 2012, only 14.7% or 72 persons presented for parole were granted release. More inmates were referred for MRIS than in past years, but fewer were released than in previous years. Of those who were released under MRIS, 40.5% (308) were terminally ill, 15.5% (211) required long-term care because of their condition, 15.7% (8) were physically handicapped, 0.05% (7) were elderly, and 0.05% (7) were mentally ill (Ethridge & White, 2015). During the study period, 1,484 inmates were released under MRIS. Of those inmates, 66% (986) died post-MRIS release; 18% (268) had their sentences discharged; 8% (121) were currently on MRIS supervision; 3% (42) were in Texas Department of Criminal Justice (TDCJ) or federal custody, absconded, or were deported; and 5% (67) were returned to TDCJ, released under another supervision program, or deceased in custody (Ethridge & White, 2015).

Knowledge Gaps

The research community can fill knowledge gaps in several different ways. First, researchers should move beyond the overview and descriptive research study and raise serious research questions including (1) What is the recidivism rate of those paroled by geriatric release? (2) Why are states not using geriatric release at greater levels? (3) Are risk, needs, and recidivism assessments (i.e., Level of Service Inventory–Revised [LSI-R]) valid for elderly inmates? (4) What is the quality of life of those in units with age-appropriate programing compared to those in the general population? (5) Are in-house geriatric units and geriatric programming more cost-effective and do they provide an adequate level of care for the elderly population, as compared to “free society” programs? And (6) Do inmates who participate in age-specific programming, or who are in age-segregated units, report higher quality of life? Descriptive and overview studies are helpful in defining the problem, but there has been an inadequate number of studies on elderly populations in prison. Rigorous research is needed on providing age-specific care, understanding quality-of-life issues, providing appropriate end-of-life care, and investigating the barriers to geriatric parole as a release mechanism.

The research community can fill knowledge gaps on understanding the costs associated with programs, segregated units, and geriatric release as well. This can occur on two fronts: (1) What are we willing to pay to release an elderly inmate based on terminal illness, long-term illness, or otherwise? (2) If we are willing to pay, how can we ensure a low recidivism rate? In 2008, the Appropriations Committee conducted a cost–benefit analysis to determine whether it would be more cost-effective for the state to operate assisted living or nursing facilities for geriatric inmates as compared to contracting for privately operated services (Virginia Department of Corrections, 2008). They conducted a cost assessment of private area nursing homes in the Richmond, Virginia, area compared to the cost of operation at Deerfield Correctional Center. Results indicated that the state could deliver more cost-effective services as compared to private facilities. The average 2006 nursing home costs of $56,940 to $66,430 do not include the costs of hospital visits, which is twice as costly as housing an inmate at Deerfield Correctional Center with hospital care ($25,395; Virginia Department of Corrections, 2008). In FY2004, the average inmate over 50 years old had annual off-site medical bills totaling $2,490. The following year, costs increased to $3,350. Therefore, it is fiscally more prudent to keep and provide treatment to inmates in-house. Furthermore, private nursing homes have long wait lists and are not willing to take inmates, and there are few public facilities. In addition, it is very difficult to release inmates through geriatric release due to the severity of their offense. Of the 489 inmates who were eligible for geriatric release, 52 applied and only 2 were released, which is less than 1% of those eligible for parole (Virginia Department of Corrections, 2008).

Financial costs are not the only costs to consider. There are social costs of releasing inmates as well as external costs. What is the social cost of releasing an inmate who reoffends, elderly or otherwise? Ostermann and Caplan (2016) found that state-level bottom-up costs in New Jersey associated with criminal activity for the years 2005 to 2007 totaled more than half a billion dollars for all inmates in the 3 years following release from incarceration. The public’s willingness to pay to have these crimes prevented (by not releasing inmates) is almost $6 billion; therefore, this social cost becomes a burden on society that reduces our overall well-being (Cohen & Bowles, 2010). This study does not differentiate between elderly and nonelderly inmates, but if we had information on the costs of releasing elderly inmates, it may influence the decision to release. One must also consider the external costs. Cohen and Bowles (2010) refer to external costs as costs imposed on one person by another, where the latter person does not voluntarily accept this negative consequence. This would be particularly relevant in a property crime as compared to a personal victimization. The fact of the matter is some released elderly inmates do reoffend, and as a society or an individual, do we want to accept the “externality” of these costs?

Second, if we are willing to risk the costs, how can we minimize risk of recidivism? Assessment tools such as the LSI-R assess risk of recidivism, but this tool may not necessarily be relevant to elderly populations, so efforts should be made to validate assessment tools that specifically address the over 50 inmate population to ensure public safety. The LSI-R (Andrews & Bonta, 1995) was developed on probationers and inmates with sentences of 2 years or less (Manchak et al., 2008, p. 479). The 54 items of the LSI-R assess 10 overarching “risks–needs” factors: (1) criminal history, (2) education/employment, (3) financial, (4) family/marital, (5) accommodation, (6) leisure/recreation, (7) peers/companions, (8) drug/alcohol problems, (9) emotional/personal, and (10) attitudes/orientation (Manchak et al., 2008, pp. 478–479). Research by Manchak et al. (2008) begins to examine the issue by assessing the validity of the LSI-R on long-term inmates (LTI). This may be of interest to the question at hand as many elderly inmates are serving long-term sentences. LTI are typically convicted of serious offenses, less likely to be trained occupationally, have few outside ties, and are more “embedded in criminality” (Petersilia, 2003, p. 40) but are also older and “closer to ‘aging out’ of crime” at time of release (Manchak et al., 2008, p. 478). LTIs are less likely to have instances of institutional misconduct and tend to engage in prosocial activities and are psychologically stable at release. These dynamic risk factors change over time as does risk level. Manchak et al. (2008) examines the utility of the LSI-R for those who have served more than 10 years in prison. The analyses included the 1,144 inmates, the 555 LTI, and a general inmate sample including both LTIs and non-LTIs. Results indicated that the LSI-R moderately predicts general recidivism but not necessarily violent recidivism. Results for violent recidivism found that high-risk group inmates had a lower risk for recidivism than those in the medium-/high-risk group (Manchak et. al., 2008). This is a first step to assess LTIs’ likelihood of reoffense, but there is a lack of research that specifically validates the LSI-R with elderly inmates.

How Can Implementation of Solutions Be Facilitated/Impeded by Outside Factors?

Implementation of programs or facilities such as True Grit or the Fishkill Unit for the Cognitively Impaired are viewed as successful because of the interconnectedness of correctional, health, and community-based resources. Administrators should continue to collaborate with outside agencies to ensure inmate reentry is successful. This would include parole, social workers, nursing homes, and other agencies and organizations that handle the needs of the elderly. Security is always of concern in this population, as Stone (2011) posits that security is a collective responsibility to protect against future harm. Geriatric release is politically and socially unpopular for this reason. The polis is a collective vision of the public interest, and the liberty of individuals is limited by the obligations of the community (Stone, 2011). The concern here is harms perpetrated by released inmates; therefore, concern about protection against those harms is an impediment to geriatric parole as a solution. These concerns are not just of structural harms, that is, the ability of the community to function, but also of accumulative harms, meaning the more inmates released, the more devastating to a community it would be. This would explain the low numbers of inmates released through geriatric parole. If states expanded the scope of who was released and monitored them, they would get a better idea of whether this program could be effective. Furthermore, proper assessment tools for risk of recidivism, coupled with proper community support, should ensure low reoffending behavior considering age of release and any underlying health conditions. Politics and marketing have created a barrier where politicians are clearly more concerned about being “tough on crime” versus “smart on crime.” Politicians should consider implementation practices that balance public safety with freedom for the sick and elderly.

Increase Levels of Research on Practices and Programs Affecting Elderly Prisoners

There are strikingly few empirical research studies on elderly inmates, considering the burgeoning populations. Programs and policies are implemented without the benefit of scientific research. Collaboration between correctional agencies and academic and research institutions can be pivotal in advancing “what works” for the geriatric inmate population. Practice–research partnerships can be successful when researchers are dedicated to rigorous scientific studies, align with practitioners who are committed to the delivery of services, and work collaboratively to identify the issues that create mutually acceptable research conditions in agency settings (Secret et al., 2011).

Increasing the quantity of research is necessary, but quality of research through empirical evaluations is necessary as well. Studies that evaluate the effects of criminological interventions differ in methodological quality (Farrington et al., 2002); therefore, research should consider the Maryland Scientific Methods Scale (SMS) to determine the strength of their research studies. The True Grit study (Kopera-Frye et al., 2013) tested a causal hypothesis (Level 2 on the SMS). By adding a control group to compare the effects of the True Grit Program to those not receiving the treatment, the SMS level would have increased to 3, which is the minimum interpretable design that is acceptable for drawing conclusions about what works. Other researchers should consider the level of research design to deliver reliable results that rule out threats to validity and increase reliability.

Validate Risk Assessment Tools for Elderly Inmates

Risk needs and responsivity assessment tools such as the LSI-R must be validated for inmates over the age of 50 to accurately assess risk. A more accurate assessment can be key to determining true risk for reoffense if released. This would be an important tool for the parole board to consider both dynamic and static factors including age, criminal history, and current attitudes about crime. With valid assessment tools, parole boards can make valid decisions about who is likely to reoffend and perhaps release more people through geriatric parole. Parole boards could convey the accuracy of these tools to politicians and the public to create understanding about the elderly population and the true level of threat to the community. More concise risk measurements can instill a community’s confidence in correctional institutions and those released. Furthermore, valid tools can prevent those who are still a public threat from being released, ensuring the safety of the community.

Consider the Needs of the Elderly Inmate

One shouldn’t assume that the inmate population, or the elderly population, is homogeneous. Each person has individual needs; elderly inmates’ living environment, end-of-life care and advanced directives, and even whether they want to be considered for release should be considered. It can be beneficial to confine elderly inmates to a specific prison or ward, but this may not be in their best interest or in the interest of the prison. Elder inmates have been known to stabilize the general prison population, and they may not want to leave general population as they may have friends or like being around younger inmates. Others, particularly those who are frail or advanced in age, may prefer a less rowdy environment and want to live in a “senior-only” facility. Still others who have cognitive impairments would benefit from a model of confinement that is more protective than punitive.

The U.S. Centers for Disease Control and Prevention (CDC) defines aging in place as the ability to live in one’s own home or community safely, independently, and comfortably regardless of age, income, or ability level; being able to have any services (or other support) they might need over time as their needs change (CDC, 2009). This concept should apply to elderly inmates as well. The concept of aging in place is popular in the “free world” as most people want the right to age on their own terms (National Aging in Place Council, 2019). Self-determination as one grows older provides quality of life and the feeling that one has control over their life circumstance.

Consideration should also be given to inmates’ denial or acceptance of end-of-life treatments. They should be surveyed on quality-of-life issues and how confinement impacts them. Prisoners have rights, and some level of self-determination in their senior years should be among them. A study by Phillips et al. (2011) found that inmates’ ideas regarding end-of-life care varied greatly based on expectation of parole, not present functional ability. Specifically, researchers found a significant three-way interaction between illness, race/ethnicity, and parole expectations (Phillips et al., 2011). Caucasian inmates’ desire for active treatment was not impacted by expectation of parole for those who had cancer, Alzheimer’s, or emphysema (Phillips et al., 2011). However, minority inmates who felt they would be paroled had a greater desire for life-sustaining treatment regardless of illness (Phillips et al., 2011). This study highlights the importance of considering the needs and wants of prisoners; this study may be relevant in considering the ramifications of geriatric parole and who is considered as well as their likelihood of release. The willingness to receive treatment if paroled may indicate more positive ideation, which could make them a good candidate for parole.

Increase Partnerships With Community Organizations

Community partnerships with correctional agencies are key for program success. Programs such as True Grit and the Fishkill Unit for the Cognitively Impaired have many outside contributors extending group membership to those who likely need that kind of support to enhance their lives and chances at reentry. Restorative justice comprises the idea that because crime hurts, justice should heal (Braithwaite, 2007). This is a popular idea, one that politicians could use to improve their marketing (Berman & Fox, 2010) to gain support for release for elderly and infirm inmates. This model of justice is particularly popular because it hands power back to ordinary people (Braithwaite, 2007) as they decide what is in the public interest. Restorative justice practices include both the victim and offender, with the aim of restoring the community to the state prior to the crime (Zehr, 2002). Restorative circles are based on Native American and Mennonite practices and include three core principles: “1) the wrong or injustice must be acknowledged, 2) equity needs to be restored, and 3) future intentions need to be addressed” (Claassen as cited by Zehr, 2002, p. 45).

Restorative circles would be beneficial in decisions to grant or deny geriatric parole. Circles can invite the parole board to present evidence to grant or release to relevant stakeholders, which can help overcome the propensity to punish over effective prevention or rehabilitation (Braithwaite, 2007). This change in strategy can be effective in helping shift politics and policies toward a more evidence-based rational response to crime and incarceration.

Establish Standard Best Practices for Geriatric Programs and Facilities

Finally, departments of corrections should aim to establish best practices across state correctional institutions and then share those practices across other states to ensure conditions are acceptable for this population. As these practices are empirically tested, other institutions without comparable programs can serve as the control group(s) to deem practices and policies effective. Short of empirical research, correctional agencies should consider modeling their practices on what works in a community setting. Hospice care at Louisiana State Penitentiary is a good example of this. In 1996, Angola collaborated with University Hospital Community Hospice, a program of the Medical Center of Louisiana at New Orleans, to create a hospice within the prison (Evans et al., 2002). This informal agreement was intended to improve the quality of life for inmates at this prison. University staff provided the prison with education, training, and counsel needed to create and sustain a hospice program (Evans et al., 2002). Patient care is provided by inmate volunteers and prison staff. This level of care and partnership is possible only with highly collaborative relationships between community and corrections.

A Final Word

Policy-making for the elderly inmate is a complicated endeavor. There are major gaps in research, studies are largely descriptive or general in nature, and programs are implemented without assessment. Through collaborative efforts, more rigorous practices can be advanced, and what is known about this population can drive policy determinations. Increased efforts to validate and create risk assessments will aid in recommending inmates for release, and restorative justice programs can aid in healing old wounds and accepting inmates back into the community. Partnerships with leaders in community medical treatment can ensure these individuals receive proper end-of-life care, while considering what might be in the person’s best interest.