Digging into data: California cyclist fatalities

The U.S. National Highway Traffic Safety Administration released their 2012 traffic fatalities numbers yesterday, and the wonkier blogs make note that cyclist and pedestrians fatalities increased faster than those of car occupants. Between 2011 and 2012, we had a 3.3% increase in traffic fatalities in the United States, while cyclist and pedestrian fatalities increased by 6.5% and 6.4% respectively.


Highway 17 Crash

How does the Golden State fare in this accounting? 124 people were killed in crashes while biking in California, representing 4% of the 2,857 people killed on California roads and highways. This is a 7% increase over the 116 people who lost their lives while biking in 2011.

In my home county, zero cyclists lost their lives out of the ten traffic fatalities in Santa Cruz County. We had a single loss of life in 2011.

Let’s look at the San Francisco Bay Area, where many Cyclelicious readers hail from:

  • Alameda: 3 cyclists killed out of 61 total traffic fatalities in 2012. 5 cyclists killed in 2011.
  • Contra Costa: 7 out of 60. 2 cyclists killed in 2011.
  • Marin: 1 cyclist out of 9 total fatalities. 0 in 2011.
  • Napa: 2 cyclists out 7 fatalities. 0 in 2011.
  • San Francisco: 1 cyclist out of 29 total traffic fatalities. This matches the single fatalities in 2011.
  • San Mateo: 3 cyclists, 38 traffic fatalties in 2012. 3 cyclists also lost their lives in 2011.
  • Santa Clara: 3 cyclists killed in 2012 with 84 total traffic fatalities. No change from 2011.
  • Solano: 0 fatalities out of 19 total. A single cyclist was killed in 2011.
  • Sonoma: 2012 was a bad year with 5 cyclists killed out of the 37 traffic fatalities. In the previous year, a single cyclist was killed. A mode for Sonoma County is zero cyclists killed.

We like to talk about numbers like this in terms of Vehicle Miles Traveled. We have pretty good automotive data and understand that, nationwide, the increase in total fatalities tracks with the increase in increased mileage people put on their cars in 2012 compared against 2011. When people drive more, you’ll get more collisions on the road.

We only have road use data for cyclists in very limited form in the USA, unfortunately, so we have little idea if a 6.5% increase happened because there are more cyclists on the road, and if there are more, how many more? If we have 50% more cycling miles in the USA and a 6.5% in fatalities, this means cycling has become safer year over year. If cycling has increased only 5% year on year, however, then we know cycling has become comparatively more dangerous.

I allude to this statistical problem in a comment over at Biking In LA.

Look at the numbers for LA County, which shows a 30% *decrease* from 2011 to 2012, in spite of an estimated 13% *increase* in estimated the number of bike commuters year over year.

Does this show there are safety in numbers? We really don’t know. It only shows that cyclists are so few that any small difference in absolute numbers results in large percentage changes.

(“Estimated bike commuters” is my proxy for the number of cyclists, since it’s the only number we have that I’m aware of for LA County. The US Census estimates 0.9% of commuters (or 38986 rode bikes to work in 2012. In 2011, the estimate is 0.8% / 34622 people. Both of those figures are ±0.1%. I use the absolute number of people on the road vs the percentage of population, since we’re looking at absolute changes in fatalities.)

I leave it as an exercise for the reader to determine where those extra 44 fatalities nationwide came from.

Santa Cruz County (where we’ve had zero to three cycling fatalities per year since 2008) does an annual bike and pedestrian count where we’ve seen a modest uptick in bicycle use over the past few years. A high percentage of Santa Cruz County residents use bikes to get around and for recreation.

Cyclist fatalities increased more than car occupant fatalties

Cyclist fatalities increased faster than car occupant fatalities year over year, and some observers suspect this is at least partly due to improved safety for car occupants with air bags, crumple zones, anti lock brake systems, traction control, and other innovations to improve survivability in car collisions. People feel safer, but risk homeostasis encourages us to keep the risk level constant so we mitigate this “safer” car by driving more dangerously. The result: car occupants are indeed safer, but at the expense of those of us outside of the motor vehicle.

With thanks to the mystery tweeter behind People Power Santa Cruz, who notes Data for Santa Cruz County shows drivers more likely to die than bikers & peds and pointed me to this county by county breakdown of 2012 traffic California fatalities. I also look forward to further analysis from data wonk Dan S. 🙂

11 Comments

  1. With that line of logic, one can also argue that cycling is so safe, that encourage cyclist to take more and more dangerous behaviors.

  2. Hi Richard, I think the fault lies in assuming too much in just that few lines of statistic. No where is there any data that says that safer cars makes people drive more recklessly. If that is the case, are Volvo drivers the one we should avoid?

    Those statistical increases are more likely the influences of gas prices making people driving more, and putting more cars on the road, or the number of cyclist have increased.

    You are also assuming that all bicycle fatalities are the fault of the drivers, which is not the case. In one end of the spectrum you do have reckless driver, but on the other end of of the spectrum you get that Strava guy that rides into oncoming traffic to shave a few seconds, or the guy in SF that kills a elderly pedestrians. In the middle, you get riders that bombs down on La Honda @ 50mph, and a driver that is talking on the phone, where the faults lies in both parties.

    While I appreciate the advocacy that keeps me safe on the bike, it is not necessary to paint every car as a killer, or twisting a few numbers to make a story.

  3. Although I would agree that being in or driving a high safety rated vehicle can make me feel more empowered, I’m not sure I’d equate that to an increased likelihood of reckless driving. I’ve always equated these types of vehicles from a defensive position, i.e. that there is a high probability I will encounter or be impacted by a reckless driver.

    An example of this was when my daughter became of driving age, I found a nice used Volvo 240DL that she drove while in college. To my knowledge, she was never a reckless driver, although, another female type of her age slammed into the back of her while stopped at a stop sign. She was driving a nice BMW with a bumper sticker that said “Santa Barbara Brat”. Her mother showed up on the scene and berated my daughter accusing her of causing it. Turned out, the husband and father of that nice family was also a local cop. My daughter bought another Volvo soon after. Needless to say, I was really glad she was in the Volvo!

    Although, it can be an interesting dilemma, and one which insurance companies would love to embrace, as another excuse to hike our rates!

    Here is a recent attempt by a Canadian insurance company to do just that…

    http://www.kanetix.ca/canadas-worst-drivers-and-best-drivers

    Our friends at the NTHSA used behavioral science in a research study. They seem to indicate it’s more related to risk/reward/consequence factors.

    http://www.nhtsa.gov/Aggressive

  4. Speaking of more vehicles and bicycles on the roadways resulting in an increase of injuries…do you know what times of day the greatest incidents occur most?

    Morning school/commute – 7am to 9am
    Lunchtime – 12pm to 2pm
    Schools out – 3pm
    Evening commute – 5pm to 7pm

    Top FIVE type of incidents for the Bay Area:

    BICYCLIST FAULT (% of Total)
    1 – Wrong side of road = 20.62% (Most injury for kids)
    2 – Traffic signal / signs = 8.67%
    3 – Auto right of way = 8.15%
    4 – Improper turning = 4.81%
    5 – Unsafe speed = 3.80%

    AUTO FAULT (% of Total)
    1 – Auto right of way = 14.44%
    2 – Improper turning = 10.23%
    3 – Other hazard violation = 4.51%
    4 – Traffic signal / signs = 2.87%
    5 – Unsafe speed = 2.53%

  5. The below study points to a possible cause of the trend. Current rules are not even able to slow down the rapid increase in cell phone use while driving. Much harsher penalties are needed. A simple infraction is just not cutting it. Cell phones have accelerometers and GPS that can easily tell when someone is driving, and automatically disable incoming/outgoing calls/texts/surfing use of the phone until the vehicle has stopped. Its time for legislation to help make people ‘outside’ of a vehicle safer, not just the driver inside of the vehicle.

    http://www.statefarm.com/aboutus/_pressreleases/2011/december/13/2011-State-Farm-Distracted-Driving.pdf

  6. Can these data be trusted and what does it really mean to you and me?

    One of the things that has always ticked be off, when it comes to how they treat most of these heinous crimes against cyclists and pedestrians, is how they label them with their definition of “intent” or “fault”.

    This is at the core of the problem with how this type of statistical reporting is acted upon. How do we (they) really know the intent, pre-meditated or who’s at fault, before they are proven as such, i.e. a confession, investigative judgment, legal decision, etc.? Yet, this is the very thing, in most cases, they do all the time with their recording and reporting.

    As I dig more into the various systems data types, that at some point roll-up into these larger national systems, I realize how damaging these systems can be in helping to understand and prevent traffic injuries.

    One in particular is EPIC, which in part gets rolled-up into the NHTSA system. When you generate a report from their system, you’ll see they use a designation of “Cause of Injury”. All of the bicycling related deaths are categorized as “unintentional”. How can that be?

    Maybe that’s why we are technically, never really making any true progress. We’re like an institutionalized roller-coaster from year to year.

    As an example, taking the NHTSA (Pedalcyclist Fatalities) numbers for all CA Counties for years 2008-2012, we get the total numbers of 111, 102, 100, 117, 124 respectively. The result is a yoy increase of 13%. Is that considered progress?!

  7. I see where you’re coming from now.

    This post was compiled in about 10 minutes and was meant to be a “just the facts” informational posting. The bit about vehicle safety features and driving safety was meant as an editorial aside, but since it was in the closing paragraph I can see how that can be viewed as a conclusion. My apologies about this miscommunication.

    I agree: there’s really no way to determine why we have 6.5% more cycling death in 2012 vs 2011. I suspect it’s mostly because we have more people on the road overall (and kind of hint at that in my comments about more working people in LA County).

  8. Also found this an interesting read. Published in October 2013.

    2012 National Survey of Bicyclist and Pedestrian Attitudes and Behavior

    http://www.nhtsa.gov/nti/811841

    “While NHTSA encourages bicycling as an alternate mode of transportation to motor vehicle travel, an increase in this behavior often results in increased exposure to vehicles and other risks, accentuating the criticality of designing and implementing effective safety programs. Despite all preventative measures, crashes and collisions continue to occur. While bicyclists on the road have the same rights and responsibilities as motorists, motorists and bicyclists do not necessarily abide by the same rules.”

  9. Has anyone looked into whether headlight glare may be a factor? It is clear to most people I have talked with that the new HID and LED headlights can be temporarily blinding and very distracting. The uptick in fatalities seems to correspond with the recent rise in popularity of these lights.

Leave a Reply

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

This site uses Akismet to reduce spam. Learn how your comment data is processed.