Evidence-Based Sentencing to Improve Public Safety and Reduce Recidivism

1. 1 Evidence-Based Sentencing to Improve Public Safety and Reduce Recidivism � Judge Roger K. Warren (Ret.) President Emeritus National Center for State Courts

2. 2 The Three Basic Principles of EBP Risk Principle (Who) Needs Principle (What) Treatment/Responsivity Principle (What Works)

3. Risk Principle(Who) Lecture (3-4 minutes) Risk: risk of committing another offense, not risk of committing a violation of probation, not risk of violence or dangerousness, nor of committing a particular kind of offense. Distinguish between level of risk of re-offense and level of seriousness of an offense, e.g., a low risk offender committing a high level (serious) crime, or high risk offender committing a low level crime. The basic concept here is we do not want to expend or waste resources on low risk offenders who are not likely to re-offend in the first place, or on extremely high risk offenders who are not amenable to treatment and are likely to re-offend no matter what we do. Lecture (3-4 minutes) Risk: risk of committing another offense, not risk of committing a violation of probation, not risk of violence or dangerousness, nor of committing a particular kind of offense. Distinguish between level of risk of re-offense and level of seriousness of an offense, e.g., a low risk offender committing a high level (serious) crime, or high risk offender committing a low level crime. The basic concept here is we do not want to expend or waste resources on low risk offenders who are not likely to re-offend in the first place, or on extremely high risk offenders who are not amenable to treatment and are likely to re-offend no matter what we do.

4. Needs Principle(What) 4 The targets for interventions should be those offender characteristics that have the most effect on the likelihood of re-offending.

5. Risk of Heart Attack 1) Elevated LDL and low HDL levels 2) Smoking 3) Diabetes 4) Hypertension 5) Abdominal obesity 6) Psychosocial (i.e., stress/depression) 7) Failure to eat fruits and vegetables 8) Failure to exercise Lecture (3 Minutes) Some criminogenic needs are more criminogenic than others, i.e. more highly predictive of the likelihood of recidivism. By analogy, there are some risk factors for heart attack that are more highly predictive of having a heart attack than others. The risk of heart attack for individuals who had all of these factors, amazingly, was almost 130 times higher than for somebody with none of them. The first two of these risk factors however, (bad lipid readings and smoking), predicted 2/3 of all heart attacks. Adapted from slides of Chris Lowencamp. 1. International study that studied the risk factors associated with heart attack 2. Gathered data on all heart attacks 3. Compared to case-matched controls Lecture (3 Minutes) Some criminogenic needs are more criminogenic than others, i.e. more highly predictive of the likelihood of recidivism. By analogy, there are some risk factors for heart attack that are more highly predictive of having a heart attack than others. The risk of heart attack for individuals who had all of these factors, amazingly, was almost 130 times higher than for somebody with none of them. The first two of these risk factors however, (bad lipid readings and smoking), predicted 2/3 of all heart attacks. Adapted from slides of Chris Lowencamp. 1. International study that studied the risk factors associated with heart attack 2. Gathered data on all heart attacks 3. Compared to case-matched controls

6. Dynamic Risk Factors (Criminogenic Needs) Anti-social attitudes Anti-social friends and peers Anti-social personality pattern Family/marital Substance abuse Education Employment Anti-social leisure activities

7. Use of Actuarial Risk/Needs Assessment The engine that drives evidence-based recidivism reduction strategies Much more accurate in predicting recidivism Intended to inform not replace professional judgment Lecture (5-8 minutes) Under third bullet, mention that the risk/needs assessment tools we are discussing were not developed or intended to be used for the purpose of determining the severity of the criminal penalty or as a major criterion for deciding whether an offender should be imprisoned. (Email correspondence of Roger Warren with Don Andrews and Steve Wormith, developers of the LSI.) See, LSI-R Manual, Page 3, last paragraph: �This instrument was designed to assist in the implementation of the least restrictive and least onerous interpretation of criminal sanction, and to identify dynamic areas of risk/needs that may be addressed by programming in order to reduce risk. This instrument is not a comprehensive survey of mitigating and aggravating factors relevant to criminal sanctioning and was never designed to assist in establishing the just penalty.� Lecture (5-8 minutes) Under third bullet, mention that the risk/needs assessment tools we are discussing were not developed or intended to be used for the purpose of determining the severity of the criminal penalty or as a major criterion for deciding whether an offender should be imprisoned. (Email correspondence of Roger Warren with Don Andrews and Steve Wormith, developers of the LSI.) See, LSI-R Manual, Page 3, last paragraph: �This instrument was designed to assist in the implementation of the least restrictive and least onerous interpretation of criminal sanction, and to identify dynamic areas of risk/needs that may be addressed by programming in order to reduce risk. This instrument is not a comprehensive survey of mitigating and aggravating factors relevant to criminal sanctioning and was never designed to assist in establishing the just penalty.�

8. Treatment (Responsivity) Principle: Part I Lecture (1 minute) Importance of the treatment setting or environment: pro-social community environments are usually more conducive to behavioral change than more criminogenic institutional settings. If you take away only one thing from this unit, it should be this. There are virtually no serious competitors for the above proposition when it comes to changing criminal behavior. The goal is behavioral change, compliance with probation conditions and court orders, and, ultimately self management.  Under social learning theory (e.g., B.F. Skinner) social behaviors are learned over time through processes of social consequences, both positive and negative, carrots and sticks, rewards and punishments; consequences of past behaviors influence future behaviors: antecedents, behaviors, consequences (ABC�s). Target supervision, personal contacts, treatment and controls at dynamic risk f Social learning = positive reinforcement and swift, certain, and fair consequences Cognitive & behavioral Lecture (1 minute) Importance of the treatment setting or environment: pro-social community environments are usually more conducive to behavioral change than more criminogenic institutional settings. If you take away only one thing from this unit, it should be this. There are virtually no serious competitors for the above proposition when it comes to changing criminal behavior. The goal is behavioral change, compliance with probation conditions and court orders, and, ultimately self management. Under social learning theory (e.g., B.F. Skinner) social behaviors are learned over time through processes of social consequences, both positive and negative, carrots and sticks, rewards and punishments; consequences of past behaviors influence future behaviors: antecedents, behaviors, consequences (ABC�s). Target supervision, personal contacts, treatment and controls at dynamic risk f Social learning = positive reinforcement and swift, certain, and fair consequences Cognitive & behavioral

9. Treatment (Responsivity) Principle: Part II Lecture (1 minute) If you take away only one thing from this unit, it should be this. There are virtually no serious competitors for the above proposition when it comes to changing criminal behavior. The goal is behavioral change, compliance with probation conditions and court orders, and, ultimately self management.  Under social learning theory (e.g., B.F. Skinner) social behaviors are learned over time through processes of social consequences, both positive and negative, carrots and sticks, rewards and punishments; consequences of past behaviors influence future behaviors: antecedents, behaviors, consequences (ABC�s).Lecture (1 minute) If you take away only one thing from this unit, it should be this. There are virtually no serious competitors for the above proposition when it comes to changing criminal behavior. The goal is behavioral change, compliance with probation conditions and court orders, and, ultimately self management. Under social learning theory (e.g., B.F. Skinner) social behaviors are learned over time through processes of social consequences, both positive and negative, carrots and sticks, rewards and punishments; consequences of past behaviors influence future behaviors: antecedents, behaviors, consequences (ABC�s).

10. 10 What Doesn�t Work:Traditional Sanctions Punishment, sanctions, or incarceration Specific deterrence, or fear-based programs (e.g. Scared Straight) Physical challenge programs Military models of discipline and physical fitness (e.g. Boot Camps) Electronic monitoring Intensive supervision, without treatment

11. Evidence-Based Sentencing The application of EBP to the sentencing process for the purpose of reducing recidivism and holding offenders accountable 11

12. 12 State Sentencing Reform: The Recent History Pre-1975: the �Rehabilitative Ideal� Rapid rise in violent crime Disparities �Nothing works� 1975-2005: Determinate Sentencing, Punishment & Incapacitation Tectonic plates underlying criminal justice system are shifting: fundamental changes going on The implications of EBP for improving our current state sentencing and corrections systems is best understood in the context of the recent history of state sentencing and corrections. Pre-70�s IS rehabilitative model TX & threats Abolished ISL; transitioned to DSL (eliminating discretionary release) Greater reliance on incarceration and incapacitation Offense v. offender information Tectonic plates underlying criminal justice system are shifting: fundamental changes going on The implications of EBP for improving our current state sentencing and corrections systems is best understood in the context of the recent history of state sentencing and corrections. Pre-70�s IS rehabilitative model TX & threats Abolished ISL; transitioned to DSL (eliminating discretionary release) Greater reliance on incarceration and incapacitation Offense v. offender information

13. 13 Sentencing Reform: The Recent History (cont.) The Consequences �Nothing works�: a self-fulfilling prophecy Unprecedented recidivism rates Highest incarceration rates in the world Diminishing benefit of incapacitation Rapidly escalating costs Continued disparities Same violent crime rate as mid-70�s INCAR: 6 fold increase; 5-8X Europe; prop. crimes: 2-7X longer; less than 5% of world pop, but � of all prisoners; 14 states even higher than national average RECID: evid of increase in 80�s & 90�s: Nationally almost 60% have at least one prior conviction ;43% fail to successfully complete probation; About half (47% in �97) of prisoners were on probation or parole at time of new arrest. COSTS: Fastest growing area; 4X pub ed; 60-70X higher ed DISPARITIES: 2/3 white-2/3 people of color; 8:1: 9:1 DIMINISHING BENEFIT OF INCAPACITATION: 25% of crime red; 80% of impact on NV; � NV & serve short terms 10% inc incar=2-4% dec in crime rate 1 study found crime rate increases over incarceration rate of 325/100K $1 Mill investment to improve HS grad rates=decrease crime four times $1 Mill in prisons SAME VIOLENT CRIME RATE AS IN MID-70�S From naivet� to disbelief; lurched from one untested assumption to the opposite untested assumption Present opportunity to discard both untested assumptions and align with best available research evidence Exciting opportunity for system overly reliant on punishment and incar. overly fixated on past accountability, not future; Restore balance INCAR: 6 fold increase; 5-8X Europe; prop. crimes: 2-7X longer; less than 5% of world pop, but � of all prisoners; 14 states even higher than national average RECID: evid of increase in 80�s & 90�s: Nationally almost 60% have at least one prior conviction ;43% fail to successfully complete probation; About half (47% in �97) of prisoners were on probation or parole at time of new arrest. COSTS: Fastest growing area; 4X pub ed; 60-70X higher ed DISPARITIES: 2/3 white-2/3 people of color; 8:1: 9:1 DIMINISHING BENEFIT OF INCAPACITATION: 25% of crime red; 80% of impact on NV; � NV & serve short terms 10% inc incar=2-4% dec in crime rate 1 study found crime rate increases over incarceration rate of 325/100K $1 Mill investment to improve HS grad rates=decrease crime four times $1 Mill in prisons SAME VIOLENT CRIME RATE AS IN MID-70�S From naivet� to disbelief; lurched from one untested assumption to the opposite untested assumption Present opportunity to discard both untested assumptions and align with best available research evidence Exciting opportunity for system overly reliant on punishment and incar. overly fixated on past accountability, not future; Restore balance

14. 14 State Chief Justices Top concerns of state trial judges in felony cases: High rates of recidivism Ineffectiveness of traditional probation supervision in reducing recidivism Absence of effective community corrections programs Restrictions on judicial discretion

15. �What is done [today] in corrections would be grounds for malpractice in medicine.� (2002) Latessa, Cullen, and Gendreau, �Beyond Correctional Quackery�� 15

16. State Chief Justices Top two sentencing reform objectives: Reduce recidivism through expanded use of evidence-based practices, programs that work, and offender risk and needs assessment tools Promote the development, funding, and utilization of community-based alternatives to incarceration for appropriate offenders

17. EBS & Purposes of Sentencing �Just Deserts:� penalty or punishment proportionate to the gravity of the offense & blameworthiness of the offender Public Safety Rehabilitation Specific Deterrence Incapacitation/Control General Deterrence Restitution/Restoration 17 Lecture (3-4 minutes) EBS focuses only on the sentencing purpose of rehabilitation or recidivism reduction. Recidivism reduction is not the only, nor probably even the most, important purpose of sentencing. � Broadly speaking, there are three purposes of sentencing in an individual case: �Just deserts�: to punish in accord with the gravity of offense in light of the blameworthiness of the defendant and the extent of injury and damage done. Public safety: utilitarian strategies to promote public safety: reduce recidivism through rehabilitation and/or specific deterrence; prevent future crime by the offender through incapacitation or lesser behavioral controls; or deter crime by others through general deterrence, sending a message, etc. Restitution/restoration of victim, community See ALI Model Penal Code on Sentencing, Tentative Draft No. 1, Section 1.02 (2)(a) (2007). � Offense v. offender info Emphasize that recidivism reduction is only one purpose of sentencing, and that recidivism reduction and punishment should not be viewed as an either/or proposition. Use of EBS practices designed to reduce recidivism must be integrated with any other provisions of the sentence intended to carry out other applicable sentencing purposes. Lecture (3-4 minutes) EBS focuses only on the sentencing purpose of rehabilitation or recidivism reduction. Recidivism reduction is not the only, nor probably even the most, important purpose of sentencing. � Broadly speaking, there are three purposes of sentencing in an individual case: �Just deserts�: to punish in accord with the gravity of offense in light of the blameworthiness of the defendant and the extent of injury and damage done. Public safety: utilitarian strategies to promote public safety: reduce recidivism through rehabilitation and/or specific deterrence; prevent future crime by the offender through incapacitation or lesser behavioral controls; or deter crime by others through general deterrence, sending a message, etc. Restitution/restoration of victim, community See ALI Model Penal Code on Sentencing, Tentative Draft No. 1, Section 1.02 (2)(a) (2007). � Offense v. offender info Emphasize that recidivism reduction is only one purpose of sentencing, and that recidivism reduction and punishment should not be viewed as an either/or proposition. Use of EBS practices designed to reduce recidivism must be integrated with any other provisions of the sentence intended to carry out other applicable sentencing purposes.

18. 20 Evidence-Based Sentencing Practices 1. Don�t over-intervene with low risk offenders 2. The principal objective is to reduce the risk of recidivism 3. Encourage counsel to argue risk reduction at sentencing 4. Risk is dynamic; the risk score is static 18 Common sense but we do it all the time; low recidivism rates v. reduce recidivism; triaging; avoid over-intervention with extremely high risk also Intensive supervision; piling on probation conditions; mandating intensive treatment programs; enrolling in drug courts; prison where other purposes of sentencing (e.g. punishment) don�t require it 2. Reduce: a. not just manage, incapacitate, but change attitudes/behaviors so the risk is diminished even after or in absence of supervision or controls; b. desired outcome is reduced recidivism: RR is performance measure, not low recidivism rates; programs/PO�s admit and hold onto low risk. Risk: We do not control the outcome; the offender does. But we can influence. Changing the odds. The offender is in charge here, unlike in determining appropriate punishment. How can we motivate, encourage, facilitate, enable attitude and behavioral change? 3. Focus counsel�s attention on the goal of recidivism reduction; OR judicial conf res. 4. This insight is the reason we are here. Risk changes and can be changed. The problem is that risk prediction/risk score is static. Some low risk offenders recidivate; and some high risk offenders do not-typically b/c of things that happen after the assessment is conducted. Keep some minimal reporting responsibilities on low-risk and determine the amenability of high risk to effective intervention and target dynamic factors. Don�t assume high risk offender needs to be incapacitated; extremely high risk yes; only ? is where/how. Common sense but we do it all the time; low recidivism rates v. reduce recidivism; triaging; avoid over-intervention with extremely high risk also Intensive supervision; piling on probation conditions; mandating intensive treatment programs; enrolling in drug courts; prison where other purposes of sentencing (e.g. punishment) don�t require it 2. Reduce: a. not just manage, incapacitate, but change attitudes/behaviors so the risk is diminished even after or in absence of supervision or controls; b. desired outcome is reduced recidivism: RR is performance measure, not low recidivism rates; programs/PO�s admit and hold onto low risk. Risk: We do not control the outcome; the offender does. But we can influence. Changing the odds. The offender is in charge here, unlike in determining appropriate punishment. How can we motivate, encourage, facilitate, enable attitude and behavioral change? 3. Focus counsel�s attention on the goal of recidivism reduction; OR judicial conf res. 4. This insight is the reason we are here. Risk changes and can be changed. The problem is that risk prediction/risk score is static. Some low risk offenders recidivate; and some high risk offenders do not-typically b/c of things that happen after the assessment is conducted. Keep some minimal reporting responsibilities on low-risk and determine the amenability of high risk to effective intervention and target dynamic factors. Don�t assume high risk offender needs to be incapacitated; extremely high risk yes; only ? is where/how.

19. 20 Evidence-Based Sentencing Practices 5. Risk needs assessment (RNA) tools can provide critical information to the court in sentencing and probation revocation proceedings 6. Use RNA information to target proba-tion conditions (both treatment and control) and judicial interactions at the offender�s specific and most critical dynamic risk factors 19 5. RNA = the engine that drive RR strategies: accurate, reliable; esp. the dynamic risk info 6. a.Probation conditions estab.teh legal framework; direction and authority; b. improper prob. conditions are ineffective and needlessly distract and impede c.Judicial interactions include judge�s contacts with offender 5. RNA = the engine that drive RR strategies: accurate, reliable; esp. the dynamic risk info 6. a.Probation conditions estab.teh legal framework; direction and authority; b. improper prob. conditions are ineffective and needlessly distract and impede c.Judicial interactions include judge�s contacts with offender

20. 20 Evidence-Based Sentencing Practices 7. Wherever possible, defer to probation on level of supervision, monitoring, and control, and with respect to appropriate treatment, especially in the absence of reliable RNA information 8. Discourage plea bargaining of probation conditions 9. Inform parties that negotiated conditions are subject to modification 20 7. a. In absence of RNA info one cannot rationally intervene b. Circumstances change; c. PO can use for incentives & sanctions; d. testing/electronic monitoring-depending on state law 8. By rule, policy, training, refusal to accept plea 7. a. In absence of RNA info one cannot rationally intervene b. Circumstances change; c. PO can use for incentives & sanctions; d. testing/electronic monitoring-depending on state law 8. By rule, policy, training, refusal to accept plea

21. 20 Evidence-Based Sentencing Practices 10. Ensure defendant�s RNA responses are inadmissible if plea bargain is not accepted (Fed. Rule 410) 11. Use available RNA information to determine the defendant�s suitability or amenability for probation 12. Don�t assume that high risk offenders are not amenable to probation supervision 21 11. Use RNA info not only to establish prob. conds., but also to determine amenability. Two ways: a. low risk assessment score may be important factor in determining not to imprison; probationary status is by definition a trial process to determine whether a person can perform the job satisfactorily, or be safely and effectively supervised in the community b. the opposite is not true: high risk score does not necessarily mean that defendant is not amenable to probation. Unlike probation, imprisonment is not a status that allows us to assess whether the offender could have been safely and effectively supervised in the community and in most cases is not a status that can be reviewed or reconsidered. It has substantial long-term impacts on the offender. 1. b/c risk is dynamic 2. as a general proposition high risk (not extremely high risk) offenders are often good candidates for RR interventions 3. imprisonment based solely on what they might do, risk of committing a future crime is inconsistent with our CJ jurisprudence�at least absent express statutory authorization and DP protections. c. But court can and should consider the dynamic risk factors in determining amenability; a qualitative assessment 11. Use RNA info not only to establish prob. conds., but also to determine amenability. Two ways: a. low risk assessment score may be important factor in determining not to imprison; probationary status is by definition a trial process to determine whether a person can perform the job satisfactorily, or be safely and effectively supervised in the community b. the opposite is not true: high risk score does not necessarily mean that defendant is not amenable to probation. Unlike probation, imprisonment is not a status that allows us to assess whether the offender could have been safely and effectively supervised in the community and in most cases is not a status that can be reviewed or reconsidered. It has substantial long-term impacts on the offender. 1. b/c risk is dynamic 2. as a general proposition high risk (not extremely high risk) offenders are often good candidates for RR interventions 3. imprisonment based solely on what they might do, risk of committing a future crime is inconsistent with our CJ jurisprudence�at least absent express statutory authorization and DP protections. c. But court can and should consider the dynamic risk factors in determining amenability; a qualitative assessment

22. 20 Evidence-Based Sentencing Practices 13. In determining amenability, conduct a qualitative assessment of whether the offender can be safely and effectively supervised in the community 14. RNA tools were designed to assist in managing and reducing offender risk, not to determine the appropriate penalty or punishment for the offense committed 22

23. Malenchik v. State of Indiana(928 N.E.2d 564 (2010)) For the goal of reducing recidivism, �the concept of evidence-based sentencing practices has considerable promise.� �Evidence-based assessment instruments can be significant sources of valuable information for judicial consideration in deciding whether to suspend all or part of a sentence, how to design a probation program for the offender, whether to assign an offender to alternative treatment facilities or programs, and other such corollary sentencing matters.� 23

24. Malenchik v. Indiana The court noted, however, that such tools were �never designed to assist in establishing the just penalty� and ruled specifically that risk assessment scores cannot serve as aggravating or mitigating circumstances in determining the appropriate length of a prison sentence. 24

25. 20 Evidence-Based Sentencing Practices 15. Use of incentives, positive reinforcement (rewards), and swift, certain, and proportionate (fair) sanctions in response to violations promotes offender compliance with conditions of probation and reduces recidivism 25

26. 20 Evidence-Based Sentencing Practices 16. Courts should support probation�s appropriate use of incentives and rewards, and ensure (with the probationer�s consent or proper due process protection) that probation has the administrative authority to impose appropriate sanctions including short periods of incarceration without returning the probationer to court 26

27. 20 Evidence-Based Sentencing Practices 17. In probation revocation proceedings, courts should consider the available RNA information, as well as the offender�s overall criminal history and record of probation compliance and non-compliance, in determining whether the offender can continue to be safely and effectively supervised in the community 27 6. a. In most instances, technical violations and commission of new misdemeanor or low-level offenses will not warrant termination of felony probation and removal from the community b. what is required is a thoughtful re-assessment of the likelihood of success in continuing to manage the probationer in the community without incurring further serious criminal behavior. c. In order to avoid unnecessary court appearances, and maintain a climate of trust and cooperation between the court and probation, it is important that the court and probation agency achieve a clear, consistent, and shared understanding about how these factors and objectives will be weighed by the court and the agency. 6. a. In most instances, technical violations and commission of new misdemeanor or low-level offenses will not warrant termination of felony probation and removal from the community b. what is required is a thoughtful re-assessment of the likelihood of success in continuing to manage the probationer in the community without incurring further serious criminal behavior. c. In order to avoid unnecessary court appearances, and maintain a climate of trust and cooperation between the court and probation, it is important that the court and probation agency achieve a clear, consistent, and shared understanding about how these factors and objectives will be weighed by the court and the agency.

28. 20 Evidence-Based Sentencing Practices 18. Judges should �educate themselves about the effectiveness of community-based corrections programs in their jurisdictions and � advocate and, when appropriate, make use of those programs shown to be effective in reducing recidivism.� 28 4. CCJ 2007 Resolution; judges don�t have operational responsibility, but duty to ensure that orders & decisions are effectively implemented to achieve desired purposes 4. CCJ 2007 Resolution; judges don�t have operational responsibility, but duty to ensure that orders & decisions are effectively implemented to achieve desired purposes

29. 20 Evidence-Based Sentencing Practices 19. Intrinsic motivation is a critical precondition for sustained behavioral change. Threatening, lecturing, shaming, or arguing with the defendant is counter-productive in promoting intrinsic motivation 20. Procedural fairness also promotes law-abiding behaviors 29 19.This doesn�t mean judges should never lecture or argue, but only that such communication does not generally promote pro-social behaviors on the part of criminal offenders. 19.This doesn�t mean judges should never lecture or argue, but only that such communication does not generally promote pro-social behaviors on the part of criminal offenders.

30. 30 Evidence-Based Sentencing to Improve Public Safety and Reduce Recidivism � Judge Roger K. Warren (Ret.) President Emeritus National Center for State Courts