Quantifying Civilian Casualties in the Northeast during the Chinese Civil War
The Korean War generated tremendous human suffering, a fact that we have covered on Sino-NK from multiple angles, from the psychological effects to Kim Dong-choon’s work on atrocities. But how often do we remember the Chinese civil war just across the border? That conflict is often seen as relevant only for having spun off some tens of thousands of Korean soldiers to augment Kim Il-sung’s armies in the steely winter of 1949-50. But there are other layers of connectivity between the Chinese civil war and northern Korea, the boundaries of which the Korean War served to blur. One such area is the trauma faced by civilians in the two conflicts. This new essay by Adam Cathcart begins to move the discourse in that direction.
The essay also engages with a local matter, germane to the politics of doing research on East Asian history today. In its detailed approach to sources and their connection to occasionally dubious claims, Cathcart arrives at the civil war in the lengthening aftermath of the J. Mark Ramseyer affair, a controversy centering on a disputed paper from the Harvard Law School professor which was nominally about “comfort women” contracts during the Pacific War. As Tessa Morris-Suzuki argues, the Ramseyer case serves as a stringent reminder about the ethics that researchers should employ in an era of “fake news,” or, we might add, in the case of Charles Armstrong, faked sources.
Finally, the essay brings us back to how we ought to write about dictatorships or their construction. How uniquely malevolent was Kim Il Sung? What about Mao? How deadly the road to power? Frank Dikötter’s new book on dictatorship speaks with great assurance on these topics, but Cathcart drags us back to the sources undergirding one claim of “hundreds of thousands” of civilian deaths which are ultimately meant to be tallied up to Mao.– Christopher Green, Senior Editor
Quantifying Civilian Casualties in the Northeast during the Chinese Civil War
by Adam Cathcart
Introduction | Who were the victims of the Chinese civil war (1945-1950), and how many civilian deaths were caused by that conflict? There is no widely-available body of statistics which clearly show how many civilians died, and when, and where, or of what causes. Even a very good book focused on China’s civilians and their difficulties during the civil war can only provide the vaguest of civilian casualty estimates (“millions”), and does so without even a footnote. We are therefore left at the mercy of reportage, anecdotes, images, films and stories to illustrate that the Chinese civil war was not a good conflict in which to be a civilian — particularly in Northeast China, one of the main theaters.
Frank Dikötter cuts through the uncertainty, stepping forth with casualty estimates for any range of events or mass movements in China. The University of Hong Kong historian has made something of a cottage industry — in a series of widely-published works — out of adding up civilian deaths in China since 1945, the responsibility for which he tends to account to the Chinese Communist Party. In Mao’s Great Famine, Dikötter asserted that “at least 45 million people were worked, starved or beaten to death in China” during the Great Leap Forward. Having garnered the Samuel Johnson Prize for non-fiction for Mao’s Great Famine, in 2013 Dikötter published a prequel focusing on the Chinese civil war and the early years of the PRC, entitled The Tragedy of Liberation: A History of the Chinese Revolution, 1945-1957.
In this book, Dikötter extends his case for a predatory relationship between the CCP and the Chinese people, and makes a damning assessment of the interaction between the CCP and the civilian population under conditions of civil war.
In a chapter of Tragedy of Liberation entitled “The Hurricane” which analyses the CCP land reform campaigns of the late 1940s, Dikötter writes:
Land reform pitted villagers against each other, and as they denounced on another in ferocious meetings, the actual holdings in the countryside finally came to light….now the party knew exactly how much land existed. It determined how much each strip could produce and demanded that each household hand over a designated amount of grain.
As one observer noted about Manchuria: ‘Heavy grain requisitions to support the Communists’ armies of 3 to 4 million men have in many areas not only stripped the countryside of food surpluses but have eaten into subsistence stocks.’
On top of tax, foodstuffs including soybeans, corn, rice and vegetable oils were traded with the Soviet Union for industrial equipment, motor vehicles, oil and manufactured supplies, increasing the overall food deficit. Hundreds of thousands of people starved to death in Manchuria as a result.1)Frank Dikötter, Tragedy of Liberation: A History of the Chinese Revolution, 1945-1957 (London: Bloomsbury, 2013), 69. Emphasis added.
In this short excerpt we have a kind of representative sample of the Dikötter method. It involves, first, a description of the CCP’s methods as involving “ferocity” or “frenzy” alongside a calculating and controlling nature, unlike the inept but basically moral Nationalist regime which Mao’s communists displaced. A demonic description of the CCP’s effect is then followed by direct quotations from a primary source; and it ends on a dramatic estimate of deaths caused by the CCP.
The rest of this essay is an attempt to understand what Dikötter is doing with these five quoted sentences, and the sources that underpin them.
Sources | First, it may help to rephrase what Dikötter is saying in the above excerpt. In short, the Chinese Communist Party enacted a rural revolution, and then taxed peasants in Northeast China on their new land holdings, and then exported the grain to the Soviet Union in return for industrial goods and requisitioned grain to support their armies, again in the Northeast. This led not just to peasant anxiety or unrest but hundreds of thousands of deaths by starvation in the region.
If you haven’t heard previously of mass CCP-induced starvation in the Northeast during the civil war, you are not alone. Whole generations of civil war scholarship have avoided this discussion. So what sources does Dikötter base his claims upon? And do those sources say what he claims they say?
Dikötter cites two sources. The first, the “observer” on Manchuria, is easy enough to find, although Dikötter’s citation is too generic: It is A. Doak Barnett in a September 1949 letter from the US Consulate in Hong Kong to the US Institute of World Affairs in New York City. The full text of the letter is here. One could spend many hours leafing through these materials, and compare them with Barnett’s book on China from 1949-1955.2)The full run of Barnett’s letters is here — and well done to Dr. Dikötter for this find; I certainly had not been aware of the source, so my complaint about a “generic” citation should be taken in the context of a certain pleasure of having been alerted to the documents in the first place.
For readers inclined to click through to read the original source, it becomes immediately clear that something is off. In the first full paragraph of Barnett’s letter cited by Dikötter, a point of confusion arises: the document is not focused on Manchuria, or Northeast China, at all, but rather on “North China.” Writing from Hong Kong in 1949, Barnett provides no estimate of deaths from starvation anywhere in China in this source, let alone of “hundreds of thousands” in a specific region. There is no discussion of the barter trade with the Soviet Union and how it was grumbled about by peasants in Northeast China or, more dramatically, taking food out of their mouths.3)For a fuller discussion of this theme, see Steven Levine, Anvil of Victory: The Communist Revolution in Manchuria, 1945-1948 (New York: Columbia University Press), 176-179. However, Levine’s sole source on the exports to the USSR is a US State Department intelligence report from January 1948, which does not appear to stipulate starvation; Levine only notes “considerable belt-tightening” in areas controlled by the CCP in the Northeast. Better Chinese sourcing on the Party’s balance between grain exports, grain supplies, military hardware, military morale, peasant taxation ratios, grain markets, and Nationalist counter-moves can be found in Harold Tanner, Where Chiang Kai-shek Lost China: The Liao-Shen Campaign, 1948 (Bloomington: Indiana University Press, 2015), 133-139. Basing his views on how things have always been in China prior to 1949 and his observations of weather and crops for 1949, Barnett only speculates about the coming winter: “before the winter [of 1949-50] is over, starvation will probably be widespread throughout many parts of China.”
Naturally, Dikötter is not basing his factual assertion of “hundreds of thousands of people starved to death in Manchuria” on this single document, given that Barnett’s language about starvation is explicitly speculative. Instead, Dikötter has an ace up his sleeve when it comes to chronicling the inhumanity of the CCP in the northeastern theater of the civil war: Zhang Zhenglong’s Xue bai, xue hong [White Snow, Red Blood], the text written by a scholar inside of the People’s Liberation Army and which was published — and promptly banned — in China in 1989.
Dikötter makes ample use of the Zhang Zhenglong book throughout the relevant sections of Tragedy of Liberation, including its portrayal of the 1948 siege of Changchun which is taken as the ultimate symbol of the CCP’s bloodthirsty ruthlessness.4)Andrew Jacobs of the New York Times went to the northeast in the autumn of 2009 and wrote about the siege of Changchun as a blemish on the CCP’s glorious master narrative, an article which clearly affected Dikötter as he was preparing Tragedy of Liberation. Therefore, it must be Zhang’s text which Dikötter uses to undergird his factual assertion of civilian deaths in Northeast China, having used Barnett essentially as a bit of color and a direct quote about grain requisitioning.
And indeed, that is the case; Xue bai, xue hong is the core source for the assertion. Unfortunately, here we run into some additional concerns. While the section cited in Zhang’s book does engage in a powerful meditation on civilian deaths in warfare writ large, its statistical data do not venture at all into large-scale civilian casualty estimates, and most of the grain requisitions discussed in northeast China are Nationalist/KMT responsibility, not communist, and never are they connected to grain exports or the Soviet connection.
Because Zhang’s text is in Chinese, exists in multiple editions, and has never been fully translated, some further explanation is warranted.5)Because the book was banned and has been reprinted by various presses in both Hong Kong and the mainland since 1989, page numbers are not consistent across editions. The two versions consulted for this essay has different page numbers from those cited by Dikötter. The editions consulted here are Zhang Zhenglong, Xue bai, xue hong (Chengdu: Sichuan Wenyi Chubanshe, 2007) and the more authoritative and original Zhang Zhenglong, Xue bai, xue hong (Beijing: Jiefangjun Chubanshe, 1989). The section in the book cited by Dikötter is entitled “河两岸是百姓” or, “The Chinese people are on both sides of the river,” a poetic allusion. This is a typical passage in Zhang’s book — written with a surfeit of emotion, packing in some miniature oral histories, and citing CCP directives from the civil war period. It is a different formula from Dikötter’s, but no less effective and even cathartic in its way, particularly if the reader embraces rather than recoils at the moralistic tone.6)Zhang has a humanistic outlook which suffuses his entire book. Xue bai, xue hong conveys his essential sorrow at the conflict as a whole, and the section summarized here is no different.
Asking explicitly how many civilians died during the Chinese civil war at the outset of this section, Zhang does not attempt a numerical answer at all. Instead, he launches into a number of anecdotes — in this case, interviews. He provides snapshots from interviews with two civil war survivors from Jinzhou (who witnessed approximately 40 and 50 deaths, respectively, in urban fighting). This is followed by an interview in Harbin about destroyed houses in Siping, and an interview in Anshan about two civilian deaths involving the communist army. The anecdotes conclude with a story of a village that fell victim to alternating waves of banditry and anti-banditry campaigns after 15 August 1945, and whose population fell from 500 to “some tens of people.” Only in the Anshan case is it clear that the CCP is responsible for the deaths, but this again is a quirk of Zhang’s book — so often, the state or party perpetrators of violence are not at all clear.
Calculating | Then, let us be generous to Dikötter’s vision and assume the communists killed absolutely every victim described in Zhang’s chapter. In the materials which Dikötter has cited claiming “hundreds of thousands” of civilian deaths in Northeast China due to the CCP, that’s about 100 deaths and the disappearance or death of about 450 people so far.
Zhang then moves on to discuss the harvest and we are at last on common ground with Barnett and Dikötter. Zhang describes how 1947 was not a good harvest, and that 1948 was the worst year of the civil war for agriculture in Northeast China. Villages on both sides of the Liao river, he notes, lost much of their crop production to insufficient rainfall in the spring — “the worst in 30 years,” he notes. Still no mention of famine deaths, however.
Getting into the problem of grain and the armies, Zhang quotes a 10 May 1948 CCP order about unified transport which deals with 抢购食粮 (qianggou shiliang, which means a rush to purchase, or a stampede for certain goods). The document notes that if CCP armies need grain as they move, they should first get it from landlords and rich peasants, and then from middle peasants, and if no other way is available, from poor peasants. Similarly, the armies are recommended to take large hogs first, then smaller hogs, etc., and given a hierarchy of feed options for horses being used by CCP armies.
No assertions are made of famine or hunger amid the population, nor of the CCP forcibly taking grain from the peasants.7)Perhaps Dikötter has mistaken 抢购 (qianggou) in the document subject for 抢劫 (qiangjie, to rob, loot, or plunder)? However, Zhang does discuss “burdens” or fudan (负担), of grain tax for people (人民), rich peasants (富农), middle peasants (中农), and such, where the CCP used its baseline numbers largely from the War of Resistance in the areas then under its control. The only civilian deaths chronicled by Zhang in this section are voiced via a CCP document from 20 March 1948 which describes how two whole families in the northeastern steel town of Anshan committed collective suicide the day after the CCP’s arrival, feeling the burden would be too much.
This is a tragic instance, to be sure. Given that Anshan was a model city for foreign ideological tourists in the 1950s — including UK Labour Party delegations, among others — it does become clear that the CCP’s relationship with the citizens of Anshan was not without scars. But, since we are focused on the figures here rather than emotional payload or shivering juxtapositions (the corpses under Potempkin roads, as it were), the Anshan incident does not move our body count appreciably higher. Realistically, we can only move the civilian death tally up to 500 thus far. We are still waiting for confirmation of Dikötter’s assertion of massive famine due to CCP grain requisitions in northeast China.
The final section of the Zhang materials he has cited does ultimately deal in depth with population loss, and draw from a CCP document from August 1948. It is exceptionally detailed — but, readers may be disappointed to learn, it unfortunately has nothing at all to do with deaths by starvation in Manchuria. Here is a rough translation:
At the time [in autumn 1945 when] our units went over the [Shanhaiguan] pass into the Northeast, we had not not more than 110,000 soldiers, and 20,000 cadre (including Party, administrative and military cadre); in total this was not more than 130,000 people. In the three years of the War of Liberation in the Northeast, we annihilated 106,000 enemies and added 1,300,000 to our troop numbers. In three years of the War of Liberation, 1,440,000 sons and daughters of the Northeast [were recruited and] participated in the PLA, taking 649,600 prisoners (of whom about a third, 200,000 replenished our own troop levels), adding to our 110,000 troops who originally entered the Northeast. This made the peak of our military strength in the region 1,755,907. In three years of the War of Liberation, we sustained 301,719 losses, including 194,813 badly wounded (about one third of whom, 58,444, became disabled), 55,439 who disappeared, 50,827 who were taken prisoner, and 4% (7744 troops) who died in hospital. 8)Zhang Zhenglong, Xue bai, xue hong (Beijing: Jiefangjun Chubanshe, 1989), 444-445.
Indeed, those are striking numbers in the “hundreds of thousands,” but these are CCP battlefield losses in the civil war, not civilian deaths by starvation caused by excess communist grain appropriation and exports. So what are we left with?
It may be that the author simply does not care what the sources cited actually say and is hoping that his assertion of “hundreds of thousands of deaths by starvation” will transition from bold claim to factual terrain without being unduly challenged. Certainly that is how it was transmitted into my own students’ essays. It is also possible that Dikötter is mentally counting tens of thousands of starvation deaths from Changchun and Shenyang during the respective sieges of those cities in summer and autumn 1948. Unfortunately, this is another topic altogether, and CCP grain exports to the USSR would have little to no bearing on starvation in Nationalist-controlled cities. That would be a slight of hand, certainly, but may well be the only way to drive the statistics up into the territory claimed.
|↑1||Frank Dikötter, Tragedy of Liberation: A History of the Chinese Revolution, 1945-1957 (London: Bloomsbury, 2013), 69. Emphasis added.|
|↑2||The full run of Barnett’s letters is here — and well done to Dr. Dikötter for this find; I certainly had not been aware of the source, so my complaint about a “generic” citation should be taken in the context of a certain pleasure of having been alerted to the documents in the first place.|
|↑3||For a fuller discussion of this theme, see Steven Levine, Anvil of Victory: The Communist Revolution in Manchuria, 1945-1948 (New York: Columbia University Press), 176-179. However, Levine’s sole source on the exports to the USSR is a US State Department intelligence report from January 1948, which does not appear to stipulate starvation; Levine only notes “considerable belt-tightening” in areas controlled by the CCP in the Northeast. Better Chinese sourcing on the Party’s balance between grain exports, grain supplies, military hardware, military morale, peasant taxation ratios, grain markets, and Nationalist counter-moves can be found in Harold Tanner, Where Chiang Kai-shek Lost China: The Liao-Shen Campaign, 1948 (Bloomington: Indiana University Press, 2015), 133-139.|
|↑4||Andrew Jacobs of the New York Times went to the northeast in the autumn of 2009 and wrote about the siege of Changchun as a blemish on the CCP’s glorious master narrative, an article which clearly affected Dikötter as he was preparing Tragedy of Liberation.|
|↑5||Because the book was banned and has been reprinted by various presses in both Hong Kong and the mainland since 1989, page numbers are not consistent across editions. The two versions consulted for this essay has different page numbers from those cited by Dikötter. The editions consulted here are Zhang Zhenglong, Xue bai, xue hong (Chengdu: Sichuan Wenyi Chubanshe, 2007) and the more authoritative and original Zhang Zhenglong, Xue bai, xue hong (Beijing: Jiefangjun Chubanshe, 1989).|
|↑6||Zhang has a humanistic outlook which suffuses his entire book. Xue bai, xue hong conveys his essential sorrow at the conflict as a whole, and the section summarized here is no different.|
|↑7||Perhaps Dikötter has mistaken 抢购 (qianggou) in the document subject for 抢劫 (qiangjie, to rob, loot, or plunder)?|
|↑8||Zhang Zhenglong, Xue bai, xue hong (Beijing: Jiefangjun Chubanshe, 1989), 444-445.|