Cambodia Section Home Mekong Network Home
Mekong.Net
About this site...

Noam Chomsky on the CIA Demographic Catastrophe Report

The article below is compiled from postings in the alt.fan.noam-chomsky newsgroup, written by Nathan Folkert. For a more detailed examination of Chomsky's comments on Cambodia, see Averaging Wrong Answers: Noam Chomsky and the Cambodia Controversy.


Recently Noam Chomsky made a claim regarding a CIA demographic study of Cambodia published in 1980, which he claims was a "whitewash" of the Khmer Rouge. I was curious about the veracity of this claim, mainly because in years past he made similar claims on ChomskyChat in response to readers' questions regarding the Khmer Rouge (possibly in reaction to debate here in the newsgroup, as they were swiftly posted here). I am still occasionally allowed to enter the Stanford University library, so I decided I'd check out these particular claims. I was swiftly able to obtain a copy of the report, and have decided to transcribe it here, because it has often been the subject of conversation in the past. (The report is now available online at http://www.mekong.net/cambodia/demcat.htm.) I would like to cite a few references here. These are recent citations (mostly originating at ChomskyChat, which is a relatively recent innovation, I believe, at least relative to the crimes they discuss), so I assume that these are Chomsky's current views:

Chomsky: [...] I do not, incidentally, agree with the later whitewash of the Khmer Rouge by US government scholars and the CIA -- who did in fact deny KR atrocities after the evidence was in, but are ignored, of course, on the internet sites you are exploring (I'm guessing, but I suspect it's a fair guess), because that would not serve the purpose of justifying US atrocities. You might recall, perhaps, that we were probably the only commentators to rely on the most knowledgeable source, State Department intelligence. Our conclusion at the time was that it was probably the most reliable as well as by far the best informed, and subsequent revelations support that tentative judgment. They were avoided in the mainstream commentary because their conclusions didn't fit the propaganda line that was required to exploit the misery of the Cambodians to justify subjecting millions of other people to comparable misery, in Central America and elsewhere. Presumably that is also why the CIA demographic study of 1980, regarded as authoritative by US government specialists, is totally ignored (I've mentioned it, and there is an extensive review by one of the major Cambodia historians, Michael Vickery -- an interesting story, but a separate one).
(http://groups.google.com/groups?selm=4b923300.0203072332.7055eb86%40posting.google.com)

[...] The CIA, in its demographic study in 1980, claims that Pol Pot killed 50-100,000 people and attributes most deaths to the Vietnamese invasion, also denying flatly the atrocities of 1978, which were by far the worst (that's the source of the famous piles of skulls, etc.; these became known after the Vietnamese invasion in 1979, and were certainly known to the CIA). Michael Vickery has written about the CIA study, suggesting that it was tailored to fit the fact that the US was tacitly supporting Pol Pot in '78 and later. There's a careful analysis in Vickery's "Cambodia." He's a very serious Cambodia scholar, and his analysis is taken seriously by other reputable scholars (e.g., Australian scholar Robert Cribb, in his standard scholarly work on the Indonesian massacres with comparative evidence). Vickery estimates about 700,000 deaths "above the normal" in the Pol Pot years -- which, if accurate, would be about the same as deaths during the US war (the first phase of the "Decade of Genocide," as 1969-79 is called by the one independent government analysis, Finland). For that period, the CIA estimates 600,000 deaths. The Yale Genocide project (Ben Kiernan and others) gives higher estimates, about 1.5 million. In fact, no one knows. No one ever knows in such cases, within quite a broad range. When numbers are put forth with any confidence, and without a big plus-or-minus, you can be sure that there is an ideological agenda, in any such case. Demographic analyses are very weak. If we wanted to be serious, we would also ask how many of the post-1975 deaths are the result of the US war. [...]
(http://www.zmag.org/forums/chomcambodforum.htm)

You're quite right that the worst atrocities were in 1978, at the time of US support for Pol Pot (the famous piles of skulls, etc.). That's now undisputed among serious analysts. Presumably US support is the reason why the CIA, in its demographic study of Cambodia in the 70's, denied totally the atrocities of 1978, and claimed that the early atrocities (which it estimated at 50,000-100,000 killed, a ludicrous underestimate) had ended by 1977, and that few peasants really suffered. Published in 1980, and regarded as definitive by top US government specialists (including the leading State Department "Cambodia watcher," later US Ambassador to Phnom Penh), the CIA study was completely suppressed, because the conclusions were plainly intolerable to the doctrinal system. I have a brief comment about it in "Towards a New Cold War." Vickery, one of the very few serious Cambodia scholars, reviewed it in detail in the "Bulletin of Concerned Asian Scholars" and his book, and (if I recall) gives this explanation for the CIA's curious behavior, which is not alone. In late '79, when the facts were in, the leading US government scholar, Douglas Pike, also produced quite positive comments about the Pol Pot regime. For Lewis and the like, such apologetics are not useful, so they are suppressed. His comments are quoted in various places, including (if I recall) Ed Herman's and my review of all of this in "Manufacturing Consent."
(http://www.zmag.org/forums/chomchatarch.htm)

[...] Much more extensive evidence became available in 1980, with the huge flow of refugees after the Vietnamese invasion drove out the Khmer Rouge. Based on that, the CIA issued a demographic study that was completely ignored in the media, though praised by the highest US authorities as the best account available. The study was outlandish. It attributed most of the deaths to the Vietnamese, and largely exonerated the Khmer Rouge, claiming that 50-100,000 people were killed (mostly military and government officials), that the atrocities were mostly in the early period, and that most peasants didn't make out badly under the KR. [...]
(http://groups.google.com/groups?selm=36ACFC7A.3565%40columbia-center.org)

Remember these claims by Chomsky concerning the CIA study:

1. It flatly denied the atrocities of 1978.
2. It blamed the majority of the deaths on the Vietnamese invasion.
3. It claimed that the CIA estimated 50,000 - 100,000 were killed by Pol Pot.
4. It claimed that few peasants really suffered.

Now, to review and respond:

1. It flatly denied the atrocities of 1978.

This is a lie. There was no "flat denial" of the atrocities of 1978. They didn't even claim that they "doubted" the atrocities of 1978 occurred. They simply weren't included in the calculation, if they were, in fact, known by the author (no atrocities are included for 1977, either, nor, for that matter, for 1979). Chomsky claims that they should have been included, that the CIA should have known about them. This may well be true, but this does not support Chomsky's claim that the atrocities were "flatly denied".

I suspect that this, and the related statement that the report "claimed that early atrocities [...] had ended by 1977" was taken from Vickery's account, his "careful reading", as he would put it. Throughout the report, there is reference to the estimate that 50,000 - 100,000 new people were likely executed. No other executions are included in the estimates, though other killings of civilians are mentioned, as noted below. On the fold out table in the back cover of the report, in listing the events and their estimated demographic impact, the entries on these executions, which clearly, in context, refer to these executions, were abbreviated to "executions begin", "more executions", and "final executions". Vickery interprets this "final executions" to mean not that the last of the specific 50,000 -100,000 executions of new people included in the estimates had ended, but that all executions had ended, and apparently Chomsky has accepted this "careful analysis" (which appears to me to be a deliberate misinterpretation) as claiming that the atrocities ended by 1977. Of course, it never made this claim. Again, one might argue (as Chomsky and Vickery do) that the report *should* have included atrocities in 1977 and onward, which the CIA *should* have known about, and this may well have been true (and it is entirely possible, though I think not plausible, that the CIA was being deliberately dishonest in leaving them out), but it is patently a lie to state that the CIA *claimed* that the atrocities had ended.

2. It blamed the majority of the deaths on the Vietnamese invasion.

This is an outright lie. Using 18 deaths per 1,000 as the "normal" death rate, the CIA report suggests that under the KR regime there were about 1,300,000 deaths over the normal in their Medium Series (my calculation, not theirs). Using the same assumption, the CIA report suggests that under the Vietnamese regime there were about 240,000 deaths over the normal in their Medium Series (again, mine), and they make the explicit claim of 600,000 - 700,000 deaths over the normal under the Lon Nol regime (about 120,000 per year average, though there's no reason to suspect that these were distributed evenly). In their account of the Vietnamese regime, they do not specifically blame these deaths on the Vietnamese, and in fact suggest that the peasants suspected of "collaboration" suffered retaliation from the "marauding Pol Pot forces", and that "the Heng Samrin/Vietnamese forces were seldom able to protect [the peasants] from sporadic attacks by the Pol Pot guerrillas". Far from blaming the Vietnamese invasion for the majority of the deaths, they attributed only a fraction of the deaths to 1979, and suggested that some unspecified number of the deaths in that year were the responsibility of Pol Pot forces.

3. It claimed that the CIA estimated 50,000 - 100,000 were killed by Pol Pot.

This is a lie, or, if we are to be charitable, a misleading half-truth. It claimed that 250,000 new people (specifically military, administration, and educated people) were targeted for execution and estimated that roughly 40% of these were actually executed (20% died of hunger or disease, and 50% of remainder executed). It made no claim about these being the only executions (much less the only killings), though these were the only executions it specifically included in estimates. The report does claim, as I note above, that marauding Pol Pot forces targeted peasants for retaliation, though it makes no estimate of the numbers killed, so it is clear that there are atrocities that the report simply doesn't even attempt to quantify, whether they know about them or not. (Before someone points this out as evidence that the CIA were whitewashing Pol Pot, I must note that it likewise makes no attempt to estimate the number of people killed by the Vietnamese).

Chomsky called the CIA's figure a "ludicrous underestimate." It should be noted that he does not make similar complaints about the Finnish Government Study, "Kampuchea: Decade of the Genocide", conducted several years after the CIA study. It stated that "realistic estimates" of executions were between 75,000 - 150,000, within which the "most likely" estimate for the CIA, 100,000, fell. Chomsky has not, so far as I know, claimed that the Finnish study is a "ludicrous underestimate", even considering that it estimates that less than one million died (total) during Pol Pot's reign, which is not significantly more than double the expected (18 per 1,000 death rate) numbers, and considering that it, like the CIA study, claimed that some 2.5-3 million people who lived in rural areas before the revolution (the "old people" in the CIA study) "did not suffer nearly as much under the Pol Pot administration" (which is still, of course, different from claiming that they did not suffer, or that they didn't make out badly).

4. It claimed that few peasants really suffered.

This is a lie, or, at best, a dishonest semantic interpretation:

"The old people -- those living in rural areas when Pol Pot gained control -- fared somewhat better. They too, however, suffered from food shortages, lack of medical care, and hard labor."

Those are just the peasants who remained in the countryside when Pol Pot took control. The peasants who fled to the cities (implied from the estimates in the paper, approximately 2 million people, or about a quarter of the population, based on UN estimate that about a quarter of population would be urban in 1975, when in fact more than half was) and were then, by definition, "new people", suffered significantly more, but the report in no way implies or suggests that "few peasants really suffered" or "most peasants didn't make out badly".

Whatever the merits or demerits of this report, it is clear that Chomsky's description of the report is a string of lies. Chomsky made four easily-checked lies -- lies that can be verified as such simply by reading the CIA report and require no further deep knowledge on the events in Southeast Asia.


W3C Validation