Survey Shmurvey, actual bike counts up in Madison

In yesterday’s blog post I reported that the Census Bureau’s American Community Survey results showed fewer people biking to work in Madison  in 2011 compared to 2010.  The ACS survey results for Milwaukee, on the other hand, showed more people biking to work, although the numbers were up only slightly. A number of readers commented that the changes for both cities were either statistically insignificant because of the margin of error and small sample size. People also said that the journey to work question is poorly worded. I agree on both points, and noted bike counts are a much more accurate.

Well, Madison does automatic bike counts at a number of locations on the trails and bike lanes.  I got ahold of those numbers and guess what?  From 2010 to 2011, the number of bicycles counted increased at every count location.  Those real numbers are not only statistically significant, they are consistent with the long-term upward trend in bike counts since Madison began doing them.


The graph above shows the average annual Average Weekday Traffic count for bicycles at the count locations. While there are peaks and dips from one year to the next, the graph clearly shows an overall upward trend.  It also shows an uptick from 2010 to 2011. The top two rows of the table below show the increases from 2010 to 2011 at 6 of the count locations in Madison. There are other locations in Madison, and last year Madison added an additional 7 count locations, so moving forward we will have even more data from other locations.


Remember these bike counts show all bike trips, not just journey to work trips, as the ACS is trying to measure. That said, based on anecdotal evidence and these counts, I think it is reasonable to intuit that if the overall numbers of people biking isare up, so are the number of people biking to work.

Does that mean the ACS is bunk? I don’t think so.  First, the long term trends for the ACS are consistent with the long term counts.  On an individual year they may differ, but they have value when looked at over time.  Second, the ACS is the only apples to apples way we have to compare cities and states.  Portland does counts one way, Madison another, Boulder another and many cities don’t do bike counts at all. The ACS asks the same (poorly worded) question in the same way of people across the country.

About Dave Schlabowske, Deputy Director

Dave was the first full-time staff member hired to open the Bike Fed's Milwaukee office 15 years ago. A former professional photographer and life-long Milwaukee resident, Dave likes wool, long rides, sour beer, and a good polar vortex once in a while.

5 thoughts on “Survey Shmurvey, actual bike counts up in Madison

  1. Another thing to consider: during the great recession, trips across all modes of transportation dropped–because fewer people were working and making trips to work. That may explain the big drop in bike counts between 2007 and 2009.

  2. Can you tell us exactly where the bike counts are from? Some of the locations in table are not very descriptive. Also can anyone explain the drop in University ave counts? Is this because the Southwest path was completed? I think that was in 2006 but the count drops rapidly between 2003 and 2004.

  3. Actually, the story is simpler. I ran the statistical test: there is no statistically significant difference in bike share of commuting in Madison from 2010 to 2011. The apparent difference is due to sampling variability, not to any actual changes. The ACS data from the Census reports margins of error which — with the right formula — can be used to compare differences in proportions/percentages.

Leave a Reply

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