In order to show that helmets *cause* accidents, they need to create a randomized study where they force employees to flip a coin to decide whether to wear a helmet or not.
Otherwise, here's one plausible scenario: Employees who work in tiny suburbs with small roads and very little traffic feel safer, and this are less likely to wear a helmet. They get in less crashes because their town has fewer and safer drivers. Employees working in the city have more crashes simply because of being in a busy city, so they are more likely to wear helmets.
Right now, you cannot prove that helmets cause crashes and not the above.
I have some issues with that study. One major one is they basically rule out half the effects of helmets as a "logical fallacy":
> Risk compensation, as it is typically defined and understood, is only one of six possible scenarios, namely a usual non-helmet wearer puts on a helmet and increases their risk taking. Importantly, evidence in the opposite direction, i.e., taking a helmet off leads to less risky behaviour, is not evidence in support of risk compensation as it is a type of logical fallacy
After reviewing that article none of the studies are convincing either way. The only studies that actually look for causality are the ones which only measure speed to asses risk. Those are also the one that I would qualify as positive results but were listed as negative results because of that above mentioned logical fallacy.
So while is is possible to untangle these effects, it has yet to be done properly to show a clear result either way.
No, I really don't untangle anything. Safer riders ride safer, but they can and do sometimes have accidents and the road will not check your safety record before impact to see if it should hurt your more or less.
Research on helmets has been ongoing for 40 years, and has even led to ANSI standards for helmet design and protection. The UCI requires hemets in amateur and professional events. This isn't about risk taking behavior, it is simply about if you do have an accident, you won't be killed, turned into a vegetable or concussed when you hit your head.
> The effects of helmet wearing are higly contingent based on the geography and demographics.
I'm pretty sure that hitting your head on Ugandan cement will damage your head roughly the same as American Cement or European cement. S
Additionally, I've never seen any research showing any kind of demographic relationship to severity of head injuries in bicycle accidents. Nor have I once saw research that did anything other than present some statistical noise about distance cars give you based on helmet or not. Close shaves are not accidents or injuries, so even the basis of the research is questionable.
> Additionally, I've never seen any research showing any kind of demographic relationship to severity of head injuries in bicycle accidents.
The effects of a helmet on overall safety when ridden at low speed on dedicated bike paths is very different from when ridden at high speeds in traffic with no bike lane.
Thus the the design of the city and streets (geography is perhaps not the perfect term for this) and the what/how of the local culture's bike riding behavior (perhaps demographics is a bad term for this, not sure of a better one.) have huge impacts on how much a helmet affects your safety simply because the risk profiles are very different.
The data is messy due to regional variability plus the difficulty of reliably removing the confounds mentioned above. I would never discourage someone from wearing and will actively encourage it when riding in bicycle hostile areas. At the same time, I think the push for helmet laws and helmet education is often a cop out to avoid talking about how we need to redesign cities to support safe bicycling. If we did the later, we would see much larger safety gains and the former would be much less necessary.
In order to show that helmets *cause* accidents, they need to create a randomized study where they force employees to flip a coin to decide whether to wear a helmet or not.
Otherwise, here's one plausible scenario: Employees who work in tiny suburbs with small roads and very little traffic feel safer, and this are less likely to wear a helmet. They get in less crashes because their town has fewer and safer drivers. Employees working in the city have more crashes simply because of being in a busy city, so they are more likely to wear helmets.
Right now, you cannot prove that helmets cause crashes and not the above.