Back from the Brink, an oral history of the introduction and implementation of Broken Windows theory policing in New York City

I have been remiss in reviewing Back from the Brink, the book by Peter Moskos that is an oral history of the New York City Police Department during the Bratton period, where new strategies and tactics were introduced to suppress crime in significantly measurable ways.  (I first mentioned the book last September, in "Murder rate in Chicago is bad: "correct" applications of broken window policing.")

Unfortuately, I have a bad habit of if I don't write the review as soon as I read the book, I'm not likely to get back to it, partly because I have another bad habit of writing my notes on various different pieces of paper, some get mislaid etc.

But the review of the book by Michael Fortner, a professor at Claremont McKenna College in Washington Monthly, "How New York City Got Safe: A historical reconstruction of the Big Apple’s crime decline, told from inside the institutions responsible for public safety," reminds me that it is important to do so.

A New York Subway car before Broken Windows policing practice was introduced in the late 1980s.

Fortner covers the book pretty well, but I think he misses some points.  

I lived in DC during the bad times in the 80s and 90s, and am a fervent believer in broken windows policing theory.  

In general, especially in progressive academia, it has a bad rep because most departments interpreted it as "zero tolerance policing" with a lack of focus on major crimes and the creation of a siege mentality by arresting people for the least transgression, like lack of a seatbelt (also see "Looking Through Broken Windows: The Impact of Neighborhood Disorder on Aggression and Fear of Crime Is an Artifact of Research Design" Annual Review of Criminology, "Why the Fraudulent 'Broken Windows' Theory of Policing Refuses to Die," Current Affairs).

What I got out of Moskos book are a few key things.  

First, the interviews make clear that Bratton and the top people believed that crime could be suppressed.  CompStat was about (re)focusing police time towards proactiveness.  

Although as Fortner quotes one commander, it was a philosophy.  Moskos cites a cop talking about how they had a problem with bike-based robberies in his precinct and how they addressed it.  But others took out of CompStat meetings about this precinct "arrest bicyclists" rather than analyze whether or not bicycles were used repetitively in perpetrating crime.

Bratton's team believed that by addressing pattern crime, and arresting perpetrators crime would drop.  

While generally, the police department wasn't focused on "broken windows" or improving public spaces per se, the application of focused policing on the subway system was complemented by a refurbishment of subway stations and subway cars--not just buying new equipment--but by addressing graffiti, etc.

I really got a sense of the incrementalism of the effort.  The idea of "moving the success of broken windows from underground/the subways" to above ground wasn't just about the CompStat tracking system and delving deeply into crime patterns, precinct by precinct.. 

Two stories in particular cover the integration of policing with public space improvement were the campaigns to significantly improve the  Port Authority Bus Terminal and Bryant Park (although the program to improve BP began before the Bratton period of policing).  

But as the president of the subway system told Bratton during his interview "if we don't fix stations and subway trains in association with better policing people won't take the train," the success of Bryant Park likely would have taken much longer to come to fruition without better, focused policing in public parks and Bryant Park in particular.

Another element of public space improvement at the time had nothing to do with policing, but with the creation of business improvement districts, funded by local property tax surcharges, these groups provided extranormal cleaning services and public space improvements and monitoring in places like Times Square and around Penn Station.

And the proof is in the pudding.  Sure multiple cities had significant crime drops starting in the late 19890s.  But eg DC versus NYC.  NYC had a 75% drop in homicides, DC 38%.  I'm sure this quantum difference in results between NYC and other cities was across the board.

Living in DC at the time, there wasn't a lot of broken windows approaches, but there was a focus on involving neighborhood groups, the way that Patrick Sharkey in Uneasy Peace attributes some of the fall in crime during that period to civic and collective action ("George Kelling, co-creator of the "Broken WIndows" thesis, dies," 2019).

I myself participated in one such neighborhood group that was started in response to a major drug area, and over time it was negated.  The cops came to monthly meetings, problems were discussed, there were occasional walks with police and elected officials, etc.  (I also did small experiments in my neighborhood, with litter pick up and measurement of its impact on certain streets and in bus shelters.)

A blog reader once made the point that as crime dropped, maybe stop and frisk was in part a response of needing something to do, but with the lack of restraints applied during the post Bratton Guiliani years (although Bratton in his later stint did support stop and frisk; later he recanted).  

Obviously, Stop and Frisk was proven to be not productive.  Unlike the value of arresting fare jumpers and checking them for outstanding warrants, an early initiative of Bratton's on the subway system, less than 2% of the people stopped under Stop and Frisk had warrants.  

So for me, the biggest issue with BW is that it hasn't really been applied in a true fashion in very many places.  And with how the Bratton people were pushed out by Giuliani, my lesson from DC is that you can't ever stop being vigilant when it comes to crime and order versus disorder.  Once disorder returns it's hard to reverse.

The rise in social justice concerns, deserved somewhat, as crime dropped and fewer elected officials and academics had direct experience of the mess of disorder during the worst of 1990s, did lead to a rise in crime in places like DC ("The Coming D.C. Crime Boomerang," Atlantic), as crime decriminalization was seen by perpetrators not as a social justice move to help people "driven to crime because of life circumstances," but as a signal to commit more crime (by extension, "The role of physical environment in the ‘broken windows’ theory," University of Chicago)

BART's new fare gates, seen at the Coliseum station in 2025, have generated more money for the transit agency. Lea Suzuki/S.F. Chronicle

Recently there was a super interesting BW finding in SF with BART stations with new fare evasion resistant gates ("BART’s long fight against fare evasion is finally paying off. Here’s how much," San Francisco Chronicle).  

Not only did rider attitudes improve, ridership and fare revenue grew, and instances of vandalism dropped significantly.  Like Bratton said, turnstile jumpers commit other crimes.  If they aren't on the subway because they didn't pay, well, then there is less likelihood for vandalism of subway infrastructure.

The final lesson with BW in NYC is that it requires constant vigilance too.  That it's hard to maintain because people get burned out.  But also when elected officials want the credit, they push out the top cops who bring the changes about.  That definitely happened in NYC.  Most of the top implementers of BW at NYPD were gone within a few years of Bratton's departure.  Although fortunately, further down the ranks were commanders, inspectors, and captains who took the precepts to heart and continued to implement the approach.

This is seen by considering Chicago, a city that has continued to have high murder rates, although they are dropping, despite purportedly adopting BW approaches.  (The city has had success with community intervention programs.)

"The federal help Chicago really needs," Crain's Chicago Business

At the University of Chicago Crime Lab, we have spent over fifteen years studying which interventions actually reduce gun violence and maximize the impact of government spending. We think we’ve found a clear path forward: Build the capacity of police commanders managing jurisdictions with the highest rates of violence. The Policing Leadership Academy has a simple premise — treat policing as a profession using the formula that has had tremendous success for private sector companies: investing in people’s leadership and management skills.

Also see "Policing and management," NBER, which makes the point about how the use of management time shapes outcomes positively or negatively.