Wars, Massacres and Atrocities of the Twentieth Century

Year-by-Year Death Toll:

Overview of Twentieth Century Wars, Massacres and Atrocities:

  1. Grand Total
  2. Magnitude
  3. Intensity
  4. Propaganda
  5. Exaggeration

Grand Total:

Well, what can you say about a century that begins and ends with killing in Sarajevo? "Good riddance" springs to mind. Somewhere around 180 million people have been killed in one Twentieth Century atrocity or another -- a far larger total than for any other century in human history.

Now before you let this number wash over you as being too big to comprehend, let's put it in perspective: Let's say that you're the receptionist in the Afterlife (a 9 to 5 job, 5 days a week, with two weeks vacation -- which comes to 40 hours per week and 50 weeks per year), and it's your job to simply ask the name of each victim, enter it into a computer and direct them to Room 504 for processing (a task that takes 5 seconds, which means that you can process 720 per hour), and these 180 million people were to approach your desk one after another without letup. At this rate, it would take you only one hour to decide that this is a really depressing job, and you would have been better off working as a checkout clerk at the Food Lion instead.


By my calculation, there have been 165 wars or tyrannies of the 20th Century which have killed more than 6,000 people. Five of these events claimed more than 6 million victims. Twenty-one events claimed between 600,000 and 6 million lives. Sixty-one events claimed between 60 and 600 thousand, and seventy-eight events killed between 6 and 60 thousand.

Of course, all these numbers are subject to a wide margin of error and a rancorous debate. Wars are messy, and tens of thousands of people can easily disappear without a trace. The worst atrocities take place in the dark, unseen and unrecorded. Estimated death tolls can therefore vary wildly, spanning several orders of magnitude at once. It's not at all unusual to have an upper estimate be two, maybe three, times the lower estimate. To arrive at the numbers illustrated on the maps, however, I have tried to be extremely mainstream. I tried to find the most commonly quoted body count, and then I rounded the number to the nearest triangular number.


Before we get carried away condemning the century as a whole, we should keep in mind that the enormous body count has come about largely because there are so many more people available to kill. For example, the St. Bartholemew's Day Massacre in France in 1575 killed some 50,000 people, which, by 20th Century standards, is hardly enough to rate a place on these maps; however, considering that there were only 15 million Frenchmen at the time, this massacre would be the equivalent of 800,000 modern Americans -- a very frightening number indeed.

I calculate that somewhere between 4 and 5 percent of all human deaths in the Twentieth Century (or something like one in 22) were overtly caused by other people. The "Deaths per Million" maps should illustrate this concept on a nation by nation basis. By "Deaths", I mean all deaths caused by political violence. They run the gamut from terrorist bombings to executed dissidents to battlefield casualties to starvation among refugees to hard labor in concentration camps. I have tallied them where they happened, regardless of who killed whom. Americans killed in Vietnam are counted in the Vietnamese totals, while Israeli athletes killed in Germany by Palestinian terrorists are counted in the German totals. The Magnitude maps shown the body count from specific events which are united by cause and participation but which often spread into several different countries. The Intensity maps often include many separate events which are united merely by place and period.

"Per Million" refers to the population of the country at the midpoint of the period, not per million participants in the event. Also, if you haven't noticed already, the highest category (10,000 deaths per 1,000,000 people) equals one percent.

Percentage of national populations killed in specific episodes of mass brutality:

(click chart for more detail)


Keep in mind that it's customary to manipulate these numbers for political gain. Obviously, most of this manipulation falls into the basic accusation and denial categories. The victims will shout huge numbers which emphasize their suffering, while the accused shout lower numbers to emphasize their restraint.

Less obvious is the manipulation of the numbers in order to set an example. There are plenty of social activities which are fine in moderation (such as free enterprise) but can have rather unpleasant consequences when taken to extremes (such as slavery). In fact, just about any atrocity you name can be seen as an example of something normally harmless taken too far. For example, a little bigotry (the belief that "we" are special) is considered good when we call it patriotism, ethnic pride or close family ties, but too much bigotry is what led to the Holocaust. Christianity and atheism are both fine in small doses -- I've never know a Christian who wasn't improved by a little doubt, nor an atheist who wasn't improved by a little faith -- but with too much of either, you'll end up with crusades, witch hunts and gulags.

The problem with atrocities -- aside from the fact that millions of people die in them -- is that there's nothing you can say about them in polite company. Once you've condemned them, what next? Learn a lesson from them? Maybe so; sometimes a historical horror story can be a morality tale which spurs the people to action -- it was no coincidence that the American generation that saw the Holocaust came home and freed its own minorities -- but your opponent is not going to appreciate being compared to a mass murderer when he hasn't killed anyone. Since every ideology can lead to unpleasant extremes, he can always find a way to retaliate in kind, and pretty soon we've got one of those "I'll see your Cambodian Massacre and raise you one Trans-Atlantic Slave Trade!" arguments going. Although we all know that supporting environmental regulations doesn't make someone a Commie any more than opposing affirmative action makes someone a Nazi, we still feel threatened whenever anyone starts waving the bloody shirt, so we either try to snatch the shirt away ("oh come on; it wasn't that bad.") or we wave a shirt of our own ("at least it wasn't as bad as ..."). And if we shrug and say, "So what? That's not my shirt," then we'll be accused of being uncaring.


Let's face it. Big numbers are impressive, and there's a very real human tendency to exaggerate. While lurking on Usenet discussions of historical atrocities, I've seen plenty of outrageously large claims (accusations of Pol Pot killing 10M, which is more than the number of Cambodians living at the time; accusations of international Communism killing "billions", which comes close to being more than the number of people who have ever lived under Communism.) and far fewer outrageously low ones. Even the most infamous exceptions to this tendency, the Holocaust Deniers, shouldn't blind us to the fact that exaggeration (slight exaggeration) occurs in the Holocaust numbers. Most scholars put the death toll between 5 and 6 million, but most popular accounts emphasize the high end alone -- 6 million. [n.1]

Generally speaking, I've seen more numbers shrink than grow under critical scrutiny. The tally for the Spanish Civil War began at a million and dropped to less than half that once someone looked into it. The Persian Gulf War gets less and less bloody with each new study. The riots over India's partition took a million lives in the earliest reports, but now are considered to have cost about a quarter of that. This is not to say that some death tolls haven't stood their ground -- the Holocaust killed just about as many Jews today as it did in the Nuremberg indictment. Also, there have been plenty of counts that have gone up -- Stalin's body count is higher today than it was when he was alive -- but here again, the counts have been going down from the peak estimates of the Cold War Era.

I hate to sound callous, but humans are tougher to kill than they get credit for. Both the Titanic and the Holocaust are considered synonymous with disaster, but in both cases, nearly a third of the at-risk population survived what seemed like hopelss situation. Two thirds of the passengers and crew of the Hindenburg survived. At Gettysburg, 150,000 angry men hammered each other viciously for three days, and when it was all done, 95% were still alive. Basically, you should be careful in accepting any unsubstantiated estimate which seems to deny human tenacity to survive.

Detailed Essays:

Contemporary Context:



Even relatively impartial scholars have three types of built-in bias which nudge them toward high numbers rather than low.

  1. No scholar, regardless of his field, ever minimizes the importance of his research. Whether it's AIDS, Chaucer, beetles or genocide, he will always emphasize that his subject is a big deal. He certainly is not going to dismiss AIDS as a minor disease or Chaucer as an overrated scribbler. Similarly, an atrocitologist goes hunting for data which proves how common genocide is, rather than how rare it is. When faced with a debatable, borderline case of genocide (such as whether a famine was caused by bad weather, human mismanagement or deliberate oppression -- or whether an air raid was a necessary act of war or a spiteful act of vengence), he's more inclined to accept the answer that brings it into his field of study, rather than the answer that removes it from his consideration.
  2. Most scholars prefer to err on the side of caution, but in atrocitology, "the side of caution" is actually found up among the bigger numbers. After all, the danger is not that the scholar will be shunned by decent folk for slandering the good name of Pol Pot or Idi Amin by falsely accusing them of killing twice or three times the number that they actually did kill. No, the real danger is that old skeletons will surface in some Third World hellhole, and that all the scholars who had earlier minimized the number of victims will be branded deniers, dupes and apologists. Therefore, academic reputations are best preserved by anticipating the worst right from the start. (hint: I've done this with my estimates for the number of deaths under Mao Zedong.)
  3. It's more psychologically rewarding to prosecute evil rather than defend it, and no normal person wants to belittle human suffering. (For example, do you think I like saying things like "No, Stalin only killed 20 million -- 30 million, tops"? "Only"!? Usually, 20 million murders is considered a bad thing, but in some circles, blaming him for 20 million deaths comes across as downright pro-Stalin. Do you think I like being accused of defending Stalin simply because I don't immediately accept the higher estimates?)


to Table of Contents

Last updated November 1999

Copyright © 1999 Matthew White