I was fascinated, and a little bewildered, by a post at Velojoy yesterday.  Susi writes:

“It was refreshing to discover in the League’s most recent Bicycle Friendly America Guide a list of 10 communities where female commuters outnumber their male counterparts.”

Then the post gives a list of of the 10 cities with a higher percentage of females riding, including places such as Las Cruces, New Mexico, and Vancouver, Washington. It’s a little hard to trust in these figures. Intuitively and anecdotally, I know there are more women riding in places like Portland than the official stats (approximately 68% males to 32% females) say. Yet, this list has some very funky data.

For example, in Birmingham, AL, there are, according to the data, 108 bicycle commuters. (First off, that doesn’t mean that only 108 people ride their bike to work. It means only 108 people answered in the Census that they rode their bike exclusively to work the week of the Census survey.). OK, it’s a small number, 108, but let’s go with it. Yet one column over is the claim that all of those 108 Birmingham bike commuters are women! 100%.

That’s just not credible. Or? I then remembered that the (now defunct) blog BikeSkirt, is from Birmingham. Could it be that the numbers reflect the strength of Elisa’s and Anna’s absolute commitment and passion to spreading the zen of commuting for women?

I continued scrolling. The numbers just look funny. 481 bike commuters in Anaheim, California. 0% of them women. One hundred and twenty eight bike commuters in Richmond, California. All of them women? Then I looked at the ‘margin of error’ column, which seems to bounce all over the place. How could there be so much error in determining on a Census form whether a commuting cyclists is a woman or a man?

The mystery will remain until I get a callback from the League of American Bicyclists, which created this amazing, but ultimately frustrating chart from American Community Survey (Census) data. For now, we can surmise that it’s true that in Vancouver, Washington, right across the river from Portland, my town, the women actually rule, with 51% of them cycle commuters. I can only hope.

Addendum: Deflation. Spoke with Darren Flusche at the League of American Bicyclists. He said, basically, that you can ‘drive yourself crazy’ looking at this Census data, because in small cities, the sample sizes are so small, and the margins of error so large, that, well, “it becomes pretty absurd.” Of course, Flusche said, it’s not credible that 100% of the bike commuters in Birmingham are women. Sample size is the culprit. Boo. Hoo.