limn

Public Safety and Wall Street

In 1994, Rudolph Giuliani was elected Mayor of New York following a campaign centered entirely on the problem of public safety.  He promised New Yorkers that he would restore the “quality of life” which had been destroyed by criminality. This promise was certainly directed at the lower and middle-classes, who were suffering so badly that they started to leave Manhattan (in preparing for the campaign, Giuliani ordered studies on the reasons behind this sort of migration, both inside and outside New York, that showed just this). More implicitly, this promise was directed at major international business players in the “big FIRE” sector (“Fire, Insurance, Real Estate”). True, this latter group had the means to protect itself. They lived and worked in ultra-secure areas. They were not personally threatened. But their status as the global elite gave them a special quality of life comprised of excellent restaurants, high-end clothing stores, elegant French bakeries, etc. Now, these services are rendered by none other than those very lower and middle classes who were directly suffering from crime. If the middle class left, the powerful elite of New York might do so, and risk that Wall Street would pack up and move to London or Tokyo. The city would no longer be the “sustainable” Mecca of world finance. The fight against street crime was therefore also indirectly about permitting New York to continue to amass immense capital. As surprising as it is, the argument is not mine, but that of a number of actors in and analysts of New York (Sassen  2007, Vitale 2008, Interview with O’Boyle 2010).

After the September 11 attacks, all the attention was shifted to the fight against terrorism, which completely overshadowed the link between Wall Street and the regulation of public safety in New York. Nevertheless, this regulation has undergone two successive and important revolutions revolving around the very statistical and information technologies it depends on. Almost immediately after being elected, Rudolph Giuliani recruited William Bratton, who put into place a technology called Compstat which has been widely talked about – even among the public at large since it was featured in the television series The Wire, set in Baltimore, which like other cities the world over had starting using it. The Compstat revolution reconfigured crime into rates of variation from statistical indicators (basically crime statistics), and the acting police chief was responsible for decreasing or increasing these numbers—and proving just that before the city council in public meetings.

Then, in 2005, new mayor Michael Bloomberg willingly accepted police chief Ray Kelly’s implementation of another device called the Real Time Crime Center (RTCC). The tool is not strictly statistical, in that rates and quantities are replaced by giant, widely varied databases with which detectives do data mining and engage in real-time identification of individuals who have just committed crimes.

The conference center of the Real Time Crime Center

Another view on the RTCC Conference Center

Despite these temporal coincidences, the connection between finance in New York and the new public security measures is rarely noted. Having focused for several years on the implementation of Compsat in Paris (Didier 2011a, 2011b), I went to New York in April, 2010 to observe the original version of this technology, which we have adopted in France and there conducted this ethnography on police matters.  I met numerous NYPD officers of all ranks, visited a police precinct headquarters in Manhattan, and talked with the upper echelon in the force, including Stephen Goldsmith, then deputy mayor, who later allowed me to visit the RTCC in NYPD headquarters at the legendary 1 Police Plaza. I argue here that these two technologies complement one another by reconfiguring the connections between public safety policing and the traditional collectives it is generally associated with: suspects, policemen, the masses. And, at the same time, to try to support the unexpected hypothesis that these transformations are also connected to Wall Street.

Compstat

Right after his election, Rudolph Giuliani recruited William Bratton, whom he had met when Bratton was Chief of Police in Boston and later Chief of the NYC Transit Authority. As police commissioner, he was in charge of reducing crime both quickly and forcefully. Bratton’s response was to invent Compstat (well analyzed by Silverman, 1999), which he said came from two principal sources of inspiration. The first, which was extremely well received in business circles, was Hammer and Champy’s Reengineering the Corporation (1993), a management theory from the private sector that defended the idea of placing individual responsibility with the officers and shortening hierarchies whenever possible. The other was the “broken window” theory of James Q. Wilson and George L. Kelling (1982), members of the very conservative Manhattan Institute. It stated that police must respond to all urban disorder, as trivial as a broken window, or else inhabitants would get the impression that their neighborhoods were neglected and would in turn neglect it themselves, abandon it, and finally leave it to real criminals.

In concrete terms, the precinct commanders are at the heart of Compstat. Halfway up the chain between beat patrolmen and police headquarter masterminds, these commanders were given as much power as possible to help implement their own crime reducing strategies. Thus, they had to show initiative. But all this freedom came with serious responsibility: every two months they had to go to a “Compstat meeting” to give an accounting before a mixed assembly  sometimes including the mayor himself, always the head police chief, other precinct commanders, and finally local residents and local business representatives.

The report basically meant preparing number charts and quantitative indicators for projection onto a giant screen. This stage was difficult because initially the data for each of the precincts was simply not available. The first step was therefore to become computerized (thus the name COMPuterized STATistics). Once this was accomplished, precinct commanders could use information much more quickly to show the public that police activity had increased (the number of patrols, time spent on the street, “stop-and-frisks” etc.) while crime had abated (whether for small infractions – for example, the campaign began with a war on squeegee men in front of Grand Central Station – or any of the “7 major felonies”).

A white board at the Mahattan Precinct Commandors tracking daily data on the "7 Majors"

One must understand that criminal investigation divisions, which fight big crime, international networks, and the mafia, have never produced any important numbers because the enemy, although powerful, is a very small crowd. Arresting Al Capone is striking only one arrest. On the contrary, the broken window theory, armed with Compstat, allows for one to compile, add, and present very important misdemeanor numbers.

Now, Compstat serves not only to represent crime, but also – and especially – to act upon it, transform it, and reduce crime. What’s more, Compstat was often considered a miracle because these misdemeanors, which initially appeared in staggering quantity, truly began to subside. Many who lived in New York in the 1990s will tell you today that they feel much safer. This is what the police have to show: that small-time crime exists, that they actively fight it, and that in stemming its tide they also help to reduce violent crime. The collective police force is thus also reformatted towards an increased subjectification in the sense that they must show that they are taking initiatives, that they are inventive and proactive, and at the same time responsible for both successes and failures. And responsibility is not an empty word because a good fraction of the commanders were disciplined after the first year.

The Precinct Commander's binder, where he has all his affairs classified by type of crime.

He calls them "the statistical crimes"

Finally, these numbers are also clearly associated with a spectacle that makes the public take notice of municipal action. Compstat is not a system of dissimulation for the police, but just the opposite: it’s an echo chamber. The neighborhood in question is invited to personally attend the precinct commander’s presentations (though this is not true anymore, as they felt that humiliation was not bearable when they were being shouted at). The press is also invited to attend whenever it desires. Police officials from around the world crowd together and attend strategy meetings (in regards to France, consult de Maillard, Le Golf, 2009, which contains an impressive list of political  representatives who have come to experience Compstat firsthand). Finally, the fact that Compstat appears several times as a setting in the TV series The Wire further illustrates its popularity. It’s not enough for crime to subside; it’s also necessary to furnish citizens with proof that helps convince and reassure them.

He explains that he must know exactly where every crime takes place, who frequents the place, who makes the report, etc. He can't get it from the database, but has to register it himself.

On might recall that scene from John Ford’s The Man Who Shot Liberty Valence, when a journalist declares about the eponymous crime: “This is the west, sir. When the legend becomes fact, print the legend.” Today we might say: “This is New York, sir. There is no longer any difference between fact and legend.” The numbers are compiled in order to meet three objectives at once: first, to produce facts; second, to act upon and transform these facts; third, to create a spectacle around them.  It’s not about accusing the NYPD of lying, because in spite of the criticism, which has for the most part attributed the lower crime rates to factors other than police innovation, the crime rate actually has dropped. It’s just that the tool used to accomplish this also allows for the creation of a legend that pleases everyone, including middle-class New Yorkers, the global elite, and police administrators the world over (Salmon, 2007).

Still, for a while now Compstat has been having problems (Eterno and Silverman, 2012). The main trouble is that the need for continually lower numbers means that these very numbers have become less meaty. On the one hand, not doing better than in the previous year means admitting failure. One the other hand, truly doing better seems to have become more and more difficult – more difficult, in any case, than… cheating with the math. The Schoolcraft scandal is a reminder of this. It concerns an NYPD commander who grew weary of orders that led to an erosion of real police work at the price of obtaining better statistics (engaging in unfounded arrests and unjustified identity surveillance, refusing to take complaints). The commander therefore decided to record all his activity including unwarranted orders from superiors for several months, and after failing to get the attention of the upper echelon of the police department, he decided to give the recordings to The Village Voice, a New York publication with huge distribution, which published a series of very noteworthy articles, and became later the subject of an episode of This American Life. He is currently at odds with the NYPD in a court trial. One might wonder if this wasn’t the beginning of the end for all this technology. But Bloomberg’s team seems to have anticipated these difficulties, for in 2005 it developed a new police technology.

Real Time Crime Center

Bloomberg at the Mayoral Office, he takes a seat and pulls up... Bloomberg Network.

Elected mayor in 2002, Michael Bloomberg, founder of the finance information network that bears his name (and also, incidentally, the city’s wealthiest taxpayer), can not be accused of having left Wall Street in the lurch. He in no way questioned the link established by Giuliani between public safety and finance. On the contrary, he used the latter as a model for his own administration and physically transformed the mayor’s office into an open floor based on the trading floor (photo). He also did not question the use of quantification of officer activity as a means of evaluation. In fact, he expanded this technique for use in every department and sub-department in his administration (cf. Mayor’s Management Report). Finally, as a journalism expert, he did not question the bombast with which the statistics were published. Still, he and Police Commissioner Raymond Kelly wanted to implement their own anticrime apparatus. They didn’t really abolish Compstat, which was still operational after some minor modifications. Rather, the two technologies evolved together.

Bloomberg and Kelly, for their part, did not rely on a theory of management, as Bratton did. They went to an expert, Jim Onalfo, who had over thirty years of experience in enterprises ranking among the top 100 in Fortune IT. It was he who was charged with the task, and in 2005 the Real Time Crime Center (RTCC) was inaugurated.

The Plan of RTCC

Physically speaking, the center is a large single room at the police headquarters on 1 Police Plaza where work stations, all in a row, sit facing a giant screen (“data wall”). The 45 “investigators” and 16 “supervisors” working there were recruited after an internal job search at NYPD. They were hand-picked and presented as an elite team, and then split up in the room according to a hierarchy: detectives up front, then sergeants and finally lieutenants. They all serve as back-up for the field detectives. When the RTCC was created, these detectives were obliged to use make use of it, to get familiar with the system. Now it is apparently at their beck and call, but there’s a waiting list.

The power of the RTCC lies in the fact that all these work stations are connected to a giant data warehouse (which I was not permitted to see). The data comes first of all from the police force itself. One of the innovations at the heart of the center is that since Giuliani’s tenure, the data for all New York precincts has been interconnected. Thus, requests can be made in all 911 calls, complaints, arrests, and narcotics files. 23 supplementary system files, some from New York’s city and state records, are also interconnected: 311 (government directory assistance), Stars (parking tickets), e-justice, parole and probation, etc. Plus, the RTCC purchases public databases, like Lexis/Nexis, and also uses Google Earth. Finally, access to the files of telephone operators has been made possible without even requiring a subpoena from a judge. In the civil framework, this file exchange is comprised of “memoranda of understanding” between contacted parties. Facial recognition from video may still not be entirely possible, but the quantity of data available since implementing the RTCC is simply staggering.

The RTCC uses this information not for minor misdemeanors, which was Compstat’s chief concern, nor against white collar crime, which is handled by another sector. It tackles mainly what is called the “7 majors”: (homicide, rape, robbery, assault, burglary, grand larceny, arson). When a crime is reported, the detective on the scene can call the RTCC, which starts an investigation as quickly as possible in order to connect the observable elements at the crime scene to all pertinent information on file. This process of “connecting the dots” helps identify the suspect. A program called Cognos allows for queries into all the databases. For example, one can do a “pattern analysis” and connect cases that are similar in terms of modus operandi. One may also do “entity analysis,” which is to say looking at cases that are similar while not yet identical. For example, one might compare the case files of “Jim Beam, born 3/12/1945” to that of “Tim Beat, born 2/12/1945” and that of “Joachim Beam, born 3/12/1944.” All this data is sent back to the RTCC and projected onto the screen so that everyone working the case can take advantage of it. IBM, which furnishes the majority of the programs and hardware for the center highlights the fact that its tools can now handle the semantics of data (the meaning) and not simply the syntax (the organization). According to Captain Godek, who showed me the RTCC, it’s nothing less than the reproduction of human thought, only faster and made up of a mutual pool of investigators.