[nooneelse]

> "...if you adjust for accidents and homicide, two things that aren't affected by healthcare."

Healthcare is quite involved in dealing with the effects of accidents on humans. So I would think that death rate from accidents, broadly speaking, is not independent from level of healthcare. Or is that a technical term that already distinguishes between those that survive the accident long enough to possibly receive treatment and those that don't?

[anamax]

> So I would think that death rate from accidents, broadly speaking, is not independent from level of healthcare.

There are two numbers - the accident rate, which is independent of healthcare, and the "how good was treatment for accident victims" number, which does depend on healthcare. Their product is the death rate from accidents.

Here are two populations that demonstrate these effects.

In population 1, everyone has an accident each year and 90% survive. That's 10% deaths due to accidents.

In population 2, 10% of the population has an accident, and only 50% of accident victims survive. That's 5% deaths due to accidents.

If the accidents are the same, it's silly to argue that population 2 has better "accident healthcare" even though its death rate due to accidents is half of population 1's.

If the accidents are different, it's impossible to say which population has better accident healthcare without more information.

Here's another way to look at it. The death rate for automobile accidents in 2000 in the US was significantly higher than the death rate for automobile accidents in 1850. Does that imply that 1850 accident healthcare was better?