Why did women in Chicago stop bicycling to work? And other stories about data

By On February 18, 2011 · 7 Comments

Why did women in Chicago stop bicycling to work?
Or is our data unreliable?

Array

Showing relative cycling-to-work rates between 2005 and 2009 in Chicago. Data from table S0801 in American Community Survey, 1-year estimates. Read the comments on this post for why this is not the best data source – 3-year estimate shows same decline in women cycling to work.

Note: The sample size is puny – data was collected from 80,613 housing units in Illinois. I don’t know how many of those were in Chicago (and we have 1,063,047 housing units). The American Community Survey only collects data on transportation modes to work for ages 16 and up.

But we simply have no other data! Maybe the Chicago Metropolitan Agency for Planning can release the Chicago data they collected for the 2008 household travel survey to show us bicycling rates for all trip purposes (they divided the report into counties). The sample size would still be small, but we could compare the work rates to find some support between the datasets.

We should look into how New York City counts bicycling as an additional way to gauge trends in Chicago (it has limitations of geography and area).

They conduct two types of counts. The first is the screenline count for bridges, Staten Island Ferry, the Hudson River Greenway, and all Avenues at 50th Street. They do this three times per year. Then, seven more times a year, they count at the same places (except the Avenues) from April to October.

While this data does not give them information on who cycles in the boroughs, it does give them a good indicator of cycling levels in Manhattan. It also disregards trip purpose, counting everyone going to work, school, or for social activities.

Sidenote: The New York Police Department will begin making monthly statistical reports on bicycle crashes in the city.

Tagged with: • • • • • • • • • • • •
 
If you enjoyed this article, please consider sharing it!
Reddit Delicious Mixx StumbleUpon Digg
  • Neil

    Yes, your data is unreliable. The decline you cite is statistical noise.
    The bike screenline count isn’t a terrific measure, either. it also depends on Manhattan’s island geography, nothing like that would be appropriate for Chicago.

    • http://www.stevevance.net/planning Steven Vance

      What do you suggest?

      The loop has a natural screenline: The Chicago River bridges, and also the few entry points of the Lakefront Trail.

      We have to change something.
      I don’t want to rely on this data – I’m using the shortcomings of this data as a way to get people talking. Not sure how much good that will do. The Chicago Police Department doesn’t collect bicycle crash data and we rely on IDOT for that (which has its own many shortcomings).

      The decline is also present in the 3-year estimate which has a higher sample size. I am looking forward to the 2010 1-year estimate, the 2008-2010 3-year estimate, as well as the 2006-2010 5-year estimate.

    • http://www.stevevance.net/planning Steven Vance

      Here’s some more data about Chicago, but it’s only the first timepoint in, hopefully, a series of timepoints we’ll see in the future.

      Bike counts in 2009.

    • http://www.stevevance.net/planning Steven Vance

      2010 data is out. 
      Stupid new American Fact Finder doesn’t have permalinks like old version did. This is an awful change.

      Women are now at 0.7% up from 0.4%.
      Men are now at 1.8% up from 1.7%.

  • Mark Elliot

    >we simply have no other data!
    Unfortunately, even the solid data that we have, injuries and deaths attributable to collisions, is often without ancillary information to make it very useful (like accident fault). At least the injury data is a cudgel – safety – that we can use to demand accommodations due us on public roadways. The ACS, yes, is just too imprecise. 
    Doesn’t Chicago have a bike count going on? At least we can have a benchmark for bike use, and then impute work trips and distance cycled. 
    My suggestion is to work with local economic firms to include a bike-to-work question like, “On average, how many employees biked to work last week?” These firms gin up reports to grease the wheels of community development with figures about over-regulation and the like, but maybe we can hitch a ride and extract some data useful for trips and for advocating employer subsidy programs.    

    • http://www.stevevance.net/planning Steven Vance

      Chicago has a bike count, but the extent and quality of the data and collection is unknown – it took 18 months for the 2009 bike count survey to be released to the public. The city’s history of secrecy seems to go right on down to the bike program. 

      Active Transportation Alliance, a local advocacy and consulting organization, manages the Bike Commuter Challenge which asks those questions. But perhaps it should be year round, like with a survey for self-reporting. The company would add data like the health insurance it provides (and the cost of providing it) along with the number of employees it hires and fires (to get knowledge on the percentage of its employees who bike). 

      • Mark Elliot

        BTW, I share your frustration with the Fact Finder. I came back to it for data after not visiting for a year, and it was all different – and somewhat difficult to navigate. I promptly turned to local libraries for a guide to the site/data, but only found a very expensive volume, but not in a library. 

        Re: counts, what we need, I think, is to treat active transportation folks (road users all the same) like we do  motorists: count them to project demand for  facilities like racks, for example.  Identify the tipping point where the bike community will be able to support a bikestation. Collect baseline data so we can begin to measure the effect of pro-bike incentive programs, or to measure the effectiveness of public health campaigns. How many cities have the baseline data to make that call? Probably none.     

        That effort should also (as you say) include employer-side surveys. The more data the better. If we put 1% of the funds and effort into changing travel behavior that we apply to changing general consumer behavior, why we’d have as many fans ‘liking’ Trek or whatever as they like Kraft Cheese.     

        Then there are the environmental documents. Let me stretch my legs on this one. 

        Where is the count data when an EIR is prepared? Nowhere. It’s people, not cars per se, that should be counted in a traffic study, but EIRs today include only reliable estimates for motor traffic – because that’s all that we measure. That’s fine if parking capacity is the only aspect of a project worth discussing. (Hat tip to Shoup’s cogent planning critique.) But bracketing out bike and ped effects not only makes traffic impact analysis a fiction, but it also sidesteps the potential to view development in something other than a negative light. 

        We need is a total understanding of what effect a large development might have, positive and negative, and not just on parking demand but on all facilities (from bike racks to mass transit). Without a good, regular, routinized bike & ped count, how will we know what effect a large TOD project, say, actually has? Or what  effect an incentive program has on shifting mode choice? And if we can’t measure it, we can’t really persuasively argue for funding it. 

        If we conduct surveys and convene focus groups about Kraft Cheese, seems to me that we can do that to understand user and potential user attitudes toward alternative transportation too! This to me is a classic case of misallocation: too many folks working on cheese, not enough on efficient transportation options.