The Architecture of Data-Driven Policing

This part introduces the LAPD architecture, weaponry, and experiments that data-driven policing is built on. By “architecture” we mean not just technologies but also police strategies and legal instruments. Data-driven policing enables unprecedented coordination of these weapons.

Our story here begins with Bill Bratton, LAPD’s chief from 2002 and 2009. Under Bratton’s leadership LAPD built up a number of practices that are now woven together as data-driven policing, including “broken windows” policing; “intelligence-led” policing; collection and mining of mass data; and behavioral surveillance. After examining those practices, this section turns to LAPD’s testing of surveillance technologies that came to form the inputs of data-driven policing, as well as the Palantir data platform and the build-up of “fusion centers” that exchange police intelligence and coordinate mass surveillance. Next we examine LAPD’s Operation LASER and PredPol “predictive policing” programs; LAPD’s notion of “Designing Out Crime”; and the Citywide Nuisance Abatement Program (CNAP), which was used to target properties and displace people within Operation LASER.

“Broken Windows” and Intelligence-Led Policing

“The department is not organized to fight crime. It's organized to circle the wagons. I'm trying to have it move towards being a proactive, risk-taking organization.”
—LAPD Chief Bill Bratton, December 2002

In the late 1990s through the post-9/11 era, LAPD began adopting tactics and technology centering around “intelligence-led policing,” a model of speculative and hunch-based policing that encompasses behavioral surveillance and police data-mining. These changes accelerated with the appointment of Bill Bratton in 2002 as LAPD chief. Bratton looked to intelligence-led policing as one of the “reforms” he would use to claim that police violence was scientific and based on data.

As many of us remember, the 1990s featured both extreme police brutality and a presidential administration that accelerated mass incarceration through the infamous 1994 Crime Bill, written by then Senator Joe Biden. As L.A. watched LAPD’s violent beating of Rodney King and the Christopher Commission attempted to lay a path for “reform,” LAPD became engulfed in the Rampart Scandal, exposing further police brutality and corruption. This and other LAPD scandals culminated in a 2001 consent decree with the U.S. Department of Justice, placing LAPD under federal oversight due to patterns of false arrests, extreme violence, and other illegalities. [13]

Bill Bratton was appointed police chief following those scandals. Before taking that job, Bratton was a “leading member” of the consulting team hired under a five-year, $11 million contract with the city to implement the consent decree reforms. [14] After just five months on that team, Bratton applied for the job of LAPD chief. Though police reformers declared the consent decree a victory and turning point for an embattled police department, ultimately this moment served as another LAPD power grab for new police resources and powers. Using the consent decree as a justification for new investments, Bratton pursued a series of experiments that expanded LAPD’s data-collection and surveillance powers, including community policing, “broken window” policing, behavioral surveillance programs, and integration of military surveillance technologies. Skid Row in particular became a testing ground for how counterterrorism and counterinsurgency programs would become integrated into everyday policing.

Two significant changes came to LAPD before “predictive policing” launched, both forming the foundation for LAPD’s data-driven policing programs: “broken windows” policing and Compstat. The term “broken windows” was popularized by a 1982 article published by George Kelling and James Q. Wilson. [15] The basic idea is that draconian criminalization of trivially harmful behaviors like public intoxication, littering, and jaywalking “prevents” larger crimes. This practice was nothing new to Skid Row, as discussed in detail in part 5 of this report. But turning this policing into a criminological theory popularized the idea that poverty should be dealt with by criminalizing individual “behavior,” ignoring the systemic failings of white supremacist capitalist society. Bratton made this view explicit in a 2007 interview:

The one thing I have learned—and now strongly advocate—is that the police, properly resourced and directed, can control behavior to such a degree that we can change behavior. Many social scientists are wedded to what I believe to be the failed and never proven idea that crime is caused by the structural features of a capitalist-based democratic society—especially demographics, economic imbalance, racism, and poverty.” [16]

Bratton also led the effort to build Compstat, a program he first popularized as chief of NYPD in the 1990s. Compstat is a data system that generates regular summaries of what a police department is doing. These summaries are reviewed at meetings where police leadership determine policing strategies and priorities. Compstat helped organize “broken windows” policing. George Kelling, who popularized “broken windows” policing, once called Compstat “the most important administrative policing development of the past 100 years.” [17]

Compstat also furthered the collection and legitimization of crime data, treating it as scientific fact rather than a reflection of racist and subjective enforcement choices of police. Bratton’s confederation of Compstat and “broken windows” policing also laid the groundwork for expansion of behavioral surveillance, which refers to police practices of speculatively monitoring behaviors that may be indicative of future crime. This kind of intelligence-gathering, according to Bratton, is “what police have always done, to observe and identify changing patterns of behavior.” [18]

Bratton was not only successful in promoting police brutality as public policy, he also wedded police data to the notion of “community policing.” The objective here was to strengthen cooperation, communication, and partnership between police, city officials, and private “stakeholders” like real estate developers and businesses. In fact, the U.S. Department of Justice’s Community Oriented Policing Services (COPS) office used data collected in part from LAPD while Bratton was chief to promote the "co-implementation" of Compstat and community policing, calling them both "powerful engines of police reform." [19] The broader relationship of police data to “community policing” is examined in detail in part 6 of this report, where we dissect LAPD’s current “Data-Informed Community-Focused Policing” framework.

Surveillance Inputs: LAPD’s Eyes and Ears

LAPD has long been one of the most technologically resourced police forces in the country, with a yearly budget of over $3.2 billion that consumes half the city’s discretionary spending. Over the years, LAPD’s expansion of its architecture of surveillance has taken the form both of new digital technologies and mass surveillance equipment as well as campaigns of racial profiling, pretextual street and vehicle stops, and “broken windows” policing that generate mass troves of police data for LAPD’s Record Management System and other databases. These surveillance inputs fuel data-driven policing. For more of our research into LAPD’s architecture of surveillance, visit

In part 5 of this report, we explore in greater detail how Skid Row is a place where LAPD has long experimented with new forms of policing. This too is characteristic of colonization. As we’ve written elsewhere, LAPD’s testing, refinement, and “reform” of surveillance technologies “continues the legacy of colonial experimentation in service of white supremacist political ends — much in the same way colonial administrators used legal bureaucracy, surveillance technology, and even scientific study to dominate people they considered seditious or politically threatening.” [20]

Skid Row along with MacArthur Park are some of the places where LAPD first tests new surveillance technologies. For example, in November 2007, LAPD implemented a Closed Circuit Televisions (CCTV) camera system, paid for by the Central City East Association (CCEA), which is funded by businesses interests downtown. [21] Skid Row was the third location LAPD put up CCTVs, following deployment in MacArthur Park [22] and Hollywood . [23] Later, in 2011, the L.A. Police Foundation “donated” $210,173 worth of Automatic License Plate Readers (ALPR) for LAPD to use in Skid Row. These systems, which scan license plates to track people’s movements, [24] were previously tested around MacArthur Park along with facial recognition and “intelligent’ video capabilities. [25] Another camera technology tested in Skid Row was Body-Worn Video (BWV or body-cams). In 2014, LAPD launched a BWV pilot run with officers assigned to the Safer Cities Initiative (SCI) “broken windows” policing program. These officers were chosen because their policing includes foot beats and constant enforcement contacts with people in the community. [26] LAPD also tested phone surveillance technology in Skid Row, including Stingrays and Digital Receiver Technology (aka “Dirt Box”). These devices mimic cell phone towers to connect and monitor mobile devices. In 2011, the Los Angeles Police Foundation gave almost $25,000 to upgrade “Stingray” devices placed in Skid Row. [27]

Of course, technology isn’t the only way LAPD gathers mass surveillance data. Some of the most critical and dangerous surveillance occurs face to face. Every time police stop or question people, they can fill out Field Interview (FI) cards that generate data for LAPD’s records systems. A person does not have to be suspected of any crime to have an FI card filled out. A large number of these cards are filled out by LAPD’s proactive policing Metro units as well as the HOPE (Homeless Outreach Proactive Engagement) and RESET (Resources Enhancement Services Enforcement Team) units that target unhoused communities. In the first half of 2018, HOPE and RESET completed over 7,800 FI cards on unhoused people. [28]

FI cards used to be paper but became digitized in 2015 under the Connected Officer Initiative, which also expanded the Skid Row body cam pilot department-wide. [29] That same year, LAPD also started using the cards to collect social media, social security numbers, and email addresses. [30] Police use this data to map people’s social networks, including who you were seen or stopped with as well as social media friend lists and who you appear with in photos online. This information about your associations can show up in “conspiracy” prosecutions as well as for gang labeling. Craig Uchida, who built the “predictive policing” program Operation LASER, described in more detail below, explained that the purpose of LAPD stops is not always to investigate or determine anything but to generate data: “Most of the time it didn’t lead to anything,” he explained about police stops producing FI cards, “but it was data that went into the system, and that’s what I wanted.” [31]

Another system of critical importance is the Homeless Management Information System / Coordinated Entry System (HMIS/CES), a database of information collected from people who receive services from federally funded nonprofits and agencies. This is the first federal database of people experiencing homelessness, storing extremely personal information like social security numbers, race and ethnicity, prior residency, what services a person has used, health status (including disabilities, pregnancy status, HIV status, mental health), education, employment, and whether they have experienced domestic violence. Anyone using the services is placed into the database, no matter the length of time one was in a program, and the information is kept for years. All this information is shared to federal agencies like HUD, HHS, and the VA without personal identifiers, and the risk of these databases being accessed by local law enforcement and DHS including ICE is very high. In 2017, LAPD told the U.N. Special Rapporteur on Extreme Poverty and Human Rights that CES was a “gold mine” of information that was “only one policy decision away” from police mining. [32]


Palantir is the data-mining platform that LAPD uses to run automated profiling, hotspot maps, AI models, and other search tools on data from the surveillance inputs listed above, including FI cards, along with data from other government agencies as well as commercial brokers. LAPD built its Operation LASER “predictive policing” program on the Palantir platform, and Palantir generates the data-driven “missions” of racial terror that we examine in part 5 of this report. Today LAPD is “the only major metropolitan force using Palantir,” after police in New York City and New Orleans ended their use. [33] LAPD also uses Palantir Mobile, a phone application that puts mass surveillance data at police fingertips and sends officers real-time notifications, for example about nearby people to monitor or stop.

Palantir is a private company founded by billionaire white supremacist Peter Thiel with funding and guidance from the CIA, which used the software for intelligence missions and drone assassinations. [34] The system first arrived in Los Angeles in 2007, when “Bill Bratton persuaded Target to donate $200,000 for LAPD to buy Palantir technology.” [35] As LAPD built up their architecture of racial profiling and mass surveillance using the software, the City Controller’s website shows payments of $15.8 million to Palantir from 2012 to 2021.           

Over the years, Palantir has “helped the LAPD construct a vast database that indiscriminately lists the names, addresses, phone numbers, license plates, friendships, romances, jobs of Angelenos.” [36] LAPD uses Palantir to identify people’s friends, relatives, colleagues and other relations, “pulling people into the LAPD’s surveillance system who otherwise wouldn’t have been,” [37] as well as to “track vehicles using data from the Automated License Plate Reader” and “to examine social networks.” [38] Officers are trained to search for people to target using attributes like race, gender, and physical features, as well as to track any phone call’s “recipient phone number, the date, duration of the call, and the latitude and longitude of all the cell towers used.” [39]

Fusion Centers

Another key part of the infrastructure that data-driven policing is built on are so-called “fusion centers,” the data analytic centers police use for real-time processing, analysis, and sharing of surveillance data. These spy garrisons received heavy investment in the post-9/11 expansion of local policing to encompass mass suspicion, data-gathering, and surveillance. These tactics relied on collecting and integrating law enforcement data into what federal spy agencies have called the Information Sharing Environment (ISE). [40]

There are over several dozen federal fusion centers in the U.S. The one used to police Los Angeles is named the Joint Regional Intelligence Center (JRIC) and located on Imperial Highway – a fitting name – in Norwalk. Fusion centers were part of police efforts to neutralize the Occupy Movement including through infiltration, monitoring political activity, and tracking social media. [41] In 2012, the U.S. Senate Committee on Homeland Security and Governmental Affairs released a highly critical report warning about “waste at state and local intelligence fusion centers.” [42] Our 2013 People’s Audit of LAPD Special Order 1 documents the operation and harm of federal fusion centers. [43]

LAPD has also built a network of its own fusion centers called Community Safety Operation Centers (CSOCs). These spy centers feature conference rooms with rows of computers and a large screen displaying a color-shifting map of the city. [44] In 2016, the first CSOC was established in LAPD’s South Bureau to coordinate and concentrate surveillance across the 77th, Southwest, Southeast, and Harbor Divisions. Today, there is a CSOC in each of LAPD’s four operational bureaus.

LAPD recently launched even more localized fusion centers called Area Crime and Community Intelligence Centers (ACCICs). These hyper-local spy garrisons produce “daily mission maps” (the brutal violence of which we document in part 5 of this report), perform “social media monitoring” and “surveillance camera video pulls,” and search the Palantir, Lexis-Nexis, ParoleLEADS, LInx, CalGangs, VeriTrack, and CCHRS data systems. [45] An internal LAPD document we obtained indicates that ACCICs produce daily hotspot maps with symbols marking locations as “gang related,” “gang member - suspect,” “gang member - victim,” “transient - victim,” “transient - suspect,” and “domestic violence.” [46] It’s hard to imagine why LAPD needs daily maps of domestic violence complaints and unhoused residents besides to speculatively criminalize neighborhoods and communities.

Operation LASER

“The program is analogous to laser surgery, where a trained medical doctor uses modern technology to remove tumors . . . . Extraction of offenders takes place in a ‘non-invasive’ manner (no task forces or saturation patrol activities), and the result produces less disruption in neighborhoods. Continuing with the medical analogy, by extracting offenders surgically, recovery time of the neighborhood is faster.” 
—Operation LASER’s architects, including Craig Uchida (2012)
“I don’t think the public is too concerned with us using technology to prevent crime. People don’t get upset when doctors use technology to prevent illness.”
—Former LAPD Chief Bill Bratton (2015)

LAPD’s homegrown predictive policing program Operation LASER (Los Angeles Strategic Extraction and Restoration) was first launched in 2009 with a Smart Policing Initiative (SPI) grant of $899,959 from the Bureau of Justice Assistance. [47] LAPD used the grant to hire its long-time research partner Justice and Security Strategies (JSS), the police consulting firm headed by Craig Uchida. At LAPD, Deputy Chief Sean Malinowski was the lead investigator for the SPI research grant. LAPD and JSS worked together to develop LASER, which targeted both specific locations (which police deemed “LASER zones” or “Anchor Points”) and specific community members (who police deemed “chronic offenders”).  

LASER Zones and Anchor Points

In 2018, we filed a PRA requesting LAPD’s policies, analysis, maps, and reports for targeting locations under LASER. The records we obtained confirmed what the community had suspected: while promoted as a scientific "breakthrough,” LASER was yet another campaign of racial profiling and territorial enforcement.

To select LASER zones, LAPD and JSS analysts combined hot spot mapping of crime reports with recommendations from local officers. Within each zone police also identified Anchor Points, which were locations considered “possibly being responsible for crime.” [48] While this designation was potentially devastating, there was no standard process or criteria for selecting the locations, which included businesses, residences, and community gathering places. JSS even noted that “public transportation centers, shopping centers, fast food restaurants, and other types of establishments . . . can be ‘crime attractors’ or ‘crime generators’ and are often at the center of hot spots in crime.” Below is a map from LAPD’s 77th Division marking LASER zones and Anchor Points. [49]

77th Division Violent Crime Map Generated by LAPD's LASER PROGRAM.
LAPD 77th Division “Violent Crime Map” showing LASER zones and Anchor Points (AP).

Locations marked Anchor Points became data-driven targets for displacement. According to LAPD training documents, officers were directed to “prepare strategies to address the issues” including through the Citywide Nuisance Abatement Program, which is described in more detail later in a part this section, as well as “Crime Prevention Through Environmental Design” and changes to “licensing/conditional use permits” – all potential levers of gentrification. [50] Former Deputy Chief Dennis Kato confirmed that once a location was designated as an Anchor Point, LAPD could then employ “ all the other city tools and the City Attorney's office to really abate the location, and to gain compliance by those property owners to really fix that problem.” [51]

These data-driven “hotspots” also became targets for extreme police brutality, as detailed in part 6 of this report. By speculatively criminalizing areas and then sending officers on missions to “suppress crime” in these areas with vague profiles of who to look for, these programs ensured lethal and racialized police violence. LAPD killed 21 people in 2016, the year Operation LASER expanded. Of these, we have identified six killings linked to LASER zones in just a short six month period in 2016. All of the men and boys killed were Black or Latino, four were shot in the back, four were teenagers, and two were under the age of 18:

●      February 6: 16-year old José Juan Méndez was killed by Hollenbeck Division officers Josue Merida and Jeremy Wagner [52] during a traffic stop in what LAPD designated as the Estrada CLEAR area, a “problem area” and a LASER zone. [53]

●      May 13: 28-year old Robert Diaz was killed by Hollenbeck Division Gang Enforcement Detail (GED) officer Miguel Ruano on a “crime suppression” mission also in the Estrada CLEAR area. Robert is remembered for his ability to defuse tense situations with his warmth and his love of animals.

●      June 10: 31-year old Keith Bursey was killed by 77th Division GED officer Charles Kumlander on Slauson Avenue, in a LASER zone and Anchor Point. Police shot him in the back. [54]

●      July 25: 18-year-old Richard Risher was killed by Southeast Division GED officers Francisco Zaragoza, Isaac Fernandez, and Joseph Chavez on a “crime suppression” mission inside a LASER zone. [55] Police shot him in the back. [56]

●      August 9: 14-year-old Jesse Romero was killed by GED officer and two-time killer cop Eden Medina inside a LASER zone, following a “vandalism call.” [57] Police shot him in the back. [58]

●      August 16: 18-year-old Kenney Watkins was killed by officer Evan Urias, assigned to the South Bureau CSOC in LASER Zone 5. [59] Urias claimed he was making a traffic stop for a missing front plate, tinted windows, and “possibly a paper back plate.” [60] Police shot him in the back.

As these murders harm a community, LAPD and its collaborators characterize this violence as necessary simply because an algorithm labeled our people’s communities, streets, stores, or homes a place where crime “will” happen.

LASER’s “Chronic Offender” Bulletins

In addition to the location-based aspects noted above, LAPD’s data-driven programs also include person-based targeting. True to the term “Extraction” in the LASER acronym, the program used data-mining to target individuals for banishment. LAPD generated hit lists called “Chronic Offender” bulletins, which police appeared to use as targets whenever there was an open case or suspicion. Police subjected these individuals to criminalization, coercive prosecution, and enforcement of parole and probation conditions to remove them from their communities, including the neighborhoods where people had grown up or had always lived with their families.

The way LASER’s individualized targeting was purported to work was that LAPD’s Crime Intelligence Details (CIDs, division units comprised of three sworn officers and a crime analyst) used Palantir technology to harvest data collected through FI cards, arrest records, the CalGang database, and “crime” history to give individuals risk scores. The scores generated through Palantir were used to assemble Chronic Offender bulletins, essentially “most wanted” fliers showing a person’s picture and personal information that were shared across LAPD divisions and units as people the police should target and harass proactively. Even though LAPD claimed to use a point system to standardize this process, a 2019 audit by the Office of the Inspector General uncovered that 112 of 637 people who were in the “Chronic Offender” database at the time had zero points under the LAPD criteria. [61]

Once people were marked as Chronic Offenders, special units were to go knock on their doors and tell whoever was there that police are watching them. [62] These units also served warrants, conducted parole and probation checks (even though they were not probation and parole officers), and pulled over or stopped people whenever they would see them. In 2019, a lawsuit we filed forced the release of lists of the people who LAPD targeted. Our analysis of these lists revealed that nearly half the people targeted are Black (even though Black people are 9% of the city’s population), some were as young as 16, and many are unhoused.

We recently began reaching out to the individuals on these lists and speaking to them. None knew they were being targeted, but they all lived lives of daily police abuse. After we explained what LASER was, the first thing nearly all of them told us was some version of “it all makes sense now.” Police constantly harassed them and broke into their homes, terrorizing their families and arresting them over nothing. “Why can’t I feel safe in my house?” one person said, “I just feel helpless. I can’t even walk to the bus stop without being pulled over.” Another person on the lists told us officers breaking into his home warned him that they “wouldn’t rest until they sent me away for good.” 

LASER was intended to make life intolerable for people in their own communities to the point where they would either abandon their neighborhoods or end up in prison. In this way, these programs apply the “attrition through enforcement” and “self-deportation” approaches that have long been the basis of federal treatment of Indigenous people as well as how state and local governments treat migrants. [63] We have long known that policing of Skid Row applies this same tactic, coercing people through a flood of citations. [64] More broadly, these tactics were the latest form of speculative criminalization of entire communities, in the lineage of gang injunctions, gang databases, and stop-and-frisk – all of which have been instrumental to dispossession, banishment, and racial segregation. Many of the people we reached out to from the lists are now incarcerated in state prisons far from home, leaving behind families who have told us the stories of how their loved ones were targeted.

We are continuing to research LASER’s personal toll, meeting with and learning from people whose names were in the Chronic Offender database when the program was suspended in summer 2018 before it was completely ended in 2019. We have heard nothing from the majority of the people on the Chronic Offender lists. LASER’s experimentation with “removing” people no doubt ended in death for some.


"The behaviors that a hunter-gatherer uses to choose a wildebeest versus a gazelle are the same calculations a criminal uses to choose a Honda versus a Lexus."
—Jeffrey Brantingham (2010)

Another location-based algorithmic policing program was PredPol, first deployed by LAPD as “field tests” in 2011 and ended in 2020 following a campaign of community pressure. [65] PredPol is a for-profit business co-founded by UCLA professor Jeffrey Brantingham and funded early on by UCLA Ventures as the company’s “lead investor.” [66] A PRA request we submitted to UCLA in 2021 revealed that the UCLA Foundation (the university’s endowment) purchased over 80,000 private shares of PredPol stock at the behest of UCLA Ventures (the university’s venture capital arm). [67] PredPol now tells customers that its business “grew out of a research project between the Los Angeles Police Department and UCLA” and that their products are today used to police “one out of every 33 people in the United States.” [68]

PredPol developed its algorithms through Branthingham’s academic research and field tests in collaboration with LAPD, along with technology developed to predict “insurgency” in Iraq and Afghanistan. [69] An anthropologist, Brantingham’s earlier research focus was data-driven paleoanthropology. From that work, he began using predictive analytics to compare the choices of ancient hunter-gatherers looking for food and current day patterns of how people “forage for opportunities to commit crimes.” [70] In 2006 Brantingham and his colleagues, including a UCLA mathematician named Andrea Bertozzi and a postdoctoral fellow named Greg Mohler, extended the foraging analogy to insurgents in Iraq. That year the researchers obtained the first of multiple U.S. Army Research Office grants to develop algorithms to predict insurgent activity. Brantingham has claimed that “the mathematics underlying the insurgent activity and the criminal activity is very much the same.” [71] In slides presented to the Air Force Research Laboratory, he drove this point home through images comparing “Afghan men and other Arab or Muslim men with their faces wrapped in scarves, gathered around a cache of automatic rifle” to “images of Latino youths in Los Angeles, labeled ‘gang members.’” [72]

PredPol’s guesses about future crimes were driven by three elements of historical crime data: crime type, crime location, and a crime timestamp. This reported crime data originated from crime reports filed by police along with calls for service, both of which can be entirely unsubstantiated allegations. LAPD provided PredPol “an automated data dump that happened every 2 hours,” also characterized by LAPD as “a ‘direct injection information transfer,’ meaning the automated data stream was compiled and sent directly to the PredPol website for consumption.” [73] PredPol used this data to make algorithmic predictions about which 500 x 500 square foot areas, called “hot spots,” had the highest expected crime rate within the city. These reports were distributed to officers to guide their patrols.

The mainstream critique of PredPol has been centered on the feedback loop that results from collecting biased and “bad” data, automating the same results. As this theory goes, when police target an area it generates more crime reports, arrests, and stops at that location and the subsequent crime data will lead the algorithm to direct police back to the same area. While this feedback effect from policing’s own activity perpetuating itself has been proven (and should be no surprise), other logics also appeared when we began to map PredPol hot spots, LASER zones, and Anchor Points. We discuss these findings in part 4 of this report, situating them in the broader history of how Skid Row has been policed.

The academic community has begun to reject the harmful research behind PredPol. In 2019, a group of 68 UCLA professors and graduate students wrote to LAPD condemning Brantingham’s work on “predictive” policing technologies. [74] And in June 2020, over 1,400 academic mathematicians joined a public letter condemning mathematical research that contributes to racist policing, singling out Brantingham’s work with LAPD, and naming PredPol’s “racist consequences.” [75] In part 6 of this report, we discuss how PredPol recently now markets itself as a “community policing” and “data accountability” company, a transformation that had first been promoted by other academics who are working to expand data-driven policing.

“Crime Prevention Through Environmental Design”

Another way land is controlled to criminalize communities is what police call “Crime Prevention Through Environmental Design” (CPTED). Also promoted by LAPD as “Design Out Crime,” the goal is usually to install or remove physical and natural structures in order to facilitate policing and surveillance. According to an LAPD press release from 2001, “Design Out Crime makes ‘community policing’ more than a mere buzzword.” [76] The same press release also proposes CPTED as a solution to problems like “decaying public housing projects” and “abandoned public spaces.” Instead of actually funding investments in public housing or shared green spaces, the city’s answer is changing the environment to facilitate policing and surveillance.

In a 1997 video about CPTED  described by LAPD as “a landmark video on this subject,” [77] an LAPD “Crime Prevention” officer who is shown surveilling an area explains how greenery makes his work less efficient:

Design Out Crime principles help me as a patrol officer by affording the opportunity to see into areas that would normally be concealed. I’ll give you an example. If I’m patrolling past this building, I’m going to be able to patrol by here and look into this mall area and be able to see a greater area in a shorter amount of time, which translates into me going down the street and patrolling another area. If I have to stop, get out of the vehicle, get out of my vehicle and walk over to this area because it’s concealed by walls and shrubbery and [unintelligible] then it’s going to take more of my time. [78]

Trading trees for cops can be deadly in more ways than one. In addition to all the direct harms of policing and surveillance, environmental research has confirmed that historical lack of tree cover is a reason why segregated Black neighborhoods face worse temperatures and air quality today. [79]

This environmental racism has now become part of data-driven policing, with LAPD’s new Data-Informed Community-Focused Policing framework including a section on CPTED that specifically recommends “the removal of hiding spots or physical barriers” to facilitate data-driven policing. [80] Likewise, Craig Uchida, the consultant who built LAPD’s LASER program, asserted in 2016 that “Crime Prevention Through Environmental Design (CPTED) assessments are necessary ingredients for data driven crime prevention” and recommended that this information “should be collected and analyzed especially within specific neighborhoods.” [81]

After LAPD’s Data-Informed Community-Focused Policing framework launched in April 2020, we filed a PRA request seeking any CPTED assessments produced under the program. In response, LAPD disclosed nine assessments conducted in late 2020 for Community Safety Partnership (CSP) “community policing” sites, which are discussed in greater detail in part 6 of this report. On first glance, large portions of these assessments hardly seem tailored to “specific neighborhoods,” instead repeating boilerplate recommendations about the need to install more surveillance. [82] Even where the assessments note that some of the infrastructural changes were requested by local residents, we have learned that these same proposals had been made years earlier by residents who were told by city officials that the changes would not be possible. For example, a community organizer who worked in the Pueblo Del Rio public housing complex told us that residents had years ago convened walkthroughs with city officials to recommend some of the same proposals for crosswalks and curb cuts that now show up in LAPD’s new CPTED assessments. At the time these proposals were denied or ignored, with city officials citing costs or other excuses. But now LAPD sweeps in to propose delivering these same solutions via the institution of policing. This dynamic of “community policing” programs holding communities hostage by offering vital resources and services that the city otherwise denies is examined in more detail in part 6 of this report.

Citywide Nuisance Abatement Program

What is known today as the Citywide Nuisance Abatement Program (CNAP) is an extension of a decades-long history and evolution of nuisance abatement laws and programs. CNAP targets buildings identified by LAPD to target for nuisance abatement actions. These enforcement actions aim to coerce the owners of the property to make significant changes to the property that further the LAPD’s surveillance programs and often displace tenants in the building, which can break up communities and fuel gentrification of the area.

CNAP is staffed by five core agencies: LAPD, the City Attorney, the Department of Building and Safety, the Planning Department, and the Housing Department. Three of these agencies (LAPD, the City Attorney, and the Department of Building and Safety) had worked together since 1990 through the FALCON (Focused Attack Linking Community Organizations and Neighborhoods) Narcotics Abatement Unit, funded by federal grants. FALCON expanded on a previous city abatement program that relied on the Uniform Controlled Substances Act of 1972 and California Health and Safety Code dealing specifically with narcotics abatement. Those laws were, in turn, an extension of the Red Light Abatement of 1953. [83]

Today, CNAP enforcement operates through data-sharing between the city agencies who run the program along with whatever stakeholders and private interests these agencies each engage. LAPD works to criminalize, track, and surveil residents. This data goes to the City Attorney, which can open an investigation and then send a demand letter or meet with the property owner to discuss changes to the building. [84] If the City Attorney determines the landlord or property owner is compliant with their proposals, then the building has “self-abated.” But if the City Attorney is unsatisfied, then they can file a nuisance abatement lawsuit against the building in state court.

Much of CNAP’s coercion occurs without a lawsuit ever being filed. Between 2018 and 2019, the City Attorney opened 479 nuisance case files but only filed 30 lawsuits. [85] Once a building is targeted for this kind of investigation, it can be subject to random inspections for compliance with municipal codes or become subject to other rules, restrictions, and changes requested by police. CNAP can also be used to make threats about a building’s conditional use permits, which can exercise extreme leverage over commercial properties that need those permits to operate. The Planning Department can also initiate administrative Nuisance Abatement proceedings to enforce municipal codes. This enforcement extends policing powers to non-police entities, expanding the city’s ability to police vulnerable communities.

Nuisance abatement allows police and developers to strategically banish residents or businesses whose buildings are targets for gentrification. The evidence of criminal activity used to target properties can be highly vague, and the CNAP process lacks even the minimal transparency that occurs with criminal prosecutions. Details of the “criminal activity” used to target properties are not made public, and most abatement efforts settle without litigation, which means the city rarely advances or defends evidence before a court. [86]

Although CNAP enforcement against a building can originate with information from police, politicians, neighbors, and property owners, the program is driven by LAPD allegations more than any other source. According to emails obtained through the PRA, a Deputy City Attorney (DCA) noted in 2019 that regardless of the source of a referral, an LAPD “gang officer’s communication would be sufficient for the DCA to open a case on the property.” [87] This is where data-driven policing’s role becomes apparent. Coercion, targeting, and displacement through CNAP were one of the “strategies” that officers were trained to use for Operation LASER’s Anchor Points, along with “licensing/conditional use permits,” another element of CNAP coercion. [88] And as we explain in part 3, CNAP prosecutors used LASER data to determine enforcement priorities. Part 3 also looks at how developers use CNAP to displace local businesses. Further, our discussion of data-driving policing in Skid Row in part 4 reveals how locations targeted as Anchor Points were key sites of displacement.

Not only is CNAP another way the city weaponizes police data, it also helps expand LAPD’s surveillance architecture. The City Attorney uses CNAP to require that residential landlords allow for LAPD walkthroughs and patrols as well as warrantless and keyless 24/7 access. CNAP prosecutors have also required buildings to hire private security who coordinate with police to screen visitors and check identification, excluding and banishing people’s family and associates, as well as enforce coercive rules on how people can spend time or gather in common areas. For residents, all of this can amount to a takeover of their homes by police, bringing the violence of “broken windows” policing and stop-and-frisk right onto people’s doorsteps, all under the threat of eviction.

Sometimes CNAP enforcement can require installation of CCTV cameras and other LAPD surveillance infrastructure. In part 3 below, we discuss an example of CNAP targeting a business in the rapidly gentrifying neighborhood of West Adams, closely coordinated with a real estate firm that was trying to buy the building that the business occupied. Along the same lines, the CNAP lawsuit filed against the owner of the Chesapeake Apartments, a 425-unit complex spread out across 17 acres in Baldwin Village, asked the judge to order “the establishment of extensive security systems at the property with direct access by the Los Angeles Police Department to these systems of monitoring and surveillance.” [89] These CNAP-compelled surveillance systems expand LAPD’s vast architecture of surveillance, which includes fusion centers, facial recognition, and other data-mining initiatives. The lawsuit also demanded that the landlord implement “improved tenant screening and lease enforcement procedures” as well as “armed, licensed security guards.” [90]

Just as the Chesapeake Apartments CNAP lawsuit helped expand LAPD’s surveillance and data-mining powers, the buildup to the case featured data-driven policing. Below is an LAPD mission sheet generated through the Palantir data-mining platform one week prior to the CNAP case being filed. The City Attorney describes the apartment complex as an "alleged stronghold for the Black P-Stones street gang," and LAPD’s mission sheet directs police to target "BPS gang areas,” even specifying that "heavy pedestrian and traffic enforcement on gang members belonging to this gang is encouraged." [91] The recap lists 105 police stops, 21 arrests, and 8 field interview cards during a one week period.

Patrol Mission Report generated by Palantir for LAPD shows hundreds of arrests in a designated target zone.

LAPD’s links to CNAP show how data-driven policing directly plugs police into enforcement of property laws, multiplying LAPD’s powers of banishment and displacement. Beyond programs like LASER, nuisance abatement is increasingly being coordinated and automated using centralized data systems. For example, City Council in 2017 enacted a motion presented by council members Joe Buscaino and Bob Blumenfield directing the agencies that comprise CNAP to improve the tracking and information sharing on these properties through “the use of shared databases and other technological tools to assist in the tracking of these properties.” [92] The motion also invited LAPD to create “additional training on identifying and tracking nuisance properties.”

In all these examples, CNAP is another way the city uses policing and police data to coordinate displacement and racialized banishment. These practices are expanding to become more automated with the growth of data-driven policing. In part 3 of the report, we dive into different examples of how CNAP and the other policing programs and tactics introduced in this part enable coordination, collusion, and affinity between police and real estate developers.