Risk Assessment at Sentencing

This article was written by guest author, Dr. J.C. Oleson. Dr. Oleson is a Senior Lecturer in the Department of Sociology, University of Auckland. Prior to joining the University of Auckland in June of 2010, Dr. Oleson served as Chief Counsel to the Criminal Law Policy Staff of the United States Federal Courts. His full-length article on risk assessment in sentencing is forthcoming in the November 2011 issue of the SMU Law Review.

The imposition of a criminal sentence is an excruciatingly difficult task, requiring sentencing judges to carefully balance different (and frequently competing) theories of punishment. In some jurisdictions, judges enjoy nearly unfettered discretion; in others, judges impose initial sentences but parole boards make the final determinations about when prisoners are ready for release; in other jurisdictions, judicial authority is channeled through sentencing guidelines.

Although some sentencing guidelines have been feted as “the exemplar of smart political and rational reform,”[i] other guideline systems have been condemned as failures.  They have also proved to be legally problematic. In Blakely v. Washington,[ii] the United States Supreme Court struck down Washington State’s sentencing guidelines as violating the right to trial by jury, and one year later, in United States v. Booker,[iii] the Court invalidated the federal sentencing guidelines on the same grounds. Today, the federal guidelines are advisory: federal district judges must calculate them, but treating the resulting guideline sentences as presumptively reasonable is grounds for reversal.[iv]

Given this confusion, some jurists have turned their attention to alternative sentencing models. In particular, evidence-based sentencing has received considerable attention. Its popularity is unsurprising, given the manner in which evidence-based practices (EBP) have transformed the field of community corrections around the world.[v]

In many locations, judges are now asking if the risk assessment instruments used so effectively by community corrections also can be used at sentencing. After all, research indicates that statistical assessments outperform clinical judgment of even trained experts.[vi] If actuarial assessment instruments could accurately distinguish recidivists from those who will not reoffend, then law enforcement resources could be reallocated to minimize crime. Individuals who are likely to commit future crimes might be incapacitated for longer periods (selective incapacitation) and provided with more EBP programming, while low-risk defendants could receive non-custodial or alternative punishments. Already, this approach is used by the State of Virginia, and it may prove to be attractive to other jurisdictions struggling with swelling prison populations and dwindling corrections budgets. Actuarial sentencing could be made easy-to-use. Indeed, a sentencing information system might “chart” optimal sentences:

The severity of sentence would be plotted on the horizontal axis (representing the entire spectrum of terms of imprisonment available under the statute) and the duration without a new arrest (“survival”) would be plotted on the vertical axis. Each point in the cloud of the scatter plot would represent a previous case (offenders matched for offender and offense characteristics), and by clicking on any single point with a mouse, the judge could pull up the specifics of that case: the name and photo of the offender, the offense of conviction, the characteristics of the offender, and the particulars of the sentence imposed.[vii]

Evidence-based sentencing is based upon social science. Criminological meta-analysis has identified fifteen key variables that are significantly related to recidivism: 1) criminal companions, (2) antisocial personality, (3) adult criminal history, (4) race, (5) pre-adult antisocial behavior, (6) family rearing practices, (7) social achievement, (8) interpersonal conflict, (9) current age, (10) substance abuse, (11) intellectual functioning, (12) family criminality, (13) gender, (14) socio-economic status of origin, and (15) personal distress.[viii] If those variables can be used in sentencing, it may be possible to safeguard public safety while reducing the financial and social costs associated with mass incarceration.

Despite these advantages, there are problems with evidence-based sentencing. Indeed, there are three distinct classes of problems: legal, logistical, and philosophical. These challenges can no longer be avoided. In 2010, in Malenchik v. Indiana,[ix] the Supreme Court of Indiana upheld a sentencing judge’s use of the Level of Service Inventory (LSI-R) instrument, but other courts have reached contradictory conclusions. Yes, criminologists may be able to identify statistical predictors of recidivism, but should judges be permitted to enhance sentences based upon risk factors that, in other contexts, would constitute impermissible discrimination (e.g., race, gender, age, class, or religion)? Looking at the Court’s reasoning in Korematsu v. United States[x] and Grutter v. Bollinger,[xi] I believe that courts applying strict scrutiny analysis to evidence-based sentencing will uphold the imposition of enhanced sentences based upon risk factors, but the mere legality of the practice does not address the many thorny questions implicit in evidence-based sentencing.


[i] Daniel F. Wilhelm & Nicholas R. Turner, Is the Budget Crisis Changing the Way We Look at Sentencing and Incarceration? 7 (2002), available at http://www.vera.org/publication_pdf/167_263.pdf.

[ii] 542 U.S. 296 (2004).

[iii] 543 U.S. 220 (2005).

[iv] See Nelson v. United States, 555 U.S. __ (2009).

[v] See Faye Taxman, et al., Tools of the Trade: A Guide to Incorporating Science into Practice (2004) (describing principles of evidence-based practices).

[vi] See William M. Grove, et al., Clinical versus Mechanical Prediction: A Meta-Analysis, 12 Psychol. Assessment 19 (2000).

[vii] J.C. Oleson, Blowing Out All the Candles: A Few Thoughts on the Twenty-Fifth Birthday of the Sentencing Reform Act of 1984, 45 U. Rich. L. Rev. 693, 745 (2011).

[viii] See Paul Gendreau et al., A Meta-Analysis of the Predictors of Adult Offender Recidivism: What Works! 34 Criminology 575 (1996).

[ix] No. 79S02-0908-CR-365 (Ind. June 9, 2010).

[x] 323 U.S. 214 (1944).

[xi] 539 U.S. 306 (2003).