@RLD I have read recently that requiring special child seats for kids over 2 years old has resulted in NO benefit to those children in an accident vs. them wearing a standard seat belt. These child seats used on children over 2 years old (when the law allows seats to “face forward”) increases injuries to small children’s spine & neck. I was surprised by this and only read one article about it, it can be google searched. Unlike many studies that are funded by or lobbied heavily by those that stand to benefit from passing a law (manufacturers of child car seats) I could not determine a conflict of interest in the article But that doesn't mean it was totally objective. Another interesting article pointed out that all car accidents where alcohol is involved - does Not mean that alcohol caused the accident. It listed that even a car that is stopped at a light and “rear ended” by a sober driver - will be listed as “alcohol related” if the driver of the Stopped vehicle tested positive for alcohol. It certainly does not dispute that no one should drive drunk but that the statistics showing alcohol related accidents are somewhat inflated.
It's actually a really interesting subject, in both cases.
Full disclosure, let me say I have been a plaintiff side PI lawyer for over 30 years now, and while I only do child sex abuse cases these days, I spent more than two decades litigating MVA's and similar cases, with a speciality in product and road design litigation cases.
Child seat data is very good on reducing and preventing injuries even in the 6-8 year old range, although the real determining factor is the height of the child (4'9" so so being a nice cut off) in that the physics or vehicle/occupant dynamics of car seat vs. seat belt really do favour car seats. And there are plenty of studies (and experiments) done by pure academic groups or government only institutes that support that, inside and outside of the United States.
Alcohol use and injury (not just from car accidents as you can imagine, think violent crime, domestic abuse, falls, diving accidents) is also a very well studied area. The data is quite clear about how alcohol use causes significantly more injuries, particularly amongst males 18-21, such that we can save a lot of lives/suffering/money by trying to reduce/moderate use in that age group. The philosophical argument for why government should regulate alcohol use is longer, but I think it holds water.
In the MVA context alcohol has been studied to death, not just on the demographic side, but in the functional arena. We know that alcohol greatly slows perception/reaction time (critical for driving), and impairs judgement in a bunch of realms. For the same reason you don't want someone who is impaired operating a crane on a construction site, you don't want someone driving a car while impaired.
The point you make about how the larger epidemiological data is gathered for alcohol use and car accident cases is a good one though. In some studies if the screening is not thorough enough, the number of cases that are really alcohol related may be inflated for precisely the reasons you point out. However, in many of the studies done based on hospital records many researchers feel that the number of MVA cases related to alcohol may be underestimated because people don't tell the hospital staff they were drinking for legal/stigma reasons.