Stephen Dubner reports on an observational study of bike helmet laws, a study by Christopher. Carpenter and Mark Stehr that compares bicycling and accident rates among children among states that did and did not have helmet laws. In reading the data analysis, I’m reminded of the many discussions Bob Erikson and I have had about the importance, when fitting time-series cross-sectional models, of figuring out where your identification is coming from (this is an issue that’s come up several times on this blog)–but I have no particular reason to doubt the estimates, which seem plausible enough. The analysis is clear enough, so I guess it would be easy enough to get the data, fit a hierarchical model, and, most importantly, make some graphs of what’s happening before and after the laws, to see what’s going on in the data.
Beyond this, I had one more comment, which is that I’m surprised that Dubner found it surprising that helmet laws seem to lead to a decrease in actual bike riding. My impression is that when helmet laws are proposed, this always comes up: the concern that if people are required to wear helmets, they’ll just bike less. Hats off to Carpenter and Stehr for estimating this effect in this clever way, but it’s certainly an idea that’s been discussed before. In this context, I think it wouldb useful to think in terms of sociology-style models of default behaviors as well as economics-style models of incentives.
It is no surprise at all that helmets discourage cycling. Effectively, they impose a tax on cyclists as helmets are not cheap and periodically need replacing. Although Carpenter and Stehr assure us that medical people (real doctors!) swear by their efficacy, the population based argument that helmets prevent injuries is far from convincing. See the figure and references in this wiki article for a case study from New Zealand where helmets are mandatory for all cyclists. It may well be the case that children are a special case though.
Dubner's a bit of a cycling fan; he's commented on my blog's cycling posts once or twice in earlier, less hectic times.
As you note, the effect on lowering cycling is pretty well known. As for helmets, they have a modest effect at best — maybe even negative once the "magic hat" effect is accounted for (cyclists with helmets feeling protected).
When I was a kid, I definitely fell into the "Helmets are uncool, so if I’ve got to wear a helmet to ride my bike, I’ll find something else to do" camp. Seems kind of silly now, of course. My guess is that this effect was sharpest on kids like me who were used to riding without helmets, and then were told that they had to wear them.
the biggest hurdle to ride bike in US to me (from china) is the road conditions: few designated lanes and too many ups and downs. They make bike riding both less pleasant and more dangerous.
so when we are talking about the difference between states that did and did not have helmet laws, do we have to consider that there are some other factors, besides the helmet itself, could affect the bicycling rate in a more or less independent way. To me, a bad road could promote the helmet law and attenuate the pleasure of bicycling in the same time.
in china, many people still use bikes as the way to go to work and none of them would wear helmet.
about the cool factor… my partner convinced me with one of these:
http://safetyissexy.blogspot.com/2007/11/red-hot….
I am a sample of one, but for me personally, being urged/ required to wear a helmet discouraged me from cycling.
Its not just the financial cost of the helmet, it was just another thing to have to remember/ fiddle with when using the bike, and to find a place to store. This was when I was in grad school and making the attempt to bike to my classes. Using the bike meant hauling it down from my third floor walkup, remembering the helmet and the lock, then at the university end finding a place to park, locking it, and then figuring out what to do with my helmet, which was just to carry it but I was already carrying my books for class.
Since the university was on the highest ground in the area, it was an slow uphill ride to get there (and downhill all the way to get back, but I was less pressed for time). So I was actually finding, with all the fiddling, that it was actually quicker to walk by about 5-10 minutes. The helmet added maybe a couple minutes to that time, but it pushed the comparative calculation over the edge to favor walking.
I've considered biking to work, and in this case it is more the unsuitability of the road conditions (the commute would be through Manhattan). I've had enough close calls with nearly being hit by drivers as a pedestrian, almost always when cars were making rapid turns into crosswalks where I was crossing with the light, that I'm wary of actually trying to use a bicycle around these drivers.
The tax isn't the monetary cost of having to buy and maintain the helmet. It's the tax on your signaling how bold you are — I'm not one of those wimps who wears a helmet just to bike around the neighborhood! It's like a dunce cap, and clearly fewer people will want to send that signal than the bold and adventurous signal.
Let me add my (anecdotal) evidence to the mix. I live in Evanston, IL, which is nice and flat, and has bike lanes that run through it. Most bicyclists I see wear helmets, as do I. As for what to do with the helmet, I just put the lock through one of the helmet's air vents and lock it to whatever post I'm using. And my SF grandchildren couldn't wait until they were old enough to graduate to 2-wheelers and helmets. [They live in a fairly flat part of the city, west of Potrero.] So it may just vary by type of city and by the way helmets are introduced.
Besides this effect being well known, it is worth remembering that discouraging cycling is actually a positive result for the do-gooder brigade. One less cyclist is one less potentially injured cyclist, and this effect is far greater than any hypothetical protective effect of the helmet itself, which has generally been shown to be pretty negligible in competent studies.
Of course, they don't care about the consequences for health, traffic levels, increased risk to other road users….
Won't Someone Think of the Children?
You can lock a helmet to a bike — if the straps are cut it has no resale value, so there's no point stealing it. (If the lock is cut you've lost the bike anyway.)
I don't wear one because they're uncomfortable, of debatable value, and because I don't wear safety gear for most other everyday tasks (walking outside, taking a shower etc).