Well, maybe like 1% of deaths if you include suicides as "gun violence" (kinda weird, but is often stated), maybe 2% if you focus on only men.

But yea, it's a literal nonfactor in aggregate life expectancy. Amdahl's law applies here too.

According to [1], "homicide, suicide and gun deaths" is the 2nd highest factor in explaining the difference.

[1] https://twitter.com/jburnmurdoch/status/1641799922583326720/...

I can't read the underlying article, but that graph a minimum is not saying "2nd highest factor". It notes both drugs (lowers gap to 2.9) and metabolic diseases (lowers gap to 2.2) as both more of a factor than suicide & homicide (only lowers gap to 3.2).

Which tracks, given that suicide is like 50k per year, homicide is 20k per year (unstated: "gun deaths" that aren't homicide or suicide are orders of magnitude lower), drug deaths are over 100k per year, and metabolic diseases dwarf all of the above.

So, no, what you wrote is wrong.

Oh I've read it the wrong way around! Yes you are correct!

Also it's unclear if there are other factors worth tracking, since the graph only lists a few.

Heart disease, cancer, stroke, respiratory disease (not including covid), accidents (which I suppose includes drug OD), Alzheimer's, and diabetes all have higher death rates that the combination of all suicide+homicide (which again, is sort of strange to join together). The first two alone are 600k+ each.

A bunch of those affect primarily the elderly though, so won't hit the average life expectancy as hard.

If you track homicide separately, it's not even in the top 20 of causes of death. Suicide is not in the top 10.

That’s per annum. The rate decreases by age, but over a lifetime the cumulative odds start to add up to a more significant number.

Also, by killing men in their twenties it affects the average more than something like heart disease that usually kills men over 50.

> it affects the average more than something like heart disease that usually kills men over 50.

If you look at the linked Twitter thing in the other comment chain here, you will note that this is not true.

The numbers are just very different - heart disease (~600k) is like 40x more than all homicide (~15k, varies by year). Even if heart disease only kills people the year or two before they would otherwise die in their 70s (which is not true, it kills a lot of 50-yr-olds too), it would affect the average less than homicide.