Analyst FAQ

This page collects the questions Flockwatch gets most often from the people who actually have to do something with ALPR data: journalists, policy researchers, city council members and their staff, public-records requesters, and graduate students. The answers point back to the relevant section of the site and try to be specific about what the public record does and does not contain.

If you're approaching ALPR for the first time, read the Start Here briefing, then the Methods page and the Documented Incident Registry.

For journalists

Where do I start if my local police department just announced they're buying Flock cameras?

Three things, in order:

1. Read the procurement contract. Request it through your state's Public Records Act. Look for the data-sharing default (is the camera searchable by other agencies out of the box?), the retention period (how long are plate reads stored?), and the termination clause (what happens to data when the contract ends?). The legal landscape section walks through the high-level variables.

2. Ask the agency for a list of agencies currently in their "sharing network." This is often not in the contract. You may need to file a Public Records request specifically for this. If the agency doesn't know, that's a story.

3. Look up the vendor's heat-map / pattern-search documentation. Most major vendors offer a pattern-detection feature (called heat maps, movement analysis, or network analysis). The vendor may have marketed this feature to your agency as part of the pitch, even if the public-facing policy doesn't mention it.

The documented incidents registry has several cases where each of these steps surfaced material that the original announcement did not.

How do I file a public records request for ALPR-related documents in my state?

Your state's Public Records Act (sometimes called FOIA-equivalent or sunshine law) governs the process. Most states have a model request form; some require it. The records most often relevant to ALPR reporting:

• The vendor procurement contract and any amendments
• The agency's ALPR policy / SOP
• Audit logs (your state may exempt these under "law enforcement sensitive"; where they are released they carry the strongest findings in this study — see FW-006 and FW-029)
• Hot-list configuration documents
• Records of agency responses to vendor data requests

The Reporters Committee for Freedom of the Press maintains a state-by-state FOIA resource with request templates and exemptions. Treat each state's exemptions as a starting point, not a definitive list — exemptions are heavily litigated and jurisdiction-specific.

What's the difference between an ALPR audit log and a vendor's internal access log?

A vendor's internal access log records every query against the vendor's platform, indexed by vendor employee account. It is controlled by the vendor. It typically includes the searching agency, the query parameters, and the timestamp.

An agency audit log records every query by an agency user against the vendor platform. It is controlled by the agency (or made available to the agency by the vendor). It typically includes the officer / user ID, the query parameters, the timestamp, and (in some systems) the case number or justification.

For accountability purposes, the agency audit log is the more important record — it tracks who in your local government searched for what. Vendor internal access logs are important for understanding vendor-side access (e.g., vendor employees responding to subpoenas), but those typically don't appear in local reporting.

Flockwatch has not documented a case where an agency audit log and a vendor internal log were both released and compared — that comparison is almost never possible from public records. What the registry does document is misuse surfaced through a victim report, agency investigations, and an internal audit: FW-017, FW-018, and FW-029.

How do I verify a vendor's marketing claim?

Three patterns of verification, ordered by reliability:

1. Test the claim against the technical documentation. Most major ALPR vendors publish some technical documentation (data sheets, integration guides, developer docs). If the marketing claim is about a feature, the documentation should describe how that feature works. Discrepancies are a flag.

2. Find a procurement document that describes the same feature in procurement-grade language. Procurement documents are signed by both parties and are public records (in most states). Vendor marketing language and procurement language often diverge.

3. Test the feature in a public-facing sandbox or trial deployment. Some vendors offer trial deployments to journalists on request. This is rare, but it's the most direct way to verify a feature claim. Don't publish vendor marketing as if it were the tested version.

The vendor market context section documents one such divergence in detail (FW-022, Oshkosh) and the safeguard tracker separates announced from verified changes.

For policy researchers

What UCR data can and cannot tell you about ALPR use

The FBI's Uniform Crime Reporting (UCR) program does not have a category for ALPR deployment or use. This is a deliberate choice — UCR focuses on offense counts, not on the tools used by reporting agencies. As a result, UCR data is silent on ALPR deployment and cannot be used to estimate ALPR prevalence or correlate ALPR use with crime outcomes.

The Flockwatch study explicitly excludes UCR-based claims for this reason. The Methods page documents this exclusion.

Researchers wanting to estimate ALPR deployment need to use other sources:

• EFF Atlas of Surveillance (a volunteer-maintained database of agency-level surveillance deployments)
• Vendor-published customer lists (where available)
• Public-records-request results (the most reliable source for any specific agency)
• Procurement records (state-level, where published)

What about the FBI's new "Suspicious Activity Reporting" (SAR) program?

The SAR program is administered by the FBI's Office of Partners and Liaison. SAR reports can include ALPR-derived observations, but the SAR program itself does not require ALPR use and does not track ALPR-derived reports separately. SAR data is not publicly available in disaggregated form.

Flockwatch's evidence base contains no SAR analysis; SAR is outside this study's claim space. Treat any SAR-based figure you encounter as a case study, not a national estimate.

Is there a difference between ALPR data and "intelligence-led policing" data?

Yes, in two important ways:

1. ALPR data is generated automatically. Intelligence-led policing (ILP) data may include manually-entered tips, officer observations, and aggregated offense data. ALPR data is generated by a sensor without human intervention at the point of capture.

2. ALPR data is searchable as text. Most ILP systems store qualitative or semi-structured observations. ALPR data is structured text (the plate string) and is searchable by the plate string itself, which means cross-agency hot-list comparison is straightforward.

This structural difference is why ALPR data has a distinct governance profile from ILP data, even when the two are combined in a single platform.

How should I handle a research question that asks "does ALPR reduce crime?"

With care. The published evidence on ALPR effectiveness at reducing crime is limited and methodologically constrained. The strongest causal design available — a randomized hot-spot trial (LIT-004) — found no general deterrent effect at the tested dosage; a 2025 quasi-experimental study (LIT-005) found no violent-crime reduction but reductions in some property categories. Effects are small, outcome-specific, and hard to separate from concurrent policy change. See the evidence base.

The Flockwatch study does not take a position on whether ALPR reduces crime. It focuses on the governance, fault-propagation, and equity questions, which have stronger evidence. Researchers should distinguish "does it work?" (effectiveness) from "what does it do?" (impact) — these are different questions with different evidence bases.

For city council members and staff

What questions should I ask the chief of police before approving an ALPR contract?

A practical checklist, derived from the cases in the incidents registry:

1. What is the data-sharing default? Specifically: are cameras in our deployment searchable by other agencies by default, or only on a per-search basis with our agency's approval?

2. What is the retention period? Specifically: how long are plate reads stored? How long are plate images stored? How long are search logs stored?

3. Who has access to our agency's data? Specifically: which vendor employees, which other agencies, and under what authorization?

4. What happens to data when the contract ends? Specifically: is data deleted, returned, or transferred to a successor vendor?

5. What audit logs are available to our oversight body? Specifically: who reviews the logs, how often, and what's the escalation pathway when misuse is detected?

6. Has the vendor been subject to any data-breach disclosure in the last 24 months? Specifically: what was disclosed, to whom, and what was the remediation?

A refusal or deflection on any of these is a red flag.

What does a model ALPR ordinance look like?

The ACLU publishes model ALPR legislation and a model cancellation resolution in its "Get the Flock Out" toolkit. Its core guardrails:

• Retention in hours or days — explicitly not weeks or a month or more
• Sharing only on a warrant for a felony recognized under local/state law, for any other government entity
• Usage limited to felony investigations, missing-person cases, unregistered/uninsured vehicles, and narrow non-enforcement uses
• Annual usage reporting, plus limits on release of ALPR data through open-records requests so it cannot be repurposed for stalking

Flockwatch's own policy floor goes further on deletion (minutes, not days) and adds notice and a private right of action. The community pushback section documents contract terminations and suspensions rather than ordinance adoptions; Flockwatch does not currently track which localities have enacted model-based ordinances.

Our agency says they need ALPR to find stolen cars. Is that true?

Often partially true, often overstated. ALPR systems can support stolen-vehicle investigations, and independent research documents targeted operational value under some conditions. Public activity files do not, however, report recoveries as a share of detections. The California 2021–22 public-records file shows a 0.110% contemporaneous hot-list hit share among 3.224 billion detections—hits are database events, not verified recoveries, arrests, or solved crimes (Findings). The same vendor platforms also support pattern-search and historical-query features that go beyond stolen-vehicle recovery.

The relevant question for a council member is: what proportion of the agency's camera use is documented as stolen-vehicle recovery versus other query purposes? Most agencies cannot answer that with current public reporting.

Can we just opt out of the vendor's data-sharing network?

Usually yes, technically. Whether the opt-out holds is a separate question: FW-019 (Mountain View) and FW-020 (Palo Alto) document sharing configured or exercised beyond what the city understood it had authorized. Flockwatch has no systematic data on how often opt-out attempts meet vendor or peer-agency resistance.

Flockwatch's pushback ledger tracks terminations and suspensions, not negotiated partial opt-outs — the configuration-level record is largely non-public. That opacity is itself the finding: FW-019 and FW-020 show sharing scopes applied or exercised without the city's knowledge.

For students and researchers new to ALPR

What papers should I read first?

A short reading list, prioritized for someone approaching the literature for the first time:

1. Lum et al. 2011 (NIJ randomized hot-spot evaluation) — the strongest causal design in the literature, and it found no general deterrent effect under the tested dosage (LIT-004).
2. Shjarback 2025 — recent quasi-experimental evidence with mixed, outcome-specific effects (LIT-005).
3. The ACLU "Get the Flock Out" toolkit for the policy and advocacy landscape.
4. The EFF Atlas of Surveillance methodology for the deployment inventory approach.
5. The Flockwatch study itself, including the evidence base and the documented incidents for the case-study corpus.

For a more academic entry point, the evidence base lists the peer-reviewed and gray-literature sources used in this study.

What's a good research question I could actually answer?

A few candidate questions, ranging from smaller to larger:

• Survey of state ALPR retention statutes — most state laws have not been updated since the introduction of networked ALPR; a state-by-state comparison is straightforward and would be a useful contribution.
• Public-records-request analysis of one state's audit logs — even a small state can yield interesting patterns if you can get the logs.
• Vendor procurement contract corpus — every procurement is a public record; building a corpus of contracts across a sample of agencies would surface patterns in data-sharing defaults.
• Hot-list hygiene analysis — request NCIC entry dates and resolution dates for a sample of cases and compare to the resolution dates on the underlying case filings.

The Flockwatch study welcomes collaboration on these and adjacent questions; reach out via the GitLab project.

Where can I get raw ALPR data for research?

This is the hardest part. Public sources of raw ALPR data are limited:

• EFF Atlas of Surveillance — agency-level deployment data, not plate reads.
• Public-records-request results — varies by state; some states release audit logs with redactions, others don't.
• Vendor-published datasets — rare; most vendors do not publish raw data.
• Court records — in cases where ALPR data was used as evidence, court records may contain excerpts.

Researchers with academic FOIA support (e.g., through a university general counsel's office) have better access than independent researchers. Plan accordingly.

For the general public

Should I be worried about driving past an ALPR camera?

It depends on what "worried" means to you.

If "worried" means "the camera photographed my plate" — yes, that happens for every vehicle that passes a working camera.

If "worried" means "the camera added me to a list" — no, the camera does not "add" you to anything. It reads plates and checks them against existing hot lists. If you are not on a hot list, the read is typically retained for the configured retention period and then deleted.

If "worried" means "the camera shares my plate with other agencies" — this depends on the agency's data-sharing configuration, which varies. The legal landscape and data-sharing failures sections describe the variability.

How do I find out if my city has ALPR cameras?

• EFF Atlas of Surveillance has an agency-by-agency lookup.
• Public records request — your agency's ALPR policy and procurement documents are usually public records.
• Visual inspection — most ALPR cameras are mounted on poles or street furniture and are visually identifiable. Vendors often have signature camera shapes.
• City council meeting minutes — ALPR contracts usually come up for council approval, sometimes in open session.

How do I get my plate removed from a hot list?

It depends on which hot list. For NCIC entries, the entry is associated with a case and is removed when the case is resolved (or sooner, in some cases). For agency-specific BOLOs, the removal is at the originating agency's discretion.

If you believe you have been wrongly entered into a hot list (e.g., a mistaken identity case), the remedy depends on the underlying record. The incidents registry includes cases where a stale or lossy record kept producing alerts, including FW-002 (a recovered vehicle still listed as stolen) and FW-013 (two records collapsing to one match key).

What do I do if I think ALPR data was used against me inappropriately?

Three steps, in order:

1. Document what you know. What agency, what date, what context. The more specific you can be, the better.
2. File a public records request for the agency's audit logs covering the relevant period. Most states allow you to request your own audit-log entries.
3. Consult an attorney specializing in civil rights or police accountability. The ACLU affiliates and several national organizations offer referrals.

Flockwatch's research corpus is descriptive; the project separately takes an explicit policy position. That position is derived from the disclosed evidence and claim boundaries rather than used to select them, so readers can reject the position and still use the data. The community pushback section lists advocacy organizations that may be able to help.

What's not here

A few questions we get that this FAQ doesn't answer:

• Legal advice for any specific case. The Flockwatch study does not provide legal advice. Consult an attorney for case-specific questions.
• Vendor product comparisons. The Flockwatch study describes vendor practices; it does not compare products head-to-head for procurement purposes.
• Operational guidance for ALPR deployment. The study is on the governance and oversight side, not the deployment side.

For any of the above, the evidence base is the best starting point.

Last reviewed 2026-07-16. This FAQ is a working document; new questions are added as they come in.