Excerpts from Cormac Ó Gráda’s review of new book by Dikötter, Mao’s Great Famine: The History of China’s Most Devastating Catastrophe, 1958-1962. Full review published in Population and Development Review 37(1) : 191–210 (March 2011), as “Great Leap into Famine: A Review Essay.” Ó Gráda is a leading scholar of famine, authoring Famine: A Short History (Princeton University Press 2009) and Black ’47 and Beyond: the Great Irish Famine in History, Economy and Memory (Princeton University Press 1999).

Frank Dikötter’s Mao’s Great Famine {henceforth “MGF”} is the longest and most detailed study of the Great Leap Forward (GLF) famine to appear in English to date.{…} The tone throughout is one of abhorrence and outrage, and sometimes MGF reads more like a catalogue of anecdotes about atrocities than a sustained analytic argument. In style and approach it recalls Jung Chang and Jon Halliday’s controversial Mao: The Unknown Story (2005); indeed, Chang leads the ”praise” for MGF on the back cover. MGF may become the best-known account of the GLF famine for a while. But should it? It is not a comprehensive account of the famine; it is dismissive of academic work on the topic; it is weak on context and unreliable with data; and it fails to note that many of the horrors it describes were recurrent features of Chinese history during the previous century or so. More attention to economic history and geography and to the comparative history of famines would have made for a much more useful book. In what follows I focus on the economic context of the famine, review features of the famine treated by Dikötter but worth further study, and conclude by discussing the role in these events of Mao and the party elite.

Poor China
Famines are a hallmark of economic backwardness. It bears remembering that China on the eve of the Great Leap Forward was one of the poorest places on earth.{…} For at least a century before 1949, major famines were probably frequent enough to warrant Walter Mallory’s depiction of China in 1926 as the “land of famine.” The Taiping Rebellion is routinely reported as costing 20 million lives, mostly from famine and disease. Neither R. H. Tawney’s (1932) report that the famine of 1849 “is said to have destroyed 13,750,000 persons” nor contemporary claims that the Great North China Famine of 1876–79 took a further 9.5 million to 13 million lives should be taken literally, but such estimates accurately underline the apocalyptic nature of those famines. Famine mortality probably declined thereafter. Yet Yang (2010) claims that China’s most severe famine before the GLF famine occurred in 1928–30, killing 10 million people. Between 1920 and 1936, he added, “famine due to crop failures took the lives of 18.36 million people.” Again, these numbers seem too high. Still, Tawney witnessed the devastation that followed in the wake of the famines of the late 1920s, and famine in Anhui province in 1929 inspired Nobel laureate Pearl Buck’s The Good Earth. Nor did it end there. Famine in the Yellow River region in 1935 resulted in significant female infanticide in 1935–36, while the Henan famine of 1942 produced its own catalogue of atrocities. Again and again, what Dikötter dubs ”traditional coping mechanisms” (p. 179) had failed to prevent famine.{…}

China’s extreme backwardness on the eve of the Great Leap matters because it greatly increased its vulnerability to disequilibria, man-made or other. Had Chinese GDP per head been, say, twice as high as it was, the devastation wreaked by the Leap would presumably have been much less. Nor, on the other hand, does MGF take sufficient account of how conditions improved between 1949 and 1958. If the standard estimate of grain output of 200 million metric tons in 1958 is taken at face value (p. 132), then there was enough food to provide an average daily intake of about 2,170 kcals (Ashton et al. 1984: 622; compare Meng et al. 2010). If, however, the output data are contaminated by Leap-style ”winds of exaggeration” and refer to unhusked grain, then the picture is much less rosy and the margin for error by central planners much narrower. Nonetheless, the achievements of the pre-Leap years prompted a false optimism that much faster growth was feasible—catching up or overtaking Britain “in fifteen years” (pp. 14, 15, 73).

What did the victims die of?
Throughout history most famine victims have succumbed to disease, not to literal starvation. Weakened immune systems and social disruption allowed diseases present in normal times to play havoc during famines. Pre-1949 China was no exception: economic backwardness made infectious diseases such as cholera, typhus, and malaria endemic and most famine deaths were from such diseases and from dysentery. So what did the victims of the Great Leap famine die of? Most accounts imply death by starvation rather than by disease; Thaxton links most deaths in the village of Da Fo in 1960 to ”edema,” and this is corroborated by the most detailed study of the causes of death to date, Yixin Chen’s analysis of public health gazetteers from Anhui province (Thaxton 2008: 209, 253; Chen 2010). Although Chen argues convincingly that the faulty data in the gazetteers underestimate the death toll from diseases such as dysentery and malaria, he nevertheless concedes the primary role of edema and literal starvation. Dikötter (p. 286) concurs and wonders why disease did not carry off more ”before terminal starvation set in.” The primacy of starvation as the cause of famine deaths is rather striking and poses a conundrum for demographers studying famine. Before the 1950s only war-induced famines in economies with effective public health regimes, such as the western Netherlands in 1944–45 or Leningrad in 1941–43, followed such a pattern. Does this imply that the Maoist public health campaigns of the early and mid-1950s influenced the causes of deaths during the Great Leap famine, if not the death toll itself? Could it be that the authorities’ attempts to control migration limited, even if unintentionally, the spread of infectious diseases? Chen (2010) gives due credit to achievements registered before the Leap; by then three major killers—smallpox, plague, and cholera—had been virtually eliminated and large-scale immunization campaigns carried out. Reluctant to allow public health improvements a role, Dikötter surmises, albeit without supporting evidence, that the Chinese peasantry succumbed to starvation quickly, “reducing the window of opportunity during which germs could prey on a lowered immunity” (p. 286).

The demographic impact
MGF is full of numbers but there are few tables and no graphs. Quantification is not its strong point. So we read that “between 1 and 3 million people took their lives” by suicide during the GLF (p. 304); that in Xinyang in Henan province “67,000” people were clubbed to death by militias (pp. 117, 294); that in some unspecified location “forty-five women were sold to a mere six villages in less than half a year” (p. 261); that “at least 2.5 million…were beaten or tortured to death” during the Leap (p. 298); and that delays to shipping in the main ports during some unspecified period cost “£300,000” (p. 156). An estimate of 0.7 million deaths from starvation and disease in labor-correction camps between 1958 and 1962 is obtained by applying an arbitrary ”rough death rate” of two-fifths to a guess at the camp population at its peak (p. 289). The main basis for the claim that “up to two-fifths of the housing stock turned into rubble” (p. xii) seems to be a report describing conditions in Hunan province from Liu Shaoqi to Mao on 11 May 1959, after Liu had spent a month in the region of his birth (p. 169).2 On page after page of MGF, numbers on topics ranging from rats killed in Shanghai to illegal immigration to Hong Kong are produced with no discussion of their reliability or provenance: all that seems to matter is that they are ”big.”

The cost of famines in lives lost is often controversial, because famines are nearly always blamed on somebody, and excess mortality is reckoned to be a measure of guilt. It is hardly surprising, then, that MGF’s brief account (pp. 324–334) of the famine’s death toll arrives at a figure far beyond the range between 18 million and 32.5 million proposed hitherto by specialist demographers (e.g., Yao 1999; Peng 1987; Ashton et al. 1984; Cao 2005). Rather than engage with the competing assumptions behind these numbers, Dikötter selects Cao Shuji’s estimate of 32.5 million and then adds 50 percent to it on the basis of discrepancies between archival reports and gazetteer data, thereby generating a minimum total of 45 million excess deaths. Much hinges on what ”normal” mortality rates are assumed, since the archives do not distinguish between normal and crisis mortality. The crude death rate in China in the wake of the revolution was probably about 25 per thousand. It is highly unlikely that the Communists could have reduced it within less than a decade to the implausibly low 10 per thousand adopted here (p. 331). Had they done so, they would have “saved” over 30 million lives in the interim! One can hardly have it both ways.{…}

Three parts nature?
The role of the weather in 1959–61 remains contested. Is Dikötter right to dismiss it? Contemporary Chinese sources highlighted ad nauseam the difficulties caused by drought and flooding, while denying the existence of famine conditions. Western journalists and historians echoed this view. Time magazine repeatedly reported adverse weather, 5 and an eminent Harvard Sinologist declared as late as 1969 that conditions such as those experienced in 1959–61 “would have meant many millions of deaths in the areas most severely affected” but for the effectiveness of public policy and the transport network (Perkins 1969: 303). MacFarquhar’s pioneering account of the famine also highlighted adverse weather as a factor (MacFarquhar 1983: 322). Dikötter acknowledges the challenges posed by the weather but blames
harvest shortfalls instead on the environmental destruction caused by the GLF, which magnified damage caused by adverse weather shocks. Perhaps, but here anecdotes are an inadequate substitute for more rigorous meteorological analysis. Research on the impact of the weather hitherto has relied on indirect measures such as the proportion of the grain crop damaged by the weather or reported grain production. Using this approach Y. Y. Kueh found that droughts and flooding accounted for the bulk of the shortfalls in 1960 and 1961, although he also insisted that “even without natural disasters, the agricultural depression was inevitable” (Kueh 1984: 80–81; 1995: 224). Researchers have only begun to use some abundantly available direct measures that are not subject to misreporting.6 In the absence of systematic analysis of these data, all one can say is that data from several Chinese weather stations show signs of exceptionally adverse weather shocks in 1959–61, though hardly enough to account for the regional variation in harvest shortfalls.7 Dikötter’s sense that the weather did not matter much may well be correct, but his failure to nail the issue is a lacuna.

Human agency
Malthus and his followers underestimated the role of human factors in exacerbating and mitigating famine in the past, even in very backward economies. As John Post pointed out in his classic account of famine in northwestern Europe in the 1740s, even very poor economies could escape “famine conditions and crisis mortality [by] import[ing] grain supplies, adequate welfare programs, and… effective… public administration” (Post 1984: 17). This message is also an important implication of Amartya Sen’s entitlements approach to famine analysis (Sen 1981). Malthusian interpretations of famine in China begin with Malthus himself, and most analyses of pre-1949 Chinese famines continue to be strictly Malthusian.{…} Dikötter’s stance is the polar opposite. He repeatedly cites variants of Liu Shaoqi’s quip (picked up by Liu from peasants in his native Hunan) that the GLF famine was three parts natural and  seven parts man-made (pp. 121, 178, 335), but only to reject Liu’s ”three-tenths Malthusian” interpretation in favor of one that rests entirely on human agency.

As the examples of Ireland and Ukraine attest, the temptation to interpret famines as genocides is strong. Dikötter, perhaps rightly sensing that this approach can distort reality, does not go quite so far as Chang and Halliday’s claim that Mao ”knowingly” allowed millions to starve. Indeed, one plausible reading of MGF’s narrative chapters is that it took a long time for the leadership in Beijing to grasp the scale of the catastrophe at its height. Utopian euphoria and a revolutionary impatience to catch up quickly had prompted the Great Leap. They also neutered Defense Minister Peng Dehuai’s interventions at the Lushan ”think-in” in July 1959. Peng’s protests, in any case, were less about the famine per se than the follies of the Leap in its first phase. Dikötter’s depiction of the follies is excellent and corroborates the more theoretical
case previously advanced by economists and economic historians such as Yao (1999), Li and Yang (2005), Bernstein (2006), and Wheatcroft (2010).

How much did Beijing know when the famine was at its height? Despite MGF’s relentless anti-Mao stance, it accepts that nobody at the top realized beforehand how murderous the economic war against the peasantry would be. Mao’s private physician, repeatedly invoked by Dikötter as a reliable witness (p. 346), “doubted that [Mao] really knew” what was happening (Li 1994), and we are told that Mao was “visibly shaken” when presented with graphic reports of famine from Xinyang in Henan province in late October 1960 (p. 116). Reliable information was at a premium; even the “fabled sinologists” in the British Embassy had no clue about what was going on (p. 345). Blaming the tragedy on the usual counterrevolutionary suspects, Mao nonetheless had “abusive cadres” removed. The news from Xinyang set in train moves that
would mark ”the beginning of the end of mass starvation” (p. 118). In that same month Mao, under pressure from critics of the Leap, ordered the redeployment of a million workers from industry to agriculture in Gansu province, citing the truism that “no one can do without grain” (MacFarquhar 1983: 323). Various concessions to the peasantry followed, and in January 1961 Mao told the 9th Central Committee Plenum that “socialist construction…should take half a century” (Barnouin and Changgen 2007: 188). {…}

China lacked an all-seeing, all-knowing Soviet-style secret police during the Leap. Too much reliance was placed on poorly monitored regional agents and thuggish local cadres. Why else would it take a visit to his home village in Hunan for Liu Shaoqi to discover the dimensions of the disaster? What he saw converted him overnight from supporter to “blistering” critic of the GLF (pp. 119–121). Central-planner-in-chief Li Fuchun’s reaction to the reports from Xinyang was that misguided policies (which he had championed)
had cost lives (pp. 116–117, 122). In a speech in Hunan to party planners in mid-1961, he summarized what have become textbook criticisms of central planning: ”too high, too big, too equal, too dispersed, too chaotic, too fast, too inclined to transfer resources” (p. 122). But thanks to a form of “closed” governance of their own creation, Mao and the party leadership seem to have discovered “destruction on a scale few could have imagined” rather late in the day (p. 123).

None of this absolves Mao from responsibility for the policies that caused the greatest famine ever. But reckless miscalculation and culpable ignorance are not quite the same as deliberately or knowingly starving millions (Jin 2009: 152). Few of the countless deaths in 1959–61 were sanctioned or ordained from the center in the sense that deaths in the Soviet Gulag or the Nazi gas chambers were.8

MGF’s reliance on fresh archival sources and interviews and its extensive bibliography of Chinese-language items are impressive, but its bite-size chapters (thirty-seven in all) and breathless prose style—replete with expressions like ”plummeted,” ”rocketed,” ”beaten to a pulp,” ”beaten black and blue,” ”frenzy,” “ceaseless,” ”frenzied witch-hunt”—are often more reminiscent of the tabloid press than the standard academic monograph. If Yang Jisheng is destined to be China’s Alexander Solzhenitzyn, Frank Dikötter now replaces Jasper Becker as its Anne Appelbaum. The success of MGF should not deter other historians from writing calmer and more nuanced books that worry more about getting the numbers right and pay due attention to geography and history.