CPRA request — pre- and post-ALPR outcome data, case classification practices, and the methodology behind previously produced statistics

Re: CPRA request — pre- and post-ALPR outcome data, case classification practices, and the methodology behind previously produced statistics. Related to prior requests PR-2025-159, PR-2026-7, and PR-2026-155.

Pursuant to the California Public Records Act, Government Code § 7920.000 et seq., and Article I, § 3(b) of the California Constitution, I request copies of the records described below.

Purpose of this request. The City and the Department have publicly cited outcome figures — arrests, cases cleared, vehicle recoveries — attributed to the Flock Safety automated license plate reader (ALPR) system. Those figures cannot be evaluated without (a) a pre-deployment baseline for the same measures, (b) an understanding of how ALPR-involved incidents are classified and counted in the Department’s records management system, and (c) the query and report parameters used to generate the figures already released. This request seeks all three. I am not asking the Department to prove or disprove any conclusion; I am asking for the records that would let anyone test one.

Definitions

• “ALPR” means any automated license plate recognition system operated by, on behalf of, or accessible to the Benicia Police Department, including Flock Safety cameras (fixed and mobile), Vigilant/Motorola systems, and any ALPR data received through NCRIC or any other agency or network.
• “Baseline Period” means January 1, 2015 through December 31, 2021.
• “Deployment Period” means January 1, 2022 through the date this request is processed.
• “Hit” means any ALPR alert, hot list match, or notification generated to or received by Department personnel.
• “RMS” means the Department’s records management system, and “CAD” means its computer-aided dispatch system, including any predecessor or successor systems in use during either period.

Category 1 — System activation and coverage

1. Records sufficient to establish the date each ALPR camera became operational, including installation completion, acceptance, or activation records, and the date the system as a whole entered live service.
2. Records sufficient to show the number of ALPR cameras in active service, by location, as of December 31 of each year from 2021 through 2025.
3. Records sufficient to show any period during which the system, or any camera, was non-operational for more than 72 consecutive hours.

Category 2 — Baseline and comparison outcome data

For each calendar year in both the Baseline Period and the Deployment Period, I request existing reports, or reports generable from RMS/CAD using existing report definitions or standard query functions, showing:

1. Total reported offenses by offense classification (UCR/NIBRS Part I and Part II, or the Department’s equivalent classification scheme).
2. Total arrests, broken out by felony and misdemeanor, and by offense classification.
3. Cases cleared, by clearance type (cleared by arrest, cleared by exceptional means, unfounded, inactive/suspended), by offense classification.
4. Stolen vehicles reported in Benicia, and stolen vehicles recovered in Benicia, by year — including recoveries of vehicles reported stolen by other jurisdictions.
5. Vehicle-related offense counts: auto theft, theft from vehicle, vehicle vandalism, and recovered stolen plates.
6. Warrant arrests and arrests of persons with outstanding felony warrants.
7. Total vehicle stops, and total vehicle stops resulting in arrest, including RIPA (AB 953) stop data submissions and any local RIPA extracts for each year those submissions were made.

If any of these measures is maintained on a fiscal-year rather than calendar-year basis, produce them on the basis in which they exist.

Category 3 — ALPR-attributed outcomes and the denominator

1. Total ALPR plate reads (detections) per year, and per month if available, for the Deployment Period.
2. Total ALPR hits per year and per month, broken out by hot list source (e.g., SVS/stolen vehicle, NCIC, wanted person, missing person/Amber or Silver Alert, agency-created custom hot list, other-agency hot list).
3. Records showing the number of hits that resulted in an officer response, a vehicle stop, an arrest, a vehicle recovery, or a case report.
4. Records documenting misreads, false hits, or hits on vehicles later determined not to be the subject vehicle, including any log, memo, complaint, or after-action record concerning a stop initiated on an erroneous ALPR hit.
5. All Flock Safety-generated reports provided to the Department or the City, including but not limited to any document Flock labels an impact report, value report, ROI report, usage report, business review, or performance summary, for all years available.
6. All audit records maintained pursuant to Civil Code § 1798.90.52, including records of ALPR data access showing date, time, username, purpose, and the source of the data accessed, for the Deployment Period. If these are voluminous, produce them in native or CSV export rather than as printed or scanned pages, and I will narrow after review.

Category 4 — Case classification and report-writing practice

This category is the crux of the comparison problem: outcome counts are meaningless if ALPR-involved incidents are recorded under a different case-generation convention than pre-ALPR incidents were. I request:

1. All versions of Benicia Police Department Policy Manual § 470 (Automated License Plate Readers), and any successor or related policy, in effect at any time from January 1, 2019 to the present, including revision history and effective dates.
2. The Department’s report-writing manual, field reporting guide, or equivalent, and any portion of the RMS user guide or training material addressing when a case number is issued, when an incident is documented as an assist to an outside agency versus a Benicia case, and how supplemental reports are associated with a primary case.
3. The complete list of RMS and CAD incident type codes, call type codes, disposition codes, and clearance codes in use during each period, with definitions, including any code, flag, or field used to designate an ALPR-involved incident, a hit, an assist to an outside agency, or a recovery.
4. Records sufficient to show whether the RMS contains a searchable field, checkbox, or narrative convention indicating ALPR involvement, and when that capability was added.
5. All training bulletins, roll-call briefings, memoranda, emails, or written directives to Department personnel concerning how to document an ALPR hit in a report, including any guidance on articulating reasonable suspicion or probable cause arising from a hit, and any requirement to visually confirm a plate before initiating a stop.
6. Records sufficient to show the Department’s practice for incidents in which a Benicia ALPR hit leads to action by an outside agency, or an outside agency’s hit leads to Benicia action — specifically whether a Benicia case number is generated, an assist-to-outside-agency entry is generated, or both.

Category 5 — Provenance of previously released statistics

1. The query parameters, saved report definitions, filter criteria, date ranges, report templates, or screenshots used to generate any ALPR-related statistics released by the City or Department in response to PR-2026-7, PR-2026-155, or any other request, or presented to the City Council or the public.
2. All communications, including email and messaging, between or among Department personnel, City staff, and Flock Safety concerning the preparation, selection, framing, or review of those statistics, including drafts and edits.
3. Any written methodology, data dictionary, or caveat document describing the limitations of those figures.
4. Any record in which Department or City personnel identify an ALPR-attributed outcome figure as an estimate, an approximation, or as excluding or including particular case types.

Category 6 — Evaluation, comparison, and claims

1. Any before-and-after analysis, effectiveness study, performance evaluation, internal assessment, or crime trend comparison concerning the ALPR system prepared by or for the Department or the City.
2. All staff reports, presentations, and supporting materials on ALPR presented to the City Council, any commission, or any community meeting from January 1, 2021 to the present.
3. Any grant application, grant report, budget justification, or funding request that cites ALPR outcomes or effectiveness.
4. Any comparison of Benicia crime or clearance trends to neighboring or peer jurisdictions, whether or not those jurisdictions operate ALPR.

Category 7 — Data sharing

1. Records sufficient to show which outside agencies have access to Benicia ALPR data, and the number of searches of Benicia data conducted by outside agencies per year.
2. All memoranda of understanding, data-sharing agreements, letters, or approvals governing that access, including any authorization made by letter rather than by formal agreement.
3. Records sufficient to show the number of outside-agency ALPR hot lists to which Benicia cameras are matched.

Format, fees, and process

Format. Where records exist in an electronic format, I request production in that native electronic format — CSV, XLSX, or database export as applicable — rather than as PDF, scanned images, or printed pages (Gov. Code § 7922.570 et seq.). Converting tabular data to a static image format degrades a record the Department already holds in usable form and increases, rather than reduces, the burden on both parties.

Statistical and aggregate records. The law enforcement investigatory records exemption does not reach aggregate or statistical compilations that do not disclose investigative technique or identify individuals. Where a responsive record contains both exempt and non-exempt material, I request that reasonably segregable non-exempt portions be produced with the exempt portions redacted and the basis for each redaction identified (Gov. Code § 7922.525).

Database queries. Extracting existing data from an existing system using existing report functions is production of a record, not creation of one. If retrieval would require new programming or a new query beyond the Department’s ordinary use, please advise me before incurring any cost so I can narrow the request; I am aware the Act permits recovery of certain data-extraction costs and I do not wish to trigger them unnecessarily.

Fees. I request a waiver of fees on the basis that this request serves the public interest in evaluating a publicly funded surveillance program. If fees cannot be waived, please notify me before processing if costs will exceed $25.

Assistance. If any portion of this request is unclear, overbroad, or does not reasonably describe an identifiable record, I ask the City to fulfill its duty to assist under Gov. Code § 7922.600 by contacting me to discuss scope. I am readily available to narrow date ranges or categories, and I would prefer a rolling production beginning with the most readily available categories over a single delayed release.

Nonexistent records. If no responsive record exists for a given item — for example, if the Department does not track hits that result in no stop, or does not flag ALPR involvement in the RMS — please state that affirmatively for that item rather than omitting it. A statement that a measure is not tracked is itself responsive and material to this inquiry.

Timeline. I look forward to your determination within 10 days as required by Gov. Code § 7922.535. If the Department invokes the 14-day extension, please identify the specific unusual circumstances relied upon.