How to make New York’s cyclists safer

By Felix Salmon
June 26, 2012

It’s becoming something of a trend these days: good report, bad press release. The latest example comes from John Liu, the New York City comptroller, who is warning about New York’s bikeshare program. “LIU: BIKE SHARE PROGRAM PEDALS PAST SAFETY MEASURES” says the release (geddit?) — and certainly that’s the message received by the New York Times, which wrote up the news under the headline “Bike-Share Program May Mean More Accident Suits Against the City, Liu Warns”.

The report itself, by contrast, is much less alarmist, and mostly extremely sensible. Biking in New York is dangerous, for cyclists and pedestrians both, and it’s important to make it safer. Especially as thousands of new bikeshare riders are going to start wobbling their way around largely-unfamiliar streets. Here’s the scary chart:


The blue curve is the well-known safety-in-numbers effect: as biking becomes more popular, it also becomes safer. New York is an outlier here, and not in a good way.

Charles Komanoff has some on-point criticisms of Liu’s report, and if you read his report closely you’ll notice one big flaw in the chart. The x-axis shows bikers as a percentage of total commuters, while most bike trips in New York are not home-to-work commutes at all. If you included all New York cyclists, New York would have a higher ratio of cyclists, and fatalities per cyclist would go down. Put it this way: the chart is taking the total number of bike fatalities, and dividing it by the total number of bike commuters, rather than the total number of bicyclists as a whole. That results in low numbers for cities like Portland, where cyclists are much more likely to commute to work, and high numbers for cities like New York, where they’re much more likely to just be running errands, or shopping, or meeting friends.

That said, New York needs to become safer for cyclists and pedestrians both, and Liu has some very sensible proposals for helping it do that. The city should put a lot of effort into maintaing signage, bike lanes, and intersections, especially the most dangerous ones: the effect of that could be huge. It should expand the Safe Streets for Seniors program, which helps older New Yorkers navigate safely around vehicles of all types. It should educate bikers and drivers both on bike safety and the rules of the road; drivers in particular should look out to make sure they don’t cut in front of cyclists when making a turn, and also leave extra space when passing a cyclist just in case the biker has to swerve around a pothole.

The recommendations continue: kids should get taught bike safety at an early age. The “5 to ride” pledge should be promoted to all businesses with bike messengers or delivery people. There should be more police on bikes, and they should start handing out tickets to cyclists speeding through red lights or dangerously salmoning. On top of that, they should start ticketing cars and vans in bike lanes. And just generally be tougher on traffic. As the report says:

New York’s roads are an interactive, multi-modal system; increased enforcement from any surface modes will increase safety across all other modes. Through greater enforcement of speed limits and greater traffic signal compliance, the roads will be safer for all users.

Liu also wants to beef up New York’s overworked and largely ineffective Accident Investigation Squads; that’s a great idea. And he wants to collect lots of data on biking in New York and make it public. Which is a no-brainer.

Liu is also pushing to make helmets mandatory; I’m not such a fan of that idea. For one thing, I have yet to see any empirical data showing that mandatory helmets increase safety. And in general, insofar as a mandatory helmet law would reduce the number of cyclists, it would also reduce the safety-in-numbers effect. And as the chief fiscal officer of New York, he’s worried that increased biking might mean increased liability in terms of settlements paid out by New York City to injured cyclists. That worry seems small to me: as Komanoff says, the number of new cyclists will only increase the total by about 6%, and the $10 million of insurance that the bikeshare program has is much bigger than the $2 million to $3 million that New York has paid out annually in the past three years.

Overall, however, I’d say that the report is a very positive thing. And in that it stands in contrast to the press release, which quotes John Pucher of Rutgers University as saying that he “would expect at least a doubling and possibly even a tripling in injuries and fatalities among cyclists and pedestrians during the first year of the Bike Share program in New York”. I’ll happily take that bet: it’s ridiculously alarmist, such a rise hasn’t happened in other cities with bikeshare programs, and no such projection is made in Liu’s report. Liu also wheeled out the media-relations guy from AAA New York, of all people, to say that the best way to prevent cyclists incurring serious injuries is to force those cyclists to wear helmets. That’s just depressing: one would hope that a car-drivers’ organization might at least pay lip service to safer driving, rather than putting the onus entirely on the bikers.

I’m very excited about New York’s bikeshare program, and look forward to using it regularly. I hope that the increase in the number of cyclists will help bring a bit more civility to New York’s biking community, especially in terms of stopping at lights and riding in the right direction. Meanwhile, my biggest fear is that we’ll see the opposite: a bunch of people who have no idea what they’re doing, riding on sidewalks, salmoning, and generally causing chaos. I don’t think that’s probable, but it’s possible, and I look forward to Citibike and NYC doing everything they can to prevent it from happening. As they do so, Liu’s report — if not his press release — is likely to be quite helpful.

More From Felix Salmon
Post Felix
The Piketty pessimist
The most expensive lottery ticket in the world
The problems of HFT, Joe Stiglitz edition
Private equity math, Nuveen edition
Five explanations for Greece’s bond yield
5 comments so far

If you remove New York and Chicago from that graph, I see no association between the two variables at all.

On the other hand, the Canadian cities are generally lower than the US cities.

Very hard to draw any sensible statistical conclusions from such a small sample.

Posted by TFF | Report as abusive

TFF – that is a fair point. There are 8 cities here, which is not a lot of data. I remember the McKinsey analysis of US healthcare costs compared to OECD countries – they used 12-18 comparison countries in their regression analysis (although for some regressions they used 6-8).

That said, there is reason to think that increased numbers of cyclists will make motorists more aware, which can make them safer (as in, accident rate per individual cyclist goes down). In Singapore, we’re always taught to turn our heads to check our blind spots when making a turn. There are a lot of motorcyclists and a significant number of bicyclists on the road. For better or worse they tend not to take the lane. So, you check your blind spot. I don’t see American drivers doing the same (folks born in the US – is this true?). They’ll hopefully learn to do so if there are more cyclists on the road. Or most drivers tend to underestimate cyclists’ speed, because the last person they saw on a bike was a kid, so they think they can do things like overtake a cyclist and immediately make a right turn. Not true, but hopefully people will learn with more cyclists on the road.

The AAA guy obviously is too dumb to know, but helmets don’t protect against all injuries. They will mitigate any collision involving the head, though. So, while I can see Felix’s point, I’m wearing mine, and I think all high mileage cyclists should. We’ll probably have to learn the hard way about what works.

Posted by weiwentg | Report as abusive

Felix “bets” that injuries and fatalities will “double or triple”. I’m calling my bookie, because I’m finding Komanoff’s argument persuasive on the issue:

“Since the increase in bicycle trips from bike-share is 1 part in 15, for the program to double the daily rate of collisions, whether with cars or pedestrians, or icebergs for that matter, each Citi Bike rider would have to be 15 times more likely than a cyclist on a regular bike to crash into something or someone.”

Posted by morshpotato | Report as abusive

Thank you Felix for your smart editorial.
This is not the first time when our media outlets ignored the most important substance of a many-page report and highlighted the most sensational, conflict-driven narrative.
I agree that mandating helmet use is absolutely the wrong idea. All over the world, the cities with highest rates of cycling, and the lowest rates of injuries, have very low rates of helmet wearing. It is because the governments of these cities (i.e. Copenhagen, Amsterdam, Berlin, etc. etc.) have actively re-engineered city streets to make them safer for all users, particularly the most vulnerable road users (cyclists and pedestrians). One thing that Copenhagen does is to send traffic engineers to study the scene of every traffic accident involving serious injury of death, to determine if an improved road design might reduce the likelihood of another accident occuring in that spot. In this way, hundreds of dangerous intersections have been systematically re-engineered for safe passage for cyclists, pedestrians, and motorists. Mr.Liu mentioned in his report that the city should improve the safety of dangerous intersections, so perhaps he might endorse NYC adopting this smart practice.

Posted by BenfromBedStuy | Report as abusive

the x axis – cyclist commuters as a percentage of all commuters – presumably includes train and subway riders, which form a much higher percentage of NY commuters than of any of the other listed cities. Isn’t the relevant figure cyclists to drivers?

Posted by Blox | Report as abusive
Post Your Comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see