How many injuries (some likely to be very serious) are prevented or otherwise mitigated by car seats?
I also seem to recall some old story about the introduction of helmets to combat and the "surprising" result of injury increasing (but can't find it now). As in of course injuries increase when the previous result would have been death.
This is definitely possible - we've only looked at deaths.
One point to note about the 57 deaths number. There's also an argument to be made that you should apply the same burden of proof to claims that car seats reduce fertility, and that car seats reduce car crash fatalities. In other words, some people are skeptical of our methodology, which is totally fair. But if you subject the hypothesis that child car seats reduce car crash deaths to the same level of skeptical prior, the evidence in favor of saving lives is much, much weaker than the evidence on fertility, in our opinion.
The 57 deaths is our best point estimate from the regressions, but in the vast majority of specifications you can't rule out an effect of zero. By contrast, we find pretty strong evidence on the fertility reduction, that we think is fairly tightly identified compared with other competing explanations.
https://medium.com/@penguinpress/an-excerpt-from-how-not-to-...
How many injuries (some likely to be very serious) are prevented or otherwise mitigated by car seats?
I also seem to recall some old story about the introduction of helmets to combat and the "surprising" result of injury increasing (but can't find it now). As in of course injuries increase when the previous result would have been death.