NBER Paper is misleading
Summary: Two authors have produced a working paper that implies that bicycle helmet laws are reducing cycling.
The authors do not show that helmet laws are actually the cause of any of the effects shown in the data.
The National Bureau of Economic Research, a private nonprofit organization based in Cambridge, Massachusetts, has
released to the press a working paper, not yet peer reviewed, with what we regard as erroneous conclusions about what
injury data show in helmet law states.
The paper's abstract says: "Cycling is popular among children, but results in thousands of injuries annually. In recent
years, many states and localities have enacted bicycle helmet laws. We examine direct and indirect effects of these laws
on injuries. Using hospital-level panel data and triple difference models, we find helmet laws are associated with
reductions in bicycle-related head injuries among children. However, laws also are associated with decreases in non-head
cycling injuries, as well as increases in head injuries from other wheeled sports. Thus, the observed reduction in
bicycle-related head injuries may be due to reductions in bicycle riding induced by the laws."
The key words in the abstract are "associated with." But the misleading part of the abstract goes on to say injury
reductions
may be due to reductions in bicycle riding induced by helmet laws. The authors fail to make a case for
that, relying on reported skateboard, skate and scooter injuries as their proof. In fact, after exhaustive statistical
analysis, the conclusions of the paper all rest on their assumptions about interpreting the relationships of the
data.
Similar studies have been done for fatalities, for which there is much more reliable data, but this one attempted to
analyze hospital-reported injuries. The results can be interpreted many ways. Although the authors took great care to
clean up their data, they included many factors in their models that have questionable effect on crashes and injuries:
unemployment rates, per capita income and others. Once the model was constructed they seemed to believe that it was
showing the correct conclusion. the do note in one place: "Given this, it is not clear whether an effect of the law
reflects the effects of the actual law itself or whether it reflects any education, public campaigns, or attitudes
towards risk that accompany the helmet laws." We would agree with that statement, and question why they came to believe
that somehow the data indicate that helmets discourage riding, a conclusion not supported by any on-the-ground
observational study anywhere in the US.
The authors found as others have that there was a reduction in head injuries following helmet laws. But injury statistics
can not establish causality. There are many factors unaccounted for that also influence injury rates. The effectiveness
of the laws should be first measured by changes in helmet use rates. If that does not change, any changes observed in
data would be from other factors.
The authors seem unaware of the changes to bicycle infrastructure in urban areas over the past decade. One visit to a
conference of bicycle-pedestrian planners would have alerted them to the efforts being made to make cycling and walking
safer in our cities since about the turn of the century. Given the number of riders not using helmets, the results of
those efforts could swamp the helmet effect. They also seem unaware that without any laws requiring ski helmets, the
National Ski Areas Association reports that over half of the skiers at their resorts are wearing helmets. The study found
no change in ski injury rates.