26

I read this claim somewhere recently, and a google search shows that it's occasionally made around the web, but was perhaps more common in the 1970's (many of the stories I find online are people reminiscing over old protests).

At first glance it seems probable to me: wars historically haven't killed all that many people (mostly because there were not all that many to be killed). In the US alone, wikipedia gives a figure of about 40,000 deaths due to motor vehicle accidents this year. Assuming that the US death rate is a reasonable estimator for the world (assuming poor countries have fewer cars, but more accidents), I get the total figure should be 40,000*20 = 800,000 this year alone. The accident rates were actually higher in the past, although population was lower, so my first guess is that we had ~10 million deaths in the last 15 years, and probably another 15 million or so in the 10-15 years before that. My guess is that overall deaths are probably something like 45 million, but this might be off by +/- 15 million.

My guestimate for war is that WWI and WWII combined saw about 40,000,000 deaths, so that gets us into the same ballpark.

What I'd really like to see is an authoritative source on the total figures for both types of death, and maybe something on the origin of the claim. Can't seem to find either. Might also be interesting to see if/when this will become true.

5
  • Hmm. A little more reading suggests that my estimates for WWI and WWII deaths is too low. I was looking at the figure for combatants only. Mar 6, 2014 at 16:38
  • 10
    Wars have been around much longer than cars. A more interesting question (to me) would be whether cars currently kill more people than wars. Mar 9, 2014 at 19:47
  • @Konrad: Agreed. Both answers have tried to address that as well :) Good thread! Would be interesting to get more information specific to regions with unusual circumstances, e.g., a comparison of war vs. traffic death rates in current or recent armed conflict zones. Both road safety and military violence vary drastically across the world. Mar 10, 2014 at 0:06
  • 2
    @KonradRudolph Did a little searching. The report at isn.ethz.ch/Digital-Library/Articles/Detail/?id=156350 indicates that battle deaths per year are on the order of hundreds of thousands. The WHO report cited in the answers below suggests 1.2 million deaths due to automobiles. Mar 10, 2014 at 0:49
  • 1
    @KonradRudolph Of course, civilian deaths can be much higher, but as long as there's no ethnic cleansing or the like taking place, it seems like a safe bet that there are fewer civilian + military deaths than deaths caused by automobiles, in a given, modern, year. Mar 10, 2014 at 0:50

2 Answers 2

36

I did some more research on this tonight, and was able to answer my own question, at least to some extent.

The conclusions seem to be:

  1. For the world as a whole, this is a very unlikely claim.
  2. For the United States of America in particular, this is likely true, at least for the deaths of Americans in wars since the country was founded. This is possibly the origin of the claim, at least in my region, which I`ve often seen made by Americans (i.e. More Americans have been killed by cars than by all wars combined becomes More People have been killed by cars than by all wars combined.) The claim is also likely to be true in many other individual countries.
  3. For this claim to become true for the world in general in the future, the bulk of the world's population would have to be at peace for a very long time, and average car safety would have to decline.

Evidence against the worldwide claim:

  1. This similar question has an answer which puts a lower bound on the number of humans killed in war in the 20th century at an estimate 200 million.

  2. A reference cited by this quite interesting WHO report (page 33), puts the cumulative total number of deaths by automobile at 25 million, conservatively, in 1997. Unfortunately I can't get access to the reference directly to see what methodology was used.

  3. A recent WHO report estimates the number of fatalities due to automobiles in 2013 at 1.2 million (worldwide).

  4. If we made a very generous estimate based on 2 and 3, doubling the conservative estimate of 25 million by 1997 to 50 million, and assume that 2 million a year were killed 1997-2014, rather than the fewer than 1.2 million that seems probable, we still get less than 90 million total deaths due to automobiles. This is less than the number of people killed in all wars, worldwide, in the 20th century alone, and in fact, is less than half the number required.

Evidence for the claim being true in America:

  1. Assume that the original claim was more people have been killed by cars than by warfare, in the history of the country of the United States of America.

  2. This table on Wikipedia (which seems to be sourced from the US government's NHTSA FARS database) gives a list of recorded US traffic fatalities going back to 1899. The total number of deaths is 3,581,306 (my computation via a spreadsheet).

  3. This table on Wikipedia gives the total number of American military deaths as 1,321,612, though the estimates come from a lot of different sources. Major deployments that were not "proper" wars are also included.

  4. The only major conflict in which US civilians died in large numbers (to my knowledge) is the US civil war. There is considerable debate on how many civilians died in total, and no official records were kept at the time. Wikipedia (4th paragraph) suggests the number is unknown, though it cites a poor source. However, I cannot find anything better than wild speculation on enthusiast websites beyond this. The highest estimate I see anywhere is around 800,000 dead, mostly slaves. This would have been nearly 25% of the enslaved population, and about equal to the total number of soldiers that died. and seems an overly high estimate. If anyone has better data, I'd really like to see it!

  5. In recent wars in Iraq and Afghanistan, US contractors have been killed in fairly substantial numbers. However, Reuters puts the total number of contractors killed in Iraq by 2007, who were also US citizens, at just 224. I cannot find a similar number for Afghanistan, but this at least gives us a ballpark.

  6. In total then, I estimate the number of American civilians and military personal who have died as a direct result of warfare at no more than about 2.5 million. This is less than the total number of Americans killed in motor vehicle accidents, for which we have very good data.

Evidence that the worldwide claim is unlikely to be true in future:

  1. At current rates, we'd need a century of world peace for cars to catch up to the 20th century death toll alone. This seems very unlikely.

  2. The current tread is for car safety to improve over time. With inventions like self-driving cars, this tread may even accelerate. Therefore, it is likely that more than a century of world peace is needed.

Final notes:

  1. It is quite likely that in other countries that have not seen large scale total wars fought on their soil, the claim is true as well. Most of the 20th century death toll is concentrated in specific countries, especially Russia, China, and Germany. Countries that remained neutral during major conflicts like Holland, Sweden, or Switzerland, are probable candidates.

  2. Air pollution from vehicles is also worth considering as a possible factor. This MIT study suggests the annual deaths in the USA from air pollution exceed 50,000 in the US. If these are counted, then the total number of deaths due to automobiles in the USA is markedly higher, but still unlikely to exceed 8 million (emissions related deaths were probably lower in the past because there were fewer people and fewer cars, and cars have only been around for about 100 years in significant numbers). This factor is not likely to significantly change the answer for any country. It approximately doubles the rate at which cars kill people, but in countries with total wars fought on their soil, or with large scale ethnic cleansing or genocides that were part of wars, the death tolls from wars still dwarf the tolls from cars.

4
  • 2
    This is simply splendid! It also answers a question I had about whether our yearly death rate surpasses 20th century average.
    – user85190
    Jan 15, 2015 at 7:40
  • you forget one thing: if the number of cars goes up more rapidly than the number of people engaging in war, the gap is likely to close. Otherwise, great work
    – jwenting
    Nov 27, 2019 at 4:45
  • @jwenting I think that's potentially true, but as I mentioned in point 2 of my evidence for the third claim, the current trend is for car safety to get better over time. As you can see here (en.wikipedia.org/wiki/…) , even though US population is increasing, and vehicle miles traveled (VMTs) is increasing even faster, total fatalities are actually dropping. Nov 27, 2019 at 14:19
  • You must also note the car/people ratio and how much people use cars to commute instead of public transport. Both numbers can be hight in the US, lower in the EU and even lower in most of the world.
    – jean
    Nov 29, 2019 at 20:03
18

"Wars haven't killed all that many people," eh? I think you want this Wikipedia page on that topic. Simply summing all these estimates (using lower bounds where ranges appear) gives 344,998,537 as the death toll of this "very incomplete list of wars (an estimated 3,010 wars have taken place in recorded human history and 438 are recorded here).[1]" Plenty of these numbers are questionable, but I won't make a point of questioning them myself; hence I've used lower-bound estimates.

Compare a comically inaccurate upper-bound estimate of automobile accidents based on the World Health Organization's "Global status report on road safety 2013" and the year of "the first recorded automobile death", 1869. Even if 2010's estimated 1.24M automobile deaths was a decent estimate of the average annual death toll over those 145 years (and it's not – it's obviously way too high), that would still fall far short of the war toll over the much longer history of war (179.8M). Automobiles simply haven't been around for anywhere near as long as war.

I also wouldn't be so sure that the USA is a good representative of the rest of the world in terms of traffic-related death rates. I agree an authoritative source on either figure would be great, but this is not that kind of answer, and I wouldn't hold my breath for one. I've tried briefly to find a grand total of automobile deaths, but haven't been able myself. That much seems feasible, but with war deaths, not so much.

The Wikipedia page on direct conflict deaths gives a good intro to the problems with recording war deaths. It also offers a couple useful statistics for annual death rates due to war between 2004–2007: violent deaths due directly to war ≈ 52K; due indirectly to factors caused by war ≈ 200K (Geneva Declaration Secretariat, 2008). This is much lower than the automobile death rate in 2010, but it's an indirect comparison.

I see you just posted your own answer on very similar information...

Reference

Global burden of armed violence. Geneva Declaration Secretariat, 2008. Retrieved from http://www.genevadeclaration.org/fileadmin/docs/Global-Burden-of-Armed-Violence-full-report.pdf.

3
  • Thanks for the answer. Can you provide some more details on why the WHO`s report provides a comically inaccurate upper bound? Mar 7, 2014 at 4:06
  • 2
    The "comical inaccuracy" is my doing, not WHO's: for the sake of argument, I used the absurd assumption that the WHO's estimate for 2010 could represent the auto deaths rate for every single year since the first recorded fatal auto accident. Of course, this death rate does not actually apply to the many decades in that 145 year period during which there were fewer than one million automobiles, let alone accidents. The point was that, even when overestimating this much, we still can't approach the death toll of all wars combined – not even an underestimate of it. Mar 7, 2014 at 4:15
  • Ah, I see. I read to hastily the first time through. Mar 7, 2014 at 4:20

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .