After 17 Months of Shifting Excuses, LAPD Published Block-Level Crime Data the Week After the Primary

For seventeen months, the Los Angeles Police Department offered a series of reasons why it could not produce block-level crime data for the public. Each reason was different. Each was eventually superseded by the next. Then, on June 9, 2026 — one week after the Los Angeles primary election — LAPD published the data it had said it could not produce, complete with latitude and longitude coordinates, on the city's open data portal. The same Police Commission agenda that week included a $50.9 million Motorola maintenance contract. The data and the contract appeared together. The excuses had run out.

The four justifications

LAPD replaced its legacy records system with Motorola's PremierOne RMS on March 7, 2024. From that point, its public crime data feed on the city's open data portal degraded and then stopped. The legacy feed, which had published block-level crime data continuously since 2010, ended with an incident date of December 30, 2024. For all of calendar year 2025, the portal published zero new crime records in real time.

Over the following seventeen months, LAPD offered four distinct justifications for the absence, documented in correspondence and public statements:

The vendor-only claim failed on the simplest possible test: LAPD's own Crime Mapping page, captured February 17, 2026, states the department “will directly feed its crime data to the Omega Group” and runs records through a process that geocodes “to a hundred block.” A display vendor can only map addresses the department transmits. LAPD holds them. And in the June 9 dataset, the hundred-block coordinates appear in fields called hndrdth_lat and hndrdth_lon — populated by LAPD, not by any vendor.

The blackout, quantified

The scale of the data gap is measurable from LAPD's own portal. Annual crime records, by date of occurrence, combining both feeds:

Three facts about this table bear emphasis. First, the portal showed zero new records for the entire 2025 calendar year — not a slow trickle, not incomplete data, but a complete absence of published crime records for a city of four million people. Second, the 150,431 cases now visible for 2025 were backfilled around May 27, 2026 — after SpotCrime filed suit and in the weeks before the primary. Third, those backfilled records still represent a 35% drop from 2023, which is far larger than any real-crime benchmark warrants.

The benchmark problem

Crime did fall nationally between 2023 and 2025. But LAPD's published drop far exceeds any comparable measure:

Indexed to 2023 = 100: LA tracks its peers through 2023, then diverges sharply at the 2024 system switch. Chicago went 100 → 90.1; NYC went 100 → 102.6. LA went 100 → 64.7. The step-change lands precisely on the system transition, not on any crime event.

Independent analyst Jeff Asher examined this directly in August 2025, finding that “LAPD is reporting a 52 percent drop in property crime to the FBI due to NIBRS transition issues,” and that “some crime types (namely theft and motor vehicle theft) weren't fully recorded when the NIBRS switch was turned on in May 2024.” Asher noted the distortion “accounts for nearly a full percentage point of the national drop in property crime.” The NIBRS transition, as we have previously documented, should raise aggregate counts through the removal of the hierarchy rule — not lower them. The counting change cannot explain a swing of 35-42%.

A controlled comparison confirms the gap is not real

A day-of-week and seasonality-controlled comparison — second Saturday of each month — makes the distortion visible without any statistical adjustment:

During 2025, the portal showed zero on every one of these dates in real time. The 2025 figures above are from the later backfill — proof that crime was occurring (400–500 offenses per Saturday) and the data existed, but was not published. The two legacy years average roughly 633 incidents per second Saturday; the two NIBRS years average roughly 385 — a step-change of approximately 39% that lands on the system transition, not on any underlying change in crime.

The shooting victim anomaly

One additional data problem stands out in the published 2026 dataset. The victim_shot field — a new NIBRS-era flag indicating whether a victim was shot — is marked “Yes” only 178 times across the first five months of 2026. Annualized, that is approximately 430 shooting victims per year. Los Angeles has historically recorded well over 1,000 shooting victims per year. The published count is implausibly low by a factor of two or more.

How LAPD defines “victim_shot” in PremierOne, and whether that count reconciles with the department's own CompStat shooting-victim totals, remains an open question. If it does not reconcile, the new system is losing the most serious category of violent-crime data — not overcounting, but systematically undercounting the incidents that matter most.

The contract and the timing

The June 9 data publication coincided with a Police Commission agenda item that received almost no public attention. BPC 26-168, heard June 9, 2026, was a request to approve a ten-year Professional Services Agreement with Motorola Solutions for “PremierOne Essential Maintenance and Support” — covering CAD, RMS, Mobile, hardware refreshes, ongoing licensing, jail module build-out, and on-site Motorola system administrators. The total contractual obligation: $50,927,266.05.

Combined with the initial 2022 PremierOne Records contract ($13,772,614 over five years), the city's total Motorola commitment reaches approximately $64.7 million. At that price, the claim that the system could not produce a standard data export is not a technical limitation — it is a procurement failure. LAPD pays for on-site Motorola system administrators as part of the agreement. The vendor was in the building.

If the system could produce location-tagged crime data in June 2026, it could presumably have produced it at any point during the prior seventeen months. No technical explanation for what changed has been offered. The BPC 26-168 contract recitals add a further anomaly: the document states “the Department implemented PremierOne RMS in 2022,” which is two years before the March 2024 public launch. The system was apparently operational internally for years before the department declared itself unable to produce public data from it.

What the suppression cost

The effects of a seventeen-month blackout are not abstract. SpotCrime's contemporaneous archive — built from daily blotter captures — preserved a record of what was happening in Los Angeles while the portal showed nothing. That archive constitutes evidence of the gap. Most residents, journalists, researchers, and public safety platforms had no such archive. They operated without crime location data for a major American city for well over a year.

The accountability function of public crime data depends on its granularity and its timeliness. ZIP-code aggregates do not reveal what block-level data reveals. A portal that shows zero records for an entire year does not allow anyone to ask pointed questions about specific places and specific decisions. Officials accountable for public safety, politicians seeking to shape the narrative on crime during an election cycle, and anyone operating in neighborhoods where crime was occurring all received a different quality of information from the gap — and they did not receive it equally.

LAPD's lawsuit with SpotCrime remains active. The legal question — whether the seventeen-month non-disclosure violated California's Public Records Act — is not answered by the eventual publication of data. The CPRA's 10-day determination deadline (§ 7922.535) and its prohibition on indefinite “processing” delays as an exemption (§ 7922.530) address the conduct during the blackout, not the state of the portal today.

What is now unambiguously established: the system works. The data exists. The coordinates are in the file. The question that remains is why four different justifications were offered over seventeen months for a capability that, on June 9, 2026, was demonstrated to have existed all along.