Using Evidence-based Practices in Sentencing Criminal Offenders

The fields of medicine, education, child welfare, mental health, probation and corrections have all been influenced by evidence-based practices. In essence, evidence-based practices are a set of guidelines—based upon rigorous research, evaluations and meta-analysis—that have proved effective in improving decision making and outcomes. In the medical world, for example, evidence-based practice refers to the “conscientious, explicit, and judicious use of current best evidence in making decisions about the care of individual patients.”1 Only recently, however, has this approach spilled over into state courts in the form of providing decision-making tools for judges at the time of criminal sentencing.

  Download the Article in PDF / E-Reader Compatible Format

About the Author
Matthew Kleiman is a senior court research associate with the National Center for State Courts in Williamsburg, Va. His work has focused on: the use of risk and needs information at the time of sentencing; assessments of consistency and fairness in criminal sentencing; the development and implementation of a set of court specific performance measures (CourTools); and the development of statewide resource assessment models for judicial officers, court staff, and public defender and prosecutor attorneys and staff. He holds a doctorate in political science and is the co-author of a recent Temple University Press book on court culture—Trial Courts as Organizations. 

Judges confront one of the most difficult decisions they have to make during criminal sentencing—the determination of what loss of liberty is appropriate for an offender and society. Evidence-based practices in sentencing hold the promise of improving judicial decision making, which can lead to enhanced public safety (reductions in offender recidivism rates) and to efficient use of scarce criminal justice resources. This essay will discuss the changing landscape of criminal sentencing; detail the risks, needs and responsivity principles of evidence-based sentencing; describe the application of evidence-based sentencing in Virginia and Wisconsin; and discuss issues important for implementation.

The Changing Landscape of Sentencing
Over the past four decades, criminal sentencing policy has undertaken several paradigm shifts that have shaped judicial discretion and judicial sentencing practices. Beginning in the mid-1970s, research concluded that programs, treatment and interventions aimed at changing offender behavior did not work. This overt criticism, along with a rising violent crime rate, provided the impetus for the adoption of policies that replaced the rehabilitative model of sentencing with a punitive approach that emphasized retribution and determinate sentencing. The new model of sentencing emphasized punishment through incarceration and general deterrence through the adoption of “tough on crime” policies, such as mandatory minimum sentences, truth in sentencing, determinate sentencing and three strikes laws. This led to a dramatic increase in the incarceration rate.

Between 1920 and the mid-1970s, the U.S. imprisonment rate remained constant at 110 per 100,000 people. With the change in sentencing philosophy in the mid to late 1970s, the imprisonment rate began to steadily increase to the current rate of 500 people per 100,000. As a result, between 1974 and 2005 the number of inmates in federal and state prisons increased from roughly 215,000 to more than 1.5 million.2 To accommodate this rapid growth, states responded by building new prisons, which greatly increased the amount of state resources devoted to corrections. For example, expenditures on corrections increased by 200 percent between 1985 and 2004, while expenditures for higher education only increased 3 percent during the same period.3

In recent years, the budgetary challenges confronting state governments and a growing body of literature supporting the effectiveness of treatment in reducing recidivism have resulted in the swing in sentencing philosophy back toward rehabilitation. Missouri’s Chief Justice Ray Price said “[t]here is a better way. We need to move from anger-based sentencing that ignores cost and effectiveness to evidence-based sentencing that focuses on results— sentencing that assesses each offender’s risk and then fits that offender with the cheapest and most effective rehabilitation that he or she needs.”4

Evidence-based Sentencing
Evidence-based sentencing refers to judges using information about offender risk, needs and responsivity to inform the most appropriate sentence for a convicted offender. This information, provided to judges prior to sentencing, improves judicial decision-making by identifying sentences and treatments that are most effective and cost efficient in reducing an offender’s future risk to the community. The information, based upon empirical investigations of past cases, quantifies the risk of an offender committing future criminal acts and the specific treatment likely to prevent reoffending.5 It provides guidance about the level of supervision and the type of interventions that are most likely to reduce future criminality.

As a risk-reduction strategy, evidence-based sentencing helps sort out which offenders should receive incarceration, intensive supervised probation, diversion, inpatient care, vocational training, etc. The risk and needs information is not intended to limit judicial discretion, but instead to better inform judicial decision-making by identifying who to target, what to target and how to target to ensure the lowest levels of recidivism and the best possibility of offender change. Evidence-based sentencing is guided by the principles of risk, needs and responsivity.6

The risk principle requires that the level of supervision and treatment match an offender’s likelihood of reoffending. Higher-risk cases require more intervention, structure, supervision and resources. Focusing supervision and treatment resources on lower-risk offenders can lead to wasted resources and, in some cases, may actually increase recidivism rates.7 In short, the risk principle tells us who to target.8 While risk assessment is a valuable tool for classifying offenders according to their risk to reoffend, the risk component alone does not provide guidance for offender treatment and supervision.9

The needs principle stresses that offender criminogenic needs be assessed and, where a need exists, targeted with treatment and interventions. Criminogenic needs are dynamic—or changeable—risk factors most associated with criminal behavior. The dynamic risk factors most strongly associated with recidivism are a history of antisocial attitudes, antisocial peers, antisocial personality pattern and a history of antisocial behavior. Additional risk factors include substance abuse, family/marital relationships, lack of achievement in education/ employment and a lack of pro-social recreational activities.10 Assessing criminogenic needs informs judges of what types of treatment services should be targeted. Chief Justice Michael Wolff, of Missouri, commented that in order to “reduce recidivism, the punishment should fit the offender as well as the crime.”11

The responsivity principle contends that any treatment intervention should be tailored to the offender’s learning style, motivation, developmental stage and cognitive abilities. The treatment intervention also should use cognitive social learning strategies. The responsivity principle addresses the question of how to target offenders to ensure successful interventions that reduce future criminality. Recent meta-analyses have confirmed that a significant relationship exists between the adherence to the risk, needs and responsivity principles and reductions in recidivism.12 Programs that directly target criminogenic factors are more effective in reducing recidivism than those that do not.13 Judge Roger Warren, former President of the National Center for State Courts and retired California Judge, highlighted these findings, saying there is “a voluminous body of solid research showing that certain ‘evidence-based’ sentencing and corrections practices do work and can reduce crime rates as effectively as prisons and at much lower costs.”14 Providing the risk, needs and responsivity information to judges can potentially allow them to divert low-risk offenders from prison settings, order treatment conditions that match an offender’s risk factors and make better informed pretrial release decisions.

A National Working Group on the use of offender risk and needs assessment information recommended that “judges have offender assessment information available to inform their decisions regarding risk management and reduction.”15 The group concluded that the incorporation of risk, needs and responsivity principles into sentencing has several advantages, including improved public safety, a reduction in prison admissions resulting from recidivism, less reliance on costly incarceration by reserving prison beds for serious and violent offenders, a reduction in subjective sentencing decisions through the use of scientifically based decision tools, a focus on offender accountability to elicit behavioral changes, and a reduction in social, economic and family costs associated with inappropriate interventions targeted at low-risk offenders.16

Evidence-based sentencing is a relatively new idea for state trial courts. Lessons learned from probation and corrections are finally beginning to influence criminal sentencing in the courtroom. Currently, a growing number of jurisdictions are providing risk, needs and responsivity information to judges prior to sentencing. States that have begun to use or explore the use of evidence-based sentencing include Alabama, Arizona, California, Colorado, Idaho, Iowa, Missouri, Texas, Virginia, Washington and Wisconsin. A brief description of the implementation of a risk assessment tool in Virginia and of a pilot program in Wisconsin follows.  

Virginia Nonviolent Risk Assessment
In 1994, Virginia abolished parole and adopted truth-in-sentencing guidelines for people convicted of a felony. At the request of the general assembly, the Virginia Criminal Sentencing Commission developed a method for diverting 25 percent of nonviolent, prison-bound offenders into alternative sanctions using an empirically based risk assessment instrument. An instrument identified which offenders, otherwise recommended for incarceration by the sentencing guidelines, have the lowest probability of being convicted of another felony crime. These offenders are recommended for some form of less restrictive, alternative punishment.

The Virginia risk assessment recommendation represents a nonbinding supplement to the preexisting sentencing guidelines and is given to judges prior to the sentencing of drug, fraud and larceny offenders. The instrument was designed to assess an offender’s risk to public safety, not gauge the needs of individual offenders and, as such, does not recommend any specific type or form of alternative punishment. An evaluation by the National Center for State Courts found that the risk assessment instrument was effective in identifying low-risk candidates for diversion. It also reduced use of prison and jail and produced cost-savings of between $2.9 million and $3.6 million for the Commonwealth of Virginia without jeopardizing the safety of its citizens.17

Wisconsin: Assess, Inform, and Measure Pilot Program
The pilot project, Wisconsin Assess, Inform and Measure—or AIM—which began in the fall of 2006, is intended to provide judges with valid and reliable information to help inform sentencing decisions. The AIM process is based upon the principles of risk, needs and responsivity. The AIM model has two stated goals:

  1. Provide the sentencing court with a valid risk, needs and unique characteristic (responsivity) and community intervention assessment, while creating a feedback loop that provides information on the success of court dispositions and community interventions in promoting offender success and public safety.
  2. Put into practice and evaluate a process that offers the court reliable information that will have value in the sentencing process. It may lead to the safe diversion of some people who may have otherwise received jail or prison time to community-based supervision and treatment. Eight pilot counties—Bayfield, Dane, Eau Claire, Iowa, La Crosse, Marathon, Milwaukee and Portage —participated in the AIM Project.

The pilot counties ranged from small, one judge courts to large urban jurisdictions. The AIM program grew directly out of concerns by members of the Wisconsin Planning and Policy Advisory Committee that jail and prison may not be the best method for changing people’s behavior and ensuring safety for the community. Committee members questioned whether the number of individuals being incarcerated was too large and how effective this type of sanction was in changing people’s behavior.

To foster participation in the AIM project, each of the pilot counties was given latitude in selecting their own target populations (Class F, G, H, I felonies; operating while under the influence offenders; misdemeanor repeat offenders), assessment tools (COMPAS; LSI-R), and the point at which the assessment would be conducted (bond hearing or pre-sentence). Despite the variations, judges in each jurisdiction received information about the current charges, criminal history, risk assessment, needs assessment, responsivity and community-based program availability. Unlike the risk assessment in Virginia, the AIM report identifies relevant resources in the community that are appropriate for the particular offender’s risk and needs profile, should the judge decide to keep the offender in the community. Wisconsin is considering expanding the AIM program statewide.18

In a speech to the American Bar Association’s Annual Meeting, U.S. Supreme Court Associate Justice Anthony Kennedy, commenting on American sentencing, said “[o]ur resources are misspent, our punishments too severe, our sentences too long.”19 Evidence-based sentencing applies risk, needs and responsivity principles to begin to redress Kennedy’s concerns. The use of these instruments provides judges with information to better inform sentencing decisions, which can lead to more effective and efficient use of scarce resources, reduction in offender recidivism and reduced prison admissions.

The continued diffusion, adoption and utilization of evidence-based practices by trial court judges is contingent on educating judges and justice system stakeholders (prosecutors, defense bar, policymakers) about the theory and use of risk- and needs-assessment information.20 Educational sessions should highlight the science and research behind the risk, needs and responsivity principles and the interpretation and use of assessment information.  The National Working Group suggests that judges and other stakeholders need to understand “that ‘high risk’ does not necessarily translate to ‘need to incarcerate.’ They also need to understand what dynamic risk factors are and … recognize that RNA (risk and needs assessment) tools are intended to enhance, not replace, judicial decision making.”21 Finally, judges and stakeholders should be exposed to the latest research on what works. Wolff, Missouri’s chief justice, proclaims that “[w]e must pay particular attention to which sentences make recidivism more likely, which sentences are ineffectual at reducing recidivism, and which programs and punishment-treatment regimes have the best outcomes.”22


1 Sackett, D., Rosenberg, W.M.C., Gray, J.A.M., Haynes, R.B., Richardson, W.S. (1996). “Evidence-based Medicine—What it is and what it isn’t.” British Medical Journal 312.
2 Warren, Roger K. (2009). “Evidence-Based Practices and Sentencing: The Application of Principles of Evidence-Based Practice to State Sentencing Practice and Policy.”
University of San Francisco Law Review 43: 585–634.
3 Ibid.
4 Price, Ray, Jr. (2010). State of the Judiciary.
5 Extensive research has shown that actuarial algorithms are more accurate and consistent than subjective clinical judgment or human decision-makers. See for example, Grove, W.M., & Meehl, P.E. (1996). “Comparative Efficiency of Informal (Subjective, Impressionistic) and Formal (Mechanical, Algorithmic) Prediction Procedures: the Clinical-Statistical Controversy.” Psychology, Public Policy, and Law; Meehl, P.E. (1996). Clinical versus statistical prediction: A theoretical analysis and a review of the evidence. Northvale, NJ: Jason Aronson. (Original work published 1954).
6 The risk, needs , responsivity framework derives from the arena of correctional  programming. See, Andrews, D., & Bonta, J. (2006). The Psychology of Criminal Conduct,
4th Ed. Cincinnati: Anderson.
7 Lowenkamp, C., & Latessa, E. (2004) “Understanding the Risk Principle: How and Why Correctional Interventions Can Harm Low-risk Offenders.” Topics in Community Corrections.
8 Duran, L., & D’Amora, D. (2011) “Responsivity in the Risk/Need Framework” (PPT  Presentation). 2011 Justice and Mental Health Collaboration Program National Training
and Technical Assistance Event. Washington, D.C.: The Council of State Governments.
9 Actuarial risk assessment uses an explicit set of factors that correlate with re-offending to classify offenders into groups that re-offend at similar rates within-group, but at different rates between groups. Traditionally, the types of factors used include the offender’s criminal history, the nature of the offense, social variables like the offender’s age, educational and employment history, and history of substance abuse. See Domurad, F. (1999). “So You Want to Develop Your Own Risk Assessment Instrument.” In Topics in Community Corrections: NIC Annual Report 2009: Classification and Risk Assessment. Longmont, CO: National Institute of Corrections.
10 Andrews, D.A., Bonta, J., & Wormith, J.S. (2006). “The Recent Past and Near Future of Risk and/or Need Assessment.” Crime & Delinquency, (52), 1, 7–27.
11 Wolff, M.A., (2008). Brennan Lecture Evidence-Based Judicial Discretion: Promoting Public Safety through State Sentencing Reform (November 2008). New York University Law Review, Vol. 83, No. 5; Saint Louis U. Legal Studies Research Paper No. 2008-14.
12 Andrews, D. & Bonta, J. (2006). The Psychology of Criminal Conduct, 4th Ed. Cincinnati: Anderson.
13 Cullen, F.T., & Gendreau, P. (2000). Assessing correctional rehabilitation: Policy, practice, and prospects. In J. Horney (Ed.), NIJ criminal justice 2000: Changes in decision making and discretion in the criminal justice system (pp. 109–175). Washington, D.C.: U.S. Department of Justice, National Institute of Justice.
14 Warren, R.K. (2009). “Arming the Courts with Research: 10 Evidence-Based Sentencing Initiatives to Control Crime and Reduce Costs.” The Pew Center on the States.
15 The National Working Group was chaired by Alabama’s Chief Justice Sue Bell Cobb. Members included an appellate court justice and representatives from parole, public defense, prosecution, probation, court administration, and academia. See, Casey, P.M., Warren, R.K., and Elek, J.K. (2011). Using Offender Risk and Needs Assessment Information at Sentencing: Guidance for Courts from a National Working Group. Williamsburg, VA: National Center for
State Courts.
16 Ibid.
17 See, Ostrom, B., Kleiman, M., Cheesman, F., Hansen, R., & Kauder, N. (2002). Offender Risk Assessment in Virginia. National Center for State Courts.
18 See, Tallarico, S., Cheesman, F.C., Kirven, M., & Kleiman, M. (2011). Effective Justice Strategies in Wisconsin. A Report to the Director of State Courts and the Wisconsin Supreme Court, Planning and Policy Advisory Committee, Effective Justice Strategies Subcommittee.
19 Kennedy, A.M., (August 14, 2003). Speech at the American Bar Association Annual Meeting: An Address by Anthony M. Kennedy, Associate Justice, Supreme Court of the United States.
20 The National Center for State Courts, in collaboration with the National Judicial College and the Crime and Justice Institute, has developed a model national judicial educational curriculum. The curriculum, “Evidence-Based Sentencing to Improve Public Safety and Reduce Recidivism,” is available at
21 Casey, P.M., Warren, R.K., & Elek, J.K. (2011). Using Offender Risk and Needs Assessment Information at Sentencing: Guidance for Courts from a National Working Group. Williamsburg, VA: National Center for State Courts, p. 21. 
22 Wolff, M.A., (2008). Brennan Lecture Evidence-Based Judicial Discretion: Promoting Public Safety through State Sentencing Reform (November 2008). New York University Law Review, Vol. 83, No. 5; Saint Louis U. Legal Studies Research Paper No. 2008-14.


matthew_kleiman_2012.pdf68.48 KB