Drawing Conclusions from Narrow Cycling Data Sets

It is very encouraging that more and more data is being collected concerning both the safety of cyclists and the overall adoption of cycling.  Most recently, Oregon Health Sciences University released their study of bicycle commuters in Portland.  While offering some specific and valuable insights about Portland, this type of study only offers limited value for drawing broader conclusions about the safety and risks of bicycle commuting as a whole.  The discussion that has followed this study, as well as many of these recent studies, has been quite heated. Generalizations and broad conclusions have been flicked back and forth incessantly (along with tempered discourse offsetting those generalizations).

Bike Data PuzzleSome of the studies seem quite tailored to draw conclusions in order to support a specific agenda.  While these agendas are generally agreeable to most bike commuters, the perception that “the agenda strongly motivates the study” runs the risk of giving the study a biased feel, which detracts from the studies overall authority and therefore its usefulness in the end.

I will state emphatically that I applaud the efforts put out into the various recent cycling studies, such as the recent Australian helmet study, the London assertiveness study and numerous others.  The information gathered, organized and presented is an arduous task that is instrumental in supporting bicyclists and our efforts to develop a safer environment for cycling.  However, while these efforts are tremendous, I feel that it is important that efforts be made to contextualize the resulting studies within a broader spectrum of existing and yet-to-be-gathered data.

Following the release of most of these studies, it would be wise to begin tempering the conversation with the acknowledgment that, at the moment, we mainly are arguing with our instincts and assumptions rather than comprehensive data. The data sets so far are not broad enough to offer any level of certainty. I can acknowledge that the arguments and debates that follow the release of a study leads to some very interesting discussions.  But it also feels like we are spending a lot of time chasing our own tails.

A broader objective in discussing the collection of cycling data sets is attempting to piece the puzzle together.  This would mean that data from a broad variety of bicycling studies would be gathered and, compared.  The goal is to organize a broader data set presenting as much available information about bicycle commuting in as many scenarios as have been studied.

It is clear that the groups behind the current studies would conduct broader levels of research if they could.  The limitations are certainly financial.  And I’m sure that corresponding data about auto and airplane use is abundant due to the financial impetus to understand this.  I wonder if the Dutch or the Danes have not already conducted broad studies on bike commuting with the purpose of supporting the momentum of their continued exemplary commitment to bicycling infrastructure development.  I wonder if the Dutch and the Danes would ever consider studying the US as an example of how not to implement transportation infrastructure. Probably not, they probably prefer to stay focused on what works, not on what doesn’t.

In piecing together the current limited resources supporting bicycle commuter studies, one of the goals should be to build momentum towards encouraging and motivating larger scale, broad based research.  Bike commuting studies must acknowledge that they show a slice of the picture, thereby defining the need for whole spectrum to be displayed.  An effort to contextualize all of the existing studies within a shared framework would be a very fruitful effort as well.  In fact, if anyone is considering conducting a study on bike commuting in their region, I would instead encourage that they consider looking at the value of gathering the existing data and finding the points of cohesion between the various data sets.  CommuteByBike.com and UtilityCycling.org would certainly offer support if such a compendium study was conducted.

With an appropriate amount of convincing, perhaps either local or federal government will come around to seeing the value in conducting broader studies of bicycling safety and adoption.  The motivation to better understand how cycling can better fit into our transportation infrastructure is clear and it would only take a very small slice of the BP reparations to conduct a very thorough, broad based and most importantly useful study of cycling safety, usage and implementation throughout the world.

There also could be a financial motivation to gather cycling usage and safety information.  Google comes to mind.  While it is unclear how committed Google is to improving their cycling map data (i.e. the Bike-There feature), it does seem clear that collecting data about cycling would help improve the functionality of these maps.  Location specific information about bicycle accidents could potentially assist Google in recommending safer cycling routes.  At the same time, this data could be recompiled for other uses.  Google is certainly masters of both data gathering and reporting, and if they discovered a motivation to focus their expertise in data towards cycling, I’m quite certain that they could compile some very relevant and useful information that would support decisions about the implementation of agendas supporting the wider adoption of cycling.

7 thoughts on “Drawing Conclusions from Narrow Cycling Data Sets”

  1. I wonder if the Dutch and the Danes would ever consider studying the US as an example of how not to implement transportation infrastructure.

    The answer is “yes.

    Some background: California and New York were doing some of those things with cycletracks and other ideas like parallel parking to the left of bike lanes, advance stop lines and bike boxes back in the early 1970s! The experiments, unfortunately, were a little bit disastrous and increased danger for cyclists. Instead of working to try to fix the shortcomings, however, traffic engineers and cycling advocates abandoned the facilities altogether. This is a lot of the reason you get so much resistance to facilities from many of our more seasoned cyclist advocates — they’ve seen it before and it didn’t work out so well.

    When the Low Countries finally jumped into the act a couple of years later, they looked at the US experience. Instead of giving up, however, they looked for ways to improve the American ideas, and they haven’t looked back since.

    BTW, I’m am really really really impressed with the new ownership for this site!

  2. Josh Lipton says:


    Thanks so much for filling in this part of the puzzle. The historical perspective on the collection and interpretation of bicycle commuting data should be brought to the forefront in more studies.

    Now, I wonder how much data gathering the Low Countries did in developing their bicycle infrastructure. They certainly made some good decision whether it was based on in-depth analysis or just boiled down reasoning.

    Thanks for the encouragement of our new ownership here at Commute By Bike. We really appreciate the support and look to your blogging at Cyclelicious as a great resource for cycling and writing inspiration.

  3. Kevin Love says:

    The City of Toronto conducted two surveys in 1999 and 2009, asking the same questions about cycling. Very interesting. One of the most interesting findings was that for the City as a whole, cycling rates were:

    29% Utilitarian Cyclists
    25% Recreational Cyclists
    46% Non-Cyclists

    For the first time, the non-cyclists were in the minority.

    Source is page 15 of:


  4. Richard: Not everything was inspired by American work in the 1970s. The first recognized separate cycle path in the Netherlands was built nearly a hundred years earlier, in 1885. There were many other cycle paths in many parts of the country a long time before the 1970s

  5. Amsterdamize says:


    The Dutch Center of Expertise on Bicycle Policy ‘Fietsberaad‘ has great data. The English version of the online knowledge bank (see top menu) is a bit limited compared to the Dutch version, but it has plenty to offer. Also, upon request this org will supply you with just about anything.


  6. That Dutch path predates even the Pasadena Veloway! That was an eight mile, multilane, elevated cyclepath with streetlights and gazebo turnouts! It was marvelous.

    1. Ted Johnson says:

      Was marvelous? Is it gone now?

Leave a comment.

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

Scroll to Top