The problem with Census data

By now, many of you have heard that the U.S. Census released new data (based on 2012 surveys) recently that shows bike commuting up by around 60% over the past decade. This sounds great, doesn’t it?

But seasoned bike advocate and author Elly Blue has some concerns about the way those numbers are collected, interpreted, and presented. Like her, I’ve often felt that the Census data collection on bicycle travel left a lot to be desired, and Blue sums up her concerns very eloquently in a recent Bicycling Magazine blog post:

New census numbers are out for 2012 (press release). Any time you hear anything in the next few years about the number of people who bicycle in the US or in a specific city and state, there is a very good chance the numbers will be from this survey. Proceed with caution: They don’t necessarily mean what we want them to.

Spin on over, give it a read, and then come back to tell us what you think — do you sometimes feel that bike commuters get under-represented? Are there more of us than the numbers show? Any other thoughts — we’d love to hear them!

1 comment

  1. Matt May 27, 2014 5:02 am 

    Yep. Agree the numbers are not as accurate as needed, not by a long shot.

    Not to mention, all census statistics are collected with a margin of error. If you look at a sheet of census data, there will be a column of data followed by estimated margin of error for that data. When you’re looking at smaller (than they should be) numbers, the margin of error ends up being quite significant. For example, you might have census data showing 300 people used bikes as their primary commute mode – but a margin of error of +/- 250. You can’t construct accurate representations when the census bureau itself is basically saying, “we’re not sure if this number represents reality at all.”

Leave a comment

Your email address will not be published.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