THE GENERAL ASSEMBLY
COMMISSION ON SENTENCING
Proposed Sentence Risk Assessment Instrument for 204 Pa. Code Chapter 305; Responses to Public Comments; Request for Proposals
[48 Pa.B. 5445]
[Saturday, September 1, 2018]Act 2010-95, effective October 27, 2010, directed the Pennsylvania Commission on Sentencing to ''. . .adopt a sentence risk assessment instrument for the sentencing court to use to help determine the appropriate sentence within the limits established by law for defendants who plead guilty or nolo contendere to, or were found guilty of felonies and misdemeanors.'' Following years of research and development, the Commission approved for public comment two initial versions of an instrument:
• On April 11, 2015, the Commission published in the Pennsylvania Bulletin (45 Pa.B. 1751) a preliminary Sentence Risk Assessment Instrument and held three public hearings;
• On April 8, 2017, the Commission published in the Pennsylvania Bulletin (47 Pa.B. 1999) a proposed Sentence Risk Assessment Instrument and held five public hearings.
Earlier this year, following additional review and analysis, the Commission approved for public comment a revised instrument and on April 28, 2018 published this proposed Sentence Risk Assessment Instrument in the Pennsylvania Bulletin (48 Pa.B. 2367). In accordance with 42 Pa.C.S. § 2155, the Commission held the following public hearings to afford an opportunity for individuals and organizations to testify:
• Allegheny County Courthouse (Pittsburgh), Monday, June 4, 2018
• Juanita Kidd Stout Center for Criminal Justice (Philadelphia), Wednesday, June 6, 2018
• Pennsylvania Judicial Center (Harrisburg), Wednesday, June 13, 2018
The Commission received testimony and written comments from more than 30 individuals and organizations related to this most recent proposal. Copies of all documents received are posted on the Commission's website (http://pasentencing.us). During the June 14, 2018 quarterly meeting, the Commission approved a motion to postpone adoption of the proposed Sentence Risk Assessment Instrument and to direct staff to do the following:
(1) prepare and publish a summary of the issues raised during the public hearings and responses to those issues;
(2) publish the findings of the Urban Institute's external review of the proposed Sentence Risk Assessment Instrument upon receipt;
(3) solicit proposals from the public for a Sentence Risk Assessment Instrument as described in 42 Pa.C.S. § 2154.7; and
(4) schedule a minimum of three public hearings in December 2018 to receive proposals and to review comments on any revisions to the Commission's proposal.
Consistent with this decision, the Commission herein provides information on each of these directives.
(1) Response to Public Comments
A summary of the issues raised during the public hearings and responses to those issues are provided in Annex A.
(2) External Review
In December 2016, the Commission contacted the Bureau of Justice Assistance/Urban Institute Risk Assessment Clearinghouse to initiate discussions of an external review of the Sentence Risk Assessment Instrument under development by the Commission. At that time, consistent with the Commission's current proposal, the Commission described the project as follows: ''. . .we have developed a second generation instrument, and seek to limit its use to identifying those cases where a more comprehensive and individualized RNR (Risk-Needs-Responsivity) report be prepared for consideration by the court; the information contained in the report would be used to address treatment and intensity rather than disposition and duration of the sentence.''
Following the publication of the Commission's proposed Sentence Risk Assessment Instrument in 2017 and subsequent public hearings, the Commission delayed action on the external review. This allowed staff to make substantial changes to the instrument as directed by the Commission in response to public comment. In April 2018, following publication of a revised proposal, staff provided the Urban Institute with a complete set of files related to construction of the instrument (e.g., data, syntax, etc.) to begin the external review.
The Commission anticipates receipt of the external review no later than the next regularly scheduled meeting of the Commission (September 6, 2018). The external review of the proposed Sentence Risk Assessment Instrument will be published in the Pennsylvania Bulletin (anticipated publication date: October 6, 2018) following the September Commission meeting.
(3) Request for Proposals
The Commission is soliciting proposals from all interested persons or organizations for a Sentence Risk Assessment Instrument to be considered as an alternative to the Commission's proposed instrument. The Commission requests all proposals be submitted in writing and received by November 30, 2018. Details related to the requirements of the solicitation are provided in Annex B.
(4) Public Hearings
The Commission will hold a minimum of three public hearings in December 2018 to review and receive comments on all proposals received, as well as to review and receive comments on any staff revisions to the April 28, 2018 proposal. The Commission will attempt to schedule hearings in various geographic locations and in various classes of counties during the weeks of December 3 and December 10. In accordance with 42 Pa.C.S. § 2155, the Commission will publish the dates and locations of the public hearings as well as any revisions to the Commission's proposal in the Pennsylvania Bulletin (anticipated publication date: November 3, 2018). The public hearings will be held no earlier than 30 days and no later than 60 days after publication to afford an opportunity for individuals and organizations to testify.
PRESIDENT JUDGE SHEILA A. WOODS-SKIPPER,
Chair
Annex A
Responses to Public Comments Act 2010-95, effective October 27, 2010, directed the Pennsylvania Commission on Sentencing to adopt a Sentence Risk Assessment Instrument for the sentencing court to use to help determine the appropriate sentence within the limits established by law for defendants who plead guilty or nolo contendere to, or were found guilty of felonies and misdemeanors.
Following the publication of the most recent proposed Sentence Risk Assessment Instrument on April 28, 2018 (48 Pa.B. 2367), the Commission held three public hearings to afford an opportunity for individuals and organizations to testify and to submit written comments. Copies of all documents received are posted on the Commission's website (http://pasentencing.us).
During the June 14, 2018 quarterly meeting, the Commission approved a motion to postpone adoption of the proposed Sentence Risk Assessment Instrument and directed staff to prepare and publish a summary of the issues raised during the public hearings and responses to those issues. This summary is provided below, with the issues organized into four categories based on the provisions of the legislation: (1) adopting a sentence risk assessment instrument to help determine the appropriate sentence and evaluate relative risk; (2) incorporating the instrument into sentencing guidelines; (3) using the instrument to determine whether a more thorough assessment is necessary; and (4) using the instrument to determine appropriate candidates for alternative sentencing.
(1) Risk Assessment Instrument (general rule)
42 Pa.C.S. § 2154.7
(a) General rule.—The commission shall adopt a Sentence Risk Assessment Instrument for the sentencing court to use to help determine the appropriate sentence within the limits established by law for defendants who plead guilty or nolo contendere to or who were found guilty of felonies and misdemeanors. The risk assessment instrument may be used as an aide in evaluating the relative risk that an offender will reoffend and be a threat to public safety.
(e) Definition.—As used in this section, the term ''risk assessment instrument'' means an empirically based worksheet which uses factors that are relevant in predicting recidivism.
Comment
This general rule authorizes three actions by the Commission: (1) a requirement that the Commission adopt a risk assessment instrument, defined as an empirically based worksheet which uses factors that are relevant in predicting recidivism; (2) a requirement that the instrument be constructed to help determine the appropriate sentence within the limits established by law; and (3) a provision that permits the instrument to be used as an aide in evaluating the relative risk that an offender will reoffend and be a threat to public safety.
(a) Requirement to adopt an instrument
1. Construction of an empirically based worksheet (development)
a. Comments received
Concerns were raised about the selection of a 2nd generation instrument (static risk factors) instead of 3rd or 4th generation instruments which also would have considered dynamic and protective factors, offender needs and responsivity. Specific to the proposed instrument, several who testified criticized the differences in the base rates of recidivism for Black and White offenders and questioned the selection of cut points in determining high and low risk based on false positive/false negative ratios. Due to the evolving research in this area, the Commission was encouraged to proceed with caution and to continue to investigate racial bias.
b. Staff response
The text of Act 2010-95 which directs the Commission to adopt a 'risk' assessment instrument, and the legislative history of the initial drafts of the bill in which 'needs' was removed, informed the Commission's decision to adopt a 2nd generation instrument. The Commission sought to automate the instrument, relying factors available electronically at the time of sentencing, and to then incorporate it into the existing JNET-based SGS Web sentencing guidelines application. The Commission used data from the Administrative Office of Pennsylvania Courts (AOPC) and SGS Web due to reliability, accuracy and availability in all counties at the time of sentencing. Because of this approach, counties avoided the costs associated with licensing fees for a proprietary product and with a requirement to prepare a 3rd or 4th generation assessment in every case.
Regarding disproportionality, the risk assessment development data reveal differences in base rates of recidivism by race group, a typical finding in the risk and recidivism literature. Additional information on the construction and racial impact analysis of the proposed scales is posted on the Commission's website. Staff will continue to investigate disproportionality, including analysis of other recidivism measures.
Regarding cut points for determining high and low risk, the Commission abandoned the false positive/false negative ratio approach following Phase I of the project (preliminary analysis) in favor of the current standard deviation approach. The Commission determined that any decision regarding appropriate or acceptable false positive/false negative ratios involves a trade-off between over-predicting recidivism and under-predicting recidivism and may not address 'relative risk' of reoffending. The Commission instead relied on a statistical model to determine the risk of a 'typical offender' (68% of offenders defined by one standard deviation from the mean risk score), and then to identify those below and above this group as low risk and high risk. Equally important in its development and coupled with this approach were the following: (1) the creation of the Sentence Risk Assessment Summary to provide detailed information on all risk factors, recidivism rates, and risk categories for each scale; and (2) the decision to recommend additional individual information rather than a specific disposition for certain low risk and high risk offenders.
2. Data
a. Comments received
Concerns were raised about the age of the data set (2004—2006 sentences), the quality of the data (e.g., tainted due to policing practices, racial bias 'baked into the system'), the historical disproportionate use of punishment reflected in practices (e.g., mandatory minimum sentences, three-strikes), and the failure to exclude stale prior records (i.e., dated prior convictions). Other more specific comments related to risk factors and outcome measures are addressed separately in this document.
b. Staff response
Creating a data set is a resource-intensive and time-consuming process involving the matching of data across multiple agencies. The task is magnified when dealing with a very large statewide sample. The Commission continues to work with other criminal justice agencies to enhance practices and to improve the quality of the data, while also working with academic research partners to investigate advanced methods to address data quality issues.
The risk scales used for the proposal were generated to analyze the risk of all offenders using a sample of over 131,000 offenders sentenced in Pennsylvania between 2004—2006. These older data sets are required for a study of recidivism involving long prison sentences imposed for more serious offenses. The original sample included 135,000 offenders sentenced during this period, but 4,000 offenders were removed because they had not yet been paroled from the DOC or were less than three years post-release (tracking period). This represented a loss of 4% of the DOC sample in 2004 and a loss of 6% of the DOC sample in 2006. Loss of 10% of a sample (or subsample) creates a substantial validity problem when data are missing 'not at random.' Using a 2008 sample would result in a loss of 9% of the DOC sample, while using a 2009 sample would result in a loss of 15% of the DOC sample. A more recent data set will be constructed for any re-validation of the instrument to ensure the continued accuracy with current offenders.
3. Risk factors relevant in predicting recidivism
a. Comments received
Concerns were raised about the use of 'tainted' criminal justice data, the inclusion of demographic factors such as age and gender (equal protection), the failure to consider dynamic factors, and the inclusion of ZIP code and factors that serve as proxies for race and poverty. Other concerns related to criminal justice factors include the use of juvenile adjudications and the failure to exclude stale records, especially for individuals with long post-conviction periods of law-abiding behavior. Several who testified noted positive changes from previous versions of the instrument, including the use of conviction in place of arrest as a risk factor.
b. Staff response
The first steps taken by the Commission as part of the risk assessment project were reviews of factors used in risk assessment instruments (Interim Report I, 2011) and factors that predict recidivism (Interim Report III, 2011). The purpose was to identify static factors that were automated, reliable and accurate, and readily available statewide at the time of sentencing. Only factors that were statistically significant were considered in the development of each scale. Two factors were statistically significant and were controlled for but were not assigned points in the scale: race and size of county. ZIP code was never used as a risk factor. Dynamic factors such as employment, education and marital status were not included because, even if significant, were not automated nor readily and reliably available in all counties. The use of juvenile adjudications is limited to consideration of whether an offender has a juvenile adjudication in the prior record score of the sentencing guidelines. This is included because it is statistically significant, and because it is limited to an express finding of an adjudication for a felony or certain M1 offenses that occurred on or after the offender's 14th birthday. The Commission is investigating policies to address stale records, and if incorporated into the sentencing guidelines, could be considered in the future.
4. Recidivism outcome measures
a. Comments received
Concerns were raised about the use of arrest instead of conviction as the outcome measure of recidivism, and of including technical parole violations. Concerns were also raised about the definition of recidivism including any offense, suggesting instead that risk should be linked to dangerousness, and therefore risk should not equal arrest for any offense. Several speakers referenced the Philadelphia Reentry Coalition recidivism data, both because of a lower recidivism rate than identified by the Commission and different racial impacts.
b. Staff response
Recidivism definitions vary by instrument, agency, and policy. Consistent with the legislation, the Commission has developed an instrument ''which uses factors that are relevant in predicting recidivism,'' that the instrument ''help determine the appropriate sentence within the limits established by law,'' and that the instrument may be used to evaluate ''the relative risk that an offender will reoffend and be a threat to public safety.''
Arrest is the most common outcome measure used in the prediction of recidivism. Use of arrest as an outcome measure may over-estimate reoffending, while use of convictions may under-estimate reoffending. Technical violations that result in re-commitment to prison may be for behavior other than reoffending, although detention and recommitment is limited by statute (see 61 Pa.C.S. § 6138, relating to Violation of terms of parole). Technical violations are at times used in lieu of arrest and prosecution for a new offense, and at times an offender is charged with both a new offense and a technical violation. In determining recidivism, the use of convictions and the exclusion of technical violations resulting in recommitment raise a common methodological issue: how to measure follow up periods when interrupted by periods of confinement.
Risk of any offense, both high risk and low risk, is an important consideration in determining an appropriate sentence and addressing other provisions of the legislation. The recidivism rates reported in the Philadelphia Reentry Coalition report are based on Pennsylvania arrest data, not re-conviction, using a one-year follow-up period, and limited to those released from jail or prison with self-reported return to Philadelphia. As correctly noted in the Coalition report, comparing different recidivism definitions is not ''apples to apples.'' To address the legitimate concerns raised, Commission staff will re-analyze the data, dropping technical violations and using as a recidivism outcome measure re-arrest within three years that results in a conviction.
While dangerousness is an important outcome measure, the term is not included or defined in Act 2010-95. The Commission developed a separate scale for 'offense against a person' to address dangerousness and threat to public safety which is described in greater detail later in this document.
5. Validity/Reliability
a. Comments received
Concerns were raised about the accuracy rates of the scales, which ranged from 60% to 80%, often characterized as not much better than a flip of a coin. Other issues raised included validation and independent review of the instrument, user training and certification to assure inter-rater reliability and fidelity, an auditing and tracking of the instrument, with the ability to 'turn it off' if it is not doing what it is intended to do.
b. Staff response
The Commission developed a module within SGS Web, the Commission's JNET-based sentencing application, to automate the actuarial risk assessment instrument. The risk assessment scales contained in the instrument provide measures of likelihood of recidivism. The risk scales are not perfect predictors, and are not determinative of future outcomes, but instead are intended to complement the court's exercise of professional judgement to provide a more accurate assessment of an individual's relative risk. Accuracy rates between 60% and 75% are within acceptable boundaries described in the risk and recidivism literature. Since the instrument is automated and the risk summary is generated upon completion of the guideline sentence form via SGS Web, inter-rater reliability and fidelity are not factors, and no separate training or certification for preparation of the risk assessment summary is required beyond that required for preparation of the guideline sentence forms.
However, training and certification, as well as inter-rater reliability and fidelity, are important considerations for the use of 3rd and 4th generation assessments. These county-specific safeguards are the focus of the Evidence-Based Practice initiative of the County Chief Adult Probation and Parole Officers Association of Pennsylvania, and for which state and federal funding is available to counties. All information generated by SGS Web is captured in the database, and monthly exports of data are analyzed to monitor compliance and provide a basis for evaluation. Currently a third-party review is underway by the Urban Institute, and the findings will be published when received.
6. Other
a. Comments received
Concerns were raised about the timing of the release of the racial impact analysis, about the advertising and dissemination of information about the mandate and the public hearings, and the limited number of individuals and organizations notified.
b. Staff response
Development of the instrument has been underway since 2010, and the Commission has carefully documented the process and provided extensive information on all research and policy decisions. All documents have been posted on the Commission's website. Throughout this period, the Commission has worked with four pilot counties (Philadelphia, Allegheny, Westmoreland, Blair) to solicit feedback and test proposals, and has conducted or participated in scores of presentation and outreach efforts with judges, attorneys, probation officers, and other practitioners.
The Commission holds four public meetings each year, with detailed minutes of each meeting posted, and has held 11 public hearings specifically concerning the risk assessment instrument. Consistent with statute (42 Pa.C.S. § 2155), the Commission has published notices of all meetings, public hearings and proposals in the Pennsylvania Bulletin, on its website, and in a general circulation newspaper in the city in which it is meeting and has afforded an opportunity to testify to any interested person or organization. At a minimum, the Commission provides the following organizations, listed in the statute, with direct notice of all hearings and copies of all proposals: Pennsylvania District Attorneys Association; Chief of Police Associations; Fraternal Order of Police; Public Defenders Association; Law School faculty members; Pennsylvania Board of Probation and Parole; Pennsylvania Department of Corrections; Pennsylvania Bar Association; Pennsylvania Wardens Association; Pennsylvania Association of Probation, Parole and Corrections; and Pennsylvania Conference of State Trial Judges.
The report on the racial impact analysis was released as soon as it was completed, and in advance of the June 2018 public hearings and quarterly meeting. Sophisticated analysis regarding racial impact and an external review (validation) could not be undertaken until risk scales were finalized. The Commission finalized scales in March 2018.
(b) Use of instrument to help to determine the appropriate sentence
1. Within the limits established by law
a. Comments received
Concerns were raised that the Commission erred on the side of over-incarceration in determining risk categories and cut points, that the instrument equates risk with punishment, and that courts will use the high-risk label to increase punishment and impose longer sentences. Several individuals suggested that even if the Commission created the instrument to identify cases for which additional information should be considered, the labeling of offenders will lead to misuse and increased punishment. Some suggested that limiting the availability of risk information to the parties (i.e., sentencing judge, prosecutor, defense) and considering the Sentence Risk Assessment Summary a confidential document may diminish the labeling concerns. Others suggested limiting the use of the instrument to identify low risk offenders that may be recommended to mitigation and/or diversion.
b. Staff response
As noted previously, the Commission abandoned the false positive/false negative ratio approach in favor of the current standard deviation approach in order to better address 'relative risk' of reoffending. This approach was coupled with two other choices by the Commission: (1) the creation of the Sentence Risk Assessment Summary to provide detailed information on all risk factors, recidivism rates, and risk categories for each scale; and (2) the decision to recommend additional individual information rather than a specific disposition for certain low risk and high risk offenders.
For those offenders who are identified as low risk or high risk by the proposed instrument, the Commission recommends, but does not require, that the court seek additional information in the form of a pre-sentence investigation (PSI) report containing risk and needs information and/or a 3rd or 4th generation risk assessment. The risk assessment does not recommend a sentence to be imposed, but rather serves as a tool that identifies individuals with profiles that are higher or lower than average. Since these individuals are not typical offenders with respect to their risk of recidivism, the court will likely benefit from seeking additional information prior to imposing sentence. This approach builds on the work of the National Center for State Courts (NCSC), through its Center for Sentencing Initiatives, of recommending managing and reducing the risk of recidivism by ensuring that the intensity of supervision and treatment services is proportionate to the offender's level of risk, and that the conditions of supervision address the offender's criminogenic needs. It is for this reason that the Commission has provided an example of a Risk-Needs-Responsivity Presentence Investigation Report (RNR PSI) based on best practices in other jurisdictions.
While the text of the proposal states that the risk score or category is not intended to be used as an aggravating or mitigating factor, this language could be more explicit, and the Commission could consider whether mitigation may be appropriate and whether the risk information should remain confidential.
(c) Authorization to use the instrument to evaluate relative risk
1. Offender will re-offend and be a threat to public safety
a. Comments received
Concerns were raised by some that the instrument should focus on violent offending, while others suggested that the Commission should remove the ''crime against a person'' scale altogether.
b. Staff response
The legislation requiring the Commission to adopt a risk assessment instrument also specifically authorizes the use of the instrument to ''evaluate the relative risk that an offender will reoffend and be a threat to public safety.'' To address the threat to public safety, the Commission established a separate scale for ''offense against a person.'' The outcome measure for recidivism in this scale is an offense involving a crime of violence (42 Pa.C.S. § 9714), a personal injury crime under the Crime Victims Act (18 P.S. § 11.103), a sexual offense requiring registration (42 Pa.C.S. § 9799.14), or an offense defined as a danger to persons (18 Pa.C.S. Article B). Although ''threat to public safety'' is not defined in the statute, a common definition of ''safety'' is the welfare and protection of the general public (Black's Law Dictionary); ''public welfare'' is defined as a society's well-being in matters of health, safety, order, morality, economics, and politics (Black's Law Dictionary).
(2) Sentencing Guidelines
42 Pa.C.S. § 2154.7(b)
Sentencing guidelines.—The risk assessment instrument may be incorporated into the sentencing guidelines under section 2154 (relating to adoption of guidelines for sentencing).
42 Pa.C.S. § 2154(a)
In adopting guidelines, the Commission shall recommend confinement that is consistent with the protection of the public, the gravity of the offense as it relates to the impact on the life of the victim and the community, and the rehabilitative needs of the offender. The guidelines shall also address the following. . .
(3) Criminal behavior, by specifying a range of sentences of increased severity for defendants who pose a substantial risk to public safety, including those who possessed or used a deadly weapon during the commission of the current conviction offense.
42 Pa.C.S. § 2153(a)(14)
The Commission, pursuant to rules and regulations, shall have the power to: . . .
Establish a program to systematically monitor compliance with the guidelines, with the risk assessment instrument. . . by:
(i) promulgating forms which document the application of sentencing, resentencing and parole guidelines, mandatory sentencing laws, risk assessment instrument, . . .
(ii) requiring the timely completion and electronic submission of such forms to the Commission.
Comment
This provision permits the instrument to be incorporated into the sentencing guidelines. Reference to section 2154 addresses the consideration of risk assessment when adopting recommendations for confinement consistent with protection of the public, gravity of the offense, and the rehabilitative needs of the offender. This section also requires the Commission to specify a range of increased severity for defendants who pose a substantial risk to public safety. Reference to section 2153, which are the general powers and duties of the commission and are related to adoption of guidelines and the risk assessment instrument, requires the promulgating of forms which document the application of the instrument, and timely completion and electronic submission of the forms to the Commission.
(a) May incorporate instrument into sentencing guidelines (policy considerations)
1. Recommend confinement consistent with protection of public safety, gravity of offense, rehabilitation needs of offender
a. Comments received
Concerns were raised that, notwithstanding the authorization provided in the legislation, risk should not be part of the sentencing process as it: applies group characteristics to individual offenders; may be used to increase sentences; and may contribute to further racial disproportionality. Several individuals suggested that the Commission expressly prohibit use of risk assessment as an aggravating or mitigating factor, while others suggested using low risk to mitigate sentences and to select alternatives. Support was expressed for permitting an offender to waive completion of a risk assessment, and for the confidentiality of the risk assessment information.
b. Staff response
A combination of 1st and 2nd generation instruments (i.e., professional judgement informed by actuarial data) has been found to perform better than the use of either alone. Disproportionality rates are a concerning part of the criminal justice system. Properly validated risk assessments have not been found to increase racial disparity within the system, however they may reflect disparities present. Through the use of additional assessments and the gathering of more individualized information, courts are better informed of the needs associated with the particular risk level, and better able to balance the competing purposes of public safety, offense seriousness, and offender rehabilitation.
2. Specify increased severity for defendants who pose a risk to public safety
a. Comments received
As noted previously, concerns were raised about the use of risk in the sentencing process, and the equating of risk with harm and linking it to punishment. Some suggested that a risk assessment should only focus on violent offenders and defining threats to public safety as threats of violence, with outcome measures limited to new convictions for violent offenses.
b. Staff response
Threat to public safety is an important consideration in the sentencing guidelines and is demonstrated through the linking of sentence recommendations to the seriousness of the conviction as well as the offender's criminal history and criminal behavior. In addition to the authorization provided in the risk assessment legislation, an offender's risk to public safety and likelihood of committing a future offense is required to be considered by the court at sentencing (see: 42 Pa.C.S. §§ 9714(a)(2); 9721(b); 9722; 9725).
The Commission developed a separate risk assessment scale for predicting risk of recidivism for a crime against a person. As compared to general recidivism, recidivism for a crime against a person is rare. Under the proposal, only information on low risk of an offense against a person would be presented. The reason for this limited use of information relates to the accuracy rates: low risk designations had a high rate of accuracy (94% accurate), while high risk designations had a low rate of accuracy (18% accurate). While the scale was effective at identifying offenders who committed crimes against persons, it also identified many more offenders who did not (false positives), and the concerns raised about the potential harm of mislabeling these offenders led the Commission to limit the use of the crime against a person scale to recommendations for low risk offenders.
(b) May incorporate instrument into sentencing guidelines (technical/implementation considerations)
1. Requirements to document the application of risk assessment instrument
a. Comments received
Concerns were raised about resource requirements, the processing of cases, and the technology associated with the application if incorporated into the sentencing guidelines and the local sentencing process. Resource concerns included increased workload for courts and/or need to change existing practices, unfunded mandate to prepare and consider additional information at sentencing, and unnecessary burden on the system. Process concerns included interfering in local practices, limitations on judicial discretion, use of group characteristics to drive individual sentences, and the reliance on computers to replace human decisions. Technology concerns included system programming, ownership of computer code, troubleshooting of problems and errors, disclosure of information, and the tracking/auditing of the system.
b. Staff response
The Commission's JNET-based SGS Web application has been available to county users since 2002, with the preparation of guidelines and the electronic reporting of sentences via SGS Web required as of July 1, 2005. Deploying the risk assessment instrument as an automated module of SGS Web adds no addition work to counties using SGS Web as recommended. Counties that fail to prepare guidelines using SGS Web prior to sentencing will be required to change procedures in order to generate a Sentence Risk Assessment Summary prior to sentencing. The risk assessment instrument is automated, but uses SGS Web to compile information obtained from the sentencing guidelines and from AOPC to generate the offender-specific scales. All existing application development/support and data management protocols governing operation of SGS Web apply to the risk assessment module and are maintained by the Commission in cooperation with JNET and the AOPC.
2. Require timely completion and electronic submission
a. Comments received
As noted above, concerns were raised about the timing/processing of cases.
b. Staff response
In 2016, over 142,000 sentences were reported to the Commission via SGS Web. This represented more than 96,000 criminal incidents, 87,000 judicial proceedings, and nearly 74,000 unique offenders. SGS Web has been a reliable guideline preparation and reporting system for more than 15 years, and the risk assessment module is able to leverage the existing technology and information to provide an automated Sentence Risk Assessment Summary upon completion of the pre-sentence guideline forms. However, this automated module requires county users to prepare the guideline forms using SGS Web prior to sentencing. This procedure was adopted by the Commission as part of the 6th Edition Sentencing Guidelines (2005). In a number of counties, guidelines are determined informally prior to sentencing, and then entered into SGS Web following sentencing along with the sentencing details. Commission staff is available to work with counties to improve processes to take full advantage of features of SGS Web, so as to generate guideline forms and the Sentence Risk Assessment Summary prior to sentencing using existing county resources.
(3) Presentence Investigation Report
42 Pa.C.S. § 2154.7(c)
Presentence investigation report.—Subject to the provisions of the Pennsylvania Rules of Criminal Procedure, the sentencing court may use the risk assessment instrument to determine whether a more thorough assessment is necessary and to order a presentence investigation report.
Pa.R.Crim.P. Rule 702 (Aids in Imposing Sentence)
(A) Pre-sentence Investigation Report
(1) The sentencing judge may, in the judge's discretion, order a pre-sentence investigation report in any case.
(2) The sentencing judge shall place on the record the reasons for dispensing with the pre-sentence investigation report if the judge fails to order a pre-sentence report in any of the following instances:
(a) When incarceration for one year or more is a possible disposition under the applicable sentence statutes;
(b) When the defendant is less than 21 years old at the time of the conviction or entry of a plea of guilty; or
(c) When defendant is a first offender in that he or she has not heretofore been sentenced as an adult.
(3) The pre-sentence investigation report shall include information regarding the circumstances of the offense and the character of the defendant sufficient to assist the judge in determining a sentence.
(4) The pre-sentence investigation report shall also include a victim impact statement as provided by law.
(B) Psychiatric or Psychological Examination. After a finding of guilt and before the imposition of sentence, after notice to counsel for both parties, the sentencing judge may, as provided by law, order the defendant to undergo a psychiatric or psychological examination. For this purpose the defendant may be remanded to any available clinic, hospital, institution, or state correctional diagnostic and classification center for a period not exceeding 60 days.
Comment
This provision permits the instrument to be used to determine whether a more thorough assessment is necessary and to order a pre-sentence investigation (PSI) report. Provisions related to the ordering of a PSI report and other aids in imposing sentences are governed by the Pennsylvania Rules of Criminal Procedure. Similar and related provisions enacted in statute have been suspended by Pa.R.Crim.P. Rule 1101(6) as being inconsistent with Chapter 7 of the Rules of Criminal Procedure.
(a) Additional information
1. May be used to determine whether a more thorough assessment is required
a. Comments received
Similar to the policy considerations of incorporating risk into the sentencing guidelines, concerns were raised about the impact of preparing additional presentence investigation reports or requiring additional assessments on local resource (unfunded mandate) and in the processing of cases. In addition to the specific issues raised earlier, the focus of concerns related to the capacity of probation departments to conduct 3rd and 4th generation assessment for the targeted high and low risk cases, the certification of probation staff to conduct such assessments, the local validation of such instruments, and the resources required.
b. Staff response
As noted previously, training and certification, inter-rater reliability and fidelity, are important considerations for the use of 3rd and 4th generation assessments. These county-specific safeguards are the focus of the Evidence-Based Practice (EBP) initiative of the County Chief Adult Probation and Parole Officers Association of Pennsylvania, and for which state and federal grant funding is available to counties. In addition to these efforts, the current Justice Reinvestment Initiative (JRI-II) has recommended the expanded use of risk, needs and responsivity to inform sentencing decisions and to improve the efficacy of sentencing decisions, especially related to duration and intensity of community supervision. This recommendation, along with increased funding to counties to support these services, is included in legislation currently being considered by the General Assembly (see: SB1070, SB1071, SB1072).
The proposed instrument is viewed as a screening tool to prioritize those cases (high risk and low risk) where additional information may have the greatest impact in terms of public safety and resource utilization. This is consistent with the work of the NCSC Center for Sentencing Initiatives and their support of the use of risk and needs at sentencing to ensure that the intensity of supervision and treatment services is proportionate to the offender's level of risk, and that the conditions of supervision address the offender's criminogenic needs.
2. May order a PSI
a. Comments received
Concerns were raised about the Commission's recommendation that courts order presentence investigation reports in high and low risk cases. Some echoed earlier concerns and indicated that judges do not need assistance in determining when to order a PSI, that the recommendation interferes with judicial discretion, and criticized the Commission for not providing data on the use of PSI reports. Others reiterated that is was an unfunded mandate in that it would increase the workload of the courts.
b. Staff response
The Commission has long worked to improve the number and quality of PSI reports to better inform sentencing decisions. During the past decade, two work groups emerged and two products were deployed: in 2006 the JNET Office created a PSI Work Group and later that year launched the JNET Statewide Presentence Investigation Index; in 2007 the AOPC established the PSI Work Group and later that year made the Pre-Sentence Summary Reports available through CPCMS. Both efforts were intended to improve accessibility to offender information, and to make relevant information more readily available at the time of sentencing. However, orders of PSI reports vary dramatically across counties, from a low of 2% to a high of 100%. In 2016, based on information reported to the Commission, 13,990 PSI reports were ordered statewide; this represented 22% of all judicial proceedings reported to the Commission.
Experiences from other jurisdictions and in at least one Pennsylvania county has found that using risk and needs information to inform sentencing decisions improve outcomes and caseload management. In discussing the 'risk principle' and the need to match the level of supervision and services to an offender's level of risk of reoffending, the NCSC's Center for Sentencing Initiatives notes: ''more intensive interventions should be reserved for higher risk offenders while lower risk offenders should normally receive minimal or no intervention. Whereas intensive interventions may decrease recidivism risk among high risk offenders, research has shown that the use of numerous or intensive interventions with low risk offenders can actually increase their likelihood of recidivism.''
As with volume, the quality of PSI reports varies substantially from county to county, and few PSI reports include detailed and individualized risk-needs-responsivity (RNR) information on the offender or contain specific recommendations for interventions to address criminogenic needs. Decisions regarding disposition or duration or intensity may be driven by retributive factors or by the sentencing guidelines or by negotiated pleas, but careful consideration of assessed risk and needs prior to sentencing, or attention to dosage or domains to target, is the exception.
The sentencing guidelines provide the court with broad discretion through a range of sentencing options, both in terms of disposition and duration. The proposed risk assessment instrument is intended to serve as a screening tool, recommending preparation and review of additional RNR information for high risk and low risk cases prior to sentencing. The RNR information may be useful in selecting an option that best addresses criminogenic needs from the recommended range and/or determining the duration or intensity of services, and may include the use of alternative sentencing.
Under the Commission's proposal, based on all sentences reported to the Commission in 2016, a RNR PSI report would have been recommended in 1,861 cases (Phase 1, high risk and low risk offenders); a total of 2,435 PSI reports were ordered statewide in 2016. While efforts are underway through the EBP Initiative and JRI-II to increase capacity and funding for a greater use of RNR, certain counties are better positioned than others to re-target existing resources. Conducting 3rd and 4th generation assessment earlier in the process and modifying existing PSI reports to incorporate RNR information in high and low risk cases may provide an opportunity to test the value of enhanced information at sentencing.
(4) Alternative Sentencing
42 Pa.C.S. § 2154.7(d)
Alternative sentencing.—Subject to the eligibility requirements of each program, the risk assessment instrument may be an aide to help determine appropriate candidates for alternative sentencing, including the recidivism risk reduction incentive, State and county intermediate punishment programs and State motivational boot camps.
61 Pa.C.S. Chapter 45 (relating to Recidivism Risk Reduction Incentive Program)
61 Pa.C.S. § 4503 (definition of eligible offender)
61 Pa.C.S. § 4505(b) (waiver of eligibility requirements)
61 Pa.C.S. Chapter 41 (relating to State Intermediate Punishment)
61 Pa.C.S. § 4103 (definition of eligible offender)
61 Pa.C.S. § 4104(a)(1.1) (waiver of eligibility requirements)
42 Pa.C.S. Chapter 98 (relating to County Intermediate Punishment)
42 Pa.C.S. § 9802 (definition of eligible offender)
42 Pa.C.S. § 9804(b) (waiver of eligibility requirements)
61 Pa.C.S. Chapter 39 (relating to State Motivational Boot Camp)
61 Pa.C.S. § 3903 (definition of eligible inmates)
61 Pa.C.S. § 3904 (waiver of eligibility requirements)
204 Pa. Code 303.12 (relating to Guideline sentence recommendations: sentencing programs)
204 Pa. Code 303.12(a) (County intermediate punishment)
204 Pa. Code 303.12(b) (State motivational boot camp)
204 Pa. Code 303.12(c) (State intermediate punishment)
Comment
This provision permits the instrument to be used as an aide to help determine appropriate candidates for alternative sentencing, including RRRI, SIP, CIP and BC. Program eligibility includes consideration of both the statutory eligibility and the eligibility under the sentencing guidelines. Additionally, for each of the programs, the prosecuting attorney is authorized to waive eligibility requirements.
(a) May be used to determine appropriate candidates for alternative sentencing
1. Subject to eligibility requirement: definition of eligible offender
a. Comments received
No specific comments received.
b. Staff response
While the definition of ''eligible offender'' for each of the alternative sentencing programs is provided in statute, in each case the prosecutor has the authority to waive the eligibility requirements. As a result, both the sentencing guidelines and the exercise of discretion by the prosecutor can be used to promote broader consideration of these sentencing alternatives. Providing the court with additional information at sentencing, particularly information linked to needs and responsivity, may further promote the consideration of these programs. In many cases (Level 3 and Level 4 of the sentencing guidelines), the standard range of the guidelines permits and encourages consideration of these programs as options at sentencing. However, consideration could be given to identification of low risk or of needs identified through a 3rd or 4th generation assessment as a mitigating circumstance.
2. Programs: RRRI, SIP, CIP, BC
a. Comments received
No specific comments received
b. Staff response
The Commission is required to conduct bi-annual evaluations of the RRRI & SIP, and previously conducted evaluations of CIP and BC. In all cases, the Commission found under-utilization of these programs. Financial incentives, in the form of funding of treatment (RIP/D&A), increased utilization. Outcome studies of CIP and SIP have consistently found lower recidivism rates for offenders who successfully complete clinically prescribed treatment. The Justice Reinvestment Initiative (JRI-II) has proposed substantial changes to these statutes to ease access and improve targeting. JRI-II also promotes the linking of duration and intensity of probation and non-confinement sentences with risk, needs and responsivity. Evaluations of the programs described below are posted on the Commission's website.
Recidivism Risk Reduction Incentive Program (RRRI) (61 Pa.C.S. Chapter 45)
Intent. This chapter is intended to encourage eligible offenders committed to the custody of the department to participate in and successfully complete evidence-based programs under this chapter that reduce the likelihood of recidivism and improve public safety.
State Intermediate Punishment Program (SIP) (61 Pa.C.S. Chapter 41)
The program shall be designed to address the individually assessed drug and alcohol abuse and addiction needs of a participant and shall address other issues essential to the participant's successful reintegration into the community, including, but not limited to, education and employment issues.
County Intermediate Punishment Program (CIP) (42 Pa.C.S. Chapter 98)
Purpose. County intermediate punishment programs shall be developed, implemented and operated for the following purposes:
(1) To protect society and promote efficiency and economy in the delivery of correctional services.
(2) To promote accountability of offenders to their local community.
(3) To fill gaps in local correctional systems and address local needs through expansion of punishment and services available to the court.
(4) To provide opportunities for offenders who demonstrate special needs to receive services which enhance their ability to become contributing members of the community.
Motivational Boot Camp (BC) (61 Pa.C.S. Chapter 39)
Motivational boot camp. A program in which eligible inmates participate for a period of six months in a humane program which shall provide for rigorous physical activity, intensive regimentation and discipline, work on public projects, substance abuse treatment services licensed by the Department of Health, continuing education, vocational training, prerelease counseling and community corrections aftercare.
Annex B
Request for Proposals Act 2010-95, effective October 27, 2010, directed the Pennsylvania Commission on Sentencing to adopt a Sentence Risk Assessment Instrument for the sentencing court to use to help determine the appropriate sentence within the limits established by law for defendants who plead guilty or nolo contendere to, or were found guilty of felonies and misdemeanors. The Commission has developed, published, and received public comment on three versions of a proposed Sentence Risk Assessment Instrument. As the Commission continues to refine its proposal, the Commission will consider other proposals from the public.
The Commission is soliciting proposals from all interested persons or organizations for a Sentence Risk Assessment Instrument for adoption by the Commission in fulfillment of the requirements of Act 2010-95 (42 Pa.C.S. § 2154.7). Proposals shall describe: the methodology proposed for development and validation of the instrument; the factors, scales and outcome measures used in the instrument; the procedures for employing the instrument; and the recommended use of the instrument. Proposals shall address all requirements described in 42 Pa.C.S. § 2154.7 (relating to adoption of a risk assessment instrument):
(a) General rule.—The commission shall adopt a Sentence Risk Assessment Instrument for the sentencing court to use to help determine the appropriate sentence within the limits established by law for defendants who plead guilty or nolo contendere to or who were found guilty of felonies and misdemeanors. The risk assessment instrument may be used as an aide in evaluating the relative risk that an offender will reoffend and be a threat to public safety.
(b) Sentencing guidelines.—The risk assessment instrument may be incorporated into the sentencing guidelines under section 2154 (relating to adoption of guidelines for sentencing).
(c) Presentence investigation report.—Subject to the provisions of the Pennsylvania Rules of Criminal Procedure, the sentencing court may use the risk assessment instrument to determine whether a more thorough assessment is necessary and to order a presentence investigation report.
(d) Alternative sentencing.—Subject to the eligibility requirements of each program, the risk assessment instrument may be an aide to help determine appropriate candidates for alternative sentencing, including the recidivism risk reduction incentive, State and county intermediate punishment programs and State motivational boot camps.
(e) Definition.—As used in this section, the term ''risk assessment instrument'' means an empirically based worksheet which uses factors that are relevant in predicting recidivism.
All proposals should be submitted in writing to:
Pennsylvania Commission on Sentencing
Attn.: Sentence Risk Assessment Instrument Proposals
P.O. Box 1200
State College, PA 16804-1200
fax: 814.863.2129
e-mail: mhb105@psu.eduAll proposals should be received no later than November 30, 2018. E-mail is the preferred method for submitting proposals.
[Pa.B. Doc. No. 18-1342. Filed for public inspection August 31, 2018, 9:00 a.m.]
No part of the information on this site may be reproduced for profit or sold for profit.This material has been drawn directly from the official Pennsylvania Bulletin full text database. Due to the limitations of HTML or differences in display capabilities of different browsers, this version may differ slightly from the official printed version.
