It has been about five months since the well-publicized police slowdown in NYC, where our officers heavily scaled back police enforcement. They reportedly did so to protest a lack of support from our mayor after the non-indictment decision in the Eric Garner case and the murder of two officers.
Immediately following the news of the slowdown, many blog readers reached out to me and asked me to do some analysis on the subject myself. But there was a major problem- a stark lack of up-to-date police data. Now, you, like many others, may ask, “how did the New York Post break the story if there was no data?” Well, the answer is that data was available if you were a reporter who went down to the precinct to get more information.
This method of reporting out of data by the City is problematic in several ways. First, it is incredibly inefficient to have people walking to police stations to get data. Forcing people to request data in this way slows down the flow of important information to the public. Second, it means that information that gets out via the news is unable to be verified, since the raw data is often not made available after the story is reported.
One thing that reporters can do to help fill in where NYC government is failing in these situations is to post copies of the data they use in their stories, whether it was obtained via FOIL or in other ways. Unfortunately, the New York Post did not follow these best data practices at the time of the reporting, so we, as citizens, were left guessing what the raw data looked like.
Even in a world where citizens were able to access data quickly via an online portal, NYC has fallen behind cities like Chicago, Baltimore and Seattle as far as the granularity of public police data goes. Data that does come out (with the exception of our motor vehicle collisions) is rarely incident level, and is instead reported out aggregated at the precinct level. One such aggregate data set gives precinct level monthly counts of moving violation summons, so lacking better data sets I decided to explore the police slowdown with that lens. The data set was recently cleaned up by the great people at BetaNYC so I was able to dive right in.
So the first question… was there a slow down in moving violations given out in NYC in December 2014 and the following January? A quick chart makes the answer clear:
There were about 71,000 violations given out in November of 2014, but only about 37,000 in December, an astounding 47% reduction for the month!
The common wisdom was that most of the slowdown was targetting quality-of-life issues, and not safety oriented infractions. To verify that, I looked to see how the rate dropped for different types of violations:
Interestingly, all 34 types of moving violations saw a drop in counts from November to December. Speeding appeared on the top 5 reduction list, showing a massive 65% drop. Given that speeding is a major safety issue in NYC, especially in the Vision Zero era, this was surprising to me.
These reductions were of course contrary to the orders of the Police Commissioner and the Mayor. So that got me thinking- were there certain parts of the city that saw bigger reductions in moving violations than others? To find out, I mapped the reduction in violations being given out per precinct and looked for any patterns that might emerge:
One finding became very obvious from the map which shows larger reductions in bright red: Staten Island had a drastic drop in summons compared to other areas. In fact, the precinct with the greatest drop was the 120th precinct, the very precinct where Eric Garner was killed as a result of a police chokehold. That precinct saw an astounding 81% reduction in moving violations in December when compared to November.
Interestingly, some neighborhoods around Bensonhurst on the Southern end of Brooklyn seemed to have some of the smallest reductions in ticketing. Bensonhurst’s 62nd precinct only saw a 14% reduction in summonses, the smallest decrease of any precinct. What a difference the Verrazano makes.
Looking at the borough by borough numbers, my suspicions about the police slowdown Staten Island were confirmed:
The Bronx only had a 41% reduction compared to Staten Island’s 69%. Among boroughs, Staten Island definitely was the outlier. These findings reveal that those precincts closest to the Eric Garner incident were some of the most likely to stray from the Administration’s directives. This finding is striking given the importance of separating politics and policing.
The high numbers are especially interesting because the New York Post reported, when it broke the story, that the slowdown was in response to the brutal execution of two police officers on December 20th. But, there was only a third of the month of December left at that point, meaning that you could only cause a 30-40% reduction for that month if you stopped writing tickets all together at that point. The numbers show that the slowdown started much earlier than they reported. (This is exactly the type of thing we could have figured out sooner if data was posted online daily, much like NYC’s 311 data or police data in other cities).
Note that part of these numbers could potentially be due to the reassignment of police associate with demonstrations following the grand jury decision. But if that were the case, we would not see the largest summons reductions in the 120th precinct.
The NYPD is not alone in dragging its feet against calls for more open data. In fact, the CCRB, the citizen agency responsible for police oversight, still issues its reports mainly in PDF. And NYCHA has been a black hole of public data, despite the fact that it houses over 400,000 New Yorkers (about the size of Miami). Until our agencies start releasing data to the public in a systematic way, we’ll be left consuming information only in the precise manner that particular reporters or agencies want us to. And we, as a city, can do better than that.
Note that I chose to do November 2014 to December 2014 instead of December 2013 to December 2014 for two reasons… First, the policing has changed quite a bit in the last year, with an increase focus on speeding. Second, the November 2013 to December 2013 numbers were very similar.