More cyclists are dying on American roads. Last year, 857 cyclists were killed by drivers, a 6.3 percent increase from the year prior, even as overall road fatalities fell by 2.4 percent, .

To try and reverse this morbid trend, the National Transportation Safety Board (NTSB) issued a series of recommendations. Most of them are beneficial if obvious ones: give cyclists their own road space physically separated from motor vehicles, designate space for cyclists in intersections where most crashes occur, and explore so-called “road diets” that remove travel lanes for turning lanes, bike lanes, etc.

But the NTSB issued another recommendation, one that very much breaks with conventional wisdom. They every single state and the District of Columbia pass mandatory helmet laws for all cyclists:

The investigators’ primary focus was on crash avoidance, but in those instances when crashes do occur, they said the use of a helmet was the single most effective way for riders to reduce their chances of receiving a serious head injury. Because research shows that less than half of bicyclists wear helmets and that head injuries were the leading cause of bicyclist fatalities, the NTSB recommended that all 50 states, the District of Columbia and Puerto Rico, require that all persons wear a helmet while riding a bicycle.

Currently, require that all cyclists wear helmets, although some localities have their own mandatory helmet laws.

To be clear, the NTSB has no power to mandate anyone wear a helmet, and obviously holds little direct sway over what individual state governments choose to do. But this is not only a misguided policy suggestion, but a worrying resuscitation of the were they a ? line of inquiry after any cyclist is killed by a driver, as if the answer to that question serves as Exhibit A about who was to blame.

It’s a patently absurd notion, that not wearing a helmet somehow causes a two-ton vehicle to crash into you. It does not. In fact, the NTSB was pretty clear about what is the most frequent cause of that vehicle to crash into cyclists: the people driving them.

In this sense, the helmet law bait-and-switch is classic victim blaming, the cyclist version of , as if somehow any cyclist killed by a driver was asking for or deserved to die because he or she wasn’t donning a Styrofoam dome.

That being said, helmet laws are a bad idea for a much more fundamental reason: they don’t work.

It’s worth noting the debate over helmet usage can take two distinct but related forms: should helmets be mandatory versus whether helmets are a good idea for cyclists to wear by their own choice. One is about helmet laws while the other is about cyclists making informed choices for themselves. We are not—I repeat, not—going to debate the latter here today. This is purely about why helmet laws, as the NTSB recommends, are not a good idea.

It’s also worth pointing out that despite the recent uptick in cycling deaths, cycling is still quite safe. Vox took at the helmet law issue in 2014 and, among other things, found no evidence from Australia, Great Britain, France, Netherlands, and Denmark, that cyclists suffer head injuries at a higher rate that drivers or pedestrians.

The U.S. doesn’t keep track of how many miles are biked and walked annually, so it’s difficult to do a similar analysis here. But, using some back of the envelope calculations, Vox hypothesized that biking and walking are roughly similar in terms of death rates per trip. So, on a strict public health logic, it’s hard to justify mandatory bike helmet laws without advocating for mandatory pedestrian helmet laws, too. Yes, this is precisely how ridiculous mandatory helmet laws seem to many cyclists.

Still, the argument in favor of mandatory helmet laws is a simple one: if you, a cyclist, and are hit by a vehicle, you’re more likely to survive if you’re wearing a helmet than if you are not. Another way to frame this argument is, if you could magically plop a helmet onto the head of anyone about to be struck by a car, the ones with helmets would be more likely to survive than the ones without. Undoubtedly true!

But, this is not how helmet laws work in the real world. The effect mandatory helmet laws have on society cannot be boiled down to such a simplistic if only he was wearing a helmet! counterfactual. Whether or not a helmet changes the outcome of any individual crash is different than the effect helmet laws have on cyclists and their health over an entire population, because helmet laws change all kinds of things about who bikes and where.

The core issue with helmet law arguments is they make a fundamental statistical error. The NTSB backed up their argument by looking at the percent of fatalities that involve head injuries, the percentage of cyclists who wear helmets, and therefore concluded more helmets would result in fewer fatal head injuries.

That may sound logical, but it’s not. Here’s , which tackled the same argument five years ago:

But this is only a part of the story — that data is only looking at the tiny sliver of bicycle trips that end in the hospital.

A more relevant question is how the use of helmets affects the total rate of head injuries and the overall accident rate.

Surprisingly, the data here is pretty ambiguous. Some analyses show the Australia law , but others .

“On the whole,” Vox summarized, “when large numbers of people begin wearing helmets, we really don’t see a benefit in the head injury or fatality rates.”

This is a really counter-intuitive finding, one that is still debated among public health experts. There are a couple of theories for why this seems to be the case, but the one that makes the most sense to me is that helmet laws only serve to dissuade people from riding bikes who have very little chance of getting into a fatal crash to begin with.

There’s strong evidence that requiring helmets is indisputably effective at one thing: getting fewer people to ride bikes. A classic example occurred in Australia, when . Ridership on their bikeshare programs plummeted. And bikeshare riders , partly because the bikes themselves simply can’t go that fast, but also because they tend to ride in the safest areas in the city for cycling with the most robust cycling infrastructure, and the people riding them tend to ride more cautiously.

The upshot to all this is mandatory helmet laws do a great job getting people who are very unlikely to ever get into a serious crash to never ride a bike in the first place while having no clear impact on whether the cyclists still on the road survive the crashes they still get into.

Again, none of this is an argument why any specific individual should go baldy. By all means, if you prefer riding with a helmet, do so! Have your kid wear them, whatever floats your boat. This is about freedom, baby, the freedom to ride however you prefer.

It’s far from clear why states should spend their time and resources passing and enforcing helmet laws when there are much clearer and effective measures to not only save lives but encourage more cycling, , rather than discouraging it. The NTSB hit on many of those in their recommendations—protected bike lanes is the most obvious one—so it’s a shame they felt the need to resuscitate what seemed to be a dead debate about mandatory statewide helmet laws.

Perhaps inadvertently, helmet laws give police, drivers, lawmakers, and the press a template for blaming a cyclist for his or her own demise. In nearly every cyclist death, the driver is unharmed, alive to tell their side of how things unfolded, which often involves some story about how they never even saw the cyclist or they “came out of nowhere.”

Rarely do drivers admit to doing anything illegal, such as looking at their phone or making an illegal maneuver, at the time of the crash. But should helmet laws get passed, and the cyclist wasn’t wearing a helmet, then they were breaking the law. The narrative is set. And, what do you know, the victim is to blame. They should have been wearing a helmet. After all, it’s the law.