Last year, you likely saw a story (or many) saying that biking was way down in Seattle. That was due to the annual release of the Census Bureau’s annual American Communities Survey data, which can vary quite a bit year-to-year. And the 2017 figures were way down.
But as we noted, these numbers were likely just outliers because they ran counter to the ten-year trend. And the 2018 survey results that just dropped seem to confirm that, with walking, biking and transit use significantly up compared to 2017.
Just as it was inaccurate last year to say that biking was down to a ten-year low, it is also inaccurate this year to say that biking increased 35% in just one year. These dramatic ups and downs are almost certainly just statistical noise. In reality, the trend line for bike commuting is on a steady upswing, and the 2018 figures fall much more in line with the trend than 2017:
More people in Seattle are biking as their primary way of getting to work than ever before. Both walking and transit use also continued their climbs in the 2018 survey. Walking to work is now up to an incredible 12% citywide, nearly double the rate in 2005. Public transit is closing in on a solid quarter of commutes, clocking in at 23%.
And all this added together means walking, biking and transit is up to 39%, and driving alone to work continues its steep decline. Now just 44.5% of Seattle workers drive alone to work, down from 53% a decade ago. Even if the 2018 estimates end up being a bit high, Seattle is on trend for walking, biking and transit use to overtake driving alone just a few years from now, a feat very few U.S. cities have accomplished.
Of course, this is not guaranteed to continue. These trends are the result of a lot of hard work by public agencies and advocates working to make walking, biking and transit more convenient. And as positive as this trend is, the state is doing their best to bring driving back. The SR 99 tunnel opened in 2019, and WSDOT has yet to build its massive waterfront surface highway. These multi-billion-dollar investments could very well bring driving numbers back up. So, uh, good job WSDOT?
But at the same time, Seattle is preparing some vital bike connections, including the south downtown bike lanes to finally provide a comfortable bike route between the International District and the 2nd Ave bike lanes. Sound Transit will open several more Seattle light rail stations in coming years. And a lot of new housing is set to open near jobs and transit, which boosts walking, biking and transit numbers. It will be very interesting to see how these numbers change in the next five years.
Speaking of these numbers, it’s always important to note some caveats about this survey. The annual American Community Survey asks residents which mode of transportation they used most to travel to work “last week.” So this is not a measure of all trips, only commute trips. Students aren’t included. Neither are trips to the grocery store or the park or a friend’s house. And there is no accounting for mixed trips, where someone bikes (or uses bike share) to connect to transit or for people who bike some days but not others. And by asking “last week,” a response will be very weather (and wildfire smoke) dependent. So, for example, many of the people who started biking to UW Station when it opened in 2016 would be filed under public transit, not biking. And “sex” data is only asked as a binary, so gender is poorly represented. But this is our most consistent nationwide transportation data set, which is why you see it referenced to often. In Seattle, real-time bike counters are probably better indicators of bike use. And those … wow. Off the charts. Stay tuned for more about what they are telling us about bike use in 2019.
Comments
13 responses to “Census survey: Biking, walking and transit up as commute data corrects itself + Driving alone down to 44.5%”
Thanks for the clear caveats on ACS methodology. The better a region’s intermodal connections, the less reliable ACS becomes.
The example I like to use — those pictures of Dutch commuter rail stations surrounded by thousands of bikes… all parked there by people ACS would say don’t bike, because their train rides are longer than their bike rides. The same is true here for anyone who has a longer commute leg via bus, light rail, Sounder, water taxi, or state ferry — we’re not cyclists in the ACS data.
Right on! ACS numbers don’t tell the real story.
I think we all remember when the 2017 ACS numbers came out and they looked really bad just as our bike counters were looking great in 2018. ACS’ 2017 numbers weren’t exactly noise, the bike counters confirmed cycling was down in 2017! Something did happen, both measurements agree! But it’s something that can be explained. 2017 had outlier-bad weather, one of the coldest and wettest winters in Seattle history. Not too long ago I joined a daily weather dataset against the Fremont Bridge bike counter data and ran some regressions. I unfortunately didn’t form an understanding of exactly how the various weather factors interact(*), but I did feel pretty confident that the 2017 dip was entirely explained by weather, at least at the Fremont Bridge.
If we want to understand what’s going on the bike counters and ACS data complement each other. Bike counters are more timely and count all trips. But they capture a small slice of the city whose trips cross a bike counter, whereas ACS tries to capture a representative slice. They can be subject to measurement error that’s hard to account for(**), whereas ACS’ main sources of error are typical of surveys . They only put bike counts in the context of their own history at some location, whereas ACS can compare among transportation modes and provide at least some demographic breakdown.
ACS was right in 2017 — biking really was down. If ACS says biking was down more in 2017 than the counters do, it’s possible even that was true(***). ACS was right but the pundits crowing about a decline of cycling in Seattle were wrong. Because bike-counter data confirming the 2017 dip and showing a strong 2018 rebound was already available as they wrote, they showed that they were at-best uninformed, and at-worst dishonest, in their punditry.
(*) I was playing with machine-learning packages; as I always say, “Machine learning is a way for people to write software they don’t understand… but on purpose this time!”
(**) This is a particular problem if you look beyond the “big three” bike counters, which are relatively heavily monitored, to all the ones the city updates monthly on their data page, to try to get a more representative slice of the city. I’ve looked at all the other counters and in every one apparent damage, maintenance, or re-calibration events overwhelm any longer-term trends.
(***) I wouldn’t have a hard time believing that people whose trips cross bike counters, especially at the Fremont Bridge, are more likely to have commutes that persist in the face of bad weather. I’d guess Fremont Bridge commutes are shorter-than-average, on better-than-average infrastructure, more likely to have expensive or scarce parking, and more likely to be served by transit that’s much slower than biking. But it also isn’t hard to believe any one number from ACS that contradicts other sources and trends is subject to the kinds of “random” error common in surveys.
Good points, Al. I shouldn’t say 2017 definitely didn’t decline, just that it appears far out of line.
I will say that the massive 2017 decline and 2018 rebound among bike commuting women looks particularly suspect, which is why I feel so confident saying 2017 was likely noise. The sample size of women who bike to work is unfortunately just small enough to get more noise from year-to-year, I suspect. I mean, there is no way half of women biking in 2015 stoped in 2017 only to largely come back in 2018. 2015 and 2017 were likely just noisy in opposite directions, making the swings look unrealistically dramatic.
I’d like to see a survey that is more oriented toward determining if bike infrastructure helps or hinders during rush hour (hopefully helps :)
Q1: do you normally go to work or school (or other regular activity) during heavy commute times ?
Q2: If so, do you bicycle on part(s) of that commute that have heavy traffic ?
Q3: If you don’t bicycle, why not ?
Probably several more questions, but that’s the gist. For Q3, some drilling down is needed to see if more bike infrastructure would help make the switch.
The problem with the existing surveys, as Tom and others point out, is they are just too arbitrary. E.g. what did you do last week ? What if you use a bike and bus combo ? What if you only bike on residential streets where there’s no traffic ? What if your commute is not during rush hour ?
Bike infrastructure can help any time of day, but to win public support, we need to show that it helps during rush hour and helps enough that it’s worth taking a lane of traffic or parked cars !
Could you provide a link to the survey data? Thanks!
So is Danny Westneat going to cover this new data point? Hello Seattle Times? The bikelash crowd have cherry picked the 2017 data extensively to support their anti-bike lane bias. Right up there with Senator “Snowball” Inhofe – “It’s snowing outside, so global warming is a hoax.”
Of course not. And there will be people recalling “that study that said cycling was on the decline anyway” for the rest of time.
Thanks Tom, great post — I think one other factor that could have made 2017 a counter-trend outlier would be the fact that the 2nd ave bike lane was under construction for 5 months, so there was no other good route to downtown from the north… would also be interesting to see a similar chart that tracked mode share, since that’s the best way to control for the change in population. Does the trend really still hold?
anecdotally I think bike commuting is doing pretty well in NE Seattle and around lake union, just judging from daily commute. ebikes are showing up more and more too, which is great.
I hate that fairview is shut down for the bridge rebuild, also I hate that there has been zero progress on a good protected bike route around the east side of lake union per http://eastlakeseattle.org/fp-content/attachs/2-clul-mp-final-parts-1-5.pdf
If we want a bike lane on Eastlake, we’re going to have to keep pushing for it or lose it due to the objections of Eastlake business owners who believe it will negatively impact them. With Alex Pedersen on the City Council (should he win the election), this could be a much harder fight, based on his previous comments regarding bike lanes. SDOT seems on track to build it for now. I’m sure nobody will be surprised to hear that this is a contentious issue in the Eastlake neighborhood.
[…] that it is the city’s most consistent barometer of biking trends. The recently-released Census survey showed Seattle’s 2018 bike commute rate rebounding from low 2017 results, though it will be another […]
[…] this reporting comes on the heels of several data points showing that biking is up in our city. 2018 Census data showed that more people are biking in our city than ever before. This data is especially […]