CVDMP Site Visit July 9-11, 2012 Site Visit Report, August 31, 2012 Prepared by Madeline M. Carter, Center for Effective Public Policy CVDMP Site Visit July 9-11, 2012 Site Visit Report, August 31, 2012 Background Work with the Colorado Department of Corrections, Division of Adult Parole, Community Corrections and the Youthful Offender System ("the Division") began in February 2010. Between February 2010 and April 2011 the consultant provided training and technical assistance to the Division in the development of the Colorado Violation Decision Making Process (CVDMP). In April 2011, the division fully implemented the CVDMP and since that time, the consultant has been providing assistance in refining the CVDMP process, including providing guidance on the design of a complimentary behavioral incentives and rewards system, and advancing the use of evidencebased practices to improve offender outcomes. CVDMP Site Visit July 9-11, 2012 This report summarizes the activities and resulting recommendations from the CVDMP site visit conducted by Madeline M. Carter ("the consultant"), Center for Effective Public Policy, from July 9-11, 2012. The purpose of the July 2012 site visit was to provide input on: ? CVDMP data analyses, April 2011- June 2012, and the Division's CVDMP quality assurance measures, ? The Division's Phase III (Community Corrections) CVDMP strategy, ? The Division's plans to implement an offender behavior management strategy, specifically developing an incentives and rewards policy, and ? The next steps the Division might take to advance their efforts to improve offender outcomes. Activities Conducted The July 2012 site visit included the following off-site, preparatory activities. ? Off-site review of CVDMP Quality Assurance reports prepared between April 2011 and June 2012. 1 ? Off-site review of recently disseminated staff survey. ? Off-site coordination (i.e., conference calls, email exchanges) with agency representatives to plan the site visit itinerary. ? Off-site preparation of meeting agendas/materials, specifically an agenda for staff focus groups and a set of focus group survey questions. The July 2012 site visit included the following on-site activities. Monday, July 9, 2012 CVDMP Site Visit July 9-11, 2012 ? Travel to Denver from Baltimore. ? Meeting with the CVDMP Core Group/Managers from 1:00 - 5:00 pm to discuss implementation progress on the CVDMP and offender behavior management strategy. Tuesday, July 11, 2012 ? Conduct staff focus groups 9:00 - 11:00 am (Englewood); 12:30 pm - 2:30 pm (Westminster); and 3:00 pm - 5:00 pm (Broadway) - each with video participation from staff representing Alamosa, Canon, Durango, Fort Collins, Grand Junction, Greeley, La Junta, Longmont, Pueblo, Sinton, Sterling, and/or W. Colorado - to receive input on the planned offender behavior management strategy. Wednesday, July 12, 2012 ? Meeting with the Division Managers from 9:00 am - 12:00 pm to debrief on the staff focus groups and provide input on the Division's next steps plan. ? Meeting with Deputy Director Hand from 12:00 pm - 1:30 pm to debrief on the site visit and provide recommendations on advancing evidenced-based practices within the Division. ? Return flight from Denver to Baltimore. The following activities were conducted following the site visit: ? Review and feedback on a presentation Division managers prepared for the American Correctional Association conference convened in July 2012. ? Preparation of this report. Findings and Recommendations 2 The following findings and recommendations were reported to the CVDMP Core Group, Managers and or Deputy Director Hand during the course of the site visit and site visit debrief. Regarding CVDMP Quality Assurance Division staff currently collects and analyzes division-wide CVDMP data on a monthly basis. These data provide useful information on a variety of factors including but not limited to the number of CVDMP's completed across the division, by region, by unit and by staff person. Data is parsed to examine CVDMP outcomes by risk level and violation severity. Adherence to expected practice is monitored, in part, through these analyses. For this reason, quality assurance is a critical component of the CVDMP work. Recommendation #1 The consultant provided feedback on the proposed process to further study overrides to jail, supporting the effort but urging the department to avoid the use of freeform fields on the data collection instrument, as is currently planned, and instead limiting responses to pre-set selections plus an "other" field. Such an approach will provide uniform data reporting and improve and streamline data analysis. Recommendation #2 In addition, the consultant recommended developing an automated report that would be generated on a monthly basis that would aid the Division in identifying parolees with ongoing behavior management problems who continually fall within the CVDMP presumptive response categories. Such analyses allow management to identify instances where officers may be over-looking or under-responding to chronic behavior problems. It was recommended that this report include the following data fields and be reviewed by unit supervisors and regional managers on a monthly basis: ? ? ? ? ? CVDMP Site Visit July 9-11, 2012 Future quality assurance measures include a planned, manual data collection and analysis effort that will provide more detailed information about CVDMP incidents that result in an override to jail. A retrospective evaluation of data will involve AD's pulling files on cases with these outcomes and supplying pre-determined information for purposes of in-depth analysis. Offender name, Offender risk level, Number of months on supervision, Total number of CVDMPs since placed on supervision, Total number of CVDMPs in the last quarter. Recommendation #3 As will be discussed later in this report, the CVDMP reinforces the use of evidence-based practices; it does not substitute for the other important, risk-reducing strategies parole agencies can and should utilize. Current quality assurance measures - which are quantitative by design - do not provide sufficient insight into the degree to which supervising officers are effectively addressing offender behavior problems. Furthermore, current quality assurance measures do not provide sufficient information to determine if officers are linking behavioral responses to offenders' criminogenic needs. For these reasons, the consultant recommended that the Division institute a routine process of convening case staffings. Case staffings may be conducted at the individual level (one-on-one between the parole officer and their supervisor) or, (ideally) in a group setting, where all officers meet together with their 3 supervisor. In either case, the purpose is to ensure that all active cases are reviewed with regularity; that problem cases are reviewed in-depth and careful strategizing around case planning and behavioral concerns are addressed; and that ongoing coaching and mentoring of staff occurs. CVDMP Site Visit July 9-11, 2012 Recommendation #4 During discussions with the Core Group/Managers it was determined that final policy regarding the CVDMP has not been written and disseminated. This seems to have resulted in some differences in the way some cases are processed (e.g., instances where multiple violations occur at the same time; instances where multiple violations occur over a period of time). To ensure consistency in staff responses to violations - which also relates to the quality of the data collected - the consultant recommended that all documents pertaining to use of the CVDMP be updated and that formal policy be finalized and disseminated. Regarding CVDMP Phase III Division staff provided an update on the efforts underway to adapt the CVDMP (or similar) methodology to parolees under the management of the Division of Criminal Justice (DCJ). These efforts appear to be progressing well and the process being used to design and implement a structured decision making tool for DCJ facilities appears sound. No recommendations were offered in this area. Regarding Plans to Implement an Offender Behavior Management/Incentives and Rewards Policy Core Group Members/Managers provided an update on subcommittee efforts to design a structured response to prosocial behavior, analogous to the CVDMP. These efforts included the administration of a division-wide survey which was completed by approximately 90 staff. The consultant reviewed the results of this survey prior to arriving on-site. To deepen understanding of staffs' views on the planned offender behavior management effort, the consultant was asked to conduct focus group meetings with staff from throughout the division. Focus group sessions were conducted over the course of a full day, in two-hour time blocks at three parole unit offices. Staff from outlying areas participated by video-conferencing. All staff who participated did so on a voluntary basis. For the most part focus group sessions were conducted absent managers and Core Group Members. Approximately 60 staff in total were interviewed through the focus group process. Focus group participants were provided a handout, prepared by the consultant, to introduce the focus group sessions. The handout defined the purpose of the session as an opportunity "to 4 discuss the division's efforts to advance the use of evidence-based practices to improve offender outcomes -- with a particular focus on recent steps to establish structured responses to prosocial behavior -- and seek staff input around these changes." Following introductions of the consultant and session participants, the consultant offered introductory remarks which included reviewing the purpose of the session; defining key ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? Prisons are not preparing offenders for parole supervision Offenders are not motivated to change Caseloads are too high to provide effective supervision Offenders no longer fear being on parole or their parole officer Programmatic services are absent; ineffective; unaffordable; or slots are unavailable due to waiting lists Tools available to officers to respond to non-compliance are inadequate Officers are under pressure not to revoke offenders in order to save the state money Staff morale is poor Forums for staff input that are "safe" are not offered Leadership does not demonstrate that they "hear" and accept input from staff The Department/Division does not have a "master plan" for change or, if such a plan exists, it is not shared with staff The Department/Division does not share the rationale for changes that are implemented; "we are just told to do it" Seek staff input with sincerity; seek staff input on the potential impact of changes before change is contemplated/implemented Brief staff before changes are implemented Do not implement policies that require a "cookie cutter approach"; retain room for individualization where appropriate Hold staff who are not adhering to policy/expectations accountable for their actions Supports could be provided that would improve morale (suggestions include flex schedules; permitting staff to donate leave to one another; hiring an external expert to conduct a staff morale assessment) Reduce the amount of time officers are required to perform data entry duties Provide incentives/rewards to staff who volunteer for extra duties Support the Employee Council (staff provided a favorable review of the Employee Appreciation event) Create an environment where all staff - regardless of their role or whether they have offender contact - feel that they are a part of the same, unified effort Increase communication between parole officers and reentry staff Support staff team building activities Do not add additional data entry requirements (expressed as "no more computer screens") Recommendation #5 The Division should develop a clear plan for implementing evidencebased practices and explicitly share this plan with staff. The plan should clearly articulate the CVDMP Site Visit July 9-11, 2012 terms/phrases that would be used throughout the session (i.e., "improved offender outcomes" and "evidence-based practices") and reviewing the handout "Applying Research to Supervision Practices: What Works in Changing Offender Behavior?" This was followed by an open, structured discussion led by the consultant. The following reflects common remarks made by staff during the course of these focus groups. While there was not unanimous agreement among all those interviewed on all statements - and therefore these do not reflect consensus statements - there was strong agreement among most staff around these sentiments. 5 CVDMP Site Visit July 9-11, 2012 Department and Division's vision for correctional management and the specific strategies that have been and are being implemented to achieve it. The plan should include timelines so that staff can determine where the Department/Division is in the implementation process, and what can be expected in the future. Recommendation #6 The Division should provide line level staff with clear and specific leadership and mentoring around evidence-based practices in parole supervision. Many staff appear to lack an understanding of the department's focus on evidence-based practices, risk reduction, and behavior change; the research that supports these strategies; or the knowledge and skills to effectively implement these techniques. Experience demonstrates that effective implementation of these strategies requires (1) advanced-level knowledge and skills on the part of managers and supervisors; (2) strong leadership on the part of managers and supervisors; (3) training and skill development among line staff; and (4) forums for coaching and mentoring of line staff. This recommendation, then, suggests that managers and supervisors themselves need training and coaching in evidence-based practices, in specific techniques for applying evidencebased practices, and in coaching and mentoring methods, while line staff needs introductory training and skill development and follow-up coaching and mentoring by their supervisors. Recommendation 7: Line staff indicates that workloads are too high to engage in meaningful contact with offenders and that a majority of their time is spent conducting data entry and other administrative duties. It is recommended that a workload study - or other method to determine the validity of these assertions - be conducted. If such an analysis reveals that staff perceptions are accurate, efforts should be undertaken to realign responsibilities so that staff engage in "work that matters" (i.e., work that results in risk reduction and improved outcomes). Recommendation 8: A majority of staff interviewed - who may or may not represent the large compliment of staff - indicate that the Division suffers from low morale. Further efforts should be undertaken to determine whether in fact this is the case, and if it is, to address it in a comprehensive, strategic manner. These efforts might include developing and administering an anonymous staff morale survey; convening unit-level meetings to discuss morale; or hiring an organizational development consultant to provide an assessment. Recommendation#9: Presuming that in fact staff lack full appreciation for the Department/Division's direction; an understanding of the research around risk reduction; and that assertions around employee morale are correct, it is recommended that the Division prioritize concentration on these issues - and the others included in this report - before further pursuing the offender incentives and rewards behavior management system. 6