The Welfare Consequences of the Great Leap Forward
Famine in China, 1959-61: The Stature of the Survivor
OUTLINE PAPER AND SELECTED TABLES AND
Stephen L MORGAN
The University of Melbourne
Session 123: Famines in History
International Economic History Congress, Helsinki, 20-25 August 2006
Past studies of the Great Leap Forward (GLF) famine in China of 1959-61 focus on
the causality or the economic consequences of the famine, but few examine the
welfare of the survivors. Between 15 and 30 million people are estimated to have died,
the worst famine in world history, which stemmed from the GLF program of rapid
industrialization. Human height, an indicator of nutrition, is used to examine the
impact on the survivors of the famine who were born from the late 1950s to the early
1960s. Data are drawn from a survey of 112,000 industrial workers born between the
1940s and 1970s that was conducted in 1994, and anthropometric surveys of children
conducted between 1950s and the 1990s. Trends in average stature are analyzed in
conjunction with data on food availability, mortality, and economic output. These data
for the first time enable us to quantify the impact on the welfare of the GLF survivors.
Dr Stephen L. Morgan,
Senior Lecturer in Asian Economic History,
Department of Management,
The University of Melbourne,
Melbourne, Victoria, 3010
Tel: +61 3 8344-5282
Fax: +61 3 9349-4293
DRAFT PAPER OUTLINE. Please do not cite without contacting the author for a
revised and extended update of the paper.
Forty-five years ago China was emerging from the worst famine in human history that
claimed between 15 and 30 million lives. Uncertainty still surrounds the estimates of
how many died as a consequence of Great Leap Forward (1958-60) famine, and the
causality and explanations for the famine are much debated. Besides the “excess”
deaths, many millions of children who ought to have been born during the period were
not, which produced an unusual narrowing in the age-specific profile of the country’s
population, followed by a post-famine bulge as birth rates rapidly rose in a
compensatory reaction, which continues to influence China’s demography. A
particular gap in the economics literature is the near absence of any assessment of the
welfare consequences for those who were born or lived through the famine as
children. Nearly anyone who has done field work in China’s countryside has
anecdotal stories about the obvious stunting of those who were infants at the time, and
the mid-to-late-40s contract workers on urban construction sites are similarly shorter
than either older or younger works. The aim of this paper is to present anthropometric
evidence that indicates the welfare consequences of the collapse in food nutrients.
My interest in the post-1949 famines or near-famines is two fold. Firstly, can we
apply an entitlements approach to the analysis of famines in ‘socialist’ societies in
which the state determines a person’s entitlements to food, even for the primary
producer of food? Until the entitlements approach of Amartya Sen (1981, 1990;
Osmani 1995), famines were explained as a product of the decline in the availability
of food that may have been brought on by war, weather failures or other impacts on
the supply of food, which results in hunger, starvation and large-scale death from the
diseases associated with hunger. This a simple story of supply-side failure – a food
availability deficit (FAD). However, famines often occur without any decline in food
availability. Sen argued that famine was the result of a change in the entitlements
regime that lessened a person’s ability to be able to acquire food. People starved
because of a failure in direct entitlements or in the mapping of exchange entitlements.
A direct entitlement failure is when a farmer cannot produce enough food because of
a change in their endowments, such as appropriation of land or illness that prevents
them tilling it. The more common failure is that of exchange entitlements where a
person is unable to trade something to obtain sufficient income to exchange for food.
A change in exchange entitlements could arise from an inability to trade cattle or
labour power for money with which to buy food, for example, or a change in the
relative prices of commodity bundles that make the exchange of one type of bundle
insufficient to buy the required bundle of food nutrients. The entitlements approach
“views famines as economic disasters, not just food crises” (Sen 1981, 162).
Understanding the different types of entitlement failures becomes the key to
understanding the causality of any particular famine. Sen’s framework, though, is
predicated on a market environment. For more than three decades until the 1980s
markets were barely relevant to the working of the Chinese economy. Grain
production, procurement and distribution – that is, the net availability of the staple
food to any group of Chinese – was determined by State policy. Therefore, I would
argue, an explication of the state-directed entitlements regime is at the centre of any
understanding of the origin and severity of any famine or near-famine occurrence in
China between 1949 and the mid-1980s.
Secondly, my interest arises from a current project to use anthropometric data
(stature and weight) to estimate changes in the relative well being of the Chinese over
the past two centuries which is not captured by conventional economic measures for
the standard of living, either because the data do not exist or because they are
unreliable (Morgan, 1998, 2000, 2004, 2005, 2006; Morgan and Liu, 2003,
forthcoming). Stature is an indicator of the history of net nutrition of a person,
especially the net nutrition during sensitive growth phases, while weight is a measure
of current nutrition. Combined in the form of the body mass index (BMI: weight in
kilograms divided by the square of the stature in metres, kg/m
) we have a powerful
indicator of the capacity of a person to engage in productive physical and mental
activity, which is nothing less than a biological indicator of the potential productivity
of a person (Fogel 1993, 2004). When we possess sufficient data we are able to
calculate average stature and other measures for the population that enables us, in
conjunction with whatever other data may be available, to say something about the a
biological well being of the population.
This paper will begin with a brief historical remark about famines in China and
the early economic development of the PRC. Next it will set out some of the recent
literature, focusing on the data, interpretations and models for explaining the Great
Leap Forward famine in particular, drawing mostly on a special issue of the China
Economic Review (Fall 1998). Thirdly, the paper will introduce the anthropometric
data for selected provinces affected by famines or near famines in China’s recent
historical past, and conclude with some preliminary results. The revised version of
the paper should be available for the conference panel session.
Famines in China
China in the western imagination was a “land of famines” (Mallory 1926), teaming
with peasants who lived, in the description of Tawney (1932), on the threshold of
survival like a man with his head barely above the water who would drown were the
water to rise another inch. There is a large literature on famines, food supply and
nutrition in China. There are various historical studies (Nathan 1965, Will 1990),
accounts of famines by missionaries (Richard c.1916) and of missionaries (Soothill
1926, Bohr 1972), numerous reports of famine and disaster relief committees (Central
China Famine Relief Committee 1912 and National Flood Relief Commission 1933
cited in Bohr 1972, the UN Disaster Relief Office Report on Drought and Floods in
Hebei/Hubei 1982 cited in Kane 1988), accounts of the demographic impact of the
Great Leap Forward (Aird 1982, Ashton et al 1984, Bannister 1987, Peng 1987, Kane
1988), discussion of nutrition and well being (Piazza 1986, Kueh 1993, Riskin 1990,
1991, Morgan 1998, 2000), and a raft of recent analyses of mostly post-1949 famine
and rural situations (Becker 1996, Bramall 1993, Chang and Wen 1998, Kueh 1995,
Lin and Yang 1998, Riskin 1998, Yang 1996, Yang and Su 1998; PLUS several
recent pieces). Brief as this list is, Chinese famines and food issues have been the
subject of many studies over the past century. Famine and food have also been central
to traditional economic policy making in China. The modern Chinese word for
economics, jingji, is but a 19
century Japanese contraction of the classical Chinese
phase jing she ji min (regulate the affairs of world to succour the people) that was the
basis of imperial Chinese administrators’ intervention in economic matters. Central to
that intervention was the need to maintain food supply through the maintenance of the
water-control systems that would enable crops to be grown and transported, and the
proper regulation of markets and employment that would enable the populace to
acquire adequate food. Were it not, as was so often the case in the closing years of
many dynasties, uprising and rebellions would progressively proliferate.
At the start of the sixth decade of the People’s Republic of China the majority of
Chinese are better off than they have ever been. Most are richer, with better nutrition
and housing than in the past. The transformation of the past 25 years is phenomenal,
despite evidence of widening disparities in income distribution. When economic
reform began there was ample evidence that most Chinese were living in a state of
endemic hunger, with per capita nutrient availability sufficient only for moderate
work intensity (Piazza 1986: 11, 71-80, 195). Although there were significant
improvements in nutrient availability from the 1950s, for most years the Chinese
average level was on or below the accepted international safe level (Piazza 1986: 77,
195). Per capita grain output, a rather crude measure of potential nutrient availability,
was below the 1957/58 peak for every year other than 1974 from the Great Leap
Forward until the start of economic reforms in l979 (see Fig 1). Our knowledge,
though, of how poorly China was doing between the late 1950s and late 1970s only
became apparent with the release of data in the early 1980s. Many scholars who had
once seen China as having achieved considerable balance in its development program
radically reversed their judgements (compare Nicholas Lardy’s writings of 1978 and
1980 with those of 1983 and 1984). China’s post-1949 development strategy, far from
being balanced or “rural biased”, was much like any other country’s industrialisation
strategy: the peasants paid for it through various and harsh extractive means
orchestrated by a communist party, rather than an indigenous capitalist class. It was
not simply “urban biased”, but “state biased”. The foundation policy of the PRC for
several decades was an urban-oriented state-biased economic development strategy.
One of the consequences of the strategy was that following the economic
recovery and consolidation of communist power in the early 1950s, the party needed
to ensure delivery of rural produce to the state. The peasant farmers’ propensity to
consume the greater output during the relative peace and stability of the 1950s got in
the way of the party’s modernisation strategy. Rapid expansion of agricultural output
was not reflected in greater delivery of that output to the state, principally for the
development of industry (to supply the new factories, feed the industrial workforce,
and barter for capital goods from the Soviet Union). The difficulties were the impetus
for the increasing pace of “socialisation” of industry and commerce, and of
“collectivisation” of agriculture from the mid-1950s (these developments are nicely
surveyed in Riskin 1987). It culminated in the Great Leap Forward, which saw the
implementation of a rapid industrialisation program in the countryside (best
characterised by the backyard furnaces) and the creation of the people’s communes,
the major institutional change of the period. The horrendous demographic outcome,
which saw China actually loose population (Fig 2), was only revealed in the early
1980s when China released data on the 1964 census in preparation for the UN-
supported 1982 national census. The data, though, are still subject to debate. There is
surprisingly little analysis of this major catastrophe, especially from scholars resident
in the PRC. “The Great Leap Famine is the least studied and understood of the major
famines that have occurred in the world” largely for reason that the Chinese
government does not encourage discussion of the period (Johnson 1998).
Insert Fig 1 and Fig 2 about here
Data problems. There is agreement among scholars that the scale of the deaths was
enormous. The exact number of excess deaths, the magnitude and timing of the rise in
mortality, and the nature of the morbidity associated with the famine remains
uncertain. This is due largely to the collapse of the statistical reporting system.
Official statistics were compiled from retrospective surveys (Kane 1988).
The problem is we cannot simply answer the question ‘how many people died?’
Estimates range widely, from 15 million implied by official statistics, to 23-30 million
among academic demographers such as Aird (1982), Peng (1987) and Bannister
(1987), to a recent former official’s estimate of 43 million (Riskin 1998). These
uncertainties complicate not only the assessment of the severity of the famine, but
also attempts to refine the dating of its onset and end, which have implications for
determining the role of government policy or other factors in the famine. There is also
inadequate information on the distribution of relief, the provision of medical care for
different populations groups, and specific causes of death (most people do not starve
to death but die from illness caused by severe nutritional deficit).
As for the accuracy of Chinese official statistics, most scholars accept these as the
best we have whatever their biases and inadequacies. Riskin (1990, 23 n6) observed:
“Checks of internal consistency and other considerations have convinced most
independent scholars that government statistics for most of the post-1949 have been
accurate expressions of what the Chinese government believed to be true.” Deliberate
falsification has rarely occurred. Rather, the problems are missing data, systemic
inaccuracies, or reluctance to disclose how statistics were constructed, the weights
used and definitions, which require scholars to exercise caution in their use of official
Causality and interpretation. There is much disagreement among scholars over the
relative weight to give to any particular factor in explaining the onset of the GLF
famine and its severity, especially the role of Chinese government policy. It is best to
begin this section with a brief summary of what is generally agreed.
The grain harvest is the major determinant of overall food availability in China,
and grain was the main source of energy and protein for the Chinese. Food grain
output has fluctuated widely. The famine began after the largest grain crop on record
and for the most part excess deaths rose before the harvest failures of 1959. From the
1958 yield of 200 million metric tonnes, grain output fell 29% to 1960. The average
per capita consumption in the countryside fell from 204 kg in 1957 to 154 kg in 1961,
and the national average per capita daily calorie intake in 1960 was about 1500
calories, well below the estimated safe limit for the period of about 2100 (Riskin
Decline in food availability (FAD) in any particular province could have been due to a decline in
production, a rise in exports or a peculiar geographic factor (eg, weather failure), or a combination of
these. The inadequacy of statistics mean we still do not know enough about grain supply and other
economic indicators in any particular province. Some, such as Anhui, one of the worst affected, remain
a statistical black hole. Important income and consumption data for Anhui for most years before the
late 1970s are missing from even the most recently published SSB historical statistics volumes.
1990, 1998; Piazza 1986). NEED TO ADD NOTE + CHECK whether the average is
total or adult equivalent. Check sources used by Riskin and Piazza.
The exaggeration of the 1958 grain yield at the time, reported up to twice the
1957 harvest, gave the impression the country had ample, and was instrumental in
government policies to raise procurement, reduce area sown to grain, and promote
exports. In 1959 the state procured 39.7% of total output (the net was 28% after
distribution to deficit areas) compared with a pre-GLF average around 25% (net about
17.5%), and exported 4.16 million metric tonnes (Riskin 1990: 29, 44). The collapse
in output produced a classic FAD-type famine, but most scholars argue that a decline
in food availability was only partial explanation of the famine. FAD does not account
for the rise in mortality that occurred in 1958, nor does it account for the great
variations in the severity of the famine among China’s provinces. Similarly an
improvement in food availability does not explain the end of the famine in 1962 – per
capita availability of grain in 1962 was below that of 1959, and about the same as
1960, the worst year of the famine. Official data estimates national per capita grain
consumption in 1960 at 164 kg and at 165 kg in 1962. Other explanations must be
sought. The only plausible explanations appear to be policy changes. Communal
kitchens were abolished, private plots were again allowed, land was turned over to the
household in some places, and the amount of grain procured was reduced, which
coincided with significant import of grain from Australia and other suppliers.
The severity of the famine was a product of many variables, of which FAD was a
factor, but not necessarily the most important. About other factors and causality there
is much disagreement. The arguments are: a game theoretic approach to model the
impact on productivity of the lost of exit rights from the collective (Lin 1990); the
modification of the exit-rights argument with a Sen-based entitlements analysis of the
urban bias of food distribution (Lin and Yang 1998); a “tragedy of the commons”
argument that focuses on food waste arising from the communal dining system,
resulting in a “failure in consumption efficiency”, that along with other radical
political policies undermined property rights governing incentive structures (Cheng
and Wen 1997, 1998, Yang 1996, Yang and Su 1998).
Lin (1990) uses a game theory hypothesis to explain the collapse of production
and ensuing famine. In the fall of 1958 the nature of the collective changed, from
voluntary to compulsory participation. This was a change from a repeated to a one-
time game. The lost of the right to exit the collective had a significant impact on the
incentive structure of the collective, which produced a decline in productivity and the
subsequent collapse in grain output. “When the freedom to quit exist, at the end of
each production round, members of a collective can decide whether they want to
participate in the collective in the next round. Those who find they are better off being
members of the collective will retain their membership. Otherwise, they will withdraw
from the collective,” argues Lin (1990, 1241). Participation in a collective allows
members to share the benefits of pooled resources, the economies of scale unavailable
to the typical individual Chinese household farmer. Due to the difficulty of adequately
supervising agricultural work, the success of the collective depends on “a self-
enforcing contract” in which each member promises to put in as much effort as they
would on the household farm. If a member shirks, others can decide to leave; the
shirker, who has benefited from the collective without due effort, needs to modify her
actions to preserve the collective if she wishes to benefit from the scale economies in
the second round. “If the discounted present value of future losses is larger than the
one-time gains in this round, the member will honor the agreement. Therefore, it is the
threat of a collective’s collapse that greatly reduces the incidence of shirking” (Lin
1990, 1242). The assumption here is that the output of a voluntary collective will be at
least as good as the sum of the households comprising it working separately.
With the compulsory imposition of collective membership in the commune exit is
not permitted. The result is the collective becomes a one-time game, and the self-
enforcing contract fails because “It becomes impossible to use withdrawal either as a
way to protect oneself or as a means to check the possibility of shirking by the other
members. … [Therefore], supervision becomes crucial in establishing work incentives
and productivity levels in the collective” (Lin 1990, 1242). Effective supervision of
agricultural work is very difficult, and the link between present effort and future
reward is weaken in the absence of ownership over output (the reward), which
undermines work incentives and consequently affects adversely productivity. Lin
argues that the game theory hypothesis explains the success of the collectivisation
movement 1952-58 and the sudden collapse of agricultural output 1959-61, which is
supported by evidence for the change in total factor productivity for agriculture (see
chart following the references).
Lin and Yang (1998) restates the exit-right hypothesis, and develop augments
drawing on Sen’s entitlements approach that recognises FAD alone is an insufficient
explanation for famine. The premise is that the causes of a production shortfall should
not be confused with the causes of a famine. Sen’s focus is on the collapse of
endowments that prevent a certain proportion of the population from acquiring food.
Lin and Yang is the only study I am aware of which attempts to explain the right food
as a determinant of the Chinese famine, and do so econometrically. They use panel
data for 28 provinces 1954-66 to explore the relationship between mortality and the
urban-biased ration system that guaranteed city residents food, and left farmers with
the residual. The effect of this system was excessive export of grain from rural areas
to the cities. They concluded that “the Chinese food entitlement system, which was
dominated by urban-biased distribution, explains a greater part of the inter-provincial
variation in mortality rates than does food availability” (Lin and Yang 1998, 131).
Factors other than FAD and food entitlements inform the conclusions of Yang
(1996), Change and Wen (1997, 1998), and Yang and Su (1998). They focus on
communal dining and radical policies as an explanation for the famine. Yang (1996)
argues that the creation of communal mess halls led to the depletion of grain, leading
to hunger and starvation. The argument is based on the tragedy of the commons: when
meals were free individuals will over consume because conservation in the present
will not ensure access in the future as others may have already consumed what an
individual had saved in the past (that is, there are no exclusivity property rights).
Yang uses evidence of a significant and negative correlation between participation in
communal mess halls and mortality rates. Change and Wen (1997, 1998) similarly
argue that famine was a product of “consumption inefficiency”, in that communal
dinning lead to over consumption and waste of food, rather than FAD or entitlements
failure. Yang and Su (1998) is an extension of Yang (1996) and Change and Wen
(1997, 1998), but emphasises the politics of the GLF period. Communal dining was
the institutional manifestation of the radical policies of the GLF, and much blame for
the famine can be sheeted home to the over-enthusiastic implementation of central
policies by provincial and lower level leaders.
This study draws on surveys conducted since the late 1970s, jointly commissioned by
several ministries and conducted under the direction of staff at the Research Institute
of Sports Science of the State Physical Culture Commission (tiyu yundong
weiyuanhui). Most surveys were of school children and youth through to the age of
22, but several surveys have been conducted of other population groups, including
mature and middle aged adults workers. This study uses both types of surveys.
Survey of Adult Workers: As part of a study of the health of workers, 112,530 Chinese
men and women employed in 17 industries plus Beijing municipal government were
measured in 1994 using a variety of indicators, including height and weight. The
62,346 males were aged 18-59 and the 50,184 females were aged 18-54, the age range
reflecting the official age of retirement of 60 for men and of 55 for women. The mean
sample size per age cohort for men was 1484 (range 1223-1904, SD 154) and for
women 1356 (range 111-1666, SD 134). They were employed in 63 large-scale
industrial or service work units, and included the railways, petroleum, coal mining,
construction, machinery, post and telecommunications, chemical industry,
automotive, aeronautics, metallurgy, textiles and government agencies. Any students
or trainees were excluded. For every industry there were about 100 subjects for each
age cohort, with the minimum size group per age for any industry of 34 subjects.
One of the aims of the study was to analyse the health of the middle-age
population to provide data for devising ways to encourage physical activity. There
were 45,726 subjects older than 41 years in the study. While there are various
problems with the reporting of the data – there is no identified weighting for region
and no information is provided about the choice of the enterprises to be included in
– the height data is probably representative of urban employed State
sector employees. Height was measured to 0.1cm intervals with an error level of
±0.2cm specified. Other anthropometric and physiological indicators recorded
included weight, chest and waist measurement, blood pressure, lung capacity, and
various measures of physical exertion.
The results are reported in summary tables. Average height, for example, is
tabulated for each age cohort, the standard deviation and the sample size. More highly
aggregated data are provided by four politico-social classifications: worker (gongren),
cadre (ganbu), scientific-educational (kejiao) and other (qita). These data might have
allowed analysis of the change in height by social background for specific political
and economic events in recent Chinese history. Unfortunately, the data are aggregated
into five-year age groups, rather than annual means, and the numbers of subjects from
each social group in each five-year interval is not specified, only the aggregate total.
A difficulty with the data for the subjects older than 40 years of age is that we need to
correct the reported height for shrinkage that begins during the fifth decade of life,
though correction is not needed for those born during the GLF. The correction method
The sample measurement card included in the field workers’ handbook does not even have a field for
provincial origin (jiguan), which is unusual given Chinese sensitivities to a person’s social background.
I considered requesting access to the underlying data, but given that provincial origin is not reported
either, further analysis of social background without controlling for region of origin would be pointless.
used for Fig 3-7 is a transposition of the estimated age-related shrinkage equation that
Chandler and Block (1991) had obtained from a longitudinal study.
Surveys of School Children and Youth: Small-scale anthropometrical surveys in single
locations have been conducted in China since the 1910s (summarized in Yeh 1959,
Piazza 1986, Morgan 2000). A quasi-national study was conducted in 1975 that
surveyed 273,735 children from birth to 17 years in urban and suburban rural counties
of nine cities (Zhang 1977). Periodic national scale studies have been conducted since
1979, covering 16 provinces and more than 200,000 students in 1979 (Report 1982),
28 provinces and nearly one million students 7-22 years in 1985 (Report 1988), and
30 provinces and about 250,000 students in 1991 (Report 1993), 1995 (Report 1996)
and 2000 (Report 2002). Only Han (ethnic Chinese) are discussed here, though data
on national minority groups were collected from 1985.
Data was collected using a stratified random sampling method. Three urban and
three rural observational districts (including one urban and rural area from the
provincial capital region) representative of poor, middle and upper socio-economic
areas were selected, and within each observational district three observational places
were selected for the 7-18 year group.
Each observational place could be a single
school or a group of neighbouring schools, from which the observational classes were
randomly selected. Finally, the student sample was randomly selected from the
observational classes and examined to ensure they met the health criteria for inclusion
in the final data set.
Preliminary exploration of the GLF
There is disagreement among scholars over the weight for any particular factor in
explaining the onset of the GLF famine and its severity, especially the role of Chinese
government policy. State control of grain production, procurement and distribution,
along with destruction of the alternative sources of income and food that Chinese had
long depended – the rural markets and household garden plot – and destruction of
incentives structures within the collectives are at the centre of any explanation of the
GLF famine. Yet, few have attempted to quantify the welfare outcomes of the GLF.
Below we use height data to measure the impact on welfare.
Final attained adult height is a measure of the cumulative the net nutrition
available for human growth from birth to adulthood (or time of measurement for a
Besides height and weight, there are sitting height, chest and physical exertion measures. Each report
includes the survey procedures and the sample size, mean and standard deviation for each measure.
Urban places were defined as those in the city area (shiqu) or its suburbs (jiaoqu), which included the
provincial capital and another two urban areas. Household registration (hukou) was used to classify
residency. Post-secondary students’ hukou prior to enrolment was used to determine their rural-urban
status. Rural populations in the text are referred to as ‘rural-classified’ population because the author
has not verified the degree to which the data are representative of rural areas.
The normality criteria excluded students with any infectious disease, growth abnormality (dwarfism
and giantism) or physical disability, as well as menstruating women.
child). The change in the average stature of a population over time indicates how well
the human organism has survived its socio-economic and epidemiological
environment. From the trend in average height of a population we can infer change in
income and income distribution, the access to nutrients as influenced by the relative
prices of food, and obtain measures of gender, social and spatial differences in
nutritional status (Eleveth and Tanner, 1990; Steckel, 1995). The GLF produced a
collapse in grain output and a sharp decline in available food energy. The
consequences, for those who were born around this time or were in their infancy, we
would expect to be revealed in reduced stature. The degree to which average height is
reduced depends on the severity and length of nutrient privation, and whether there
was more than adequate nutrition after the effect to allow “catch up” growth to take
place (Eleveth and Tanner, 1990).
Fig 3-7 show the average height of men and women based upon the survey of
adult workers in the early 1990s. Fig 3 reports the measured height of men ordered by
year born and their estimated adjusted height, and Fig 4 shows both series smoothed
using a double-centred three-year moving average.
Fig 5 and Fig 6 report the same
results for women. Fig 7 shows the smoothed series for both men and women. The
adjusted stature – which is a better estimate of actual mature adult height of the older
subjects – shows a secular downward slope during the Sino-Japanese War for both
men and women. Male stature begins to trend upwards from the last year of the Sino-
Japanese War, 1944-45, into the early years of the PRC, despite the economic
instability, hyper inflation and conflict of the Civil War. Women fared less well. The
final attained stature of men and women born during the 1950s rose quickly, with a
gain of about 0.75cm for each in six years to 1957-58. The stature of both men and
women born during the GLF declined (the GLF period is highlighted by a red
rectangle on Fig 3, Fig 5 and Fig 7). The average height for men declined about half a
cm and women about the twice the decline of men. For women, the decline began
earlier than for men, about 1957, and the fall was greater and more prolonged. For
those born from c.1962, the attained height for men and women both rose through to
1971-72, with a slight dip coinciding with the most intense period of the Cultural
Revolution, 1966-67. The decline in height after 1972 may be more due to the effect
of youth – young adult workers had not attained full height – rather than due to the
food crisis in 1972-73 after widespread drought in China’s wheat zones.
Figs 3-7 about here
Using the anthropometric surveys of students through to 25 years of age, we are
able to derive a series of adult heights covering most years of the GLF. To do this we
first needed to run regressions to estimate whether the older students had reached
adult height (stopped growing) to avoid any downward biases from immature
subjects. Based on the regression analysis of height-for-age, were able to determine
A Double-centred three-year moving average is the three-year moving average of a previously
calculated three-year moving average. It has the same span as a five-year moving average but, in my
view, more faithfully tracks secular trends than a five-year moving average while reducing stochastic
when adolescent growth had ceased for the various series and thus filter the data set to
obtain a time series of mature Chinese adults ordered by year of birth.
sequencing does not allow us to use the children survey data to look at the worst years
of the GLF – there is a gap in the constructed series for 1960-62. Nevertheless, the
children series is plotted with the 1994 adult survey data in Figs 18-19. Both series
track each other quite well up to 1959, but less so during the 1960s and 1970s, though
they do indicate severe biological stress was placed on Chinese born in these years.
Such a conclusion would fit with what we know about the impact of the GLF and the
Cultural Revolution periods on demography, grain output, the collapse in food
availability and the agricultural productivity.
Figs 8-9 about here
Famine in China after 1949 cannot be explained singularly by simple appeal to the
failure of weather or FAD or government policy, but by a complex interaction of all
of these elements. It is not plausible to argue, as Becker (1996) does, that Mao and the
CCP leadership sought to kill the population, say in the manner analogous to that of
Stalin in the 1920-30s. Mao’s “crimes” may be judged severely by history, perhaps as
the delusions of a visionary (for the “vision thing” was certainly a powerful driver of
the policy of the period), but it does not seem reasonable to argue that his actions
were, in a premeditated sense, deliberately murderous.
Reflecting on the great variety of evidence – despite the big holes in the statistical
data, and uncertainties about how those data were compiled – we need an analysis that
focuses on the institutional dynamics that set off events which created the famine of
1959-61. It seems to me that we need to map the nature of the entitlements that the
communist state permitted its people, which denied the Chinese the ability to acquire
food. These state-determined entitlements constrained the opportunities of the
population so much more than would seem to be the case for people living in a market
or quasi-market environment, and certainly in a democratic political environment that
would have allowed the dissemination of news of such tragedies that may have
prompted a change in policy or more speedy relief measures.
State control of grain
production, procurement and distribution, along with destruction of the alternative
sources of income and food that Chinese peasants had so long depended — the rural
markets and food production on the household garden plot — and destruction of
For the the 1979 survey, males and females > 19 years of aged (20-25); the 1985 survey, males 19-22
and females 17-22; the 1991 survey, males 18-22 and females 17-22; the 1995 and 2000 surveys, males
17-22 and females 16-22.
Drèze and Sen (1995, 76) observed that “no substantial famine has ever occurred in a democratic
where the government tolerates opposition, accepts the electoral process, and can be publically
criticized. A government which has to deal with opposition parties, to answer unfriendly questions in
parliament, to face condemnation from the public media, and to go to polls on a regular basis, simply
cannot afford not to take prompt action to avert a threatening famine. But a non-democratic country has
no such guarantee against famine.” Cited in Riskin (1998, 122 n.14).
incentives structures within the collectives, are at the centre of any explanation of the
GLF famine. They also explain in varying degrees the endemic hunger that
characterised China until the early 1980s. The institutional change of the late 1970s
and early 1980s that saw the re-introduction of household farming, in its success is an
indictment of the folly of the previous two decades of policy making. The tragedy of
China from the late 1950s to the early 1980s was the tragedy of state failure, no better
exposed than in the perverse mapping of the entitlements regime that denied so many
Chinese access to adequate food and nutrition for more than a generation.
Aird, John S. 1982. “Population studies and population policy in China”. Population and Development
Review, 8(2): 267-297.
Ashton, Basil, Kenneth Hill, Alan Piazza, and Robin Zeitz. 1984. “Famine in China, 1958-1961”.
Population and Development Review. 10(4): 613-645.
Bannister, Judith. 1987. China’s Changing Population. Xx: xxx.
Becker, Jasper. 1996. Hungry Ghosts: Mao’s Secret Famine. New York: Free Press.
Bramall, Chris. 1993. In Praise of Maoist Economic Planning. Oxford: Clarendon.
Bohr, Paul Richard. 1972. Famine in China and the Missionary: Timothy Richard as Relief
Administrator and Advocate of National Reform, 1876-1884. Cambridge, Mass.: Harvard
Chandler, P.J. and R.D. Bock (1991), “Age changes in adult stature: trend estimation from mixed
longitudinal data”, Annals of Human Biology, 18(5): 433-440.
Chang, Gene Hsin, and Guangzhong James Wen. 1997. “Communal dining and the Chinese famine of
1958-1961”. Economic Development and Cultural Chnage, 46: 1-34.
Chang, Gene Hsin, and Guangzhong James Wen. 1998. “Food availability versus consumption
efficiency: Causes of the Chinese famine”. China Economic Review, 9(2): 157-166.
Drèze, Jean, and Amartya Sen. 1995. India: Economic Development and Social Opportunity. Delhi:
Eveleth, P.B., and J.M. Tanner, 1976, 1990, Worldwide Variation in Human Growth, Cambridge:
Cambridge University Press.
Johnson, D. Gale. 1998. “China’s great famine: Introductory remarks”. China Economic Review, 9(2):
Kane, Penny. 1988. Famine in China, 1959-61. London: Macmillan.
Kueh, Y. Y. 1993. “Food consumption and peasant incomes”. In Y. Y. Kueh and Robert F. Ash, eds.,
Economic Trends in Chinese Agriculture: The Impact of Post-Mao Reforms. Oxford: Clarendon
Kueh, Y. Y. 1995. Agricultural Instability in China, 1931-1991: Weather, Technology and Institutions.
Lardy, Nicholas R. 1978. Economic Growth and Distribution in China. Cambridge: Cambridge
Lardy, Nicholas R. 1980. “Regional growth and income distribution in China”, in F. R. Dernberger, ed.,
China’s Development Experiences in Comparative Perspective. Harvard: Harvard University
Press, pp. 153-190.
Lardy, Nicholas R. 1983. Agriculture in China’s Modern Economic Development. Cambridge:
Cambridge University Press, chapter 4
Lardy, Nicholas R. 1984. “Consumption and living standards in China, 1978-83”, China Quarterly,
Lin, Justin Yifu. 1990. “Collectivization and China’s agricultural crisis in 1959-61”. Journal of
Political Economy. 98(6): 1228-1252.
Lin, Justin Yifu, and Dennis Tao Yang. 1998. “On the causes of China’s agricultural crisis and the
Great Leap famine”. China Economic Review, 9(2): 125-140.
Mallory, W. 1926. China — Land of Famines. American Geographical Society.
Morgan, Stephen L. 1998. “Biological indicators of change in the standard of living in China during the
20th century”. In John Komlos and Joerg Baten, eds, The Biological Standard of Living in
Comparative Perspective, Stuttgart, Franz Steiner Verlag, pp. 7-34.
----. 2000. “Richer and taller — stature and the standard of living in China, 1979-1995”, The China
----. 2004. “Economic growth and the biological standard of living in China, 1880-1930”, Economics
and Human Biology, Vol. 2, No. 2, pp. 197- 218.
----. 2005. “Health Status and Occupational Mobility of Chinese in Australia, 1890-1940s”, Chinese
Studies Association of Australia, Biennial Conference, Bendigo, 29 June-2 July.
----. 2006. “Height, health and welfare in South China over the past two centuries”, The Third
International Conference on Economics and Human Biology, Strasbourg, France, June 22-24
Morgan, S. and S. Liu, 2003. “Was Japanese Colonialism Good for the Taiwanese? Stature and the
Standard of Living in Taiwan, 1895-1945”, Economic History Society of Australia and New
Zealand, Conference, 29 Sep-1 Oct
----. forthcoming. “Japanese Colonialism, Physical Stature and the Standard of Living in Taiwan, 1895-
1945”, The China Quarterly
Nathan, Andrew James. 1965. A History of the China International Famine Relief Commission.
Cambridge, Mass: Harvard University Press.
Osmani, S. R. 1995. “The entitlement approach to famine: An assessment. In Basu et al, eds, Choice,
Welfare and Development: Essays in Honor of Amartya Sen. Oxford: Clarendon.
Piazza, Alan L. 1986. Food Consumption and Nutritional Status in the PRC. Boulder: Westview Press.
Peng, X. Z. 1987. “Demographic consequences of the Great Leap Forward in China’s provinces”.
Population and Development Review. 13(4): 639-670.
Putterman, L., and G.L. 1992/93? “The role of exit costs in the theory of cooperative teams”. Journal
of Comparative Economics. 16: 596-618.
Riskin, Carl. 1987. China’s Political Economy: The Quest for Development Since 1949. New York:
Riskin, Carl. 1990. “Food, poverty and development strategy in the People’s Republic of China”. In
Lucile F. Newman, ed., Hunger in History: Food Shortage, Poverty and Deprivation. Cambridge,
MA: Blackwell, p. 3xx-3xx.
Riskin, Carl. 1991. “Feeding China: The Experience since 1949”. In Jean Drèze and Amartya Sen, eds.,
The Political Economy of Hunger. Vol. 3 Endemic Hunger. Oxford: Clarendon Press, 15-58.
Riskin, Carl. 1998. “Seven questions about the Chinese famine of 1959-61”. China Economic Review,
Sen, Amartya. 1981. Poverty and Famine. Oxford: Clarendon.
Sen, Amartya. 1990. “Food, Economics and Entitlements”. In Jean Drèze and Amartya Sen, eds., The
Political Economy of Hunger. Vol. 1 Entitlement and Well-Being. Oxford: Clarendon Press, 34-52.
Soothill, William E. 1926. Timothy Richard of China: Seer, Statesman, Missionary and the Most
Disinterested Adviser the Chinese Ever Had. London: Seeley, Service & Co.
State Statistical Bureau (SSB). 1983- . Zhongguo tongji nianjian (Chinese statistical yearbook). Beijing:
Steckel, Richard H. 1995. ‘Stature and the Standard of Living’, Journal of Economic Literature,
SSB. 1990. Quanguo gehseng zizhiqu zhixiashi lishi tongji ziliao huibian, 1949-1989 (Compendium of
historical statistical materials for provinces, autonomous regions and municipalities). Beijing:
SSB. 1987. Guomin shouru tongji ziliao huibian, 1949-1985 (Compendium of statistical materials on
national income, 1949-1985). Beijing: Tongji.
Ricard, Timothy. 1916. Forty-five Years in China: Reminiscences by Timothy Richard, D.D., Litt.D.
London: T. Fisher Unwin Ltd.
Tawney, R. H. 1932. Land and Labour in China. London: George Allen & Unwin.
Will, Pierre-Etienne. 1990. Bureaucracy and Famine in Eighteenth-Century China. Translated from
French by Elborg Forster. Stanford: Stanford University Press.
Yang, Dali L. 1996. Calamity and reform in China: State, Rural Society and Institutional Change since
the Great Leap Forward. Stanford: Stanford University Press.
Yang, Dali L., and Fubing Su. 1998. “The politics of famine and reform in rural China”. China
Economic Review, 9(2): 141-155.
Fig 1. Per capita grain output index and China population growth
Fig 2. China GLF demographic disaster
population growth and change in total grain output
per capita grain output index
(1957=100), left hand scale
Source: ZGRKNJ 1994, 409; ZGTJNJ 1984, 137, 1995, 347
Death and famine during the Great Leap Forward, 1958-61
rate per thousand
population (m illions)
Total pop (area, RHS) birth rate, LHS death rate natural increase
mean height male
adjusted ht male
Fig 3 Height of Chinese men born 1935-76 (annual means)
Unadjusted 3 year
Adjusted 3 year
Fig 4 Height of Chinese men born 1935-76 (double-centred moving average)
mean ht female
adjusted ht female
Fig 5 Height of Chinese women born 1935-76 (annual means)
Unadjusted 3 year
Adjusted 3 year
Fig 6 Height of Chinese women born 1935-76 (three-year double-centred moving
Fig 7 Average height of adult Chinese men and women ordered by year of birth
(three-year doubled-centred moving average)
student surveys-m adult survey-m
Fig 8 Change in average height of adult men 1954-84
student surveys-f adult survey-f
Fig 9 Change in average height of adult women 1954-84