Safety on our streets is a gender equity problem

Rodrigo Davies
Strava Metro
Published in
4 min readJun 12, 2019

This is an edited version of a talk I gave a Code for America Summit 2019 on a panel with Tiffany Chu of Remix and Danielle Dai at the City of Oakland.

Have you ever wondered why there seem to be more men on bikes around our cities than women? If you haven’t noticed, I can share the news that there’s a big gender gap in participation in active transportation — in some places at least.

Kay Teschke of the University of British Columbia pointed out recently at Velo Canada Bikes in Ottawa that in typical Canadian cities, women are half the population, take half the walking trips, half the motor trips, half the transit trips, but one quarter of the cycling trips.

The pattern is similar even in other countries that have focused on promoting cycling in recent years: a 2017 study by the UK Department for Transport found that men made almost three times as many cycling trips as women.

And perhaps more concerning is the fact that there’s evidence the deterrence factor is getting worse. The last comprehensive study of women’s cycling behavior in the US found that the number of women cycling decreased by 13% between 2000 and 2010.

So what’s going wrong? In some cases it may be a problem with the way we count. But in most others, the participation gap can be directly linked to safe infrastructure.

Teschke studied rider behavior in Vancouver and Montreal relative to people’s proximity to safe bike infrastructure (i.e. protected bike lanes and dedicated paths), and found that for every kilometer they were closer to safe infrastructure, women’s participation in cycling increasing 6x. For men there was also a boost (3x), but half the magnitude.

Women’s tendency to gravitate towards safer infrastructure was also demonstrated by Sarah Kaufman of NYU, who studied the varying behaviors of New York City’s Capital Bikeshare program. She found that 40% of the stations that were most popular with women had a bike lane or protected greenway, compared to 30% of the stations most popular with men. Women also chose stations on lower-traffic streets and places that have had fewer accidents historically.

So if we know that safety is a big concern for groups that are underrepresented in cycling, what can we do about it?

The good news is that we know safer infrastructure works.

Calgary invested C$5.7M ($4.2M) in new bike infrastructure over four years and saw a 30% increase in the number of women riding.

A study by my colleague Haynes Bunn (which will be presented at the TRB Women in Transportation conference in Irvine, CA in September) quantified the gender impact of safe infrastructure in cases such as Philadelphia’s Market Street — which got a protected bike lane in June 2018 — and found that the women changed their route choices to routes previously dominated by male riders after safety conditions improved.

Unfortunately, public agencies typically don’t have the ability to measure participation by gender at scale, and counters and surveys are of course not capable of capturing route choice.

So step one is that we should adapt to the emerging consensus that protected bike lanes are a minimum requirement to make active transportation accessible to everyone, and use data sources such as Strava Metro to measure participation by gender at scale and evaluate whether interventions are working as intended.

Step two, however, of understanding the state of safety in existing transportation networks is a little more complicated.

In some cases, agencies rely on collision counts to evaluate safety. While in some places this data is well organized (kudos to the TIMS project at Berkeley, for example), using counts alone is problematic for a few reasons:

  1. Counts require reporting — typically hospitalization or a police report
  2. Unadjusted counts that don’t take into account overall volumes of people cycling (what planners call exposure) lead to safety maps that highlight the areas that people cycle most often
  3. As a result of 1 and 2, counts alone systematically underweight certain areas, such as underserved communities that may be less likely to report crimes to the police and areas where cycling participation is lower overall, perhaps due to unsafe infrastructure

In order to have a holistic view of safety risk and prioritize areas for investment, communities need an understanding both of where accidents are occurring and the overall volume of cycling and pedestrian activity in those places. Some agencies are leading the way in this type of work.

The Seattle Department of Transportation used Strava Metro to assess the effects of a bike lane pilot project on Second Avenue. They created a baseline of volumes on streets by building a model that combined Strava Metro and physical counter data, and from there could evaluate the relative risk of collisions at the street block level. Next they compared the characteristics of high risk streets, identified types of intersections or infrastructure that contributed to collisions, and found similar characteristics in other parts of the city. Street improvements can now be made before collisions even occur.

If we’re serious about helping everyone get around safely and efficiently, we need to build infrastructure intentionally, with full visibility of the impacts that those choices have on different parts of our communities. By embracing the fact that safety and equity are inherently connected, we can ask deeper questions of the data that’s available. This will enable us not only to quantify existing access and equity problems but perhaps most importantly, to demonstrate the significant benefits that interventions can have so that we can continue to do that work and inspire others to follow.



Rodrigo Davies
Strava Metro

Product @asana. Previously @strava @civicMIT @condenast, cofounder @howtobuildup. Runs on music.