COMMUNITY COLLABORATIVE COMMISSION - COMMUNITY SURVEY & RESEARCH COMMITTEE

200 E. Wells Street City of Milwaukee Milwaukee, Wisconsin 53202 Meeting Minutes COMMUNITY COLLABORATIVE COMMISSION - COMMUNITY SURVEY & RESEARCH COMMITTEE Paul Mozina, Chair. Deborah Blanks and Steve Jansen Monday, August 30, 2021 6:00 PM Virtual Join Zoom Meeting https://us02web.zoom.us/j/81193833350?pwd=Zm4zbW9KTjV4VFo0Y0paS1dhMHY5Zz09 Meeting ID: 811 9383 3350 Passcode: 7pfJbD One tap mobile +13126266799,,81193833350#,,,,*704523# US (Chicago) +16465588656,,81193833350#,,,,*704523# US (New York) Dial by your location +1 312 626 6799 US (Chicago) +1 646 558 8656 US (New York) +1 301 715 8592 US (Washington DC) +1 346 248 7799 US (Houston) +1 669 900 9128 US (San Jose) +1 253 215 8782 US (Tacoma) Meeting ID: 811 9383 3350 Passcode: 704523 Find your local number: https://us02web.zoom.us/u/kqNWgYiaK Call to order at 6:14 p.m. Present: Steve Jansen, Deborah Blanks and Paul Mozina. (Video of meeting here https://youtu.be/FP0lTFPORzw) 1. Minutes Discussion and Approval. Minutes from August 9, 2021 meeting unanimously approved 2. Update complete timeline on Resolution to conduct a Community Survey. We updated the Consultant Specifications incorporating more flexible timelines for completion of the Community Survey: “The consultant shall complete the execution of the survey within 4 months of signing the agreement/contract, at which time a preliminary report will be submitted to the Survey & Research Committee and a final report will be submitted to the Survey & Research Committee one month later with possible extensions negotiated by the Chair of the Committee and confirmed in City of Milwaukee Page 1 COMMUNITY COLLABORATIVE Meeting Minutes August 30, 2021 COMMISSION - COMMUNITY SURVEY & RESEARCH COMMITTEE writing.” 3. The MPD is required by the Charles Collins et al., vs City of Milwaukee et al. Settlement Agreement to provide data on Traffic Stops and Field Interviews to the FPC for public access. Review the scope and impact of the data integrity issues related to mismatched Computer Aided Dispatch (CAD) numbers between CAD, TraCS and RMS and discuss making a recommendation/resolution for the CCC to follow-up with the parties to the Settlement Agreement. We reviewed the MPD Data Integrity Analysis document appended to these minutes. The problem is that MPD members must manually enter the Computer Aided Dispatch (CAD) number when creating records related to traffic stops or field interviews and they sometimes enter the wrong value. The CAD number is the “key” connecting the data across MPD systems which enables getting a complete and accurate context for stop. Paul presented analysis of the numbers of contact summaries, traffic citation, non-traffic citation, warnings, and field interview records, by quarter spanning 2019-2020, that could not be matched correctly. The “missing” data includes the call type, vehicle search information, person search information, use of force, and probable cause for the stop, which are all essential for getting a complete picture of the incident. The data shows that MPD members have made steady improvement in entering the correct CAD numbers since the 1st quarter of 2021. However, more improvement is needed for the database to be sufficiently clean and accurate. Paul suggested creating a resolution to present to the full CCC to ask for their help in following up with the defendants regarding this and other issues with the data. The City Attorney is now the focal point for all inquiries regarding the Settlement Agreement and Paul explained that Assistant City Attorney Heather Hough asked him for a summary of all the issues he had been bringing up with the MPD, FPC, CJI and Common Council. He provided that summary in early June and has not heard anything back yet. Paul described the effort that the Crime and Justice Institute has made to “clean” the data by correcting the mismatched CAD numbers. We don’t know how many records they fixed and Paul repeated a suggestion he made at the last meeting that we ask if CJI could share their cleaned data. Deborah suggested adding statement explaining: What’s the impact? How does not having this data impact the community? How does it impact the lawsuit? Why is it important? She asked for a clear statement of why the data integrity issue I raised matters. How does it impact an individual or the community if these records don’t match? Is a citizen being penalized because these records don’t match? Is it true that community police relations can continue to be strained because there’s a perception of inadequate data and so the public can’t get the full picture of why things are happening, why police are taking the actions that they’re taking. So, inadequate data fosters the perception that police are acting inappropriately. Steve agreed with Deborah and added that we need to have matching data to have an accurate picture of what is going on on the ground. He explained how this data was related to the Community Survey data that will be collecting. He pointed to a possible discrepancy between whatever data is being collected and the feelings on the ground City of Milwaukee Page 2 COMMUNITY COLLABORATIVE Meeting Minutes August 30, 2021 COMMISSION - COMMUNITY SURVEY & RESEARCH COMMITTEE with the average person. He reiterated the importance of having accurate data to get an accurate picture. Steve explained that a motion is all we need to present to the full CCC for them to act. A more formal resolution for the Common Council would only be necessary if the defendants are not responsive to our request for consultation. We agreed that Paul would prepare a motion for consideration and refinement at our next meeting and Steve would make sure it got on the agenda for the next CCC meeting. 4. Discuss the research areas that the Committee should focus on going forward including: MPD Data, status of recommendations in the original Collaborative Community Committee’s “Report to the Community”, ongoing activities of the FPC and MPD (SOP Changes, New Initiatives e.g., new traffic enforcement effort), and any other suggestions you may have. This item was held until the next meeting. 5. Discuss semi-annual report of the Survey and Research Committee’s activities for submission to the full CCC. Steve clarified that the scope of the semi-annual report would include all of the CCC’s activities, not just the Community Survey and Research committee. It will be referred to as an Executive Summary report. The report will include a summary of the history of the CCC, the current membership, a synopsis of the committees and their roles, the accomplishments of the CCC, a detailed analysis of the work of committees, data collection challenges when it comes to outreach and recommendations going forward. Steve is looking for input from the rest of the committee on this report and he will present a draft for consideration soon. Meeting adjourned at 7:12 p.m. Minutes provided by Paul Mozina. MPD Data Integrity Analysis Resources Datasets are here: https://data.milwaukee.gov/organization/fire-and-police-commission It is important to review MPD SOP 085 – CITIZEN CONTACTS, FIELD INTERVIEWS, SEARCH AND SEIZURE when looking at this data. Review Appendix F: Encounter Data Linkage Charts (Data Model) in the Second Annual Report September 2020 Analysis of Traffic Stop Data, to get a better understanding of how the MPD data is structured. City of Milwaukee Page 3 COMMUNITY COLLABORATIVE Meeting Minutes August 30, 2021 COMMISSION - COMMUNITY SURVEY & RESEARCH COMMITTEE Review the MPD COMPLIANCE DATA DICTIONARY AND GUIDE that accompanies each quarterly extract and be aware that the data files included with each quarterly extract have changed over time. Why it matters CJI Second Annual Report September 2020 page 32 A key requirement of the Agreement is that the FPC publish “data on all traffic stops, field interviews, no-action encounters, frisks and searches… with the exception of any personally identifiable information” (SA IV.A.13). The FPC is required to post the stop data on an annual basis (IV.A.13) and has thus far posted on a quarterly basis. It is important for community members in Milwaukee to access and understand policing stop data but the manner in which the data is posted precludes that at the present time. page 35 The published datasets are numerous and complex, in part reflecting the systems and structures behind the data extraction. In their current form, these data files do not allow an average community member to easily review the data and get a good picture of stops in Milwaukee, which is the intent of the public posting requirement. We have encouraged the Defendants to make these data more accessible to the public, noting that this more accessible format is not an explicit requirement of the Settlement Agreement. What is wrong with the data? Crime and Justice Institute (CJI) Analysis of 2019 Traffic Stops, Field Interviews, No-Action Encounters and Frisks SEPTEMBER 2020 Page 11 (the challenges of correcting data loss) To prevent data loss, we clean the CAD number field for TraCS and RMS data to remove obvious data errors such as dashes or spaces. We also clean CAD numbers by checking for extra, missing, or transposed digits. We do this by comparing the merged CAD encounter file (location, data, time information) to the RMS or TraCS files we are cleaning (with location, date, and time information). If the associated information between the CAD file and RMS or TraCS file matches, and the CAD number differs only by missing, transposed, or extra digits, we generate a corrected CAD number so that the CAD files can match with the RMS or TraCS information as much as possible. Matching CAD information to TraCS or RMS information is essential in order to gain a complete understanding of the data elements present or missing from documentation of each encounter. [my emphasis] Page 12 (explanation of impact) The structure and association of the TraCS files requires each of the different forms (contact summary, ELCI, warning, and NTC) to relate back to the TraCS header file before creating datasets that represent all the associated information present for a person involved in a given police encounter. Invalid CAD numbers in citation and warning forms present the greatest challenge to this process in that the only way to City of Milwaukee Page 4 COMMUNITY COLLABORATIVE Meeting Minutes August 30, 2021 COMMISSION - COMMUNITY SURVEY & RESEARCH COMMITTEE associate citations or warnings to contact summaries or field interviews is to rely upon valid CAD numbers that match across the different forms. For example, if an officer makes a traffic stop and decides to issue a citation for speeding, documentation for the traffic stop would be present in the CAD files and there would be a row in the TraCS header file for the contact summary for the person involved in the traffic stop and another row for the speeding citation. Additional rows represent any warnings the officer may issue or additional contact summaries for passengers that may need to be documented. Associating all of this information in order to represent one traffic stop requires the correct CAD number to be recorded by officers on each form that matches the dispatched CAD number for that particular traffic stop. Breakdown of data integrity/loss related to mismatched, missing, or bad CAD Numbers Tableau Dashboards MPD Records With Bad CAD Numbers 2019-2020 These are counts of the numbers of records that had a CAD Number that could not be matched to a CAD Number in the Computer Aided Dispatch System (data entry error). ELCI, NTC and Warning With No Match to Contact Summary 2019-2020 These are records where we cannot match the CAD Number on a Traffic Citation, Non-Traffic Citation or Warning record to the CAD Number on a Contact Summary record. Thus we cannot find the reason for the stop, if the individual was searched, or if the vehicle was searched – we can’t tell the whole story. Traffic Contact Summaries 2019 and 2020 • 100,221 Total Records • 973 (1%) Contact Summary records could not be matched to a CAD record using DocumentPoliceNumber=CALL_NO To investigate the context of the issuance of a Traffic Citation (ELCI) to the full extent that the data schema supports, we need to be able to match the documentPoliceNumber from Tracs_ELCI_Joined to the CALL_NO in CAD_PCARSCALL_JOINED to get the CALL_TYPE. We need to match the documentPoliceNumber from Tracs_ELCI_Joined to the documentPoliceNumber in Tracs_ContactSummary_Joined to get the summaryReason, summaryOutcome and agencyAgencySpace (probable cause explanation) for the Traffic Stop. If we can match the ELCI record to the Tracs_ContactSummary_Joined record, then we can continue on to the Tracs_ContactSummary_Unit table to get the vehicleContraband and the vehicleSearchBasis and to the Tracs_ContactSummary_Indivdual table to get the individualContraband and City of Milwaukee Page 5 COMMUNITY COLLABORATIVE Meeting Minutes August 30, 2021 COMMISSION - COMMUNITY SURVEY & RESEARCH COMMITTEE individualSearchBasis. The analysis of NTC and Warning records is very similar to that of ELCI records and they suffer the same limitations when the documentPoliceNumber cannot be matched. In the case of Field Interviews, we need to match the documentPoliceNumber in the Tracs_ELCI_Joined table to the CADNumber in the Inform_FieldInterview_Joined table to pick up interesting context like the SuspectDescription, PatDownJustification, SearchJustification and StopJustification. Traffic Citation (ELCI) 2019 and 2020 • 93,404 Total Records • 23,552 (25%) ELCI records could not be matched to a CAD record using DocumentPoliceNumber=CALL_NO • 24,835 (26.6%) ELCI records could not be matched to a Contact Summary using the DocumentPoliceNumber and 68,569 (73.4%) could • 92,746 (99.3%) ELCI records could not be matched to a Field Interview using the DocumentPoliceNumber and 658 (.7%) could • 24,609 (26.3%) ELCI records could not be matched to either a Contact Summary or Field Interview using DocumentPoliceNumber and 68,795 (73.7%) could Non-Traffic Citation (NTC) 2019 Q3-4 and 2020 • 5,698 Total Records • 4,072 (71.5%) NTC records could not be matched to a CAD record using DocumentPoliceNumber=CALL_NO • 5,062 (88.8%) NTC records could not be matched to a Contact Summary using the DocumentPoliceNumber and 636 (11.2%) could • 995 (82.5%) NTC records could not be matched to a Field Interview using the DocumentPoliceNumber and 658 (17.5%) could • 4,130 (72.5%) NTC records could not be matched to either a Contact Summary or Field Interview using DocumentPoliceNumber and 1,568 (27.5%) could Traffic Warnings 2019 Q3-4 and 2020 • 34,187 Total Records • 1,204 (3.5%) Warning records could not be matched to a CAD record using DocumentPoliceNumber=CALL_NO • 1,509 (4.4%) Warning records could not be matched to a Contact Summary using the DocumentPoliceNumber and 32,678 (95.6%) could • 33,856 (99%) Warning records could not be matched to a Field Interview using the DocumentPoliceNumber and 331 (1%) could • 1,472 (4.3%) Warning records could not be matched to either a Contact Summary or Field Interview using DocumentPoliceNumber and 32,715 (95.7%) could Field Interviews 2019 and 2020 • 7,399 Total Records • 151 (2%) Field Interview records could not be matched to a CAD record using CADNumber=CALL_NO Status This information was communicated to the FPC, MPD, CJI, Common Council, and summarized in an email to the Assistant City Attorneys Heather Hough and Julie City of Milwaukee Page 6 COMMUNITY COLLABORATIVE Meeting Minutes August 30, 2021 COMMISSION - COMMUNITY SURVEY & RESEARCH COMMITTEE Wilson on June 3, 2021. No substantive response received to date. What happened to Non-Traffic Citation data in 2019 Q4 and 2020 Q1? The number of non-traffic citation records skyrocketed in 2019 Q4 and 2020 Q1 and most of these records could not be matched back to the CAD system, nor could they be matched to a Contact Summary. Status This information was communicated to the FPC, MPD, and summarized in an email to the Assistant City Attorneys Heather Hough and Julie Wilson on June 3, 2021. No substantive response received to date. City of Milwaukee Page 7