Claude working memo, July 23, 2026. Data: LAPD Crime Data from 2020 to Present (city open data portal snapshot), 1,004,989 incident records, January 1, 2020 through May 29, 2025. One caveat governs everything: LAPD switched records systems in March 2024, so the legacy dataset thins out after that date. The reliable window is January 2020 through early 2024, about four years. The study area is a bounding box approximating ZIP 90035, latitude 34.040 to 34.063, longitude -118.403 to -118.371, which covers Pico-Robertson and slivers of its neighbors. All incident counts below come from that box.<\/p>\n
Preregistered signal criteria<\/p>\n
Before filtering, the criteria for a serial signal were: a cluster of unsolved homicides sharing method or victim type; any strangulation-weapon incidents, since strangulation is the method most overrepresented in serial killing; repetition of method, victim type, or premises among unsolved assaults on women; or a homicide rate running above expectation for the area.<\/p>\n
The homicides<\/p>\n
Three deaths in the box in five and a half calendar years of data.<\/p>\n
November 13, 2021. A 26-year-old Hispanic woman died in an apartment on the 8600 block of West Olympic. Charged as negligent manslaughter, weapon field coded liquor\/drugs. Cleared by adult arrest.<\/p>\n
May 31, 2022. A 52-year-old White man died inside a pharmacy on the 1800 block of South La Cienega. Weapon unknown. Cleared by adult arrest.<\/p>\n
August 3, 2022. A 66-year-old White man died in a discount store on the same 1800 block of South La Cienega, killed by bodily force. Cleared by adult arrest.<\/p>\n
The second and third cases are the audit’s teaching moment. Same block, nine weeks apart, two older White men, both in retail stores. The pattern instinct fires on exactly this shape. The clearance data defuses it: both cases closed with arrests, so both killers are known, and known killers cannot be one unknown killer. This is what most apparent patterns look like on contact with case status.<\/p>\n
The base rate<\/p>\n
Los Angeles averaged roughly 370 homicides per year from 2020 through 2023 across about 3.9 million people, near 9.5 per 100,000. At the city rate, a population of 26,000 produces about 2.5 homicides per year, or roughly ten over the reliable window. The box produced three, only two of them criminal homicides. The neighborhood runs at about a quarter to a third of the city average.<\/p>\n
The signal criteria, checked<\/p>\n
Strangulation, choking, or ligature appears as the weapon in 50 incidents citywide in this window. In the box: zero.<\/p>\n
Unsolved homicides in the box: zero. All three cleared.<\/p>\n
Unsolved aggravated assaults on female victims in the box: 15 across the window. The weapons scatter: pepper spray, a vehicle, a BB gun, handguns, bodily force, a knife. The premises scatter: street, sidewalk, apartment, gas station, parking lot. No repetition of method, victim type, or location. Two of the 15 occurred on the 1700 block of South La Cienega in 2020, one a knife assault in a dwelling and one pepper spray in a parking lot, which share an address range and nothing else.<\/p>\n
Verdict under the preregistered criteria: no signal. The answer to the opening question is no, with the standard epistemic caveat that absence of signal in city incident data is not proof of absence, pending the coroner and corridor layers below.<\/p>\n
Context numbers for the essay<\/p>\n
Aggravated assaults in the box: 150 over the window, 116 male victims, 32 female. Unsolved rate 55 percent against 60 percent citywide (32,267 of 53,525). The most common weapons were bodily force (25), unknown (19), and pepper spray (16), the profile of street disputes rather than predation.<\/p>\n
Rapes reported in the box: 27, of which 17 remain coded Invest Cont. Even in a watched neighborhood, this is the file where under-attention lives. Most occurred in homes, hotels, and motels, the acquaintance pattern, but the unsolved fraction deserves a paragraph.<\/p>\n
Citywide criminal homicides in the window: 1,568, of which 484 remain coded Invest Cont, about 31 percent open. That number is the bridge to the essay’s third movement. The machinery found nothing in 90035 because there was nothing to find. Four hundred eighty-four open case files elsewhere in the city are where it points next.<\/p>\n
Total reported crime in the box ran 1,532 (2020), 1,575 (2021), 1,755 (2022), 1,808 (2023), with serious violence flat to declining: 85, 91, 86, 68.<\/p>\n
Data caveats to state in print<\/p>\n
The box is a rectangle, not the ZIP boundary polygon. Status codes are a snapshot; Invest Cont means open as of the data pull, and some closed cases carry stale codes. The data covers LAPD only. Beverly Hills PD and Culver City PD territory begins blocks away, and the jurisdictional seam remains unexamined. Post-March 2024 data is incomplete because of the records system transition; the apparent 2024 crime drop in this dataset is an artifact.<\/p>\n
The pre-2020 layer: Corning Street<\/p>\n
A 2010 slice of the LAPD legacy data (200,510 incident records) shows two homicides in the box that year, both on South Corning Street, one block apart, four months apart. News archives and LAPD press releases then opened a history the 2020-2025 data could not see. The 1900 and 2000 blocks of South Corning form a two-block corridor with at least four homicide incidents across a decade.<\/p>\n
November 20, 2004. A drive-by at 2035 S. Corning killed LaShawn Brunson, 21, of Palmdale, and wounded two others. Witnesses described a light minivan.<\/p>\n
May 5, 2010. Jana Collins, 26 and 18 weeks pregnant, sat in her car with two men on the 1900 block when a gunman walked up and fired repeatedly into the car. She died at the hospital with her unborn child. LAPD said she was not the intended target and called it gang related. Suspect Javier Angulo Romero was arrested in Atlanta on October 5, 2010. The incident record shows Adult Arrest.<\/p>\n
September 9, 2010. A Black man died on the 2000 block. His record is the thinnest in the file: no age, no weapon, no premises, status Invest Cont as of the dataset snapshot. General and LAPD-newsroom searches that named every other Corning victim return nothing for him. He is, so far, the unnamed one.<\/p>\n
December 2, 2013. Gunmen opened fire on a car in the 1900 block. Taalib Pecantte, 7, a passenger, died at the hospital. The FBI and LAPD arrested four suspects in February 2014, and the DA filed murder charges against Tradel Faniel.<\/p>\n
Since 2014, the corridor appears in no homicide record I can reach. The violence stopped, or moved, and the neighborhood’s memory of it faded. This reframes the essay. The safe zip code contains a forgotten two-block war that took a pregnant woman and a second-grader, and it holds at least one open case whose victim has no name in anything publicly searchable. The “old enough to be forgotten” hypothesis found its case.<\/p>\n
Follow-ups requiring hands and records access: the LA County Medical Examiner’s case files for a September 9, 2010 death on the 2000 block of South Corning, which will carry the man’s name; a CPRA request to LAPD West LA homicide on the case’s current status, since Invest Cont is a snapshot from the dataset era; the LA Times Homicide Report archive, reachable through library database access or the Wayback Machine, which likely profiled him at the time; the status of the 2004 Brunson case; and neighbors on Corning who remember 2004 to 2013.<\/p>\n
The corridor layer: the seam test<\/p>\n
Data: FBI Supplementary Homicide Report, 1976 to 2014, 716,550 records nationwide, the Murder Accountability Project file with the Solved flag. This is the only dataset in which LAPD, Beverly Hills PD, Culver City PD, and the Sheriff appear side by side. Caveat that governs this layer: the Solved flag reflects clearance within the reporting window, so cases solved years later still read unsolved. That staleness turned out to be a finding, not just a caveat.<\/p>\n
The three agencies around 90035, 1976 to 2014: LAPD 26,252 homicides, 68 percent solved, 504 by strangulation or asphyxiation with 234 unsolved. Culver City PD 106 homicides, 68 percent solved, 11 strangulations with 5 unsolved. Beverly Hills PD 56 homicides, 68 percent solved. The Sheriff’s county file: 6,308 homicides, 46 percent recorded solved. LAPD’s unsolved female strangulations concentrate in the 1980s (46) and 1990s (58), the statistical shadow of the Southside era.<\/p>\n
The flag: Culver City’s unsolved female strangulations fall in 1980 (84, White), 1982 (29, White), 1992 (39, Black), 1995 (41, Black), and 2005 (25, Black), five cases in a city of about 39,000. Run through news archives, the 1992 case resolves into Brenda Bradley, 38, found in a Culver City alley on October 5, 1992, a victim of serial killer Michael Hughes (b. 1958), who dumped four women in Culver City’s commercial alleys in 1992 and 1993, was arrested by Culver City police in December 1993, convicted of four murders in 1998, convicted of three more by DNA in 2011, and sentenced to death in 2012. Hughes was one of at least three serial killers, with Chester Turner and Lonnie Franklin Jr., behind the killings once attributed to a single Southside Slayer, in an era when as many as 90 women died in inner-city Los Angeles.<\/p>\n
Three lessons for the essay. First, the protocol works: the preregistered signal, unsolved female strangulations clustered in one small jurisdiction, pointed at a documented serial killer on the first pass. Second, the file lies by omission: Bradley’s murder has been solved since 1998, but the SHR still counts her unsolved because the arrest came after her reporting year, so national unsolved counts overstate in both directions and every flag needs news-archive verification. Third, the two later cases, 1995 and 2005, Black women strangled in the same small jurisdiction, remain unresolved in anything I can reach, and the detective who worked Hughes said publicly that other agencies should check unsolved cases against him. Those two cases are the corridor layer’s open item.<\/p>\n
The seam finding proper: Myles and Harrington, killed in November 1993, read as solved in the SHR because Hughes was arrested weeks later, while Bradley, killed thirteen months earlier by the same man, reads as unsolved forever. Same killer, same alleys, opposite codes, an artifact of the reporting-year boundary. Jurisdictional and temporal seams do the same work: they cut one pattern into pieces no single view reassembles.<\/p>\n
Open layers<\/p>\n
The corridor check: FBI Supplementary Homicide Report data via the Murder Accountability Project, covering all agencies, to test the seam between LAPD, Beverly Hills, Culver City, and Sheriff territory. The excess-death layer: LA County Medical Examiner-Coroner records for the box, deaths by cause against expectation. The missing and unidentified layer: NamUs, county-searchable. The historical layer: LAPD 2010 to 2019 dataset and the LA Times Homicide Report archive for pre-2020 cases, including any open ones old enough to be forgotten.<\/p>\n
Police, Power, and the Production of Racial Boundaries (2015)<\/a><\/strong><\/p>\n Ana Mu\u00f1iz<\/a> wrote her ethnography about Cadillac-Corning, the small tract of apartment blocks between Robertson and La Cienega south of Cadillac Avenue, which includes the exact Corning Street corridor where we found Brunson, Collins, Pecantte, and the unnamed man. Her argument<\/a>, laid out from that Force Option Simulator scene forward, runs through the Coetzee frame: the neighborhood functioned as the imagined barbarian zone for the Whiter, wealthier areas around it, and policing there worked to maintain a racial boundary rather than to protect the people inside it. Her chapter 3 covers the origins of gang injunctions, and this is the book’s historical claim to fame: the 1987 injunction against the Crips in Cadillac-Corning was the first use of a gang injunction in Los Angeles, the prototype for a tool the city eventually deployed against dozens of neighborhoods. This 90035 zip code is where the gang injunction was invented. The update since 2015 runs on three tracks.<\/p>\n The author. Mu\u00f1iz is now an associate professor of criminology, law and society at UC Irvine, working on gang profiling, gang injunctions and databases, immigration enforcement, and state violence. Her second book, Borderland Circuitry (California, 2022), follows the same instinct upstream into the data layer: it traces the information systems behind immigration enforcement since the late 1980s, using ethnography, interviews, and previously unseen documents to show how data-sharing among local police, federal agencies, and foreign partners creates digital borderlands. The through-line from book one to book two is the gang database: the CalGang designation that began as a Cadillac-Corning-era policing tool became an input to deportation machinery. Her CV also shows her involved in the injunction litigation itself, in the Rodriguez and Arellano cases against the city. <\/p>\n The policy afterlife, which is the striking part: the book’s argument substantially won. In 2018, a federal district court barred the city from enforcing its gang injunctions, with Judge Virginia Phillips holding the city likely violated the Constitution by imposing probation-like conditions on residents, mostly men of color, without any hearing or chance to deny gang membership. The preliminary injunction lifted restrictions from nearly 1,500 residents, restrictions that had limited conduct as ordinary as carrying a cell phone or standing with family members. A 2020 settlement then required the city to give any individual a day in court before subjecting them to such restrictions. The earlier Rodriguez case had already established the damages side: after a California appellate court found the curfew provisions unconstitutional, the city agreed to pay up to $30 million in benefits including job training and tattoo removal for people served with injunctions. The class covered about 6,000 Angelenos hit by 26 injunctions with unconstitutional provisions. <\/p>\n And the sour coda, very much in the spirit of her book: the remedy itself became a case study in institutional neglect. By 2023, city data showed fewer than a quarter of the 6,000 class members had been assisted and less than half of the $30 million spent, and in May 2024 a federal judge ordered the program extended, appointed a special master, and sent in a forensic auditor to determine where the money went. The city that spent decades enforcing the injunctions could not competently administer the restitution for them. <\/p>\n The book gives the Corning corridor its history: the violence we found in the data, 2004 through 2013, happened inside the shadow of the country’s oldest injunction zone, in blocks that had been under intensive legal surveillance since the Reagan administration. And it complicates the attention thesis I handed you earlier. I said the safe zip is safe because it’s watched. Mu\u00f1iz’s counter is that Cadillac-Corning was the most watched ground in the zip and the watching was containment, aimed at protecting the boundary around it. The unnamed man of September 9, 2010 sharpens that to a point: he died inside the most policed blocks of 90035, and his case sits open while the pharmacy and discount-store killings of White victims a few blocks east cleared. Watched is not the same as avenged. That tension, between your attention argument and her boundary argument, is where the essay gets its argument instead of just its data. Worth deciding whether you engage her frame by name; it’s the standard academic citation for this exact ground, and citing her is how you join that conversation.<\/p>\n","protected":false},"excerpt":{"rendered":" Claude working memo, July 23, 2026. Data: LAPD Crime Data from 2020 to Present (city open data portal snapshot), 1,004,989 incident records, January 1, 2020 through May 29, 2025. One caveat governs everything: LAPD switched records systems in March 2024, … Continue reading
\nHer Campus<\/p>\n