Max Out - The Rise in Prison Inmates Released Without Supervision, Pew, 2014
Download original document:
This text is machine-read, and may contain errors. Check the original document to verify accuracy.
A report from Max Out The Rise in Prison Inmates Released Without Supervision June 2014 The Pew Charitable Trusts Susan K. Urahn, executive vice president Alexis Schuler, senior director Program team Adam Gelb Karla Dhungana Benjamin Adams Ellen McCann Kathryn Zafft Acknowledgments Valuable research support was provided by Pew staff members Stephen Bailey and Liz Gross. We also thank our colleagues Kristin Centrella, Jennifer V. Doctors, Jennifer Peltak, Lesa Rair, Gaye Williams, and Christina Zurla, as well as former staff members Brian Elderbroom, Ryan King, and Jennifer Laudano, for their significant contributions to the research, writing, and production of this report. Jenifer Warren provided essential writing services. The Pew Charitable Trusts would like to thank the participating state departments of corrections for their help in reviewing and revising the release data, if needed. We would also like to extend our thanks to E. Ann Carson of the U.S. Department of Justice, Bureau of Justice Statistics, for her invaluable guidance in the use of the National Prison Statistics and National Corrections Reporting Program data, as well as thinking through the survey instruments. Cover photos: 1. GreyCarnation Contact: Lesa Rair, communications Email: email@example.com Project website: pewtrusts.org The Pew Charitable Trusts is driven by the power of knowledge to solve today’s most challenging problems. Pew applies a rigorous, analytical approach to improve public policy, inform the public, and stimulate civic life. Contents 1 Overview 2 Policy changes lead to spike in max-outs 3 1 in 5 inmates max out 3 State release policies vary widely California Realignment Affects Max-out Rates 5 6 Increase driven by nonviolent offenders 6 Supervised release can cut recidivism and costs 9 Policy framework to reduce max-outs Require a period of post-prison supervision for all offenders 9 Carve out community supervision period from prison terms 9 Victims’ Statement of Principles 10 Strengthen parole decision-making 11 Straight From Solitary to the Street 11 Tailor supervision conditions to risk and need 12 Adopt evidence-based practices in parole supervision 12 Reinvest savings in community corrections 13 14 Conclusion 14 Appendix A: Methodology 15 Appendix B: Jurisdiction notes 17 Endnotes Figures Figure 1: Max-outs on the Rise 1 Figure 2: Wide Variance in State Max-out Rates (2012) 4 Figure 3: Nonviolent Offenses Drive Increase in Max-outs 6 Figure 4: Poll: Mandate Supervision for All Offenders Released From Prison 8 Overview Despite growing evidence and a broad consensus that the period immediately following release from prison is critical for preventing recidivism, a large and increasing number of offenders are maxing out—serving their entire sentences behind bars—and returning to their communities without supervision or support. These inmates do not have any legal conditions imposed on them, are not monitored by parole or probation officers, and do not receive the assistance that can help them lead crime-free lives. An analysis of this trend by The Pew Charitable Trusts found that:1 •• Between 1990 and 2012, the number of inmates who maxed out their sentences in prison grew 119 percent, from fewer than 50,000 to more than 100,000. •• The max-out rate, the proportion of prisoners released without receiving supervision, was more than 1 in 5, or 22 percent of all releases, in 2012. •• Max-out rates vary widely by state: In Arkansas, California, Louisiana, Michigan, Missouri, Oregon, New Hampshire, and Wisconsin, fewer than 10 percent of inmates were released without supervision in 2012. More than 40 percent of inmates maxed out their prison terms and left without supervision in Florida, Maine, Massachusetts, New Jersey, North Carolina, Ohio, Oklahoma, South Carolina, and Utah. •• Nonviolent offenders are driving the increase. In a subset of states that had data available by offense type, 20 and 25 percent of drug and property offenders, respectively, were released without supervision in 2000, but those figures grew to 31 and 32 percent, or nearly 1 in 3, in 2011. The increase in max-outs is largely the outcome of state policy choices over the past three decades that resulted in offenders serving higher proportions of their sentences behind bars. Indeed, “truth-in-sentencing” laws, limits on release eligibility, and the outright elimination of parole in some states added nine months to the average prison time served by offenders released in 2009, compared with those released two decades earlier.2 Figure 1 Max-outs on the Rise Policy changes increased the number of inmates serving their full prison terms 1 in 5 Inmates were released without supervision in 2012 119 % Increase in unsupervised releases, 1990-2012 Source: U.S. Department of Justice, Bureau of Justice Statistics’ National Prisoner Statistics (1990 data); NPS data for 36 states and seven states that provided data directly to Pew (2012 data). See the report methodology for details. © 2014 The Pew Charitable Trusts 1 Yet new research suggests that for many offenders, shorter prison terms followed by supervision have the potential to reduce both recidivism and overall corrections costs. In New Jersey, for example, inmates released to parole supervision before their sentences expired were 36 percent less likely to return to prison—even when controlling for risk factors such as an offender’s prior record that reliably predict recidivism—than inmates who maxed out.3 In the past few years, policymakers in at least eight states took steps to ensure that offenders are supervised after release from prison. Among the measures: mandating a period of post-prison supervision, carving that time out of the prison term rather than adding it at the end, improving parole decision-making, tailoring the intensity and duration of supervision to each offender’s public safety risk level, and strengthening community corrections through reinvestment in evidence-based practices. This report examines the rise in max-outs and its policy origins, looks at states that are leading on the use of post-prison supervision to protect public safety and reduce costs, and provides a framework to help other states use evidence to inform release and supervision decisions. Policy changes lead to spike in max-outs As crime rose during the 1980s and early 1990s, policymakers across the country sought to protect public safety and punish lawbreakers by putting more offenders behind bars and keeping them there longer. The average offender released from state prison in 2009 served nearly nine months longer than two decades before. 4 Limits on discretionary parole were encouraged by passage of the Violent Crime Control and Law Enforcement Act of 1994, which authorized billions of dollars in federal grants to expand prison capacity in states with truth-in-sentencing laws. 5 Establishing certainty in sentencing is a legitimate goal, which, when combined with sanctions that are proportional to the offense, can build trust and confidence in the criminal justice system. What followed was a move toward determinate systems that require violent offenders to remain in prison for a minimum of 85 percent of their original court sentence before becoming eligible for release. Many states built on this approach by requiring all offenders to serve higher percentages of their sentences behind bars. By 2000, 16 states and the federal government had abolished discretionary parole for all offenses, and four additional states eliminated it for certain violent crimes. 6 Rather than relying on parole boards to make release decisions, a growing number of states released inmates according to a sentencing formula established under guidelines set in statute, commonly known as mandatory parole. In 1980, discretionary parole releases accounted for more than half (55 percent) of all releases of inmates serving sentences of one year or more, but by 2011 that proportion had dropped to just 26 percent. 7 Mandatory parole releases, meanwhile, grew from 19 percent of the total in 1980 to 33 percent in 2011. 8 Although mandatory parole, as the name implies, includes a period of post-prison supervision, this aspect was not at the center of the policy discussion that led to its greater prevalence. Rather, policymakers were focused on increasing prison time served and limiting the role of parole boards. As a consequence, there has been a considerable increase in the number and proportion of offenders who are incarcerated for the full duration of their sentences and transition out of prison with no legal conditions, monitoring, or reentry assistance. 2 1 in 5 inmates max out The total number of max-outs has grown remarkably since 1990, when fewer than 50,000 inmates completed their sentences in prison and returned to the community with no parole supervision or services. By 2012, that number had more than doubled. In the 43 states with data for both 1990 and 2012, max-outs increased 119 percent, from 47,519 to 103,831 inmates. In the 47 states that can report data for 2012, 115,372 inmates maxed out. This trend is even more notable when considering the rate of max-outs, the proportion of inmates released unconditionally each year. In 1990, 14 percent of inmates, 1 in 7, maxed out. That rate continued to increase over the next two decades, to 22 percent—more than 1 in 5—in 2012. This report is based on release data from 1990 to 2012. In recent years, several states, including Kansas, Kentucky, New Hampshire, North Carolina, Ohio, Oklahoma, South Carolina, and West Virginia, enacted laws to ensure that more offenders undergo a period of post-release community supervision. These policies are likely to have an impact on current max-out rates in these states but are not reflected in the data used for this analysis. State release policies vary widely Between 1990 and 2012, the max-out rates increased in 23 states. Thirteen states released more than one-third of their inmates unconditionally in 2012, and max-outs accounted for more than 4 in 10 releases in nine states: Florida, Maine, Massachusetts, New Jersey, North Carolina, Ohio, Oklahoma, South Carolina, and Utah. Florida had the nation’s highest max-out rate. The state abolished parole in 1983 and, like many states, adopted a determinate sentencing structure. It also was one of the first states to adopt a truth-in-sentencing measure, passing a law requiring all offenders who committed their crimes on or after Oct. 1, 1995, to serve at least 85 percent of their sentences. 9 Since then, the number of inmates maxing out in Florida has risen steadily. In 1990, Florida released approximately 12,000 inmates, or 32 percent of offenders, without supervision. By 2012 the max-out rate had increased to 64 percent, resulting in more than 21,000 inmates leaving prison with no monitoring or support. In New Jersey, 41 percent of inmates maxed out in 2012, the ninth-highest rate. This is notable in light of the research mentioned above showing that inmates released to parole supervision in the state are less likely to be rearrested, reconvicted, or reincarcerated than inmates who max out.10 At the other end of the spectrum, 17 states released fewer than 15 percent of their departing offenders without supervision in 2012. Among these states, Arkansas, California, Louisiana, Michigan, Missouri, New Hampshire, Oregon, and Wisconsin had max-out rates of less than 10 percent. 3 Figure 2 Wide Variance in State Max-out Rates (2012) State Unsupervised releases Max-out rate Oregon 22 0.4% California* 455 0.9% Arkansas 313 Wisconsin 468 New Hampshire† 98 * See the California sidebar on page 5. 5.0% This state has revised policies within the past two years to reduce its max-out rate. † 6.1% 6.3% Michigan 961 Louisiana 1,505 7.3% 8.8% Missouri 1,616 9.0% Indiana 1,888 Virginia 1,276 New York 2,696 Colorado 1,315 11.1% 12.0% 10.2% 10.2% 520 12.2% Maryland 1,308 12.6% Montana 284 13.6% Minnesota 1,049 13.6% Texas 11,280 13.7% 15.6% Idaho Vermont 306 Arizona 2,119 North Dakota 187 16.3% 17.4% South Dakota 345 17.6% Mississippi 1,370 17.9% Georgia 3,436 Washington 1,489 Kentucky† 3,272 Pennsylvania 3,933 State average 115,372 Connecticut 19.2% 19.7% 20.2% 20.9% 21.5% 23.7% 1,423 Delaware 374 Kansas † 1,159 23.8% 24.2% 213 24.3% Wyoming Iowa Nebraska 1,330 New Mexico 1,034 Tennessee 4,878 Hawaii 25.5% 722 26.9% 30.7% 30.7% 32.6% 320 West Virginia† 1,086 Alabama 3,740 Nevada 2,080 New Jersey 4,414 Utah 1,262 Massachusetts 1,241 41.2% 43.3% 3,176 43.6% South Carolina† Ohio† 10,008 Oklahoma† 3,884 North Carolina† 7,388 Maine 703 Florida 21,426 33.0% 33.2% 39.3% 41.2% 46.6% 55.9% 59.9% 63.4% 64.3% 0% 10% 20% 30% 40% 50% 60% 70% Source: U.S. Department of Justice, Bureau of Justice Statistics’ National Prisoner Statistics data collection; data submitted directly to Pew by state corrections agencies. See the report methodology for details. © 2014 The Pew Charitable Trusts 4 Five of the eight states with the lowest rates have retained discretionary parole. New Hampshire, where the max-out rate had hovered around 20 percent, passed legislation in 2010 to address this issue and releases most inmates to supervision at least nine months before their sentences expire. The state’s max-out rate has since declined to about 6 percent. Oregon operates under a determinate sentencing structure and abolished parole in 1989. But, for sentences of 12 months or more that were imposed after Nov. 1, 1989, the state requires a period of post-release supervision with the length determined by offense type.11 In 1980, Oregon released 10 percent of its offenders unconditionally, but by 2012 that number had dropped to 0.4 percent. California Realignment Affects Max-out Rates In 2011, Governor Jerry Brown of California signed Assembly Bill 109, the Public Safety Realignment Act. The landmark legislation transferred jurisdiction of lower-level offenders from the state Department of Corrections and Rehabilitation to the counties. Felony offenders who are classified as nonserious, nonviolent, nonsex registrant, known as “non-non-nons,” are now sentenced to county jail instead of prison, supervised by county probation departments under post-release community supervision, and sent to local jails if they violate the terms of their release. As a result, the number of inmates released from California prisons fell by more than half between 2011 and 2012, from 109,467 to 49,574. Other elements of realignment also affected the number of California prison releases. All revocations for state parolees, except those with an original sentence of life, go to county jail instead of state prison for a maximum of 180 days. Additionally, the non-non-nons are being diverted from state prison at sentencing, reducing both admissions and releases. As a result of these changes, the number of max-outs from state prisons fell in the first full year of realignment from 12 percent in 2011 to less than 1 percent in 2012. Under the new system, non-nonnons—more than 30,000 offenders who accounted for 62 percent of releases—are released to their county of last legal residence and supervised under post-release community supervision. Offenders diverted to supervision are eligible for discharge at six months, and sanctions for violators are capped at 180 days. Counties have discretion to determine the type of supervision provided. The remaining 36 percent of inmates released in 2012 were serving sentences for serious or violent crimes; they remained under the jurisdiction of state parole agents. The extent to which realignment has shifted max-outs to the local level is unclear. County judges can now exercise their discretion to impose either a straight jail sentence without supervision or a split sentence that combines a jail term with a period of mandatory supervision to follow. Current use of split sentencing varies widely among the counties. Some order it in more than 80 percent of cases, while several, including Los Angeles and Alameda counties, use it less than 10 percent of the time.* Without greater use of split sentences, large numbers of non-non-nons may be returning to California communities without supervision. * Chief Probation Officers of California (April 8, 2014), http://www.cpoc.org/assets/Realignment/splitsentencedashboard.swf. 5 Increase driven by nonviolent offenders It is commonly assumed that offenders who max out are incarcerated for the most serious crimes. Pew’s analysis of a subset of states, however, shows that nonviolent offenders were more likely than violent offenders to serve their full sentences and that inmates incarcerated for drug and property offenses drove the overall increase in the max-out rate.12 The max-out rate for violent offenders changed little over the past decade. Unconditional releases accounted for 22 percent of violent-offender releases in 2011, down slightly from 23 percent in 2000. Nonviolent offenders, on the other hand, maxed out at a much higher rate in 2011 than in previous years: One in 4 property and 1 in 5 drug offenders were released unconditionally in 2000, but by 2011 the rates for both groups had jumped to nearly 1 in 3, up 25 and 50 percent, respectively. Figure 3 Nonviolent Offenses Drive Increase in Max-outs Violent offenders less likely to serve full term, be released without supervision 40% 2000 29% Max-out rate 30% 32 % 6% 23% 22% 20% 58 % 2011 31% 25% 20% 10% 0 Violent Property Drug Offense type Note: State unconditional release rates by offense type, 2000 and 2011. Source: Pew analysis of data for 19 states from the U.S. Department of Justice, Bureau of Justice Statistics’ National Corrections Reporting Program © 2014 The Pew Charitable Trusts 2000 2011 Supervised release can cut 35% 29% recidivism%and costs 58 2000 Although monitoring offenders when they leave prison% may seem like common sense, there was little research 30% 2011 32 % % 31 until recently on the question of whether parole supervision effectively reduces recidivism. Research is now 6 25% % beginning to show that inmates released to25 parole supervision are more likely to have better public safety 20% 23%max % These results demonstrate that supervision, when implemented with fidelity outcomes than those who 22out. 20% 15% to evidence-based practice, can reduce crime and corrections costs. 10% 5% 6 0% Violent Property Drug One of the largest studies, a national analysis of offenders released in 1994, found that inmates released by operation of law to mandatory supervision before the expiration of their sentences were rearrested no more than those who served their entire sentences and had no post-release supervision. 13 It also found that inmates released by a parole board to discretionary parole performed slightly better than max-outs. A recent Pew study in New Jersey found that parolees have better public safety outcomes than inmates who serve their full sentences. Among offenders released in 2008, fewer parolees than max-outs were rearrested (51 percent vs. 65 percent), reconvicted (38 percent vs. 55 percent), or returned to prison for a new crime (25 percent vs. 41 percent) within three years of release. Even controlling for key risk factors such as age, time served, current offense, and criminal history did not diminish the benefits of supervision; parolees were still 36 percent less likely to return to prison for new crimes. A recent Pew study found that New Jersey parolees are less likely to be rearrested, reconvicted, and reincarcerated for new crimes than inmates who max out their full prison sentences and are released without supervision. Even when controlling for key risk factors such as age, time served, current offense, and criminal history, parolees were 36 percent less likely to return to prison for new crimes within three years of release.” The Pew Charitable Trusts, “The Impact of Parole in New Jersey” (November 2013). The cost savings from releasing inmates to parole supervision before the end of their sentences could have been substantial, because the cost of parole supervision generally is a 10th that of incarceration. 14 But return of parolees for technical violations offset much of the savings: Parolees were just as likely (38 percent) as offenders released without supervision (39 percent) to return to prison within three years of release, either for a new crime or a revocation. Kentucky’s mandatory reentry supervision policy, which was part of the state’s Public Safety and Offender Accountability Act of 2011 (H.B. 463), requires that certain prisoners be released to supervision no less than six months before their sentences expire if they have not already been granted discretionary parole. The policy took effect in January 2012, and more than 3,700 offenders were released through the program in the first year. The Kentucky policy was recently evaluated, and early results are encouraging. Among offenders paroled through the program, just 4.2 percent were returned to prison for a new crime, compared with 6 percent of those in a comparable group who maxed out or were released with fewer than six months remaining on their sentences before the policy was implemented. 15 According to the Kentucky Department of Corrections, the policy has already saved the state more than $29 million. 16 These results support the use of parole as part of the offender reentry process. They also suggest that with appropriate policies and practices, post-prison supervision can cut both recidivism and corrections costs. 7 Figure 4 Poll: Mandate Supervision for All Offenders Released From Prison In a 2012 national survey, voters strongly preferred that inmates be subject to a period of mandatory supervision, rather than be held until their sentences expire and released without any supervision, regardless of offense type. Violent offenders Strongly prefer When given a choice between violent offenders serving a full 5-year prison sentence or 4 years of a 5-year term plus a year of mandatory supervision, voters preferred the mandatory supervision option. Shorter sentence, plus supervision 49% Full sentence, no supervision 21% Shorter sentence, plus supervision Full sentence, Democrats no supervision Independents With supervision 72% 66% No supervision 24% 25% When given a choice between nonviolent offenders serving a full 3-year prison sentence or 2 years of a 3-year sentence plus a year of mandatory supervision, voters preferred the mandatory supervision option. Full sentence, no supervision 49% Nonviolent crime victim Law enforcement member 62% 68% 69% 62% 30% 24% Strongly prefer Democrats With supervision 72% 67% No supervision 23% 25% 23% Total prefer 18% 25% Strongly prefer Total prefer 51% 18% 8 69% 25% Household type Violent crime victim Nonviolent crime victim Law enforcement member 67% 74% 74% 67% 26% 21% 21% 28% Independents Republicans Source: The Pew Charitable Trusts, “Public Opinion on Sentencing and Corrections Policy in America” (March 2012), http://www.pewstates.org/uploadedFiles/PCS_Assets/2012/PEW_NationalSurveyResearchPaper_FINAL.pdf. © 2014 The Pew Charitable Trusts 34% 69% 51% Party affiliation Total prefer 67% Violent 26% crime victim 21% Republicans Shorter sentence, plus supervision Full sentence, no supervision Total prefer Household type Shorter sentence, plus supervision Nonviolent offenders 67% 26%Strongly prefer Party affiliation Total prefer Total prefer Policy framework to reduce max-outs With the benefit of new research showing the public safety and fiscal benefits of parole, many states are now adopting policies to reverse the max-out trend and invest in reentry and post-release supervision. Pew recommends the following framework for reducing max-outs and recidivism: 1. Require a period of post-prison supervision for all offenders. 2. Carve out community supervision period from prison terms. 3. Strengthen parole decision-making. 4. Tailor supervision conditions to risk and need. 5. Adopt evidence-based practices in parole supervision. 6. Reinvest savings in community corrections. 1. Require a period of post-prison supervision for all offenders Research indicates that supervision, when implemented properly, is likely to reduce criminal activity and can save substantial taxpayer dollars. To capitalize on this, states should adopt policies requiring post-prison supervision for all inmates. In general, offenders should be subject to a period of supervision after release, but it is especially important for those convicted of violent, serious, or repeat offenses. However, to ensure that supervision levels are tailored to the circumstances of individual cases and to make efficient use of scarce community corrections resources, there will be instances in which an offender does not warrant active supervision. In such low-risk cases, states have at least two options: a. Place the offenders on inactive supervision to conserve resources but retain a mechanism to receive reports about their residence and employment. b. Place them on the lowest level of active supervision and encourage them to earn their way to inactive supervision by complying with all rules and requirements. At least eight states have passed mandatory reentry supervision policies in recent years—Kansas, Kentucky, New Hampshire, North Carolina, Ohio, Oklahoma, South Carolina, and West Virginia. While these states vary in how they chose to expand post-release monitoring, each mandates a period of supervision for at least some inmates who would otherwise have maxed out of their sentence. As part of its omnibus Crime Reduction and Sentencing Reform Act of 2010, South Carolina requires that nonviolent inmates who have been incarcerated for at least two years be released to mandatory supervision 180 days before their release date. 17 In North Carolina, until recently, only the most serious felony offenders were supervised after release. Now, as part of its Justice Reinvestment Act of 2011, the state requires nine months of supervision for all felony offenders.18 2. Carve out community supervision period from prison terms Although adding a supervision period to the end of every prison term may seem an obvious way to ensure that all inmates receive post-release supervision, research indicates that for many offenders adding time would add costs without providing any public safety benefit. 19 Instead, the period of post-release supervision should be carved 9 Victims’ Statement of Principles Beyond the benefits to public safety and states’ fiscal health, ensuring that released offenders undergo supervision gives victims of crime an official source of information about offenders. Knowing that an offender’s activities and whereabouts are subject to restrictions and monitoring can provide some peace of mind to a victim who is concerned about potential future contact with the offender. More than 100 national and state victim advocates have signed on to a set of guiding principles, several of which address the issue of max-outs and mandatory supervision directly. Among them:* •• “An ultimate goal of public safety policy is to reduce crime, resulting in fewer people and communities who are harmed.” •• “Mandatory supervision of offenders who pose a serious risk to public safety is essential both upon their return to the community and throughout the reentry process to promote victim and survivor safety.” •• “It is important for offenders to receive just punishment. The quantity of time that convicted offenders serve under any form of correctional supervision, however, must be balanced with the quality of evidence-based assessment, treatment, programming, and supervision they receive that can change their criminal behavior and thinking and reduce the likelihood that they will commit future crimes. For many offenses and offenders, shorter prison terms are acceptable if the resulting cost savings are reinvested in evidence-based programs that reduce recidivism.” * The Pew Charitable Trusts and the Council of State Governments’ Justice Center, “Sentencing, Corrections and Public Safety Guiding Principles for Crime Victims and Survivors in America,” http://www.pewstates.org/uploadedFiles/PCS_Assets/2012/ Guiding_Principles_for_Crime_Victims.pdf out of the prison term. Shifting selected offenders from expensive prison beds to evidence-based community supervision before the expiration of their sentences can reduce corrections costs and reoffense rates. One Pew study found that a significant portion of inmates in three states could have been released sooner with little or no impact on recidivism rates or on overall crime. 20 Specifically, the study found that 14 percent of all offenders released in Florida, 18 percent in Maryland, and 24 percent in Michigan in 2004 could have served prison terms three months to two years shorter with no public safety consequences. Had these inmates served shorter prison terms, these three states would have saved up to $54 million, $30 million, and $92 million, respectively. Regardless of whether a state has a determinate or indeterminate release structure, a period of post-release supervision can be carved out of the original court sentence. States such as Oregon have established mechanisms ensuring that every inmate receives a period of post-release supervision in a determinate system without sacrificing uniformity in time-served requirements. But states with indeterminate systems, such as Kentucky, New Hampshire, and South Carolina, also have adopted policies to reduce prison time served through supervised release. 21 These policies require parole boards to release inmates to community supervision before the expiration of their sentences. 10 3. Strengthen parole decision-making Although the proportion of inmates released by the decision of a parole board has dropped significantly, more than one-quarter of all prison releases still result from discretionary decisions. In 2011, more than 161,000 offenders were released conditionally in this manner. Ensuring that these decisions are anchored in evidence about how best to assess and manage offenders is essential. In past decades, many state parole boards lacked sufficient information and training to guide their decisions about who was suitable for release and what level of supervision and accompanying conditions should be imposed. Instead, board members often relied on personal experience and subjective judgment in evaluating offenders. A growing number of states have begun to address this issue, equipping board members with modern riskand needs-assessment instruments and other tools to fortify their review and decision-making. Many parole boards are also professionalizing through the establishment of educational or subject-matter requirements for appointees and ongoing training in evidence-based practices. At least 10 states have improved the decisionmaking process by reforming parole hearings and eligibility standards. 22 Michigan reduced its max-out rate by strengthening reentry efforts and ensuring that offenders are adequately supervised. The Michigan Prisoner Reentry Initiative, launched in 2005, targets moderate- to high-risk offenders for evidence-based programming and services, helping reduce revocations of probation and parole for new Straight From Solitary to the Street Solitary housing units or administrative segregation often are used to deal with difficult prisoners or to punish infractions. Inmates are removed from the general prison population, restricted to a confined space for up to 23 hours a day, and allowed limited human contact. This practice has troubling implications for inmates who max out and are released to their communities directly from prolonged segregation. The effects of solitary confinement have been well-documented, and the lack of monitoring or support in the community is particularly problematic for this high-risk, high-need population. One study found that these inmates developed psychopathologies at higher rates than those housed in the general population, exacerbating the challenge of reentry.* Another concluded that inmates released directly from solitary confinement had higher recidivism rates compared with offenders in the general population.† Colorado’s new corrections director, Rick Raemisch, recently spent 20 hours in a segregation cell. In an op-ed in The New York Times, he recounted his experience and highlighted the need and urgency for reform.‡ Colorado has significantly reduced max-outs from solitary, from 140 releases in 2012 to 70 in 2013 to just two through February 2014. * H.S. Andersen et al., “A Longitudinal Study of Prisoners on Remand: Psychiatric Prevalence, Incidence, and Psychopathology in Solitary vs. Non-solitary Confinement,” Acta Psychiatrica Scandinavica 102, no. 1 (2000): 19–25. † David Lovell, L. Clark Johnson, and Kevin C. Cain, “Recidivism of Supermax Prisoners in Washington State,” Crime and Delinquency 53, no. 4 (2007): 633–56. ‡ Rick Raemisch, “My Night in Solitary,” The New York Times, Feb. 20, 2014. 11 crimes and technical violations. The initiative gives the Michigan Parole Board greater confidence in releasing offenders, and the results are evident. Between 2005 and 2012, parole-granting rates increased for every crime category, 23 and the number of inmates maxing out was cut in half. 24 Pennsylvania’s Board of Probation and Parole plans to cut its max-out rate by reducing administrative delays in its review and release process. As part of the justice reinvestment framework it adopted in 2012, the state expects to increase parole hearings 20 percent by 2015 and reduce the number of inmates being released without supervision. 25 4. Tailor supervision conditions to risk and need Maximizing the public safety benefits and cost savings of post-release supervision requires that parole boards and other decision-makers assign appropriate supervision levels and develop case management plans tailored for each offender. Risk and needs assessment have become a critical ingredient in this process. Using increasingly sophisticated tools, officials can estimate the probability of recidivism and identify criminal risk factors that can be addressed to reduce the likelihood of reoffending. These tools should guide release, violation, and revocation decisions, particularly the intensity of supervision. The importance of matching the intensity of supervision and programming to risk cannot be overstated. Focusing on the highest-risk offenders provides the greatest return on investment in terms of crime prevention, but it also ensures that low-risk offenders are not being over-supervised. Research has shown that intensive programs, which work to reduce reoffending among high-risk populations, can actually increase recidivism among low-risk offenders. 26 As stated above, offenders who are assessed as posing a very low risk of committing new crimes can, in some cases, be placed on an “administrative” caseload, a status that typically involves little in-person contact with parole authorities. This reduces the burden on community corrections resources and targets supervision to the highest-risk offenders. 5. Adopt evidence-based practices in parole supervision In addition to tailoring conditions of supervision based on risk and need, researchers have identified several strategies that policymakers can adopt to strengthen parole supervision. In 2008, the Urban Institute published a report highlighting best practices that have been shown to improve public safety and make the best use of taxpayer dollars. 27 Among the 13 strategies identified were “frontloading” supervision resources, incorporating incentives for good behavior, and employing swift and certain responses to violations when they occur. Frontloading supervision, both surveillance and treatment, in the days immediately after release is supported by research demonstrating the importance of the reentry period. One study found that the probability of rearrest during the first month after release was nearly double that of the 15th month. 28 Oregon, which has the nation’s lowest max-out rate, assigns a reentry coordinator to all inmates six months before release and 12 develops a supervision plan tailored to the risk and needs of the individual. In 2013, the state built on this policy, enacting legislation that, among other things, allows inmates to be released to transitional reentry—an intensive community supervision program—three months before the end of their sentences. 29 Behavioral incentives, such as offering offenders the chance to reduce the length of their supervision terms by meeting certain conditions, can be a powerful motivational tool. With parole revocation representing the “stick,” these earned discharge programs add a “carrot” to the mix, encouraging offenders to obtain and keep jobs, stay sober, abide by other conditions of release, and avoid new criminal behavior. 30 Missouri, South Carolina, and South Dakota are among the states that have passed laws authorizing community corrections agencies to reduce supervision terms by up to 30 days for every month an offender complies with court or parole board conditions. Arkansas’ Public Safety Improvement Act of 2011 gives its community corrections agency the discretion to discharge offenders halfway through their supervision terms if they have complied with specified conditions. 31 Other states have recognized the value of responding to noncompliance with graduated sanctions, thereby limiting the use of costly prison beds to revocations. States have employed a wide range of swift and certain sanctions, using everything from increased reporting to short jail stays to address violations. Several states also cap the length of time an offender can be returned to prison for a technical violation, ensuring the sanction is proportional to the violation. North Carolina, for example, established a 90-day cap on revocation time for technical violations of post-release supervision. 32 6. Reinvest savings in community corrections Requiring offenders to serve the final portion of their sentence under community supervision frees up prison beds and generates savings. It also would add thousands of offenders to the caseloads of supervision officers. States should reinvest some portion of the prison savings into strengthening their community corrections systems, ensuring that probation and parole agencies have sufficient resources to successfully monitor and support highrisk, high-need offenders. South Dakota’s Legislature embraced the strategy of reinvesting in community corrections in passing the Public Safety Improvement Act (S.B. 70) in early 2013. The law is projected to save state taxpayers $207 million in prison construction and operating costs through 2022. 33 Under S.B. 70, $8 million of the savings were redirected into evidence-based programs in the first year, with another $4.9 million expected to follow annually. The reinvestment includes $3.2 million in expanded substance abuse, mental health, and cognitive-based treatment services for probation and parole populations. Georgia’s comprehensive sentencing and corrections reforms in 2012 were projected to save $264 million over five years. The governor and Legislature reinvested $11.6 million of averted prison spending into accountability courts that focus on drug offenders and those with mental illness, and another $5.7 million into new residential substance abuse treatment programs to reduce recidivism among inmates with drug and alcohol addictions.34 13 Conclusion Policymakers are increasingly recognizing the fragility of the successful resumption of community life after incarceration. States and the federal government are committing significant resources to improve reentry planning and strengthen community supervision in response to evidence of their effectiveness in protecting public safety by preventing recidivism. Research shows that during the months immediately after their release from prison—a period consumed by finding employment and housing and reconnecting with family—offenders are at the greatest risk of committing new crimes. Recent evidence also shows that supervised parolees are less likely to engage in new criminal behavior than max-outs. By employing evidence-based supervision practices, including using graduated sanctions rather than revocations for technical violations, states can reduce costs and recidivism. Despite these developments, 1 in 5 offenders serves a full sentence behind bars and is released from prison without any supervision in the community. This rise in max-outs is not driven by serious, chronic, violent offenders. On the contrary, nonviolent offenders have driven the growth in time served and max-outs since 1990. In response, many states are crafting policies to ensure that inmates undergo a period of post-release supervision and support. This most recent trend holds promise for both public safety and public spending. Appendix A: Methodology This study uses two data sources from the U.S. Department of Justice’s Bureau of Justice Statistics, or BJS. The main source is the National Prisoner Statistics, or NPS, data, which provide annual numbers on prisoners in state and federal prison facilities. Pew utilized published release numbers from the BJS Prisoners in 2012 report as a baseline to calculate the number of inmates released with and without supervision in 2012. To keep the methodology consistent with the NPS survey, unconditional releases were used as a proxy to measure unsupervised releases. This included expiration of sentence, commutations, and other unconditional releases. Similarly, conditional releases were used as a proxy for supervised releases and included releases to probation, supervised mandatory releases, discretionary parole releases, and other conditional releases. Totals for all releases include unconditional releases, conditional releases, deaths, releases to appeal or bond, and other releases, and they exclude transfers, escapes, and those absent without leave. However, due to varying interpretations of the NPS survey instrument and subsequent reporting practices by state respondents, the unconditional releases reported to BJS by some states included inmates who were actually released to supervision. For example, in some states that employ split sentences, the inmates released to a new probation term are reported as “unconditional releases.” Other states count inmates who are no longer under their jurisdiction as “expiration of sentence,” regardless of whether they are being supervised by another agency. As a result of these discrepancies, Pew followed up with the 50 states to verify the 2012 releases reported to the Bureau of Justice Statistics and clarify potential discrepancies. To accurately capture the number of max-outs, states were asked to report on all releases to supervision, regardless of supervision type or jurisdiction. Specifically, states were asked to provide 2012 release numbers broken down by the same categories as National Prisoner Statistics, with a note to include inmates serving a split sentence or inmates released to supervision under the jurisdiction of another agency in the supervised release category. Seven states revised their numbers to reflect this clarification. Three others provided revised numbers to reflect recent system or coding changes, and one state that did not submit the 2012 NPS numbers in time for publication by BJS provided those data to Pew. 14 See the jurisdiction notes on Page 15 for additional information on these revisions. The final sample included 47 states, all of which were verified through direct communication with the state or published corrections department website reports. Alaska, Illinois, and Rhode Island’s release numbers could not be verified and were excluded from the analysis. The remaining 47 states represent 94 percent of all releases in 2012. To compare and calculate the change in releases between 1990 and 2012, NPS numbers for 1990 were requested from states that revised their 2012 numbers. As a result, seven states—Alaska, Connecticut, Illinois, Massachusetts, North Carolina, Rhode Island, and Washington—were excluded from the 1990 count. The remaining 43 states represent 87 percent of all releases. To supplement the NPS data and provide information on the offense type of those released without supervision, Pew analyzed survey data from the National Corrections Reporting Program. The survey compiles individual-level data on admissions and releases from state and federal prisons and includes such information as demographics, conviction offenses, sentence length, and type of release. This offense-based analysis was limited to 19 states that had data for 2000 and 2011, the most recent year available. The 19 states included in the National Corrections Reporting Program analysis are Alabama, California, Colorado, Florida, Georgia, Kentucky, Maryland, Michigan, Minnesota, Missouri, New York, North Carolina, Oklahoma, Pennsylvania, South Carolina, South Dakota, Utah, Washington, and West Virginia. Records were divided into four mutually exclusive offense categories—violent, property, drug, and other—using the survey’s crime classification for which each offender received the longest sentence. Using the categories, the type of release for each was calculated for 2000 and 2011. Appendix B: Jurisdiction notes California’s first full year of implementation of the Public Safety Realignment Act was 2012. As part of the legislation, jurisdiction for inmates who were classified as “nonserious, nonviolent, nonsex registrant” was transferred to California’s 58 counties. These inmates are required to be supervised on post-release community supervision by county probation agencies. The state revised its release numbers to include the “non-nonnon” inmates released to post-release community supervision by the state Department of Corrections and Rehabilitation as supervised releases. For more information, see the box on Page 5. Connecticut Department of Correction’s Division of Parole and Community Services represents the consolidated community supervision and enforcement functions of the Department of Correction and the Board of Pardons and Paroles. The Court Support Services Division functions under the judicial branch and oversees pretrial services, family services, and probation supervision. A count of inmates serving split-sentence probation terms with a mandatory, court-imposed period of supervision after their term of incarceration was requested from the support services division. These releases were included in the supervised inmates count. A similar breakout of split-sentence probation terms could not be completed for 1990, and the state was excluded from the 1990 count. Delaware operates an integrated prison and jail system. National Prisoner Statistics data included both jail and prison populations. The state revised its release numbers to include only inmates with a sentence of one year or more. Georgia revised its release numbers to reflect changes to its data system that allow for more accurate coding of release type. Specifically, in early 2013, Georgia changed its admission and release codes from its old mainframe system to its current inmate system, SCRIBE. 15 Idaho revised its release numbers to reflect a discrepancy that occurred as a result of converting offender systems. Massachusetts revised its release numbers to include inmates whose terms of incarceration ended and who were then placed under a period of supervision via probation as supervised releases. The state was unable to provide a similar count for 1990 and was excluded from the totals for that year. Nevada did not submit National Prisoner Statistics data in 2012. The numbers published in the Prisoners in 2012 report were estimated using imputation procedures. The state provided Pew with 2012 release numbers, which were used in this report. New Jersey revised its release numbers to include inmates released through the state’s No Early Release Act, which requires them to be supervised for a short period post-release, as supervised releases. The act did not take effect until 1997, so there were no changes to or impact on the 1990 data. North Carolina revised its release numbers to include inmates who were released to probation after serving a period of incarceration as supervised releases. The state was unable to confirm the 1990 numbers and thus was excluded from the totals for that year. South Dakota revised its release numbers to include only inmates with a sentence of one year or more and to account for parolees who had been double-counted in a previous National Prisoner Statistics count. Washington revised its release numbers to include only inmates with a sentence of one year or more. The count does not include inmates who were under the jurisdiction of the state Department of Corrections and released directly from supervision without serving time in prison. Before 1996, probation and parole were managed by the counties, and inmates released to these forms of supervision were not under the jurisdiction of the department. Therefore, these releases were reported as “expiration of sentence.” Because the number of inmates released to county supervision in 1990 could not be identified, the state was excluded from the count for that year. Endnotes 1 Pew’s analysis includes data on inmates released from state prisons in 47 states in 2012. Data from Alaska, Illinois, and Rhode Island could not be verified and were excluded from this report. Sentenced prisoners in the District of Columbia are the responsibility of the Federal Bureau of Prisons and are excluded from this report. 2 The Pew Charitable Trusts, Pew Center on the States, Time Served: The High Cost, Low Return of Longer Prison Terms (June 2012), http:// www.pewstates.org/research/reports/time-served-85899394616. 3 The Pew Charitable Trusts, “The Impact of Parole in New Jersey” (November 2013), http://www.pewstates.org/uploadedFiles/PCS_ Assets/2013/PSPP_NJParole-Brief.pdf. 4 The Pew Charitable Trusts, Time Served. 5 Violent Crime Control and Law Enforcement Act of 1994, 42 U.S.C. § 14141, http://www.justice.gov/crt/about/spl/policestat.php; and U.S. Department of Justice, “Violent Crime Control and Law Enforcement Act of 1994 Fact Sheet,” accessed March 4, 2014, https://www. ncjrs.gov/txtfiles/billfs.txt. 6 U.S. Department of Justice, Bureau of Justice Statistics, “Reentry Trends in the U.S.,” last modified April 30, 2014, http://www.bjs.gov/ content/reentry/releases.cfm. 7 National Archive of Criminal Justice Data, “National Prison Statistics, 1978–2011, ICPSR 34540,” http://www.icpsr.umich.edu/icpsrweb/ NACJD/studies/34540/version/1. 8 Ibid. 16 9 Florida Department of Corrections, “Doing Time: Most Florida Inmates Serving More Than 85% of Their Sentences” (January 2014), http://www.dc.state.fl.us/pub/timeserv/doing/DoingTime2014.pdf. 10 The Pew Charitable Trusts, “The Impact of Parole in New Jersey.” 11 Oregon State Legislature, Chapter 144.775–144.790 (2013), http://www.oregonlegislature.gov/bills_laws/lawsstatutes/2013ors144.html; and Oregon Board of Parole and Post-Prison Supervision, “Historical Administrative Rules and Exhibits (255)” (2008), accessed March 4, 2014, http://www.oregon.gov/BOPPPS/pages/historical_rules.aspx. 12 The data used in the analysis were obtained from the National Corrections Reporting Program and based on records from 19 states for 2000 to 2011. These 19 states accounted for 60 percent of unconditional state prison releases in 2011. 13 Amy L. Solomon, Vera Kachnowski, and Avinash Bhati, “Does Parole Work? Analyzing the Impact of Postprison Supervision on Rearrest Outcomes,” Urban Institute (March 2005), http://www.urban.org/UploadedPDF/311156_Does_Parole_Work.pdf. 14 The Pew Charitable Trusts, Pew Center on the States, One in 31: The Long Reach of American Corrections (March 2009), http://www. pewstates.org/uploadedFiles/PCS_Assets/2009/PSPP_1in31_report_FINAL_WEB_3-26-09.pdf. In 2007, the national average cost of one day in prison was $78.95. The average cost of a day on parole was $7.47. 15 Forthcoming report from The Pew Charitable Trusts on an evaluation of Kentucky’s mandatory reentry supervision policy, expected June 2014. Returns were within one year of release. 16 Savings calculated by the Kentucky Department of Corrections. These calculations are based on the number of days released inmates are not incarcerated multiplied by the daily cost to incarcerate, less the additional daily supervision cost for each mandatory reentry supervision offender. They do not reflect other implementation costs associated with H.B. 463, including the costs hiring of additional probation and parole officers. 17 South Carolina General Assembly, Omnibus Crime Reduction and Sentencing Reform Act of 2010 (S.B. 1154) (2010), http://www. scstatehouse.gov/sess118_2009-2010/bills/1154.htm. 18 North Carolina General Assembly, Justice Reinvestment Act of 2011 (H.B. 642) (2011), http://www.ncleg.net/sessions/2011/bills/house/ pdf/h642v9.pdf. 19 Paul Gendreau, Claire Goggin, and Francis T. Cullen, The Effects of Prison Sentences on Recidivism (1999), http://www.securitepublique. gc.ca/cnt/rsrcs/pblctns/ffcts-prsn-sntncs-rcdvsm/ffcts-prsn-sntncs-rcdvsm-eng.pdf; Thomas Orsagh and Jong-Rong Chen, “The Effect of Time Served on Recidivism: An Interdisciplinary Theory,” Journal of Quantitative Criminology 4, no. 2 (1988): 155–171; Washington State Institute for Public Policy, Sentences for Adult Felons in Washington: Options to Address Prison Overcrowding, Part II (2004), http://www. wsipp.wa.gov/ReportFile/879/Wsipp_Sentences-for-Adult-Felons-in-Washington-Options-to-Address-Prison- Overcrowding-Part-IIRecidivism-Analyses_Full-Report.pdf; Ilyana Kuziemko, Going Off Parole: How the Elimination of Discretionary Prison Release Affects the Social Cost of Crime, National Bureau of Economic Research Working Paper (2007), http://www.nber.org/papers/w13380; National Council on Crime and Delinquency, Accelerated Release: A Literature Review (2008); G. Matthew Snodgrass et al., “Does the Time Cause the Crime? An Examination of the Relationship Between Time Served and Reoffending in the Netherlands,” Criminology 49 (2011): 1149–1194, http://onlinelibrary.wiley.com/ doi/10.1111/j.1745-9125.2011.00254.x/pdf; and Thomas A. Loughran et al., “Estimating a Dose-Response Relationship Between Length of Stay and Future Recidivism in Serious Juvenile Offenders,” Criminology 47 (2009): 699–740, http:// nccdglobal.org/sites/default/files/%20publication_pdf/focus-literature-review.pdf. 20 The Pew Charitable Trusts, Time Served. 21 Kentucky General Assembly, Public Safety and Offender Accountability Act of 2011 (H.B. 463) (2011), http://www.lrc.ky.gov/statrev/ acts2011rs/0002.pdf; South Carolina General Assembly, Omnibus Crime Reduction and Sentencing Reform Act of 2010 (S.B. 1154) (2010), http://www.scstatehouse.gov/sess118_2009-2010/bills/1154.htm; and New Hampshire Legislature, Justice Reinvestment Act (S.B. 500) (2010), http://www.gencourt.state.nh.us/legislation/2010/SB0500.html. 22 The Pew Charitable Trusts , “Sentencing and Corrections Reforms in Justice Reinvestment States” (July 2013), http://www.pewstates.org/ uploadedFiles/PCS_Assets/2013/PSPP_Sentencing_and_Corrections_Reform_Matrix.pdf. 23 Michigan Department of Corrections, “Parole Approval Rates” (Feb. 19, 2013), http://www.michigan.gov/documents/corrections/Parole_ Approval_Rates_190318_7.pdf. 24 The change in Michigan’s max-out rate was calculated as part of the analysis for this report. 25 Council of State Governments Justice Center, “Justice Reinvestment in Pennsylvania: Dynamic Strategies for Reducing Corrections Costs 17 and Improving Public Safety” (December 2013), http://csgjusticecenter.org/wp-content/uploads/2013/12/PA_2-page_report.pdf. 26 The Pew Charitable Trusts, Pew Center on the States, “Risk/Needs Assessment 101: Science Reveals New Tools to Manage Offenders” (September 2011), http://www.pewstates.org/uploadedFiles/PCS_Assets/2011/Pew_Risk_Assessment_brief.pdf. 27 Urban Institute, Putting Public Safety First: 13 Parole Supervision Strategies to Enhance Reentry Outcomes (December 2008), http://www. urban.org/UploadedPDF/411791_public_safety_first.pdf. 28 National Research Council, Parole, Desistance From Crime, and Community Integration (Washington: National Academies Press, 2007). 29 The Pew Charitable Trusts, “Public Safety in Oregon,” last modified May 28, 2013, http://www.pewstates.org/research/state-fact-sheets/ public-safety-in-oregon-85899410540. 30 Urban Institute, Putting Public Safety First. 31 The Pew Charitable Trusts, Pew Center on the States, “Arkansas’s 2011 Public Safety Reform: Legislation to Reduce Recidivism and Curtail Prison Growth” (July 2011), http://www.pewstates.org/uploadedFiles/Pew_Arkansas_brief.pdf. 32 North Carolina General Assembly, Justice Reinvestment Act of 2011. 33 The Pew Charitable Trusts, Pew Center on the States, “South Dakota’s 2013 Criminal Justice Initiative” (June 2013), http://www. pewstates.org/uploadedFiles/PCS_Assets/2013/PSPP_SD_2013_Criminal_Justice_Initiative_.pdf. 34 The Pew Charitable Trusts, Pew Center on the States, 2012 “Georgia Public Safety Reform” (July 2012), http://www.pewstates.org/ uploadedFiles/PCS_Assets/2012/Pew_Georgia_Safety_Reform.pdf. 18 pewtrusts.org Philadelphia Washington