Is there anything cities can do to encourage cycling? Portland, for instance, has twice as many bike commuters per 1,000 people as Washington. But maybe that’s just because Portland has nicer weather or more young people. It’s not clear that there’s an actual policy issue here.
Buehler and Pucher found that the presence of off-road bike paths and on-street bike lanes were, by far, the biggest determinant of cycling rates in cities. And that’s true even after you control for a variety of other factors like how hot or cold a city is, how much rain falls, how dense the city is, how high gas prices are, the type of people that live there, or how safe it is to cycle. None of those things seem to matter quite as much. The results, the authors write, “are consistent with the hypothesis that bike lanes and bike paths encourage cycling.”
If that sounds overly obvious, the authors do note that previous research was somewhat scattered on this question. A few studies had found that more bike lanes in a city were associated with more cycling, though it was unclear which was causing which. Perhaps cities were building bike lanes because they already had a group of devoted cyclists. And this causation question still hasn’t been fully settled, but Buehler and Pucher’s regression analyses — going through a dataset of 90 of the 100 largest U.S. cities — suggest that the relationship between bike lanes and cycling is quite robust. (Previous studies on biking had often just looked at single cities in isolation.)
The question, they note, is an important one because of a “mounting body of evidence” that encouraging bike commuting can improve health, reduce pollution, and other fatalities.
Note also that the United States is still very different from Europe in that regard. Portland is the most bike-friendly major city in the United States, but just 4.2 percent of commuters in that city bike to work. In Copenhagen, Denmark, that number’s hovering around 37 percent. And Copenhagen gets a fair bit chillier than Portland for part of the year, consistent with Buehler and Pucher’s hypothesis that weather doesn’t seem to affect cycling rates very much.