# The Better Systems Project Brief #001: Prisons ## A Better Prison System Evidence-based accountability, rehabilitation, restitution, and fewer future victims Version: v0.1 working draft Date: 2026-07-02 Status: Public reviewer draft, not a final policy submission Related public series: Patch Notes for Society #001: Prisons Companion files: public evidence matrix and review hub linked from the public reviewer bundle ## Purpose This memo turns the public Patch Notes prison essays into a reviewer-ready policy artifact. It is not a final proposal. It is a structured draft meant to be criticized by researchers, correctional operators, victim advocates, policymakers, formerly incarcerated people, prison staff, and funders. The core test is practical: can the justice system protect the public, hold people accountable, reduce future victimization, and produce better outcomes than today's prison model without pretending every person, crime, institution, or state can be handled the same way? ## Executive Summary The United States already spends heavily on incarceration, supervision, courts, jails, policing, and reentry. Yet the system often fails on its own stated goals: too many victims are created before intervention, too many people leave prison without the skills or support needed to avoid reoffending, and too many communities inherit the costs of untreated trauma, addiction, poverty, violence, and weak institutional design. This memo proposes a gradual prison-system redesign built around four principles: 1. Public safety comes first, and public safety should be measured by fewer future victims, not only by the length or severity of punishment. 2. Accountability and rehabilitation are complements, not opposites. 3. Programming should be evidence-based, risk-calibrated, and matched to individual criminogenic needs. 4. Reform should be piloted, evaluated, scaled, revised, or stopped based on measurable results. The proposed strategy is not to "replace prisons" in one sweep. It is to build a parallel operating model inside selected jurisdictions and facilities: structured risk and needs assessment, individualized case plans, correctional education, vocational training, cognitive behavioral interventions, mental health and substance-use treatment, family and reentry support, victim-centered restitution where appropriate, and post-release continuity. The policy aim is a safer system, not a softer one. ## What This Memo Is Not Claiming This memo does not claim that: - every incarcerated person can or will be rehabilitated; - every offense type should be handled through the same intervention model; - risk assessment is neutral, perfect, or immune to bias; - program participation is the same thing as public safety; - victims owe forgiveness to people who harmed them; - restitution is simple to administer; - prison staff can absorb reform without resources, training, and operational changes; - correctional education, CBT, or reentry services work automatically when poorly implemented; - savings will appear quickly or flow to the same agencies that fund the pilot. The narrower claim is that the current system should be tested against a better operating model: one that keeps accountability, targets risk and need, measures results honestly, and scales only what works. ## Core Recommendation Launch a state-level Better Prison System Pilot in 3-5 facilities or reentry regions over five years. The pilot should: - prioritize medium- and high-need populations where intervention can plausibly reduce future harm; - exclude or separately study groups where standard rehabilitation models may not apply or where public-safety risk requires different controls; - use validated risk-needs assessment tools; - target criminogenic needs rather than generic self-improvement; - provide correctional education, vocational training, CBT-oriented programming, behavioral health care, and reentry planning; - measure victimization, recidivism, institutional safety, employment, housing stability, restitution completion, program fidelity, and cost per avoided reoffense; - publish results and revision notes on a fixed schedule. ## Problem Statement The current prison system has several overlapping failure modes: - It often incapacitates without repairing the conditions that drive future offending. - It creates weak feedback loops: prisons are judged by custody and compliance more than long-run public-safety outcomes. - Programming access is uneven, under-resourced, and frequently disconnected from reentry. - Victims and communities are often left with little meaningful repair. - Political incentives reward appearing tough more than proving what reduces harm. - Reform debates frequently collapse into a false binary: punishment versus compassion. The better frame is: what system creates fewer victims over time while maintaining moral accountability for harm already done? ## Baseline Data These are initial anchor facts for the v0.1 draft. The v0.2 memo should add state-by-state details, cost baselines, staffing ratios, and current program capacity. - BJS reported 5,625,400 adults under correctional supervision at yearend 2023, including 3,772,000 supervised in the community and 1,852,900 incarcerated in state or federal prisons or local jails. - BJS reported a 2023 U.S. prison population of 1,254,200, up 2% from 2022. - BJS's 10-year follow-up of people released from prisons in 24 states in 2008 found that about 66% were arrested within 3 years, 82% within 10 years, and 61% returned to prison within 10 years for a violation or new sentence. - RAND's correctional education meta-analysis found that incarcerated people who participated in correctional education had 43% lower odds of recidivating than those who did not, with an estimated 13 percentage-point reduction in recidivism risk. ## Theory of Change The system should be designed around a simple causal claim: If correctional facilities identify each person's risk level and criminogenic needs, deliver interventions matched to those needs, maintain institutional safety, connect prison programming to reentry supports, and measure results honestly, then the system can reduce future victimization while preserving accountability. This claim should be tested, not merely asserted. ## Evidence Base Working standard: every major claim in this memo should be traceable to the evidence matrix before the memo is treated as submission-grade. Claims can remain in the memo if they are explicitly labeled as assumptions, hypotheses, or reviewer questions. They should not be presented as settled facts unless the evidence supports that level of confidence. The evidence base is strongest for the broad direction of the memo: risk-calibrated intervention, targeting criminogenic needs, correctional education, reentry continuity, and careful measurement. It is weaker where the memo moves from evidence-backed principles to a full operating model. That does not mean the operating model is wrong; it means it should be treated as a pilot hypothesis. ### Evidence Confidence Summary | Claim area | Confidence | How the memo should use it | | --- | --- | --- | | Correctional populations and prison population baseline | Strong | Use as factual background. | | Long-run recidivism/rearrest after prison release | Strong | Use as evidence that current outcomes are not good enough, while distinguishing rearrest, reconviction, technical violations, and reincarceration. | | Risk-Need-Responsivity as a correctional rehabilitation framework | Strong | Use as a design principle, not as a single turnkey program. | | Correctional education reducing recidivism risk | Strong/moderate | Use as a promising evidence-backed component, with attention to program quality and labor-market relevance. | | Cognitive-behavioral interventions | Moderate | Use as an evidence-backed category that requires fidelity, matching, and implementation detail. | | Reentry continuity before and after release | Strong/moderate | Use as a core design principle and logistics requirement. | | Justice reinvestment as a funding/governance frame | Strong/moderate | Use as a plausible financing and policy framework, while noting savings timing and agency-mismatch risks. | | Victim-centered accountability and restitution mechanics | Moderate | Treat as a design requirement, but keep victim participation voluntary, distinguish restitution from victim compensation, and ask victim advocates to review details. | | Occupational licensing and blanket hiring restrictions | Strong/moderate | Treat as a practical reentry barrier where statutory and administrative review standards can matter, while avoiding one-size-fits-all recommendations across states and industries. | | Supervision technical violations | Strong/moderate | Separate technical violations from new criminal conduct in metrics and policy design, while acknowledging that some violations may signal elevated risk. | ### Risk-Need-Responsivity The Risk-Need-Responsivity model is one of the dominant frameworks in correctional rehabilitation. Public Safety Canada and the Andrews/Bonta/Wormith literature describe three core principles: - match service intensity to a person's risk of reoffending; - target criminogenic needs related to criminal behavior; - deliver treatment in a way responsive to the person's abilities, motivation, learning style, culture, and barriers. Implication for this memo: reform cannot be a generic menu of good programs. It has to allocate the right interventions to the right people at the right intensity. A v0.2 reviewer should also pressure-test the risk principle: intensive intervention for low-risk people can be wasteful or counterproductive, and risk assessment can produce unfair or overconfident classifications if used badly. ### Criminogenic Needs The memo should distinguish between needs that are morally or socially important and needs directly linked to reoffending. Examples to examine include antisocial cognition, antisocial peers, substance misuse, weak employment history, family instability, and poor impulse control. Implication: programming should be tied to risk reduction, not just personal enrichment. Education, therapy, work, and family programming may all matter, but the pilot should state which needs each program is meant to address and how those needs will be measured. ### Cognitive Behavioral Interventions CBT-oriented correctional programming is frequently used to address thinking patterns, decision-making, emotion regulation, and behavior. NIJ and CSG Justice Center materials support CBT/CBI as an evidence-based intervention category, but this memo should not imply that any program labeled "CBT" will work. Implication: the proposal should name specific intervention types and fidelity requirements rather than saying "therapy" broadly. The pilot should require curriculum selection, trained facilitators, attendance standards, completion measures, supervision integration, and post-release continuity where relevant. ### Desistance Desistance research focuses on how and why people stop offending over time. It often emphasizes identity change, social bonds, employment, family, maturity, and a credible path into lawful community life. Implication: reentry cannot be an afterthought at the prison gate. Release planning, employment pathways, documents, housing, treatment continuity, and community supervision must be designed as one system. The memo needs a deeper desistance source pass before v0.2, but the concept is useful because it prevents the pilot from treating prison programming and post-release life as separate worlds. ### Correctional Education and Vocational Training RAND's correctional education meta-analysis found substantially lower odds of recidivism among participants in correctional education programs. This is one of the clearest evidence anchors for a public audience because it connects public safety, employability, and cost-effectiveness. Implication: education and training should be treated as core public-safety infrastructure, with program quality, credential value, and post-release labor-market connection measured directly. The pilot should not count seat time as success; it should track credentials completed, employer relevance, post-release employment, wages, and whether participants can actually use the credential after release. ### Reentry and the Second Chance Act The Second Chance Act and the National Reentry Resource Center provide a practical federal-policy frame for reentry, not just a philosophical one. They point toward the kinds of services a serious pilot must coordinate: housing, employment, treatment, family reunification where appropriate, supervision, and community-based support. Implication: the pilot should be designed around a reentry pipeline, not a prison-only program. The release handoff should include documents, benefits, medication continuity, treatment appointments, housing plans, employment or training placement, supervision expectations, and follow-up accountability. ### Justice Reinvestment Justice reinvestment is a useful funding and governance frame because it starts from data, costs, recidivism, and resource allocation. The memo should not claim that savings will automatically appear or be easy to redirect. Implication: the funding model should identify which agency pays, which agency saves, how long savings take to appear, and what happens if the pilot improves outcomes without producing immediate budget savings. This is a central question for Urban Institute, CSG, Pew, and funder reviewers. ### Victim-Centered Accountability and Restitution Victim accountability should be designed as part of the operating model, not added as moral language at the end. The evidence and policy base supports a cautious distinction: restitution can be one form of accountability, but victim compensation, services, safety planning, and voice should not depend on whether an incarcerated or formerly incarcerated person can pay quickly. OVC materials show that restitution is an operational field: jurisdictions have tried statewide mandates, tracking and reporting systems, centralized restitution units, restitution courts, and juvenile restitution models. Restorative justice evidence is promising enough to review seriously, but it varies by model, population, offense type, preparation quality, and victim preference. A 2005 meta-analysis found restorative justice programs performed better than traditional non-restorative approaches on satisfaction, restitution compliance, and recidivism outcomes, while warning about self-selection bias. A later meta-analysis found a significant reduction in general recidivism odds, but not a significant effect on violent recidivism. The memo should not imply that victims owe offenders forgiveness, dialogue, or closure. Any restorative process should be opt-in, safety-screened, professionally facilitated, and separable from a victim's access to compensation or support. Implication: the pilot should track restitution obligations, payments, victim-service connections, and victim-safety outcomes where legally and ethically available. It should also include victim-advocate review before public claims are expanded. ### Occupational Licensing and Reentry Barriers Blanket employment and licensing exclusions can turn reentry into a trap: the system tells people to build a lawful life, then blocks lawful work even when the conviction is old or unrelated to the occupation. CSG Justice Center, National Reentry Resource Center, NELP, and NCSL materials all point toward a more targeted approach. NCSL's 2024 barriers-to-work report identifies automatic bans, vague "good moral character" clauses, fees, and unclear prequalification as recurring policy barriers. Implication: the legal-change section should recommend individualized review tied to relevance, risk, time since offense, rehabilitation evidence, and public-safety requirements. It should not argue that all restrictions are illegitimate. Some jobs require heightened screening. The policy question is whether a restriction is specific, evidence-informed, reviewable, and connected to the work. ### Supervision and Technical Violations Probation and parole can either support reentry or recreate incarceration churn. Technical violations should be measured separately from new criminal conduct because they answer different questions. CSG Justice Center's 2024 supervision-violations report found that in many states, a large share of prison admissions from community supervision were for technical violations rather than new offenses. A new violent offense, absconding, a missed appointment, a curfew violation, a failed drug test, and inability to pay fees should not be collapsed into one undifferentiated failure category. That said, technical violations are not automatically harmless. Some patterns can signal destabilization or rising risk. The reform question is whether the response improves safety and compliance, or whether it uses prison as the default response to low-level failure. Implication: the pilot should require graduated responses, clear violation categories, due process, service escalation where appropriate, and public reporting that separates new crimes, absconding, high-risk violations, and low-level compliance failures. ### Evidence Gaps That Should Stay Visible This memo is not yet fully sourced on: - victim-centered accountability details by offense type and victim preference; - restorative justice evidence by adult/youth population, offense type, and implementation model; - occupational licensing and employment restrictions by state and industry; - supervision technical-violation reform by state, including safeguards for violations linked to new criminal conduct or absconding; - state-level corrections costs and budget flows; - staff safety, burnout, and implementation burden; - political failure modes after a high-profile reoffense; - failed prison-programming pilots and why they failed. The reviewer process should focus heavily on these gaps. ## Accountability and Victims This proposal should not ask victims or the public to accept sentimental rehabilitation as a substitute for accountability. Accountability should include: - truthful recognition of harm; - consequences proportionate to the offense and risk; - victim input where appropriate and non-coercive; - restitution pathways where legally and practically possible; - structured obligations during incarceration and supervision; - behavioral standards inside facilities; - no automatic assumption that remorse or participation equals safety. The reform argument is that a system can hold people accountable and still ask whether the current model creates too many future victims by failing to change future behavior. ## Pilot Design ### Pilot Scope Start with 3-5 facilities or regional reentry pipelines in one or more states. Prefer facilities where leadership, data access, program capacity, and legislative support make implementation possible. ### Population Segmentation Segment participants by: - offense type; - sentence length and release horizon; - assessed risk level; - criminogenic needs; - mental health and substance-use needs; - institutional behavior; - reentry county or region; - victim-safety constraints. Do not promise equal expected results across all groups. The memo should state where evidence is strongest, where it is mixed, and where special safeguards are needed. ### Service Model Each participant receives: - validated risk-needs assessment; - individualized case plan; - required accountability and conduct standards; - education or credential pathway; - vocational pathway tied to real employer demand; - CBT-oriented programming where indicated; - behavioral health and substance-use treatment where indicated; - family and community support planning where safe and appropriate; - release-document checklist; - housing, employment, treatment, and supervision handoff; - post-release follow-up for at least 12-24 months. ### Facility Operating Changes The pilot should require: - program schedules protected from routine cancellation; - staff training in the model; - incentives for participation and completion; - quality control and fidelity checks; - data-sharing agreements across corrections, parole/probation, workforce agencies, treatment providers, and evaluators; - safeguards against gaming metrics by selecting only the easiest cases. ## Funding Model The v0.1 funding thesis is justice reinvestment: reduce avoidable costs from reincarceration, violations, emergency services, unemployment, and untreated behavioral health needs, then redirect savings into higher-quality intervention. Potential funding sources: - state corrections budgets; - federal reentry and justice grants; - workforce development funds; - Medicaid-eligible behavioral health support where legally available; - philanthropic pilot capital; - employer partnerships for training pipelines; - social-impact or outcomes-based financing only if measurement is strong and incentives are carefully designed. Funding questions for v0.2: - What does the pilot cost per participant? - Which costs are new versus reallocated? - What break-even reduction in reoffending or reincarceration is required? - What staffing ratios are needed? - Which services are already funded but poorly coordinated? - Which cost savings accrue to corrections versus local governments, victims, employers, hospitals, or families? ## Human Capital and Staffing The proposal needs a workforce plan, not just a program list. Roles likely required: - correctional officers trained to support the operating model while maintaining safety; - case managers; - educators; - vocational instructors; - behavioral health clinicians; - substance-use counselors; - reentry navigators; - victim-services professionals; - data and evaluation staff; - employer and apprenticeship coordinators. Risks: - staffing shortages; - burnout; - union or management resistance; - role confusion between custody and rehabilitation; - insufficient clinical capacity; - weak rural provider networks. ## Logistics and Supply Chain The operational plan should include: - classroom and training-space capacity; - secure technology access; - approved devices and connectivity; - tools and materials for vocational programs; - credentialing partners; - transportation for post-release appointments; - IDs, birth certificates, Social Security cards, and benefits applications; - medication continuity; - housing placement; - employer onboarding; - data systems that follow the person from facility to community. ## Laws and Administrative Changes Potential laws or rules to remove, revise, or examine: - unnecessary occupational licensing exclusions for people with records; - blanket hiring bans unrelated to job duties; - supervision rules that create technical reincarceration without improving safety; - restrictions that block education technology or credential access without a security rationale; - funding silos that prevent correctional education, treatment, and workforce systems from coordinating. Potential laws or rules to add: - independent pilot authorization and evaluation requirements; - reentry-document readiness requirements before release; - data-sharing authority with privacy protections; - evidence and fidelity standards for funded prison programs; - reporting requirements for recidivism, employment, housing, program access, victim restitution, and cost; - occupational licensing review standards tied to relevance and risk. ## Metrics Public safety metrics: - new arrests; - new convictions; - violent reoffending; - victimization estimates where available; - technical violations separated from new crimes; - time to first reoffense. Reentry metrics: - employment; - wages; - housing stability; - treatment continuity; - education and credential completion; - family reunification where appropriate; - supervision compliance. Institutional metrics: - assaults; - disciplinary incidents; - staff injuries; - program attendance and completion; - staffing levels; - lockdowns and program cancellations. Accountability metrics: - restitution orders and completion; - victim-service referrals; - participant completion of accountability requirements; - victim-safety conditions. System metrics: - cost per participant; - cost per avoided reincarceration; - prison-bed utilization; - parole/probation workload; - provider capacity; - program fidelity. ## Timeline ### Phase 0: Design and Authority, 0-6 Months - Select state or jurisdiction. - Identify facilities or regions. - Secure pilot authority. - Choose evaluator. - Build advisory group including victims, corrections staff, researchers, formerly incarcerated people, prosecutors, defense, workforce partners, and community providers. - Define eligibility, exclusions, metrics, and data-sharing agreements. ### Phase 1: Pilot Build, 6-12 Months - Train staff. - Hire or contract missing roles. - Establish assessment workflow. - Build education, treatment, and workforce partnerships. - Create reentry handoff process. - Baseline current outcomes. ### Phase 2: Controlled Launch, Year 1-2 - Enroll first cohorts. - Monitor fidelity and institutional safety monthly. - Publish early implementation notes. - Fix logistics before scaling. ### Phase 3: Evaluation and Revision, Year 2-3 - Compare participant outcomes to matched comparison groups where possible. - Separate implementation failure from theory failure. - Revise service mix and eligibility. - Publish revision log. ### Phase 4: Scale or Stop Decisions, Year 3-5 - Scale what works. - Stop or redesign what fails. - Expand to more facilities only with staffing, funding, and fidelity capacity. - Convert pilot authority into durable law only after evidence supports it. ## Failure Modes and Hard Objections The memo should explicitly address: - Some people may remain dangerous despite intervention. - Some programs fail when implemented poorly. - Risk assessment can encode bias or create false confidence. - Victims may experience reform language as minimizing harm. - Politicians may cherry-pick results. - Facilities may select low-risk participants to make outcomes look better. - Staff may experience reform as added burden without resources. - Employers may not hire participants even after training. - Savings may not appear quickly or may accrue to the wrong budget. - Public support may collapse after a high-profile failure. The answer is not to deny these risks. The answer is to design guardrails, measure honestly, and avoid overclaiming. ## Reviewer Questions 1. What is wrong or naive in this theory of change? 2. Which populations should be included, excluded, or studied separately? 3. Which evidence-based interventions must be named specifically? 4. What does this memo understate about victims, staff, politics, or implementation? 5. What existing state or federal program is the closest model? 6. What would make the pilot credible to corrections leaders? 7. What would make it credible to victim advocates? 8. What metrics would be most easily gamed? 9. What legal changes are truly necessary versus merely useful? 10. Where should a serious version of this memo be submitted? ## Source List For v0.1 - Bureau of Justice Statistics, Correctional Populations in the United States, 2023 - Statistical Tables: https://bjs.ojp.gov/library/publications/correctional-populations-united-states-2023-statistical-tables - Bureau of Justice Statistics, Prisoners in 2023 - Statistical Tables: https://bjs.ojp.gov/library/publications/prisoners-2023-statistical-tables - Bureau of Justice Statistics, Recidivism of Prisoners Released in 24 States in 2008: A 10-Year Follow-Up Period (2008-2018): https://bjs.ojp.gov/library/publications/recidivism-prisoners-released-24-states-2008-10-year-follow-period-2008-2018 - RAND Corporation, Evaluating the Effectiveness of Correctional Education: https://www.rand.org/pubs/research_reports/RR266.html - Public Safety Canada, Risk-Need-Responsivity Model for Offender Assessment and Rehabilitation: https://www.publicsafety.gc.ca/cnt/rsrcs/pblctns/rsk-nd-rspnsvty/index-en.aspx - Andrews, Bonta, and Wormith, The Risk-Need-Responsivity Model, Criminal Justice and Behavior: https://journals.sagepub.com/doi/10.1177/0093854811406356 - National Institute of Justice, Cognitive-Behavioral Therapy: An Evidence-Based Intervention for Offenders: https://nij.ojp.gov/library/publications/cognitive-behavioral-therapy-evidence-based-intervention-offenders - Council of State Governments Justice Center, Using a Cognitive-Behavioral Approach in Programs to Reduce Recidivism: https://csgjusticecenter.org/2017/01/13/in-brief-using-a-cognitive-behavioral-approach-in-programs-to-reduce-recidivism/ - Bureau of Justice Assistance, Second Chance Act Programs: https://bja.ojp.gov/taxonomy/term/second-chance-act-sca-programs - National Reentry Resource Center, Second Chance Act: https://nationalreentryresourcecenter.org/second-chance-act - Urban Institute, Justice Reinvestment Initiative: https://www.urban.org/policy-centers/justice-policy-center/projects/justice-reinvestment-initiative - National Institute of Justice, Reentry: https://nij.ojp.gov/topics/corrections/reentry - Council of State Governments Justice Center, Victim Compensation Programs and Restitution: https://csgjusticecenter.org/2017/01/13/in-brief-victim-compensation-programs-and-restitution/ - Office for Victims of Crime, Restitution: https://ovc.ojp.gov/taxonomy/term/restitution - Office for Victims of Crime, Making Restitution Real: Five Case Studies on Improving Restitution Collection: https://ovc.ojp.gov/library/publications/making-restitution-real-five-case-studies-improving-restitution-collection - Office for Victims of Crime, Financial Support for Victims of Crime: A Quick Guide for Corrections and Community Supervision Officers: https://ovc.ojp.gov/library/publications/financial-support-victims-crime-quick-guide-corrections-and-community - Office of Juvenile Justice and Delinquency Prevention, Restorative Justice Literature Review: https://ojjdp.ojp.gov/sites/g/files/xyckuh176/files/media/document/restorative_justice.pdf - Latimer, Dowden, and Muise, The Effectiveness of Restorative Justice Practices: A Meta-Analysis: https://journals.sagepub.com/doi/10.1177/0032885505276969 - Fulham, Blais, Rugge, and Schultheis, The effectiveness of restorative justice programs: A meta-analysis of recidivism and other relevant outcomes: https://journals.sagepub.com/doi/10.1177/17488958231215228 - Council of State Governments Justice Center, Fair Chance Licensing Project: https://csgjusticecenter.org/projects/fair-chance-licensing/ - Council of State Governments Justice Center, The Consideration of Criminal Records in Occupational Licensing: https://csgjusticecenter.org/publications/the-consideration-of-criminal-records-in-occupational-licensing/ - National Conference of State Legislatures, Barriers to Work: People with Criminal Records: https://documents.ncsl.org/wwwncsl/Criminal-Justice/Barriers-to-Work-People-with-Criminal-Records.pdf - National Conference of State Legislatures, Barriers to Work: Increasing Access to Licensed Occupations for Individuals With Criminal Records: https://www.ncsl.org/labor-and-employment/barriers-to-work-improving-employment-in-licensed-occupations-for-individuals-with-criminal-records - Bureau of Justice Statistics, Probation and Parole in the United States, 2022: https://bjs.ojp.gov/library/publications/probation-and-parole-united-states-2022 - Council of State Governments Justice Center, Supervision Violations and Their Impact on Incarceration: https://projects.csgjusticecenter.org/supervision-violations-impact-on-incarceration/wp-content/uploads/sites/15/2024/01/Supervision-Violations-Impact-2024_508.pdf - Pew Charitable Trusts, Probation and Parole Systems Marked by High Stakes, Missed Opportunities: https://www.pew.org/en/research-and-analysis/issue-briefs/2018/09/probation-and-parole-systems-marked-by-high-stakes-missed-opportunities - National Conference of State Legislatures, Community Supervision: Limiting Incarceration in Response to Technical Violations: https://www.ncsl.org/civil-and-criminal-justice/community-supervision-limiting-incarceration-in-response-to-technical-violations ## v0.2 Research Additions Needed - CBT correctional intervention meta-analyses. - Desistance literature review. - Restorative justice and victim-centered accountability implementation guidance by population and offense type. - Reentry employment and housing studies. - State corrections budget examples. - Justice reinvestment implementation examples. - State-by-state legal appendix for licensing, restitution, and supervision reform. - First Step Act implementation lessons. - Examples of pilots that failed, and why.