> And war (for Americans and the rest of the world) is a lot less deadly than it used to be, in large part because it got more professionalized.

This seems plausible, but I wonder if what you calculated is the best way to measure it.

First, since you mentioned "Americans and the rest of the world", does the same trend hold for the people on the other side of these conflicts? At least in the past 20 years' examples, the U.S. military wasn't fighting adversaries who were equipped nearly as well as it was.

Second, the number of people who fought in each of these wars is also pretty different. Isn't a more direct comparison something more like "deaths per combatant per month" rather than "deaths per month"? (I don't know how that comparison would come out.)

    First, since you mentioned "Americans and the rest of the 
    world", does the same trend hold for the people on the other 
    side of these conflicts?

Yeah and no. It's so hard to measure the human cost of war because so many of the deaths are indirect.

Operation Rolling Thunder (50 years ago) was probably the last time the US took mass numbers of lives directly, with the high end of estimates approaching something like an average of 5,000/month.

Estimates of deaths from the Iraq War range from 150K "violent deaths" to 1mil "excess deaths" which would include all sorts of indirect causes: https://en.wikipedia.org/wiki/Casualties_of_the_Iraq_War

Those are horrific levels of loss but nowhere near the toll of WWII.

    Second, the number of people who fought in each of these    
    wars is also pretty different. Isn't a more direct 
    comparison something more like "deaths per combatant per 
    month" rather than "deaths per month"? 

It depends. Do we want to measure how risky it is to be a combatant, or how risky it is to be a human being on Earth?

If it's the former, then "deaths per combatant per month" is what we want. That would be interesting, though I question how broadly useful it would be. It would obviously be directly useful for combatants (or those considering becoming combatants) but is probably not useful as an overall indicator of "hey, how's the human race doing in general?"

If it's the latter, the stat we probably want is "war-related deaths per human being per month" or something like that. That would be a good indicator of overall levels of war and strife. It would take into account the increased population of the world as well.

Even if (heaven forbid) we had a conflict that results in WWII-like absolute levels of deaths today, that would represent a far lower number of deaths as a percentage of the human population lost.

On the first question, there is some scholarly debate, but as far as I can tell as an outsider the general consensus is that there are fewer deaths due to armed conflict than anytime in the last two millennia. Even something as terrible as the Syrian Civil War (~half a million dead over a decade) is the sort of civil unrest that was happening at multiple places across the world during the 17th century (see _Global Crisis_ by Geoffrey Parker for how horrible the 17th century was, all around the world). It's just the numbers are a lot less precise than we have for American deaths in military operations, so there is a lot more room for debate. This can be complicated even for something well documented and recent like World War Two: do you count the Bengal famine as part of the war dead? It killed ~2 million Indians, but not directly due to the war, it's just that the ships that in peacetime would have been carrying food to India to relieve the famine were instead carrying weapons (and food) from the US to Britain. Did they die due to the war? Or due to British colonialism and evil Winston Churchill? Well, a whole lot of the deaths in 17th century wars were due to war-related famine and disease, so if you don't count the Bengal Famine but do use the higher numbers for the 30 Years War it isn't a clean comparison, etc.

On the second question it's even harder to say. Would probably be a masters degree level of work to compute. There are all sorts of confounding factors: during World War Two, for the US Army, infantry made up about 20% of the personnel (half were in logistics, various other combat arms like artillery, tankers, engineers etc. made up the balance) but did the vast majority of the dying: something like 80% of US Army deaths came from that 20% of the force. So is the combatant count just infantry (and the small number of other combat arms who were forward of the gun line)? Similarly, about half the total US deaths during WW2 came during the Northwest Europe campaign (D-Day to V-E Day, France-Low Countries-Germany). The other half of the deaths came from the entire Battle of the Atlantic, the Bomber Offensive, Mediterranean and Pacific combined. So those 11 months were much deadlier than the rest.

This is the advantage of the straightforward monthly average, it can be computed in five minutes from Wikipedia. Trying to get more detail quickly becomes a whole bunch of tricky choices that reasonable people can disagree on, but which leave you with wildly different numbers.

> First, since you mentioned "Americans and the rest of the world"

Given that we’re discussing American test pilots/fighter pilots, this makes sense.

> And war (for Americans and the rest of the world) is a lot less deadly than it used to be

My hunch for wars getting less deadly is because of nuclear weapons. There hasn't been a direct war between superpowers for the better part of a century. This sounds good, except it's possible we'll essentially wipe out the planet overnight because of a technical error.

https://en.wikipedia.org/wiki/1983_Soviet_nuclear_false_alar...

I think we need a longer timespan to say whether or not wars are getting deadlier.

Fatalities have gone down even for wars of similar intensities. More important is the improvement in battlefield medicine and body armor - the KIA/WIA ratio for the US military has plummeted, from around 1:1.5 during WW2, to around 1:3 in Korea after the introduction of helicopter evacuation and more refined frontline aid stations, to just above 1:10 in Afghanistan.

e.g. I strongly doubt that Tammy Duckworth's wounds would have been survivable on the battlefield in 1940, or even in 1970.

>I suspect that it's more a generational thing than an occupational one. I've met a couple of test pilots from the '90's and '00's and they were the some of the most by-the-book people I've ever met, they were meticulous and followed their steps precisely, even for quotidian things like making breakfast or pitching a tent.

The people who were test pilots in the 90s were people who succeeded under a system of massive "by the book"-ness.

The cowboys of the 60s don't wind up being successful test pilots in that kind of system.

The organizations changed so the people who succeed in them changed.

And test pilots regularly died in the old system, and come home to their children in the current one, so I prefer the current one.

The point is that it isn't "they were test pilots, so crazy" it is "in the 1950's and 1960's a lot more death was accepted and because of that, the standards were lower."

In the 60s you needed some crazy cowboy to try and fly a brick strapped to a gas tank. Today you simulate what happens on a Dell laptop and that’s that.

There are old pilots and there are bold pilots, but there are no old, bold pilots. Except until recently, Yeager, who passed late last year, RIP.

That would be more to computer simulation before the aircraft ever takes it first flight. Test pilots from the in the 50s and 60s did not have that option.

Those numbers are fascinating in their own right, but even more so when compared against the staggering 35,000+ deaths per month the US is suffering right now.

(467k over ~13 months of pandemic, obviously it's not evenly spread but the point is it's an order of magnitude greater than the worst ever war death rates.)

You make a very valid point about casualty rates, though I am just about certain that Jim Lovell would have been eager to go regardless.

Getting back to casualty rates, I think you have to at least put them on a log scale to get a feel for their impact on society, at least as long as they are less than the survivorship rate.

The risk isn't just from being a test pilot, there is a line in "The Right Stuff" pointing out that the chances of a Navy pilot being killed in peacetime was 1 in 4 over their career.