ADOT’s Bicycle Safety Action Plan Study ( the study’s home-page link is now dead ) is a multi-phase plan to assess and improve bicycle traffic safety, with emphasis on Arizona state highways. In urban areas that often means the interchanges. Here are direct download links to final reports:
During the five-year study period “There were a total of 9,867 bicycle crashes statewide in Arizona… crashes that occurred on state highways were extracted from the statewide data set. There were 1,089 bicycle-motor vehicle crashes reported on state highways between January 1, 2004 and December 31, 2008” from this a focus area was determined, within the focus area there were 746 bicycle crashes. PBCat was used to crash-type all 746 of these collisions.
Thus this data set accounts for a small minority of bike-MV crashes, around 11%. But Working Paper 1, table 15 offers a useful comparison between the studied data and all statewide data. For example, we see the same suspiciously-high percentage (24 to 25%) of “other” fault ascribed to bicyclists as with other studies. As I’ve written before, the “other” fault is generally the result of a poor/improper crash investigations that tends to wrongly faults cyclists who are doing nothing illegal (see Understanding Collision Summaries) — this is statistical proof of poor-quality investigations are a statewide problem for bicyclists. This is a shortcoming of the crash reports, and not the BSAP; in Working Paper 1, figure 20, something they call “primary contributing factor” by crash group is assigned overwhelmingly to motorists (67%), and only 24% to bicyclists.
There was a lengthy, front page A1, Arizona Republic article by Sean Holstege on Sept 17, 2011 which perhaps was intended to be about the plan but did wander, understandably, to general topics. For example they make great hay out of the per capital fatality stats, without any discussion of how to interpret them — e.g. how weather probably affects them, with Arizona being more of a year-round cycling state; or a higher per capita usage, e.g. Arizona has significantly higher (than US) percentage of commuters (according to census figures, see Working Paper 1, Table 1 — Arizona is 0.9% versus 0.5% nationwide).
The story, as many “bicycle safety” stories do, lacks context of traffic in general. So, for example, there was a chart of the number of bike-MV collisions (about 2,000/year total). There is no mention of the fact that that represents only a tiny fraction of all MV collisions ( which ran well over 100,000/year over the study period). And though it mentions the number of fatalites, say 25 in 2008 — it never mentioned the total number of traffic fatalities (it runs around, and lately something under, 1,000 per year).
So here are some hard numbers, over the five year period 2004 – 2008 there were 681,466 MV crashes, of which 9,730 were bike-MV — a little less than 1.5% (taken from the historical overview in the 2009 Bicyclist Fatality study, which were gleaned from AZ Crash Facts — note that the numbers a slightly different in the BSAP, but I don’t know why). The number of fatalities is 4,943 total, 132 bicyclist; or 2.67% — so bicyclist fatalities were somewhat over-represented but not dramatically so.
Note that the ADOT plan by design is aimed at the small percentage of bike-MV crashes that occur on the state highway system. “The majority of bicycle crashes in Arizona (approximately 90 percent) occur on local, city, and county roadways outside of ADOT jurisdiction”
Also, by the way, the article inaccurately stated that the BSAP recommends a mandatory taillight law. That was in an earlier draft but was since removed — I don’t believe there is adequate evidence to support the additional burden on cyclists. The article does correctly mention that the BSAP recommends state-level sidewalk law clarifications, which seem like a worthy endeavor, given the huge proportion of sidewalk-related collisions, along with the current legal murky morass that currently exists when cyclists who cycle on the sidewalk subsequently collide with vehicles in crosswalks and driveways.
The raw data from the 746 crashes studied can be viewed and mapped at this google fusion table. I also read the data into a mysql database, bsap, password access available upon request using a scheme analogous to asdm and presently only available on mysql.azbikelaw.org, and not on the godaddy server.
The crashes studied were full-blown pbcat 2.0 with crash types, groups, cyclist’s direction; that’s the good news. The bad news is it’s not clear how that relates to bicyclist crashes in general.
Fatal vs. non-fatal: The dataset mentioned above, the 746 crashes includes only 4 fatalities and is data only from the focus area. The BSAP report has several charts and graphs that refer to 24 fatalities. The distinction is that the total 24 is over the entire SHS (state highway system); whereas the 4 is only the focus area. I don’t have the raw pbcat data for the 24 (or rather , the other 20).
Here is a bunch of data extracted via query from ASDM data — currently covering years 2009-2014. In other words, it is an attempt to automate the types of data presented in the BSAP but applied to all Arizona bike-MV crashes, not just state highway system.
Summary of PBCAT results
Many of the motorist drive out’s (from stop signs, or performing a right-turn on red) involve counter-flow sidewalk cyclists. Out of the 746 crashes studied here are the top 5:
Table 5 – Top 5 Crash Types
Percentage of SHS Focus Area Crashes Crash Type Description
103 13.8% Bicyclist Ride Through ‐ Signalized Intersection
83 11.1% Motorist Drive Out ‐ Sign‐Controlled Intersection
76 10.1% Motorist Drive Out ‐ Right‐Turn‐on‐Red
71 9.51% Motorist Drive Out ‐ Commercial Driveway / Alley
61 8.17% Motorist Drive Out ‐ Signalized Intersection
746 Total SHS Motor Vehicle‐Bicycle Crashes
Also interesting title of this PBIC presentation How to Create a Bicycle Safety Action Plan: Planning for Safety. It’s from the toole design people so, as expected, is filled with nacto and facilities stuff.