Fun Facts But Little Analysis in NYU Traffic-Injury Study

There’s a lot to like in this morning’s New York Times front-pager summarizing a new study of injuries to pedestrians and cyclists in Manhattan and western Brooklyn. There’s the pull-no-punches headline, “Crosswalks in New York Are Not Haven, Study Finds.” Amen to that. And to the accompanying photo in which a bus, two cabs, and a pedestrian hang out in the bike lane, forcing a cyclist to detour within a whisker of a truck’s protruding mirror.

The study itself, by a team of trauma surgeons, ER physicians and researchers at NYU’s Langone Medical Center, is featured in the April issue of the Journal of Trauma and Acute Care Surgery. According to the abstract (the full 8-page article is behind a pay wall):

Road safety constitutes an international crisis. In 2010, 11,000 pedestrians and 3,500 bicyclists were injured by motor vehicles in New York City… [Yet] studying fatality or [hospital] admissions data [alone] fail to capture the extent of the epidemic.

The researchers aimed instead “to identify the demographics, behaviors, injuries, and outcomes of vulnerable roadway users struck by motor vehicles in New York City’s congested central business district and surrounding periphery.” They therefore teamed with the Bellevue Hospital regional trauma center, which treated more than 1,400 pedestrians and cyclists injured in the Manhattan Central Business District and western Brooklyn from December 2008 to June 2011.

That database is a potential gold mine. Alas, the Times’ story sheds little new light on patterns of endangerment to pedestrians and cyclists, and may end up perpetuating stereotypes about who causes traffic crashes. The problem doesn’t appear to be a windshield perspective; Times reporter Matt Flegenheimer has evinced refreshingly little of the Times’ habitual pro-driver bias since taking over the transportation beat last year. Rather, it’s the age-old pitfall in reporting epidemiological results: the case of the missing denominator.

For example, Flegenheimer reports that “In a finding unlikely to surprise the city’s cyclists, about 40 percent of injured riders were hit by taxis, compared with 25 percent of the pedestrians.” But given that medallion taxis account for a little more than 40 percent of vehicles in motion in the Manhattan CBD (the area in which the bulk of the analyzed crashes took place), one might conclude that, relative to other vehicles, taxis are neutral vis-à-vis cyclists and even a plus for pedestrians.

In the same vein, the Times reports that among traffic victims ages 7 to 17, more than 10 percent of pedestrians and nearly 30 percent of cyclists were using a cellphone or music player. Yet the percentages of all children and teens who use electronic devices while walking and cyclist are also high. Without knowing those “background” levels, one can’t say whether cellphone use among that population is correlated with a greater injury risk, or a lesser one.

Ditto for the “finding” that 15 percent of adult pedestrian victims and 11 percent of adult cyclist victims “were found to have consumed alcohol before the collision.” Even leaving aside that many people consume alcohol without reaching the legally-defined 0.8 percent threshold for intoxication, the victim percentages are meaningless unless they are shown to be statistically higher than for the populations of pedestrians and cyclists as a whole.

At least reckless driving gets its due in the story, though after the jump, and only by implication:

One harrowing take-away from the report is that no area, it seems, can be entirely safe. Six percent of pedestrians were injured while on a sidewalk. Of those injured on the street, 44 percent used a crosswalk, with the signal, compared with 23 percent who crossed midblock and 9 percent who crossed against the signal.

The focus on sidewalk injuries is overdue and thus welcome. The first-ever epidemiological report on NYC pedestrian fatalities, Killed By Automobile (for which I was lead author), found that five percent of pedestrian and cyclist fatalities in the five boroughs in 1994-1997 occurred on sidewalks or other places (such as parks) where driving is prohibited. That the frequency rate (for injuries) has risen to six percent validates the escalated sense of vulnerability in the city following the sidewalk killings of Martha Atwater in Cobble Hill in February and Tenzin Drudak in Long Island City in March, not to mention the sidewalk-jumping episode in East Flatbush last weekend that left a 3-year-old gravely injured and his mother in a coma. Indeed, of the eight street memorials that activists spray-painted around the city a few weeks ago, four are on sidewalks.

Both the NYU-Bellevue study and the Times story underscore the value of data in understanding and curing the aptly-termed road-danger “epidemic.” The irony is that the data that arguably could have the greatest impact on traffic danger — the NYPD Collision Investigation Squad’s supposedly meticulous crash analysis and reconstruction reports — continue to be sealed off from researchers and advocates. Our proudly data-driven mayor goes ballistic on legislators who block traffic-calming speed cameras and lavishes millions on global road-safety studies, yet he remains curiously passive about the opaque state of street safety data in his own backyard.

  • Ari

    Well said.

    However, if 15% of the adult pedestrian victims had consumed alcohol before the collision, I imagine that is statistically higher than the pedestrian population as a whole. Unless all collisions happen on a Saturday night in the Meatpacking District!

  • Dan Lazare

    I found the following paragraph in the Times article particularly incomprehensible: “In a finding unlikely to surprise the city’s cyclists, about 40 percent of injured riders were hit by taxis, compared with 25 percent of the pedestrians. More than 80 percent of cyclists rode with traffic flow, but less than a third wore helmets.” I presume they’re referring to INJURED cyclists riding with the traffic flow. But how does “less than a third” compare with cyclists as a whole? Did helmet wearers suffer greater injuries or less?

  • What I wanted to see, but didn’t, is a breakdown of the pedestrians hit in the crosswalk by:

    daytime plus rain
    nighttime plus rain

    When I bought my rain parka I bought yellow. Four dark colors are also an option, and they are listed first.

  • Anonymous

    What’s the concern with rain? I mean, I understand the theoretical concern, but is there any evidence that rain affects the frequency or severity with which motorists strike *pedestrians* (as opposed to motorists striking other motorists)? I can’t recall any serious collision I’ve read about that anyone even tried to pass off as resulting from rainy conditions.

  • AS

    I could not access the NYT article because it requires a log-in, but your response seems flawless. The NYT does not care about accuracy, and some people would even say worse things, but it is good that this large media outlet actually addressed the issue.

  • Anonymous

    I think rain diminishes visibility, particularly peripheral visibility for drivers turning through cross-walks.

  • Mike

    The 44% hit in the crosswalk is also misleading (and was a big point in the NYT piece). It seems damn likely that more than 44% of people crossing the street use the crosswalk, so it’s proportionally safer than crossing elsewhere. The exact opposite of the conclusion drawn in the article.

  • Joe R.

    Crossing at crosswalks with the walk signal is VERY dangerous because of turning cars. You literally need eyes on the back of your head to keep from getting run over. During peak times with a steady stream of turning cars you could literally be waiting 30 minutes to cross the street. Given a choice, I’ll either cross midblock or on red. Either way, I just need to look both ways, then go if it’s clear.

  • Joe R.

    The sodium vapor streetlights the city uses, besides being aesthetically horrible, also diminish peripheral vision (actually, they pretty much shut down peripheral vision). This makes it more likely you’ll hit pedestrians simply because you won’t see them until they’re right in front of you. You need bluer light for your peripheral vision to work. Also, bluer light keeps you more awake and alert. This is obviously a good thing while driving. Liveable streets advocates should sue the city to install whiter streetlights, similar to the HID lights often seen in parking lots and gas stations. LED is another option. In fact, many municipalities have already installed LED streetlights. I don’t know why NYC is so far behind the curve. It’s even more surprising we don’t yet have LED streetlights because we converted the traffic signals to LED over a decade ago.

  • Joe R.

    It’s stupid for them to even mention helmets in the same context as bike-motor vehicle collisions. Helmets are designed to protect the rider from falls at speeds a child cyclist might ride at (i.e. 10 mph or less). They’re not designed to protect the head in collisions involving much greater forces, such as bicycle-motor vehicle collisions, or falls at the much higher speeds many adult cyclists achieve. Remember in the great cycling countries helmet use is probably under 1%.

    And yes, mentioning the percentage of injured cyclists not wearing helmets without giving the percentage of all cyclists wearing helmets is a glaring omission. I suspect the figures are similar, meaning helmets don’t decrease the chances of injury in a bicycle-motor vehicle collision.

  • Station44025

    I’m sure this will be misinterpreted and cherry picked by the pro-death I mean car interests. It is impossible to interpret the various stats thrown out without context, but my takeaway is that riding with traffic doesn’t guarantee safety.
    I don’t know what stats reveal, but my anecdotal experience is that driving in rain reduces visibility and increases topping distance. Add to that the fact that people are hurrying and hiding behind umbrellas. At night it is much worse.

  • Joe R.

    The 6% killed on sidewalks is a particularly disturbing figure. Sidewalks are the one place where a pedestrian should have a reasonable expectation of total safety, at least from motor vehicles. I think it’s long past time we install sidewalk bollards, particularly in crowded areas and near schools.

  • Anonymous

    Sounds like what “scofflaw” cyclists do. You’re right, intersections were made for cars so they don’t hit each other. All other road users are just a side thought. Safety trumps legality every time.

  • Andrew

    First priority for sidewalk bollards should be near police stations.

  • Ian Turner

    The new station at 321 East 5th Street does have bollards, presumably to protect against car bombings. It may be a standard feature at new NYPD construction. I think the new 33rd precinct building has bollards too, though I haven’t seen it.

  • Andrew

    So where do the cops park?

  • Miles Bader

    You can’t add bollards everywhere, it’s expensive, and only addresses a small part of the problem. It also makes the urban environment decidedly worse, and penalizes the victims rather than the transgressors.

    The problem is that people drive like mindless idiots, American culture encourages it, and there’s little repercussion for doing so.

    Change the police culture of ignoring driving problems (that’s easy, just start firing, from the top down, until they get the hint), change the physical environment to affect driving behavior (traffic calming, narrower streets, separated bike lanes taking lanes from motor vehicles, wider sidewalks, etc, etc), and hopefully, change the wider driving culture by making it very clear that acting like a homicidal maniac isn’t OK simply because you’re in a car.

  • Ian Turner

    Look at the streetview.

  • Joe R.

    All good ideas I’d love to see yesterday, but realistically not many of those in power would support such changes? I’m sure we’ll get all those things eventually, but in the mean time we could save lives now by installing bollards long more heavily used sidewalks. Bollards aren’t horrible expensive. In fact, besides lining curbs with them, you could use them to separate lanes at intersections. That gives motorists a good incentive to go slow, even when they have a green light. The message bollards send is if you drive recklessly, your car gets totaled, you likely end up in the hospital, or perhaps even dead. That’s an effective, self-enforcing deterrent against reckless driving if you ask me. Trees along sidewalks could serve a similar function.

    We really need to reduce traffic volumes also. The constant frustration caused by congestion undoubtedly contributes to people driving like maniacs by increasing travel time variability. Consistent trip times, even if on average they’re longer than highly variable trip times, cause far less stress. Less stress means less homicidal behavior behind the wheel.

  • Anonymous

    I am shocked and deeply disappointed in the title of this article. How could the finding that 44%of pedestrians are injured while in the crossing with the light , be a fun fact? In my neighborhood it is 100%.
    It is time that all advocates focus on that issue and I am glad this study and the Times gave it exposure. The DOT has the tools to eradicate this problem (spilt phases) without resorting to the improbable approach of posting a reluctant NYPD agent at every intersection. At a minimum it should be installed along all bike lanes, where it contibutes a full 20% to the reduction in collisions.

  • JK

    “That ‘crash’ database is a potential gold mine.” Yes, could TA and Tri-State ask the study’s authors to put that database online to maximize it’s public value and potential to save lives? I’d like to see what insights Charlie, and other top experts, could find in the data. How about it TA/TSTC?

  • Jonathan Rabinowitz

    What are we supposed to learn from this report? That New Yorkers in large numbers get run down by motor vehicles? That people who have a beer or glass of wine get run down by motor vehicles from time to time? That working cyclists get run down by motor vehicles more often than people who ride less?

    These all seem pretty obvious, and you don’t need to be a Streetsblog commenter to realize it.

  • Andrew

    Angle parking on the street. How civilized! Perhaps they should try it elsewhere rather than squatting on the sidewalk.


Report: Pedestrian Injuries Caused by Cyclists Declining in NYC

Contrary to the would-be bikelash revivalists among the city press corps, a new study finds that injuries to pedestrians hit by cyclists are on the decline in NYC. Released this week, the study was authored by Peter Tuckel and William Milczarski of Hunter College, along with NYU’s Richard Maisel. Reporting for CityLab, Sarah Goodyear writes that researchers […]

Doctors Relate the Horror of Traffic Violence at Pedestrian Injury Summit

Medical professionals and transportation experts convened Thursday for the third New York City Summit on Pedestrian Injury, hosted by Elmhurst Hospital. The day-long event brought transportation officials and advocates together with doctors who witness the destruction caused by reckless drivers in the city every day. The summit is chaired by Dr. Jamie Ullman, director of neurosurgery […]

What’s Causing the Drop in Bike-on-Ped Injuries?

Well, it didn’t take long for our local media to mangle the Hunter College study measuring bike-on-ped injuries in New York state. Statewide hospital data compiled by professors Peter Tuckel and William Milczarik plainly show a drop in pedestrian injuries caused by cyclists between 2007 and 2010 — the same period when NYC doubled the […]