Finally, an Uptick in Minneapolis Bicycling Numbers

[Minneapolis is the only line going down.]
Call me a nerd, but I was excited to pore through the Minneapolis Bike Count Data released last week. It’s a bit like the Audubon Society’s annual bird count, in that in both cases you’re trying to get a statistical handle on a rare, endangered and flighty species, and in both cases you’re working to maintain and build a long series of longitudinal data.

The numbers are in and I was pleased that, finally, Minneapolis has started to see some growth in bicycling. For years, the city has had flat or even declining numbers of bicyclists. The problem is that back in ’08 and ’09 there was a bumper crop of cycling, a banner year.  (It may have had something to do with the economy.) That was the year that we caught up to Portland in our commuting percentage, launching the city into the “#1 Bike City” stratosphere.

Since then, we’ve invested a lot of time and money into expanding our network of bike lanes and trails, and seen little increase in the number of people riding. It’s been a bit of a mystery to me why Minneapolis has had flat numbers while cities across the rest of the country have shown consistent growth.

[3590 cyclists / day along the Greenway is easily an all-time record.]

The good news is that the data this year shows an uptick. In particular, I’m most excited by the large increase of traffic over on the central span of the Midtown Greenway. This is the spot just West of the Sabo bridge over the Highway 55 impasse, where the Greenway starts in its trench. It’s one of the key (non University of MInnesota) spots for bicycling in the entire city, but for some reason, there was a technical problem with the 2011 data. This year, though we saw a 30% increase in bikes on the Greenway even compared to the ’08 and ’09 years.

That says to me that bicycling really is growing in Minneapolis, and not just “shifting around” from one route to another.

Differences in Data

Part of the problem with counting and measuring bicycling is that its so difficult to get accurate data. Counting cars is relatively easy. All it takes are a couple of rubber pressure strips, and any public works department anywhere can do a traffic study of a street, measuring cars during an entire day and tracking how the flow changes over time.

For bicycles, it’s a lot more difficult. There are basically two different methodologies. The first, the American Community Survey (ACS), goes door to door as part of the US census and asks people how they get to work. This happened to me last year! I had just moved into a new apartment, and hadn’t even changed my address. One weekday afternoon, there was a knock on the door. A nice man asked me a few questions about where I worked, how much I was paid, and how I got there. I was more than happy to tell him about riding my bicycle to "work."

[ACS data for Minnesota. The yellow number is the % of "commutes to work" by bicycle.]

There are advantages and disadvantages to the ACS counts. You get a lot of interesting demographic information, and it makes sure that you are geographically thorough… because each city likely has a similar ratio of bicyclists captured by the information, you can compare cities to each other. On the other hand, the League of American Bicyclists (LAB) has a good list of the downsides on their site touting the 2011 bicycling counts:

ACS limitations, notes, and cautions
  • The ACS asks only about commuting. It does not tell us about bicycling for non-work purposes.
  • Results are based on a survey of a sample of the population. Surveys take place throughout the year. The journey to work question asks respondents about the previous week.
  • The journey to work question asks about the primary mode of transportation to work. The wording of the question undercounts the actual amount of bike commuting that occurs. It does not count people who rode once or twice a week or people who bike to transit (if the transit leg is longer than the bike leg).
  • Since the ACS is a survey of a sample, the results are estimates. The ACS releases a margin of error along with the estimate. Users can add and subtract the margin of error value from the estimate to find the top and bottom of the range within which the ACS is 90 percent confident in their estimate lies. Refer to the 2010 city tablef or margins of error.
  • Changes among years may not be statistically significant. Be cautious when drawing conclusions based on one year changes. Look at the trend over a number of years.

This rings true for me. I might ride my bike a lot around the city, but not necessarily to and from work. Or, I have a couple different jobs at any given time. I remember struggling to answer the question because I had to pick out which “job” I wanted to use to answer his question. There are real limitations to focusing strictly on the work commute, rather than how people get around on everyday trips. And if cities don’t have similar mode choice profiles (for example, if you live in a city which has a large community of “everyday” bicyclists, v. commuters, or if you live in a city with a lot of students), then you might be under-represented in the ACS survey.

[Daily bike traffic near the Lake & Hiawatha intersection.]

The second method is even more labor intensive. Minneapolis is one of few cities that sends people out to stand on street corners with clipboards counting bicycles by hand. They’ve been doing this for years, and it provides a thorough look at how many bicyclists are moving around Minneapolis during a “typical” fall day. While this survey method has some of the same problems – most notably, the focus on the ‘rush hour’ block of time – it’s a lot more able to track riders all over the city.

While there are a few differences between the data sets, for the most part the ACS data matches up pretty well with the city’s count numbers. According to the census, there are about 6,800 bicyclists in Minneapolis “commuting to work” on any given day. (There’s a high margin of error on this, about 1,300 people.) For contrast, the city wide count estimates daily traffic at its site at around 14,000, a bit more than double the commute numbers. (Makes sense, if you think that each commuter will pass by twice.)

On the other hand, looking at the numbers, I have to think that the ACS undercounts the students that make up a large percentage of Minneapolis overall bicycle traffic. For example, just on the Washington Avenue bridge alone, there are about 7,000 bicyclists each day. How many of them are considered to be “commuting to work”? I trust the overall city counts a bit more than the ACS numbers, though the problem there is that you can’t use it to compare Minneapolis to anywhere else.

[The #1 bike spot in the city, despite its lack of decent connection.]

The Uneven Geography of Minneapolis Bicycling

[Bike counts around the U of MN almost double anywhere else in Mpls.]
The main thing that jumps out at you when you glance over the Minneapolis bike count map is how unevenly distributed bicycling is through the city. Basically, almost all of the bike traffic is in a small triangle between downtown, the University of Minnesota and the Midtown Greenway. The only other key bicycling areas are the city’s off-street bike trails: the Cedar Lake Trail, Minnehaha Creek, and the Mississippi River Boulevard trails. Once you start heading north, or go across the river, bicycling drops off a cliff.

The other thing that’s easy to forget is that the majority of bicycling in the city takes place around the University of Minnesota campus, either in Dinkytown, Stadium Village, or on the West Bank bridge. And while there have been some small steps toward improving the University’s bike infrastructure, in my opinion the large numbers of cyclists around the campus are still woefully underserved by the city and the University. At the very least, there should be a cycletrack along University Avenue through Dinkytown, and as much protection as possible for the extremely busy 15th Avenue SE corridor.

Herding Cats or Counting Sheep? 

In the end, getting accurate information about urban bicycling is really difficult. Because bicycling is (for many people) an optional decision, the numbers can vary greatly depending on the vagaries of the weather or individual schedules. For any given city, much like the ACS data for Minneapolis, counts can go up or down as part of statistical noise, showing an uptick one year, and a downfall the next.

On top of that, its impossible to tell the difference between new riders and riders that have simply moved around. If you put in a new bike lane or path in your city, you’re going to see a big increase in traffic at that spot. But, how much of that traffic is just people moving from alternative routes, shifting over to the safer, more convenient path? And how much of that traffic is actual new riders deciding to bicycle?

I guess that’s one reason why bridges are such a great marker for overall bicycle counts. I’m encouraged by the large increase in bicycling just west of the Sabo Bridge, and am looking forward to see if there’s an impact from some of the new bike infrastructures that in place for next year. The Stone Arch Bridge has seen slow growth, and traffuc along Ford Parkway bridge has been flat for years. Looking at the bridge traffic below, the key seems to be to actually build new bicycle infrastructure. You can't expect that simple encouragement campaigns are going to change anything; the real driver is actual good quality bike routes.

That's why new infrastructure is so important. In the next year or two, I’ll be watching to see if the new large bike lanes along Portland and Park boost numbers and to see the Hiawatha LRT trail extension to downtown finally open. Hopefully the numbers of Minneapolis bicyclists continue to rise, and we don’t just rely on University students to drive our bike counts.

I’m looking forward to a day when Minneapolis is no longer stuck on a plateau while other cities keep adding more bicycle traffic. By investing in new, comfortable infrastructure, we can begin increasing bicycling as a way to get around. I'm hopeful that those days are just ahead.

[Very modest growth in traffic on the Stone Arch Bridge/]
[No growth in traffic since 2008 on the Ford Parkway Bridge.]

[Traffic counts on Park and Portland Avenues will be something to watch.]

1 comment:

Brendon Slotterback said...

You should encourage your readers to start using the Dero Zap program. The City and the University now have radio sensors ringing the campus and downtown core to automatically count cyclists that come in or out. Some employers are now using this system to encourage bike commuting and give health insurance discounts. Eventually, if enough people sign up, it could be a rich data source for counting cyclists (and could expand to other parts of the city).