Since the release of New York City’s much-anticipated comprehensive crash dataset earlier this month, community members, advocates, and other proponents for safer streets can more easily access traffic crash data to advocate for safer streets. Opening crash data in this format is an integral step towards Vision Zero’s success.

As the City continues to outline next steps towards eliminating traffic deaths, the public can now view the NYPD-provided dataset, which includes information on all vehicle crashes dating back to July 1, 2012. Its 332,871 records (as of May 20,2014) contains date, time, number of persons/pedestrians/cyclists/motorists involved (broken out by injury and fatality), as well as geographic information. Of the 332,871 records within this period, 28,244 represent vehicle crashes that involved pedestrians or cyclists (or both).

Of particular interest when looking at crashes involving pedestrians or cyclists was the “Contributing Factor” attribute, which is what the investigating officer indicated as a factor in what caused the crash. This attribute could help us understand the true causes of crashes, but the dataset leaves a lot to be desired.

Many crashes are not recorded with a contributing factor. The availability of data can play a major role in enabling communities to identify problems and work towards solutions — but not if the dataset is incomplete. Of the 28,244 crashes that involved pedestrians and/or cyclists, 17,627 of them were either recorded without a contributing factor, or the contributing factor was listed as “Unspecified.” This means the majority (62 percent) of the data is missing critical elements, and isn’t as useful for those who are trying to understand why crashes are happening in their neighborhoods and prevent future crashes.

“Contributing Factor” categories can be misleading. It’s also about quality — not just quantity. The bigger picture may be obscured due to overlapping categories listed as contributing factors. For example, the most commonly cited contributing factor, Driver Inattention/Distraction, was cited in 30 percent of crashes, but if you combine other similar contributing factors that often exacerbate distracted driving, we see that this number jumps to almost 43 percent. Among the factors that could exacerbate distracted driving are: Unsafe Speed (.05 percent of crashes), Backing Unsafely (2 percent of crashes), Failure to Yield Right of Way (9 percent of crashes) — yet these behaviors are their own separate categories. Similarly, Driver Inexperience (.7 percent of crashes) or Aggressive Driving/Road Rage (.5 percent) could have easily led to Unsafe Speed, Backing Unsafely, Turning Improperly, and Unsafe Lane Changing, among other contributing factors. This raises questions regarding the nuances between these categories, and how officers on the scene recording the crash may interpret them.

While the release of this newly geocoded dataset is a huge improvement, there are still some gaps to fill. As Streetsblog pointed out, the City has yet to publish 2011 crash data, and the geocoded crash locations aren’t as precise as they should be. Make Queens Safer also highlighted discrepancies in crash injury/fatality counts in the City’s Open Data database when compared to Motor Vehicle Collision Reports on the NYPD website.

In addition to amending these issues, more must be done to increase crash reporting accuracy. This effort should include increased training for the NYPD on recording contributing factors, as well as revamping Contributing Factors categories to reduce redundancy. More accurate data will further help to unveil the true causes of crashes so that the information can be used by city and agency officials, safe streets advocates and community members themselves to advocate on their own behalf for safer streets and increased enforcement.

Below is a breakout of the contributing factors to crashes involving a motorist and pedestrians/bicyclists: