Data Collection of Bicyclist Helmet Use in Crashes

I have some questions/concerns/misgivings about helmet usage as it relates to bicyclist safety and crash reports. It seems to me that it is not well-reported…

Arizona Crash Report

Curiously, given the hoopla intense interest surrounding bicyclist helmet usage, there is NO place on the ACR to report whether or not a bicyclist was using a helmet. There is a block for each traffic unit(4dd) and passenger(5c), SafetyDevice  lists things like helmet/airbag/seatbelt and so forth. However, this block is always supposed to be coded 0/Not Applicable, and is explicitly never to be 1/None Used, or 5/Helmet Used for pedalcyclists. See p.23, 26 of 2010 Arizona’s Crash Report Forms Instruction Manual.

Clearly this is often coded not according to the manual, see below for some live data e.g. from 2010. We might guess that the 10% or so that say Helmet Used probably does mean those bicyclists had helmets, and likewise probably the 35% that said None Used probably means not helmet. But the other 55% is (from the data) anybody’s’ guess.

There is likewise no place on the ACR to code for nighttime crashes whether or not a bicyclist was using required lighting equipment (a front headlight or a rear taillight/reflector).


For 1994 through 2009; all persons including bicyclists and other pedalcyclists used the P10 Restraint/Helmet Use field. The column was called REST_USE in the person table.

From 2010 and later, this info was moved to a whole new dataset, the SafetyEq table. the column name is MSAFEQMT; and it’s slightly annoying because it’s indexed by ST_CASE (i.e. the case number), and person number. There can be any number of records for any given person.

The field is called NM13 Non-Motorist Safety Equipment “This element indicates the safety equipment that was used by the non-motorist
involved in the crash”, in the FAR Manual. In the FAR Validation Manual, there’s an elaborate confusing explanation of differentiating between Not Reported and Unknown. Allowable values are:

  1. None
  2. Helmet
  3. Reflective Clothing (jacket, backpack, etc.)
  4. Protective Pads Used (elbows, knees, shins, etc.)
  5. Lighting
  6. (not used)
  7. Other Safety Equipment
  8. Not Reported
  9. Unknown if Used


For reference, you can see‘s field P23 Non-Motorist Safety Equipment from (MMUCC is a nhtsa-funded group that sets standards for data collection on traffic crash reports). This is consistent with FARS definition; except that mmucc says only two may be selected for any person, and FARS allows any number to be selected (theoretically, the vast majority have either one or two selected).


2 thoughts on “Data Collection of Bicyclist Helmet Use in Crashes”

  1. Here is some live data from the 2011 Arizona ASDM breakdown of 1,942 persons of type pedalcyclist; remember the manaul says the only correct answer is NOT_APPLICABLE (I haven’t thought through the implications of a passenger on a bicycle, specifically a passenger in a child-seat, that might be ok, and might explain not only the 1 CHILD_RESTRAINT, but also the 8 Seat belt users… in any event that’s obviously not statistically going to be significant).

    | eSafetyDevice          | count(1) |
    | NONE_USED              |      708 |
    | SHOULDER_AND_LAP_BELT  |        8 |
    | CHILD_RESTRAINT_SYSTEM |        1 |
    | HELMET_USED            |      215 |
    | NOT_APPLICABLE_0       |      712 |
    | OTHER_97               |        5 |
    | UNKNOWN_99             |      280 |
    | NOT_REPORTED_255       |       13 |

    SELECT eSafetyDevice,count(1) FROM 2010_person WHERE ePersonType=’PEDALCYCLIST’ GROUP BY 1 ORDER BY 1 ASC;

  2. Pretty good on-the-one-hand-this, and on-the-other-hand-that type article:

    I do related to the culture of fear line of reasoning, which Elly describes as “Despite the iffy statistics and eye-catching conspiracy theories in his now-famous anti-helmet talk, he makes some compelling points when he digresses from advocating against the actual wearing of helmets and focuses instead on what he calls a ‘culture of fear.’ Why, he asks, do we view bicycling as imbued with danger, but not walking or driving — when you can get killed doing those, too? If fear causes us to stay sedentary, won’t that kill us faster? He has a point.”

Leave a Reply

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