I sort of wonder if that is heavily dependent on country or area, I mean drunk pedestrian walking in Los Angeles might be more dangerous than in Copenhagen.

Kind of anadotical but the only time I have been around someone who got run into in Copenhagen they were visible drunk. Though I don't have any stats I can pull up.

Perhaps reading Super Freakonomics?

Rehashed here: https://freakonomics.com/2011/12/28/the-perils-of-drunk-walk... Google will provide people who don't like assumptions made in Levitt's math.

Disclaimer -- I did research for Levitt and John List when I was planning to go for an Econ PhD, but on nothing related to drunk driving.

Likely, but it's been long enough that I don't remember.

I'm fairly certain these exact statistics are spoken and quoted in Super Freakonomics, or maybe the first Freakanomics.