Revisiting alleged 30 million famine deaths during China’s Great Leap


Thirty years ago, a highly successful vilification campaign was launched against Mao Zedong, saying that a massive famine in which 27 to 30 million people died in China took place during the Great Leap period, 1958 to 1961, which marked the formation of the people’s communes under his leadership. The main basis of this assertion was, first, the population deficit in China during 1958 to 1961 and, second, the work of two North American demographers, A J Coale (Rapid Population Change in China 1952-1982, 1982) and Judith Banister (China’s Changing Population, 1987). No one bothered to look at the highly dubious method through which these demographers had arrived at their apocalyptic figures.

The ‘estimate’ was later widely publicised by Amartya K Sen who built an entire theory, saying that democratic freedom, especially press freedom, in India meant that famine was avoided while its absence in China explains why the world did not know that such a massive famine had taken place until as much as a quarter century later when the North American demographers painstakingly uncovered it.

The capitalist press was happy to reciprocate the compliment by repeatedly writing of “30 million famine deaths,” to the extent that a fiction was established as historical fact in readers’ minds. The London Economist had a special issue on China some years ago, which repeated the allegation of 30 million deaths in three separate articles and refused to publish the Letter to the Editor this author sent contradicting the claim. More recently, in his Introduction to the book Mao Zedong on Practice and Contradiction, which he edited and published in 2006, Slavoj Zizek also mentioned the figure of 30 millions as though it were a given fact. Well known intellectuals have to be taken seriously and the claim examined.

Two routes

There are two routes through which very large ‘famine deaths’ have been claimed – firstly, population deficit and, secondly, imputing births and deaths which did not actually take place. Looking at China’s official population data from its 1953 and 1964 censuses, we see that if the rate of population increase up to 1958 had been maintained, the population should have been 27 million higher over the period over 1959-1961 than it actually was. This population deficit is also discussed by the demographers Pravin Visaria and Leela Visaria. The population deficit was widely equated with ‘famine deaths.’ But 18 million of the people alleged to have died in a famine were not born in the first place. The decline in the birth rate from 29 in 1958 to 18 in 1961 is being counted as famine deaths. The Chinese are a highly talented people, but they have not learnt the art of dying without being born.

There is a basic responsibility that everyone, and more particularly academics, have to be clear and precise about. To say or write that “27 million people died in the famine in China” conveys to the reader that people who were actually present and alive, starved to death. But this did not actually happen and the statement that it did is false.

China had lowered it death rate sharply from 20 per thousand to 12 per thousand between 1953 and 1958. (India did not reach the latter level until over a quarter century later.) After the radical land reforms and the formation of rural cooperatives, there were mass campaigns to clean up the environment and do away with disease bearing pests while a basic rural health care system was put in place. That a dramatic reduction in the rural death rate was achieved, is not disputed by anyone.

During the early commune formation from 1958, there was a massive mobilisation of peasants for a stupendous construction effort, which completely altered for a few years the normal patterns of peasant family life. Women were drawn into the workforce, communal kitchens were established and children looked after in crèches as most of the able-bodied population moved to irrigation and other work sites during the slack season.

We find a graphic description of this period of mass mobilisation in William Hinton’s Shenfan. When this author spent three weeks in China in 1983, visiting several communes – which still existed at that time – he was told every time that “we built our water conservation system during the Great Leap.” The birth rate fall from 1959 had to do with labour mobilisation, and not low nutrition since the 1958 food grain output was exceptionally good at 200 million tons (mt).

Holes in the argument

There was excess mortality compared to the 1958 level over the next three years, of a much smaller order. Let us be clear on the basic facts about what did happen: there was a run of three years of bad harvests in China – drought in some parts, floods in others, and pest attacks. Food grain output fell from the 1958 good harvest of 200 mt to 170 mt in 1959 and further to 143.5 mt in 1960, with 1961 registering a small recovery to 147 million tons. This was a one-third decline, larger than the one-quarter decline India saw during its mid-1960s drought and food crisis. Grain output drop coincided in time with the formation of the communes, and this lent itself to a fallacious causal link being argued by the academics who were inclined to do so, and they blamed the commune formation for the output decline. One can much more plausibly argue precisely the opposite – that without the egalitarian distribution that the communes practised, the impact on people of the output decline, which arose for independent reasons and would have taken place anyway, would have been far worse. Further, without the 46,000 reservoirs built with collective labour on the communes up to 1980, the effects of later droughts would have been very severe. Recovery to the 200 million ton level took place only by 1965. Throughout, however, the per capita food grain output in China even during the worst year, 1960, remained substantially above that in India.

As output declined from 1959, there was a rise in the officially measured death rate from 12 in 1958 to 14.6 in 1959, followed by a sharp rise in 1960 to 25.4 per thousand, falling the next year to 14.2 and further to 10 in 1962. While, clearly, 1960 was an abnormal year with about 8 million deaths in excess of the 1958 level, note that this peak official ‘famine’ death rate of 25.4 per thousand in China was little different from India’s 24.8 death rate in the same year which was considered quite normal and attracted no criticism. If we take the remarkably low death rate of 12 per thousand that China had achieved by 1958 as the benchmark, and calculate the deaths in excess of this over the period 1959 to 1961, it totals 11.5 million. This is the maximal estimate of possible ‘famine deaths.’ Even this order of excess deaths is puzzling given the egalitarian distribution in China, since its average grain output per head was considerably above India’s level even in the worst year, and India saw no generalised famine in the mid-1960s.

Ideological bias

Relative to China’s population, this figure of plausible excess mortality is low and it did not satisfy the academics in northern universities who have been always strongly opposed to socialised production. Coale’s and Banister’s estimates gave them the ammunition they were looking for to attack the communes. How exactly do Coale and Banister reach a figure of ‘famine deaths’ which is three times higher than the maximal plausible estimate? Examining carefully how they arrived at 30 million ‘famine deaths’ estimate, we find that the figure was manufactured by using indefensible assumptions, and has had no scholarly basis.

In the 1982 census, there was a survey on fertility covering one million persons or a mere 0.1 per cent sample of the population, who were asked about births and deaths from the early 1950s onwards. The very high total fertility rate obtained from this 1982 survey is used by them to say that millions more were actually born between the two census years, 1953 and 1964, than were officially recorded. They ignore the birth rate of 37 per thousand derived from a very much larger 1953 sample which had covered five per cent of all households and was specially designed to collect the information on births and deaths used in the official estimates. Instead, they impute birth rates of 43 to 44 per thousand to the 1950s, using the 1982 survey. There is no justification for such an arbitrary procedure of using a much later reported high fertility rate for a long distant past. We know that a distant recall period makes responses inaccurate. These imputed extra births between 1953 and 1964 total a massive 50 million but according to them did not increase by an iota the 1964 population total, 694.6 million, the official figure which they assume as correct. Thus, although all official birth and death rates are rejected by them, the official population totals are accepted. This opportunistic assumption is clearly necessary for their purpose because it allows them to assert that the same number of extra people died between 1953 and 1964, as the extra people they claim were born.  

Fallacious claims

But the demographers are still not satisfied with the 50 million extra births and deaths that they have conjured up. Fitting a linear time trend to the falling death rate of the early fifties is done to say that deaths should have continued to decline steeply after 1958 and since it did not, the difference from the trend meant additional ‘famine deaths.’ Such straight-line trend fitting is a senseless procedure since the death rate necessarily shows non-linear behaviour. It cannot continue falling at the same steep rate; it has to flatten out and cannot reach zero in any population – not even the inimitable Chinese people could hope to become immortal. The final estimate of extra deaths in both authors is raised thereby to a massive 60 million, a heroic 65 per cent higher than the official total of deaths over the inter-censal period.

Having created these 60 million extra deaths at their own sweet will out of nothing, the authors then proceed to allocate them over the years 1953 to 1964, arbitrarily attributing a higher portion to the great leap years in particular. The arbitrariness is clear from the variation in their own manipulations of the figures. Coale’s allocation raises his peak death rate in 1960 to 38.8 per thousand while Banister is bolder and raises it to 44.6 compared to the official 25.4 for that year, and 30 million ‘famine deaths’ are claimed over the Great Leap years after all this smart legerdemain [sleight of hand]. Having violated every tenet of reason, these ‘academics’ may as well have allocated all their imaginary deaths to the Great Leap years and claimed that 60 million died – why hang themselves only for a lamb rather than for a sheep! Seldom have we seen basic norms of academic probity and honesty being more blatantly violated, than in this travesty of statistical ‘estimates.’ And seldom have  noted intellectuals, who might have been expected to show more common sense, shown instead more credulous naiveté and irresponsibility, by accepting without investigation and propagating such nonsensical ‘estimates,’ giving them the status of historical fact. In the process, they have libelled and continue to libel Mao Zedong, a great patriot and revolutionary. They have unwittingly confirmed the principle attributed to Goebbels – that a lie has to be a really big lie and be endlessly repeated; then it is bound to be believed.

Thirty million or three crores is not a small figure. When one million people died in Britain’s colony, Ireland, in 1846-47, the world knew about it. When three million people died in the 1943-44 Bengal famine, the fact that a famine occurred was known. Yet 30 million people are supposed to have died in China without anyone knowing at that time that a famine took place. The reason no one knew about it is simple, for a massive famine did not take place at all. The intellectuals who quote the massive famine deaths figure of 30 million today, are no doubt outstandingly clever in the small, im kleinen, but are proving themselves to be rather foolish im grossen, in the large. A person has to be very foolhardy indeed to say that 30 million people died in a famine without anyone including the foreign diplomats in China and the China-watchers abroad having the slightest inkling of it. And those who credulously believe this claim and uncritically repeat it show an even greater folly than the originators of the claim.

With thanks to People’s Democracy, Vol. XXXV, No. 26, June 26, 2011, Weekly Organ of the Communist Party of India (Marxist)