29 December 2011

Why RJ Rummel Shouldn't Be Taken Seriously


RJ Rummel is a political scientist best known for his contribution to the "democratic peace theory". He is also the author of the book Death by Government where he claims that states in the 20th century were responsible for killing around 174 million people. Interestingly this figure is frequently quoted by libertarians who are dismissive of the rest of his output; his "democratic peace" and his advocacy of aggressive foreign policy. It can be quickly shown his work on "calculating" the number of victims of governments calls for the same treatment.

We can ignore the fact his "most probable estimate" of deaths the responsibility of the Soviet Union presupposes 40 million deaths in the gulag, which happens to be twice the number of people who passed through the camps. We can also ignore his "high estimate" of the number of people killed by the USSR presupposes the murder of a fantastic 115 million Soviet citizens — a number of deaths sufficient to demographically break a far more populous nation than the Soviet Union (148 million people in 1926). Ignore also his having "calculated" that Soviet Union extinguished 2.38 million lives in the gulag between 1961 and 1982, and an additional 200,000 lives between 1983 and 1987. Ignore, because there is an even better way to showcase the level of his scholarship than debunking any of these. Better because it is made by Rummel himself.

In Death by Government Rummel claimed 174 million killed by government, but has since revised the figure to 262 million. This is partly the outcome of his changing his estimate of the number of people killed by colonial regimes from 870,000 to 50,870,000! Rummel explains this revision became necessary when, having read a book on Belgian colonialism in Congo, he realized he had been ignorant of a huge state-caused loss of life in that African territory. That seems like an honest explanation, but should he really be left of the hook this easily?

To begin with, the enormous loss of life in colonial Congo is hardly new (or obscure) information that was not available before. Rummel himself gives examples of sources (including Encyclopedia Britannica) that speak about it and were published well before Death by Government:
  • "Britannica, 'Congo Free State' claims that the population declined from 20 or 30 million to 8 million."
  • "A 1904 report by Roger Casement’s estimated that as many as 3 million Congolese died since 1888 (cited in Gilbert’s History of the Twentieth Century; also in Colin Legum, Congo Disaster (1972)."
  • "Peter Forbath (The River Congo (1977) claims that at least 5 million killed."
  • "John Gunther (Inside Africa (1953) estimates 5-8 million deaths."
  • "Fredric Wertham, A Sign For Cain: A Exploration of Human Violence (1966) estimates that the population of the Congo dropped from 30M to 8.5M, a loss of 21.5 million."
    Also, it is not the case Rummel found 50 million victims he was not previously aware of in Congo. No, it is the case that having read a book on one specific colonial experience he became aware of the level of atrocity colonialism rested on and so he now figures the number of victims in Congo and elsewhere could add up to something like 50,870,000. This means he could have easily picked up a study on any number of brutal colonial regimes besides the one in Belgian Congo and it could have had the same effect of convincing him his figure of deaths caused by Western colonialism was off.

    In other words, Rummel admits that at the time of writing Death by Government he was ignorant of the true nature of colonialism, but in a manner of a dilettante gave the 870,000 figure anyway. Professor Rummel can be commended that having learned something new he admitted to his delusion and revised his claim. However, if an author is going to make errors of this magnitude he probably should not be merely "reevaluating" estimates, he should be reevaluating whether he has any business writing books on the topic.

    19 comments:

    1. Huh. I guess you don't have to actually know anything any more to write a book on just about anything you choose.

      My wife keeps pushing me to give it a try myself. Perhaps I will, considering that it doesn't seem to be much work these days at all.

      ReplyDelete
      Replies
      1. Sounds like it. But I think it only works that way if you have a PhD in hot air, uhm I mean political science. ;)

        Delete
      2. This comment has been removed by the author.

        Delete
      3. Found this post through google, not sure why considering it's immensely useless and written by some unknown blogger. But you have a really solid theory, don't you? Yes, lets disregard everything Rummel has to say because he didn't include a set of statistics, the inclusion of which, actually strengthens his overall premise. He was right before he added numbers from the Congo, and he was even more validated after he added them. Clearly, time to disregard everything the man has written, my eyes have been opened. Hero of Crappy Town with a crappy intellect.

        Delete
      4. Not to be an ass, but you're an idiot. If you think reading a review about Rummel's numerous books makes you an expert enough to say he didn't do any work, then you're delusional. He has done an immense amount of research and has put years into his study. But go ahead and write your book, maybe I'll find it in a dumpster somewhere

        Delete
      5. Rummel was a raving idiot, Jarvis. Suck it up.

        Delete
    2. Coming up with an exact number of people murdered by totalitarian regimes is of course very difficult and accuracy uncertain. These types of governments don;t give out those number and don't have a free press to investigate. But Rummel did a very substantial amount of research and goes into great length on how he made his estimates. They are close enough for government work and show how amazingly bad Communist Governments were. Who knows or cares if a total is a few million high or low? Still horrible http://worldwarcoldwarvietnam.blogspot.com/

      ReplyDelete
      Replies
      1. 'But Rummel did a very substantial amount of research and goes into great length on how he made his estimates. They are close enough for government work'

        Bollocks, they are!

        'show how amazingly bad Communist Governments were.'

        Actually, they massively exaggerate how bad Stalinist regimes were, which is not only a grotesque distortion of history but, given how bad the track record of such regimes is in the absence of such exaggeration, is pretty damned superfluous.

        'Who knows or cares if a total is a few million high or low?'

        It isn't just a few million too high. Try a few tens of millions too high and you'll be closer.

        Delete
      2. PS: I don't doubt that estimating the death toll from totalitarian regimes is generally very difficult. That doesn't excuse the kind of grotesque exaggeration that Rummel engaged in, however.

        Delete
      3. What he did wasn't even close to research.

        Delete
    3. A person who describes himself as a, "Rothbardian anarchist and history buff posting from Ljubljana," ought to be praising Rummel for his original and unique contribution to the eventual demise of the State by being the first to call attention to the fact that governments murders innocent people by the hundreds of millions. Your critique sounds like jealousy of scholarship beyond your ken.

      ReplyDelete
      Replies
      1. Lies and false stats are not exactly conductive to anything but the opposite of what the liar tries to achieve.

        Delete
    4. If anything, the Congo famines weren't caused by Belgians. Given the state of Congo and the Congolese, it is very possible that the mass deaths were caused by warlords trying to make a quick profit off of the poor and needy.

      ReplyDelete
    5. «Not to be an ass, but you're an idiot. If you think reading a review about Rummel's numerous books makes you an expert enough to say he didn't do any work, then you're delusional. He has done an immense amount of research and has put years into his study. But go ahead and write your book, maybe I'll find it in a dumpster somewhere.»

      Years (or decades) of research have actually turned up figures much lower than Rummel's overblown claims on the victims of Stalinism:

      "All in all, the Germans deliberately killed about 11 million noncombatants, a figure that rises to more than 12 million if foreseeable deaths from deportation, hunger, and sentences in concentration camps are included. For the Soviets during the Stalin period, the analogous figures are approximately six million and nine million. These figures are of course subject to revision, but it is very unlikely that the consensus will change again as radically as it has since the opening of Eastern European archives in the 1990s. Since the Germans killed chiefly in lands that later fell behind the Iron Curtain, access to Eastern European sources has been almost as important to our new understanding of Nazi Germany as it has been to research on the Soviet Union itself. (The Nazi regime killed approximately 165,000 German Jews.)"

      (Timothy Snyder, "Hitler vs. Stalin: Who Was Worse?" (http://www.nybooks.com/daily/2011/01/27/hitler-vs-stalin-who-was-worse/).

      Rummel's years of research seem to have consisted in perusing secondary sources about the subject (which, as concerns the Soviet Union, include the Cold War era claims of dissidents like Solshenizyn) rather than looking up primary sources in archives.

      ReplyDelete
    6. This is an immensely cherry picking article (with from what I can determine only one significant claim) trying to cast doubt on the very thorough work of a scholar who has studied literally dozens (if not more) of major and minor governments and regimes and collects large amounts of information to come to estimates that you can calculate yourself using his methods. After all he takes you through step by step how he came to his figures. Also as for the amount of people who passed thorough the Soviet prison camps (as others have pointed out) records are often uncertain and so I don't know the extent to which is a true representation of his work's alleged inaccuracy. His step by step process also can be used to 'dismiss' your argument that he 'presupposes' his death rates, as if they are assumptions when anyone who reads his site can verify that is not the case. Finally there is a reason he publishes ranges of death figures (low and high end) because it is impossible to exactly calculate how many people a government killed, even if it keeps open records. None of this proves that he "shouldn't be taken seriously" as you claim. You would have to come up with something more substantial then that. Something that could be well, taken seriously.

      ReplyDelete
    7. Good case made that the late Rummel (may he rest in peace) didn't shrink from pulling fantastic numbers out of thin air when it suited his theses.

      Apparently some of Rummel's fans didn't like it, and it's noteworthy that their criticism of the article included ad hominem attacks. «Hero of Crappy Town with a crappy intellect», «Not to be an ass, but you're an idiot», «But go ahead and write your book, maybe I'll find it in a dumpster somewhere», «A person who describes himself as a, "Rothbardian anarchist and history buff posting from Ljubljana," ought ...», «Your critique sounds like jealousy of scholarship beyond your ken» are not exactly weighty arguments.

      Meanwhile, scholars who did something that Rummel never did (i.e. archival research, as opposed to making spreadsheets with figures from this and that secondary source and establishing maximums and minimums, or plain guesswork) found out that Stalin's killing record most probably stands at about nine million.

      Rummel was a bit more realistic regarding Hitler, whose killing record he blew by a factor of less than two (from about 12 million to about 21 million).

      In fairness it should be said, however, that Rummel was not always as much above the mark as regarding the Nazis, or as outrageously above the mark as regarding the Soviets. The numbers he attributed to the Japanese militarists, for instance, may even be considered conservative compared to claims made in other sources.

      ReplyDelete
      Replies
      1. For someone who is hell-bent on Nazifying the Soviet Union and the Russians it is rather remarkable that Timothy Snyder should be the first to point out that A.) Stalin's victims were far less than the 20 million usually thrown around (or even 60-80 million per Rummel, Solzhenitsin etc) and B.) that Hitler's victims exceed that of Stalin.

        As of his actual numbers (9 and 12 million) they are too low. Hitler caused 15 million in the Soviet Union alone and Stalin's number comes to something closer to 12 million (again mostly in the Soviet Union itself). The caveat is that different numbers aren't so much the result of working with different sets of information, but of interpretation. What does it mean for someone to be "killed by Stalin?" Do people who perished in the GULag during wartime famine count? What about victims of the 1946-47 famine? I would say yes but other sincere people could disagree.

        Delete
    8. I think you need to be a little more cautious in reading your sources. The source you cite in:

      "We can ignore the fact his "most probable estimate" of deaths the responsibility of the Soviet Union presupposes 40 million deaths in the gulag, which happens to be twice the number of people who passed through the camps."

      states that approximately 18 million passed through the gulags during the Stalin era only, i.e., up to 1953. Rummel's figures cover the period to 1987, another 34 years of gulag and forced labor. Also, he attributes 'perhaps' 43 million deaths in total to Stalin, not the 60-80 million you quote in your above comment.

      A lot of the variation in figures can also be attributed to what is 'counted' and what is not - Rummel's figures include a wide range of causes of death, including deaths due to population displacement as a result of government policy and so on. You're welcome to disagree with the empirical boundaries he puts on his data, but it doesn't mean that the data itself is incorrect just because it doesn't match other sources that are using different definitions. Cheers.

      ReplyDelete
      Replies
      1. "Rummel's figures cover the period to 1987, another 34 years of gulag and forced labor."

        Which goes to show just how little he knows. Khruschev promptly dismantled the gulag system and sent the inmates home.

        "Rummel's figures include a wide range of causes of death, including deaths due to population displacement as a result of government policy and so on."

        Most everyone includes deportation deaths including I. They come up to about 1 million out of 6 million forcefully internally exiled.

        Delete