HOPE

Cross-Disciplinary Group for Quantifying the Role of Incarceration in National Public Health and Safety

    About

    Driven by more and longer prison sentences, Australia’s imprisonment rate has now surpassed that of most in the G20 countries[1]. In the last eight years alone, it rose by more than 20 per cent [2]. The cost to taxpayers has been substantial. Real recurrent expenditure on prisons in Australia in the 2019/2020 financial year exceeded $4 billion, doubling since 2011/12 [3].

    A particular concern is the over-representation in prisons and higher rate of deaths in custody of Aboriginal and Torres Strait Islander peoples [4]. Despite the efforts to reduce the rate of Indigenous imprisonment, the age-standardised rate of Indigenous incarceration is steadily increasing relative to non-Indigenous prisoners. We have now reached the point where almost one in four Indigenous males (22.5%) aged 35-44 have been imprisoned at some point in their lives [2].

    The unequal use of incarceration across society and the potentially negative consequences of prison on inmates and their families have culminated in increased calls for justice reforms that deemphasise the use of jail as a sanction [5]. Whether less prison is optimal is an empirical question that depends on what is in the black box of prison (e.g., prison conditions, treatment programs, or peers). The answer to this question varies across countries due to institutional differences.

    Unfortunately, while incarceration consumes an ever-growing proportion of State and Territory budgets, the effects or effectiveness of Australian prisons are unknown. Existing studies are correlational and unlikely to capture causal effects. Inmates both enter and exit prison with health problems that could be driven by unobserved factors other than time served. All prominent existing studies are also based on non-Australian settings making them even less informative as Commonwealth policy guidelines.

    This grand seeks to estimate imprisonment’s effect on the inmates’ general and mental well-being and healthcare utilisation using novel de-identified linked Australian administrative datasets. We also will identify the causal effects of prison using a research design that proved successful in comparable recent studies in the US and Europe.

    Our proposal will answer whether imprisonment rates can be reduced without jeopardising public safety and what additional modifications need to be taken to ensure those leaving prison are assisted in adjusting to a normal healthy, and lawful community life. These answers will guide Australian Federal, State and Territory Governments in their commitment to reducing the Indigenous imprisonment rate of Aboriginal and Torres Strait Islander adults by at least 15 per cent by 2031 [6].

    AIM

    Rigorous evaluation of penal policy and correctional programs is in its infancy in Australia [7], notwithstanding a general commitment on the part of most Australian Governments to evidence-based policy. Governments can’t make prudent decisions about the extent of their reliance on imprisonment without a clear understanding of its effects and effectiveness relative to other sanctions. Correctional and health authorities can’t secure additional funding for post-release support without clear evidence that prison adversely affects offenders and their families. The proposed study addresses these deficiencies.

    The primary aim is to examine the impact of incarceration on the inmates’:

    (A) General and mental well-being,

    (B) Healthcare utilisation.

    The following complimentary outcomes further assist in understanding what improvements in prison conditions or treatment programs are needed:

    (C) Reoffence,

    (D) Employment.

    BACKGROUND

    The impacts of prison on the health of inmates are theoretically ambiguous, making this an empirical question. On the one hand, prisons can help improve health if prisoners can obtain better health care than they could otherwise afford, given their average lower incomes and less formal employment. Prisons can also help inmates stay drug-free. On the other hand, the lack of freedom, poor incarceration conditions (e.g., overcrowding, poor hygiene and nutrition) and increased (threat of) violence in prison can adversely affect inmate health. Beyond the inmates themselves, the potential effects of prison on inmate health can spill over to the health of close family members through increased trauma, stigma, and financial hardship when the partner, parent or child serves prison time.

    Numerous overseas studies are showing that incarceration is correlated with worse health outcomes and behaviours, including problems that prevent work [8], depression [9], fast food consumption and smoking [10], stress-related illnesses and infectious diseases [11], and higher mortality due to both natural and unnatural causes [12, 13, 14, 15, 16]. Mortality risks are even more prominent with more time in prison [17]. It is acknowledged, however, that these studies do not establish the causal effect because those engaged in criminal activity have a higher baseline health risk than the matched sample due to differences in unobservable (and thus unmatched) characteristics, biasing correlational estimates.

    In contrast to these correlational studies, we follow an emerging robust methodology that has shown remarkable success in establishing the causal effect of prison on linked administrative data from the US and Europe [18, 19, 20, 21, 22]. The feasibility of these methods for Australia has already been established in a series of smaller studies [23, 24, 25]. In a practical sense, our proposal aims to extend these earlier Australian studies by including newer outcomes from linked datasets. An alternative view is that we aim to replicate the above-mentioned ground-breaking studies from the US and Europe on the Australian population.

    For Australia, establishing the health effects of prison is particularly long overdue. In April 2015, the NSW Inspector of Custodial Services released a report according to which ‘Corrective Services [NSW] is operating under a combination of conditions that have the potential to create a dysfunctional, if not dangerous, custodial environment’ [26]. In November 2015, the NSW Audit Office released a report stating that overcrowding in NSW prisons was costing the State $200,000 a day [27].

    Overcrowding is also an Australia-wide issue. In 2016 the Age newspaper reported that, in the first seven weeks of 2016, the Victorian Corrections Department was so stretched for resources it failed to bring 455 prisoners to their scheduled court appearances, prompting magistrates to start releasing offenders on bail simply to ensure their appearance in court [28]. The Adelaide Advertiser reported in 2011 that South Australian prisons would run out of beds in two years, despite expansion [29]. In June 2015, ABC news in Queensland reported that ‘jails’ are ‘overcrowded and becoming dangerous for staff and prisoners’ [30]. The West Australian Office of the Inspector of Custodial Services has described the state’s prison system as ‘unsustainably stretched’ [31]. The Australian imprisonment rate has risen by another seven per cent since these comments were made.

    DETAILED RESEARCH PLAN

    Research question I: Establishing the effects of Australian prison on

    1. Mortality
    2. General well-being
    3. Mental well-being
    4. Treatment for drug dependence
    5. Health care utilisation
    6. Reoffence
    7. Employment

    Research question II: Identifying the subpopulations that bare most of the the established prison effects.

    Research question III: Classifying the prison treatment programs that are most effect in reintegrating convicts into communities.

    Research question IV: Performing the cost-benefit analysis of the Australian penal system

    Study design: Retrospective de-identified linked administrative data from various government departments.

    Population: The master data is the Reoffending Database (ROD). The database contains the universe of NSW offenders whose cases have been decided at Local Courts during 1994 – 2020. The approximate sample size is 3 million case-level observations. The NSW Bureau of Crime Statistics and Research (BOCSAR) manages ROD [32]. BOCSAR has provided the principal agreement for linking ROD with other datasets.

    Statistical design: We will use a quasi-experimental design that uses the allocation of judges to cases to randomise imprisonment among defendants [33, 34]. The design leverages the ‘equality before the law’ principle that stipulates that all cases should be treated impartially [35]. In practice, this is implemented by assigning judges to cases in a fashion that is uncorrelated with the defendant’s characteristics (quasi-randomly).

    Statistical method: The statistical design will be implemented using two-stage least squares regression with the judge’s propensity for incarceration used as an instrumental variable. An array of defendant-level covariates and location and time fixed effects will be included in the regression to better isolate the quasi-random component of the judge’s allocation to cases [36, 37].

    REFERENCES
    [1]Sentencing Advisory Council, “International Imprisonment Rates,” SAC Website, 2022.
    [2]Australian Bureau of Statistics, “Prisoners in Australia,” ABS Website, 2022.
    [3]Productivity Commission, “Report on Government Services 2022. Justice,” Productivity Commission Website, 2022.
    [4]Royal Commission into Aboriginal Deaths in Custody, “Final Report – Summary. Part B: The Disproportionate Number of Aboriginal People in Custody,” 1991.
    [5]Justice Reform Initiative, “Jailing Is Failing,” Justice Reform Initiative Website, 2022.
    [6]Australian Government, “Closing the gap in partnership,” National Agreement on Closing the Gap, 2020.
    [7]L. Bartels and D. Weatherburn, “Building community confidence in community corrections,” Current Issues in Criminal Justice, vol. 32, pp. 292-308, 2020.
    [8]J. Schnittker and A. John, “Enduring Stigma: The Long-Term Effects of Incarceration on Health,” Journal of Health and Social Behavior, vol. 48, pp. 115-130, 2007.
    [9]K. Turney, C. Wildeman and J. Schnittker, “As Fathers and Felons: Explaining the Effects of Current and Recent Incarceration on Major Depression,” Journal of Health and Social Behavior, vol. 53, pp. 465-481, 2012.
    [10]L. C. Porter, “Incarceration and Post-release Health Behavior,” Journal of Health and Social Behavior, vol. 55, pp. 234-249, 2014.
    [11]M. Massoglia, “Incarceration as Exposure: The Prison, Infectious Disease, and Other Stress-Related Illnesses,” Journal of Health and Social Behavior, vol. 49, pp. 56-71, 2008.
    [12]E. S. Sailas, B. Feodoroff, N. C. Lindberg, M. E. Virkkunen, R. Sund and K. Wahlbeck, “The mortality of young offenders sentenced to prison and its association with psychiatric disorders: a register study,” European Journal of Public Health, vol. 16, pp. 193-197, September 2005.
    [13]A. Haglund, D. Tidemalm, J. Jokinen, N. Långström, P. Lichtenstein, S. Fazel and B. Runeson, “Suicide after release from prison: a population-based cohort study from Sweden,” J Clin Psychiatry, vol. 75, p. 1047–1053, October 2014.
    [14]I. A. Binswanger, M. F. Stern, R. A. Deyo, P. J. Heagerty, A. Cheadle, J. G. Elmore and T. D. Koepsell, “Release from Prison — A High Risk of Death for Former Inmates,” New England Journal of Medicine, vol. 356, pp. 157-165, 2007.
    [15]T. Skardhamar and V. Skirbekk, “Relative Mortality among Criminals in Norway and the Relation to Drug and Alcohol Related Offenses,” PLOS ONE, vol. 8, pp. 1-6, November 2013.
    [16]M. Jones, G. D. Kearney, X. Xu, T. Norwood and S. K. Proescholdbell, “Mortality Rates and Cause of Death Among Former Prison Inmates in North Carolina,” North Carolina Medical Journal, vol. 78, p. 223–229, 1 July 2017.
    [17]E. J. Patterson, “The Dose–Response of Time Served in Prison on Mortality: New York State, 1989–2003,” American Journal of Public Health, vol. 103, pp. 523-528, 2013.
    [18]S. Norris, M. Pecenco and J. Weaver, “The Effect of Incarceration on Mortality,” The Review of Economics and Statistics, pp. 1-45, July 2022.
    [19]R. Hjalmarsson and M. J. Lindquist, “The Health Effects of Prison,” American Economic Journal: Applied Economics, vol. 14, pp. 234-70, October 2022.
    [20]M. Bhuller, G. B. Dahl, K. V. Løken and M. Mogstad, “Incarceration, Recidivism, and Employment,” Journal of Political Economy, vol. 128, pp. 1269-1324, 2020.
    [21]W. Dobbie, J. Goldin and C. S. Yang, “The Effects of Pretrial Detention on Conviction, Future Crime, and Employment: Evidence from Randomly Assigned Judges,” American Economic Review, vol. 108, pp. 201-40, February 2018.
    [22]M. Bhuller, L. Khoury and K. V. Løken, “Prison, Mental Health and Family Spillovers,” Mental Health and Family Spillovers (November 23, 2021). NHH Dept. of Economics Discussion Paper, 2021.
    [23]J. Williams and D. Weatherburn, “Can Electronic Monitoring Reduce Reoffending?,” The Review of Economics and Statistics, vol. 104, pp. 232-245, March 2022.
    [24]S. Rahman and D. Weatherburn, “Does Prison Deter Drunk-Drivers?,” Journal of Quantitative Criminology, vol. 37, pp. 979-1001, 01 December 2021.
    [25]S. Alexeev and D. Weatherburn, “Fines for illicit drug use do not prevent future crime: evidence from randomly assigned judges,” Journal of Economic Behavior & Organization, vol. 200, pp. 555-575, 2022.
    [26]J. R. Paget, “Full House: The growth of the inmate population in NSW,” Justice Inspector of Custodial Services, 2015.
    [27]A. T. Whitfield, “New South Wales Auditor-General’s Report: Law and Order,” 2016.
    [28]B. Hall, “Prisoners released on bail instead of remanded because of jail overcrowding,” The Age, 2016.
    [29]K. Stokes, “Warning as South Australia prison system clogged at record levels,” News.com.au, 2011.
    [30]E. Worthington, “Queensland prisons are grossly overcrowded, Corrective Services figures show,” ABC News, 2015.
    [31]T. Wildie, “Lack of dignity in double-bunked WA prison cells, report finds,” ABC News, 2016.
    [32]J. Hua and J. Fitzgerald, “Matching court records to measure reoffending,” Crime and Justice Bulletin, p. 1–10, 2006.
    [33]M. Bhuller and H. Sigstad, “Errors and monotonicity in judicial decision-making,” Economics Letters, vol. 215, p. 110486, 2022.
    [34]B. Frandsen, L. Lefgren and E. Leslie, “Judging Judge Fixed Effects,” American Economic Review, vol. 113, pp. 253-77, January 2023.
    [35]Judicial Commission of New South Wales, Equality Before the Law Bench Book, Judicial Commission of New South Wales, 2022.
    [36]N. Huntington-Klein, The Effect: An Introduction to Research Design and Causality, CRC Press, 2021.
    [37]S. Cunningham, Causal Inference: The Mixtape, Yale University Press, 2021.