Boston Bicyclist Injury Report

I’ve finally completed a report analyzing almost 1,800 bicyclist crash incidents reported by Boston Police between 2009 and 2012. The City of Boston produced a report in May 2013 summarizing this information (and also EMS data), but this new analysis uses the raw data (released by the Boston Area Research Initiative) to code the incidents by crash type, using the PBCAT system. Some highlights:

Based on data from the Mass Department of Public Health, the Boston police reports almost entirely miss 76% of bicyclist emergency department visits and 60% of bicyclist hospital admissions–the ones that do not involve a motor vehicle. Only 9% of the incidents in the BPD data did not involve a motor vehicle.

However, the DPH data provides no information about how crashes occurred, whereas the BPD data does. Looking at the police-reported data I found:

    1. The single most common crash type in Boston, with 12% of the total, is when a bicyclist hits a suddenly opened car door (“dooring”). Not only are these crashes common, but they can be fatal, especially if the bicyclist is thrown into the path of moving traffic after hitting the door. Bicyclists can avoid these crashes by keeping a door’s width from parked cars.
    2. Almost as common as dooring (also 12%) is motorist left turn. Although the turning motorist is required to yield, there are sometimes complicating factors, as when the bicyclist has no headlight after dark, is approaching from the sidewalk or the wrong side of the road, or is overtaking a car on the right and thus screened from the view of the driver of the turning car.
    3. The third most common type, also representing 12% of the cases, involves a motorist driving out from a stop sign or driveway without yielding. In nearly half of the cases, the bicyclist was approaching from the wrong way – in fact this is so common that the PBCAT authors used a wrong-way bicyclist in their illustration of the crash type.
    4. The fourth most common is motorist right turn across the path of a bicyclist (10% of cases). While some occur when the motorist is overtaking the bicyclist, in Boston it is much more likely that the bicyclist is overtaking the car (or riding beside) when the motorist turns right. There were five “bicyclist overtaking” situations identified for each single “motorist overtaking” situation.
    5. The fifth most common type was bicyclist signal violation – where the bicyclist fails to stop for a red signal (also 10%). There were 160 cases of this crash type in the data. By comparison, there were only 14 cases where a motorist failed to stop for a red signal – more than 11 bicyclist violations for each motorist violation.
    6. The sixth most common type involved a bicyclist failing to yield when riding out from a stop sign or driveway, or entering the road from between parked cars (8%).
    7. The seventh most common type was a fall due to surface conditions, mechanical problems, or alcohol use. None of these crashes involved moving motor vehicles (6%). (Keep in mind that the hospital data show that this is in fact the most common reason for bicyclist injury, it’s just that most incidents not involving cars are not reported to police.)
    8. Other crossing” collisions were the next most common type with 6% of the total. These are intersection collisions that would probably be included in one of the other categories if more information was available about the exact mechanism of the collision (e.g., who failed to yield).
    9. Motorist overtaking was the ninth most common crash type. Being hit by a car approaching from behind is probably the crash that bicyclists fear most, but it accounted for only 4% of crashes.  Nearly half, 43%, happened in the dark. Of the daylight cases, 11 involved bicyclists swerving to avoid an object ahead. Of the remaining cases, all but one involved a sideswipe – suggesting that motorists see bicyclists in daylight, and when there is an overtaking collision it is because they misjudge the passing space needed. The single daylight rear-end collision involved a motorist who was driving so recklessly and erratically before hitting the bicyclist that a witness had been watching him
    10. The tenth most common type, with almost exactly the same number of cases as the previous one, was bicyclist overtaking (4%).

Together these 10 crash types account for 82% of the incidents. Except for dooring and overtaking, all involve intersections or driveways.


  1. A single, daylight hit-from-behind out of 1800 incidents, although the 43% of overtaking crashes in the dark is of concern to me; it’s too bad police didn’t make note of the bicycle equipment in most of your cases.

    The dooring data is especially fascinating to me. When discussing doorzone bike lanes in my city, the city transportation planner claims that dooring is not a factor here.

    Do we have any exposure data at all? i.e. what percentage of Boston cyclists ride in a way to avoid hooks, doorings and so forth?

  2. The new crash reports will prompt police to indicate the type of lights etc. used if the crash happened after dark. So eventually we will get that data.
    There is data from Cambridge, Mass., Toronto, and elsewhere showing a high % of dooring crashes. Obviously in places with little or no on-street parking it is not much of a factor. There may be in some places an issue with failure to report crashes with parked cars (but not in Boston).
    No exposure data, so the next best was to compare wrong way and sidewalk rates for crashes that are (probably) unrelated to these factors such as falls. Of course in practice they may be related, if wrong way riders are more likely to have falls.

Leave a Comment

Your email address will not be published. Required fields are marked *