1 / 25

The Police Risk Assessment Instrument

The Police Risk Assessment Instrument. Captain Michael Gropman Ed.D Brookline Police Department Gina M Vincent, PhD Associate Professor UMMS Psychiatry Rachael Perrault, MA Research Coordinator, UMMS Psychiatry. Precursor to PRAI.

media
Télécharger la présentation

The Police Risk Assessment Instrument

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. The Police Risk Assessment Instrument Captain Michael Gropman Ed.D Brookline Police Department Gina M Vincent, PhD Associate Professor UMMS Psychiatry Rachael Perrault, MA Research Coordinator, UMMS Psychiatry

  2. Precursor to PRAI • Criminal Case: 1991-4880: 8-25-1991 @ 17:00 hours ~100 yards from Dana Farber in Boston. Operating on adjoining street and observe small child hanging from window of a parked MV & another acting as look out. Warm, sunny, August day in Boston, regular vehicle and foot traffic. • Hereto referred as: Harold the Hanger; William the Watcher, and Samuel the Squatter • I stop and exit vehicle and watch as Harold attempt to pull the window down. Another officer pulls up to check on me (very active area at time) and we enjoy a laugh. We approach w/o William detecting us; Harold walks away from car. The 3rd juvenile (Samuel) is in bushes squatting down as we approach. He’s ordered to come out, “I can’t I’m still going [defecating]…I have to go when I'm nervous.” We identify them: Harold is 11 y.o., William 11 y.o., Samuel 12 y.o. • Records checks for ID: Harold & Samuel have BOP’s; both have defaults for B/E of Museum of Science in 1990 (10 & 11 at time). Seek direction from Lieutenant Det. O’Leary. The 3 are brought to BPD in backseat of cruiser w/ PO back there, too. William & Samuel are released to parents; Harold’s parent can’t be located; he is sent to detention center. • Decide to track Harold as part of program; He is arrested in Brookline 3 more times by age 14. When family asks for me to, “get him out of the arrest in Boston,” a year later, research is ended. My first venture into LE research is eye-opening: • Risk factors present: 1) No parent available 2) Family on public assistance 3) Living in violent housing project (now torn down) 3) No father present 4) English not primary language 5) Older siblings have criminal records 6) Truancy ends up on BOP. • 16 years later I am assigned to the JDAI working group by Chief O'Leary

  3. Brookline, Massachusetts

  4. The Police Risk Assessment Instrument • Background: 2007 Assignment to Massachusetts JDAI • JDAI working group: DMC, DYS, DSS (DFC), Mass. Trial Court, Mass Probation • Keynotes: Daniel O’Leary-MMCA, Jane Tewksbury, Barbara Morton DYS; Honorable Jay Blitzman. • Fundamental issue: Controlling front gates of juvenile detention w/ limited resources & disjuncture services; Detention should be for high rick juveniles; Maybe we should consider a different approach? • DYS Research project undertaken after charter formed because: • 1 in 100 Adult males; 1 in 30 (20-34), 1 in 9 for black males are in jail. (Pew Report 2008) • Use of RAI widespread in Public Health: MAYSI, NYSAP, Structured Assessment of Violence Risk for Youth (SAVRY) & Pennsylvania Youth Level of Service/Case Management Inventory (YLS/CMI) • 2009 Watershed Moment: Why wouldn’t LE screen too? NO RAI for L.E. at POC; Utilizing an objective measuring tool removes subjective/discretionary nature of arrest; Structured Decision-making increases accountability at supervisory level; There are way too many low risk kids in JJS • Besides, How hard could it be to make a RAI? • Using the Dr. Kelling formula (G.C.C.S.) research began: Data compilation of Juvenile Arrest in Brookline, MA. over 1 year; Weighted Crimes: based on “Seriousness of Offense” developed by DYS; Excluded non-localized risk factors, gang affiliation, high on inhalants, etc. Completedpilot study w/ BU Dr Adam Naylor. ~5 years/8 PRAI’s later: Hard to make a RAI

  5. The Police Risk Assessment Instrument • What is Risk Assessment? The Process of quantifying probability. Use is Very common in P.H. • We make assessments on risk every day as supervisors: Should a 2nd vehicle been sent to that call? Should officer X be used in a U/C assignments? Is an area going to have sufficient P coverage in the wake of a shooting? Policies are designed to mitigate and manage negative outcomes (AKA Risks) It is a part of what in LE ; however, we don’t do that w/ juvenile detention. • What: A Risk Management tool to utilize post-arrest/pre-arraignment (Even pre-processing?) to determine risk of: reoffending, and failure to appear. • Who: 2008 started at Brookline PD,DYS, & JDAI working group (UMMS 2012) • Where: Currently: Brookline, Chelsea, MBTA, Framingham, Lynn, North Attleboro, et al. • How: Using existing data set (Pilot Study) add data from participating cities to compile & analyze attributes of juvenile population (post hoc) that increase or moderate risk. Trying to identify, w/ confidence, what attributes indicate risk, & what attributes do not indicate risk: re-offend & FTA. GOAL: to initiate a Public Health approach: ID Risk Factors, develop pathways to mitigate risk, decrease low risk youth detention and arrest. • Why: Detention & referral into the JJ system increases the likelihood of a number of social ills; risk of re-offending (1) ; dropoutrates (2); risk of suicide (3) ; Arrest decisions are often based on suspect behavior; At every stage of our CJS DMCis evident; Voluminous data indicates most juveniles are low risk and should not be held;It is counter to Rehabilitation: the purpose of the JJ system; • ** The current systems in place in Massachusetts are in need of a new approach**

  6. Massachusetts Juvenile Delinquency Lockup Flow Chart Secure detention: Physically detained or confined in a room, cell, etc. w/ ability to lock an individual within; being placed in such, or being physically secured to stationary object such as a cuffing rail. Delinquent Offense: any criminal-type charge or city ordinance/ town by-law. (non-age-related) MGL Status Offenders: Age related criminal-type charge under MGL C. 119, Sections 39H Arrest of Juvenile | Notify Juvenile Probation and Notify Parent/Guardian | Based on the juvenile offender’s criminal history, current charge(s), circumstances of the arrest, etc., decision by Probation to: A) Releasethe juvenile offender to a parent/guardian/ OR other reasonable adult/Probation Officer within 6 hours if securely detained. Hold sight and sound separate from adult detainees until released. B) Hold If decision is B, If eligible, contact Clerk- Magistrate for Bail Determination. In the meantime, or if cannot post bail: Hold sight and sound separate from adult detainees and release from secure custody within 6 hours to a parent/guardian (if bail is posted) or release to an alternative lock-up program

  7. From IACP Survey results to Action • 1) A Separate System for Juveniles An overwhelming majority of survey respondents (88%) responded affirmatively.. Nearly all of the respondents (94%) from large departments with 250 or more officers. Who’s responsibility is it to develop & maintain separate systems? “Innovation in Gov’t starts in LE.” George Kelling 2013 • 2) Efficacy of the Juvenile Justice System • Survey respondents were asked whether they thought that both the overall juvenile justice system and their local juvenile justice system: (1) improves public safety and (2) promotes rehabilitation. Just a quarter of agency executives believe the juvenile justice system overall improves public safety only about a third believe it promotes rehabilitation. Respondents expressed even less confidence in their local juvenile justice systems, with just 23% indicating their local system improves public safety and 29% saying that it promotes rehabilitation. • COMMENTS FROM AGENCY EXECUTIVES VISION FOR A SUCCESSFUL JUVENILE JUSTICE SYSTEM • A system that acknowledges the most important stakeholder in the process outside of the juvenile offender • is the parent or guardian. An acceptance that government and social agency interventions are no substitute • for parents and family. A system that assists but holds parents accountable for the actions of their children. • +++ A diversion system that addresses juveniles when they are caught or arrested on their very first offense. • ------A weekend stay at detention facility or significant community service…before they become hardened or used to the system. (NO)

  8. From IACP Survey results to Action • Executives from large agencies (250 officers or more) have a particularly negative view of their local systems’ effect on public safety, with 47%reporting their local juvenile justice system does not improve public safety, compared to 38% of all respondents. However, respondents from large departments are more likely to believe their local system promotes rehabilitation – just 29% said their local system does not promote rehabilitation, compared to 40% of respondents overall. • JJS Does NOT promote PS, but DOES promote Rehabilitation? Are Round pegs going into Round holes? • 3) Partnerships on Juvenile & Youth Issues • Law enforcement agencies are most likely to have formal partnerships with prosecutors, schools, juvenile courts, probation/parole, and other law enforcement agencies. Informal collaboration is relatively common across a range of partners and is particularly prevalent with families, social services, counseling/mental health services, and drug treatment services. A significant proportion of survey respondents said they have neither a formal nor informal relationship • Types of collaboration vary based on agency size, structure, and region: • Departments that have more sworn officers or at least one officer assigned to youth crime >> formal partnerships • Northeast Agencies more likely to have formal partnerships, least likely w/ no partnerships. Informal prevalent in West. • Without collaboration how do we ensure that Round pegs are directed into Round holes?

  9. From IACP Survey results to Action 4) Perceptions of Diversion • More than three-quarters of agency executives (76%) believe that diversion programs such as community service or drug treatment help to prevent future offenses by youth who have committed relatively minor crimes, and 67% believe such programs save tax dollars in the long run. Those in departments with youth crime officers…in the Northeast…in larger, non-rural jurisdictions are most likely to believe diversion programs prevent crime & save tax dollars. • Diversion: Prevents future crime AND saves Tax Dollars? Why not formalize as a LE best practice? • Comments from Agency Executives on…Diversion • -Kids need to be diverted from the system where appropriate in greater numbers. The system needs to really be looking at what is in the best interest of the kid… Figure out early on what's going on with these kids and their families and get them what they need. Avoid putting them in the criminal justice system in the first place. • - A successful juvenile justice system would look to limit criminalizing the conduct of youths when [they’re] better served through services or diversionary programs that help educate or better serve the juvenile. A formal system would exist for persistent violators or others.

  10. From IACP Survey results to Action • COMMENTS FROM AGENCY EXECUTIVES ON…COLLABORATION • -- You cannot do it alone as a chief – it has to be a collaboration. -- In 38 years plus experience in law enforcement, I believe we talk too much about working together in this matter of juveniles, but we definitely work independently to each other. • -- A successful “system” requires the involvement of the entire community (i.e. families, courts, law • enforcement, schools, faith-based organizations, healthcare, social services, corrections, etc.). As of now, these listed parties do not work in a coordinated fashion to offer the best services to our juvenile justice system… While certain strides have been made there is still too much finger pointing and friction • 5) Information Available Prior to Arrest or Diversion • In the majority of agencies (87%), officers have some types of information at their disposal prior to making decisions about arrest or diversion. Prior arrests or contact with their department is the information most widely available. About halfsaid they have access to probation status, prior arrests or contact with other agencies, or school status (expelled, suspended, or truant). Departments that have youth crime officers are more likely to have each of these sources. This information can be invaluable if used in a structured tool to properly classify youths on a risk spectrum so processing & detention decisions are made objectively

  11. From IACP Survey results to Action • Obstacles to Diversion • The vast majority of respondents, 81%, indicated they face at least some obstacles to diverting an optimal number of juveniles from formal processing. Legal constraints are the most frequently cited obstacle (52%). Departments in the south were particularly likely to cite legal constraints (63%). • Many pointed to an insufficient number of diversion programs (38%), diversion programs not reporting back on outcomes (35%), and inadequate knowledge of existing programs (29%) as major obstacles. • ONLY 10% of departments said diversion programs in their community are ineffective, 7% cited department culture as an obstacle. • With empirical evidence that diversion • Is a better alternative than detention • for low risk offenders, won’t building • partnerships & consensus be easy? • Won’t these partnerships be able to work • Together more easily to remove the • obstacles that exists b/w them?

  12. Design of the PRAI • Empirically-Based–items of the tool were selected because research has shown an association w/ • Re-arrest • Failure to appear for arraignment • Must be feasible for completion by police • Number of items studied: • 17 items based on offense and history • 3 behavioral observations • PRAI goal is to develop a national platform that: • Is as primary & complementary assessment instrument for juvenile screening • The other primary instrument is what George Kelling calls “GCCS” • Provides a common platform to weigh actions against risks: Action of officers on street weighed against Risk detention of a juvenile presents vs. diversion • Encourages increased supervision w/ accompanying wider set of tools to examine JJ System • Offers a standardized, objective, empirical basis for the processing of low risk juveniles down pathways to rehabilitation versus pathways to criminality • Re-direct JJS mandates towards rehabilitation vs. punishment & detention

  13. Example Items Historical Items • Prior FTAs? 0 1-2 >2 • Open warrants? 0 1-2 >2 • Prior Felony Convictions? 0 1-2 >2 Observational Items • Substance Abuse problems • Antisocial attitudes • Poor parental monitoring/environment

  14. Study Design and Progress Archival data collection • All youth arrested in Brookline from 2008-2011 • N = 96; Females = 18, Males = 78 • Race: White = 23%, Black = 49%, Hispanic = 27% • Prior arrests: 42% • Coding of items by Capt Gropman Follow-up data collection • Tracked new arrests for a minimum of one-year • Base rates: Any arrest = 57.3% (n = 55) • Violent arrests = 23% (n = 22) • FTA for arraignment = 6.3% (n = 6)

  15. Items Correlated With Violent Rearrest Race & Gender not correlated with violent rearrest

  16. Items Correlated With Any Rearrest Rates of rearrest not significant by Race or Gender

  17. Items Correlated With Any Rearrest Correctly classified 84% (64% w/o substance abuse)

  18. Items Correlated With FTA Rates of FTA not significant via Race or Gender but populations small; most FTA’s Black via residency; Surprising findings age of 1st arraignment

  19. Items Correlated With FTA Correctly classified 93.8%

  20. Risk Score for FTA Two versions • 6 item FTA risk score (not including length of residence) • Range 0 to 10; Mean = 3.12 (SD = 3.30) • Cutoff for upper third = 5 • Does not differ significantly between genders • Is significantly higher for Blacks than Whites or Hispanics [F(2, 91) = 5.54, p = .005] • 7 item FTA risk score (including length of residence) • Range 0 to 11; Mean = 3.92 (SD = 3.20) • Does not differ significantly between genders or race • 5 items (includes substance abuse) • Range of scores = 0 to 7; Mean = 1.92 (SD = 2.22) • The risk score does not differ significantly between genders or race • Predictive Accuracy (ROC curve): • AUC = .71(SE = .05), p < .001 [CI = .61-.81] • Ideal cut-off score = 3 (sensitivity = .47/specificity = .13)

  21. Risk Score for FTA (cont.) Predictive Accuracy (ROC curve): • 7 item FTA risk score (including length of residence) • AUC = .92(.03), p = .001 [CI = .87-.98] • Ideal cutoff 4 (?) sensitivity = 1/specificity = .3 • Many potential cut scores w/sensitivity = 1 so this requires cross-validation • SYRP OJJDP • or with prior convictions (12 percent). Most youth with prior convictions (74 percent) have committed another person offense or property crime7 as their most serious career offense. • Justice System Involvement • SYRP asks youth whether they were previ­ously taken into custody, on probation or parole, or convicted of any offenses— • information that enters into court deci­sions about where to place them in the • juvenile justice system. The large majority(85 percent) of youth interviewed have prior convictions. Ten percent report only prior custody, probation, or parole. Only 5 percent of youth in custody have no prior involvement w/JJ

  22. Summary Racial differences • Rearrest risk score – • No race differences among youths who were actually rearrested • Risk score not significantly different by race (albeit the p-value was close) • No appreciable differences in predictive accuracy by race (AUCs ranged .66 for Hispanics to .77 for Whites) • FTA risk score- tenuous due to low baserates and missing data • There is racial bias in the score when length of residence is NOT included • Problem – most youth who failed to appear for arraignment were Black; Baserate too low for meaningful ROC comparisons by race Next Steps • Cross-validate in other jurisdictions: Chelsea, MBTA, Framingham… • Use a sample stratified by race

  23. Police Risk Assessment Instrument (PRAI) Design: • Designed for: • Post-”arrest” decision-making by by police commander (Not arresting PO) • Determining who should be referred for arraignment & for divserion pre-arraignment & • Predicting probability of new arrests & likelihood to not appear for arraignment (FTA)>>>> • Reducing Disproportionate Minority Confinement AND Contact w/in JJ System & • Developing pathways to rehabilitation (vs. criminality) for low risk offenders • _________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________ • PRAI a structured tool that combines information about youth to classify them as low, moderate or high risk for reoffending or continued delinquent activity. The purpose is to help in making decisions about [arrest, referral and/or diversion] youths’ placement, [to] reduce their level of risk. (Vincent, et al, 2013) • RAI: Some risk assessment approaches [separate] risk and needs assessment. This may be appropriate in some settings where time is limited and a “screening out” process for low risk youths is needed…“brief” risk assessment tools typically consist of a short checklist of items, and they are designed only to sort youth into categories of low, moderate or high risk, not to offer treatment recommendations. The point of a brief risk tool generally is to sort people into categories to separate…Specifically, an individual who is flagged as high risk on a short risk tool is probably high risk. Unfortunately, there is no brief risk tool currently available that can identify low risk youth with high enough accuracy to say they “do not [problems] (Ibid.)

  24. Police Risk Assessment Instrument (PRAI) • What Does Risk Assessment DO? • It can minimize bias in judgments about youths’ risk to public safety (DMC) • Reduce costs by decreasing use of more intensive supervision, over-use of expensive incarceration, and provision of services for youth who do not need them. • Risk assessment provides a standardized method of important data collection for an agency. • Data can indicate most common areas of need among delinquent youth stopped/detained • Improves the targeting of services/interventions that would address youths’ identified risk factors. • Risk assessment tool can direct processing decisions: Detention v. Court v. Divserion, • What Risk Assessment Does NOT do? • A risk assessment will only estimate the likelihood that delinquent behaviors will continue • A trained individual can use tool to assess if youth is at relatively low or high risk for reoffending. • Can help ID what factors in a youth’s life are likely driving youth to offend but is not prescriptive • A risk assessment tool can direct intervention planning, only if it contains correlated risk factors • An assessment tool is a primary screening tool, but not an independent screening tool (Supra)

More Related