Transcript 3 to 4

The Ripple that Drowns:
Twentieth-century Famines
as Economic History
Cormac Ó Gráda
[[email protected]]
Estimated Death Tolls from Selected Famines
Year
Country
EM (m.)
% DR
Observations
1693-4
France
1.5
7
Poor harvests
1740-1
Ireland
0.3
13
Cold weather
1
12
Potato blight; policy failure
0.1
7
Poor harvests
9.5 to 13
3
Drought, floods
1876-79 India
7
3
Drought, policy failure
1921-22 USSR
9
6
Drought, civil war
1846-52 Ireland
1868
Finland
1877-9
China
1927
China
3 to 6
1
Natural disasters
1932-3
USSR
5 to 6
4
Stalinism; harvest shortfall
1942-4
Bengal
2
3
War; policy failure; supply
shortfall
1946-7
Soviet Union
1.2
0.7
Poor harvest, policy failure
1959-61
China
15-18
2
GLF; drought, floods
1972-73
India
0.1
0.03
Drought
1974-5
Bangladesh
0.5
0.5
War, floods, FAD
1972-3
Ethiopia
0.06
0.2
Drought; poor governance
1975-9
Cambodia
0.5 to 0.8
7 to 11 Khmer Rouge
1980-1
Uganda
0.03
0.3
Drought, conflict
1984-5
Sudan
0.25
1
Drought
1985-6
Ethiopia
0.6 to 1
2
War; human agency; drought
1991-2
Somalia
0.3
4
Drought, civil war
1998
Sudan
0.07
0.2
Drought
19952000
North Korea
0.6 to 1
3 to 4
FAD; policy failure
2002
Malawi
Negligible
0
Drought
2005
Niger
Negligible
0
Drought
Sources: Lachiver 1991, p. 480; de Waal 1997, p. 106; Devereux 2001, p. 6; Devereux 2002, p. 70; Davis 2001, p. 7;
Ó Gráda 2007.
CHINA 1959-61
Becker (1997) referred to ‘figures of fifty and sixty
million deaths…cited at internal meetings of senior
Party officials’. BUT:
Source
Estimate (m.)
Ashton et al. (1984)
29
Peng (1987)
23
Yao (1999)
18
Houser et al. (2003)
15
Grain Production, Grain Consumption, and
Mortality in China, 1958-65
Year
1958
1959
1960
1961
1962
1963
1964
1965
Grain
Rural
Urban
Death Rate
Production Retention Allocation (per 1,000)
(% change) (% change) (% change)
2.5
-1.8
23.2
12.0
-15.0
-22.6
14.0
14.6
-15.6
-8.0
-35.0
25.4
2.8
8.1
-16.5
14.2
8.5
10.3
-0.3
10.0
6.3
5.1
12.4
10.0
10.3
10.3
10.1
11.5
3.7
3.4
5.5
9.4
Source: Lardy 1987: 381; NBS 1999
Figure 1. Birth and Death Rates in China, 1950-1969
50
45
40
35
per 1,000
30
25
20
15
10
5
0
1950
1952
1954
1956
1958
1960
br
1962
dr
1964
1966
1968
CDR =
18.73 - 1.08T + 0.29TSQ +
(22.43) (-5.44) (3.18)
3.72D59 + 14.98D60 + 4.19D6
(3.19)
(12.84)
(3.60)
Adjusted R2 = 0.937, t-statistics in parentheses
CHINA 1959-61
• Biggest famine ever
• Great Leap Forward
• ‘The bragging wind’
• Close cropping [Lysenko]
• Diversion of labour from agriculture
• Communal dining
• Excessive procurements
• Secrecy, fear
GDP per head in China and Other Selected Countries (1990 G-K $)
Country
Year
GDP per capita
China
1950
439
China
1955
575
Africa
1980
1,538
Chad
1980
339
Guinea
1980
551
UK
1850
2,330
China
1890
540
China
1870
530
China
1820
600
c. 1950 Chinese GDP per head
<< African average in 1980
< GDP per head of all African countries
except Chad
1955: China was still behind all Africa in
1980 except Chad and Guinea
Irish GDP per head before GIF > Chinese GDP
per head in both 1950 and 1955
Chinese GDP per head in 1950 less than in
1870 or 1890, and less in 1870 than in 1820.
IMR 195 per thousand in 1950-55;
e0 39.3 years for males and 42.3 years for
females.
Such data would put China roughly on a par
with pre-famine Ireland.
Q. True or False:
‘The national scale of the famine largely
rules out natural explanations; the
climate of China was too varied for
famine to hit everywhere at once.’
A. False!
Mortality in Less Affected Provinces, 1956-62
15
14
13
Per 1,000 population
12
11
10
9
8
7
6
5
1956
1957
1958
shanghai
1959
heilongjiang
nei mengg
1960
shanxi
1961
shaanxi
1962
Mortality in worst-affected provinces, 1956-62
65
Per 1,000 population
55
45
35
25
15
5
1956
1957
1958
sichuan
1959
anhui
qinghai
1960
guizhou
1961
henan
1962
Date Source
1849
Tawney
1876-78 Tawney
1906-07 WMO
1920-21 Tawney
Toll (million)
Observations
‘said to have destroyed
13,750,000 persons’
9.5 to 13
Drought: Shanxi, Henan
‘Over 24 million perish
from starvation’
Drought; Kiangsu, Anhui
0.5
Drought: Gansu, Shaanxi
1927
D
3 to 6
NW China
1929
D
2
West China
1931
WMO
3.7
Flooding of Yangtze
1936
WMO
5
The ‘New Famine’:
Sichuan
1941-3
D, WMO
5
Drought, conflict: Henan,
Anhui, Sichuan
1959-61 CÓG
15-18
Poor harvests, politics:
Sichuan, Anhui, Henan
Sources: WMO: http://www.wmo.ch/wmo50/e/world/weather_pages/chronicle_e.html
D: Devereux, Famine in the Twentieth Century; Tawney: Tawney, Land and Labour in China
Anhui:
• Devastated in 1876-78
• Northern Anhui epicentre of 1906-7 famine
• A few years later an American missionary
described ‘Anhwei’s fame of late years [as]
only the bitter fame of her sorrow’; Anhwei
was the ‘famine sufferer ‘
• Pearl Buck’s best-selling The Good Earth
(1931) set in Anhui
Henan (formerly Honan):
• 2 million died in in 1928-9
• ‘Of all marks on my thinking, the Honan
famine [of 1943] remains most indelible’
[T.H. White, 1978]
• 1943 famine killed 3-5 million people
Sichuan:
• 1936: famine killed up to 5 million, reports of
widespread cannibalism
• 1941: 2.5 million died
Modelling the Regional Variation in Death
Rates
Variable
[1]
[2]
[3]
LNY
-0.483**
-0.587**
DAGQ
CONSTANT
N
R2
Prob>F
2.846**
24
0.203
0.0015
-1.005**
-1.181**
0.088
24
0.185
0.018
2.893**
24
0.440
0.0000
Robust t-statistics in parentheses; LNY, MESSHALL, and PARTY as given in Yang, Table 7;
** => significant at 1%
Modelling the Regional Variation in Death Rates
Variable
[4]
[5]
[6]
LNY
-0.670**
-0.638**
DAGQ
-1.155**
-1.550
PARTY
-0.210**
-0.186**
MESSHALL
0.003
-0.003
CONSTANT
0.570**
3.743**
3.016**
N
24
23
23
R2
0.283
0.593
0.464
Prob>F
0.008
0.000
0.001
OLS
OLS
Robust t-statistics in parentheses; LNY, MESSHALL, and PARTY as given in Yang, Table 7.
** => significant at 1%
IVREG
party
3.14
.71
-.698953
-.137833
DAGQ
procur5961
44.08
13.1
-.698953
-.137833
DAGQ
messhall
97.8
16.7
-.698953
-.137833
DAGQ
Demographics
[1] ‘Lost’ births:
• Total cost = Excess deaths plus ‘lost’
births?
• Drop in BR is a common phenomenon
• Famine amenhoerria, spousal separation,
lower libido
• Yao [1999] claims 31 million
• But problematic arithmetic
Birth Rates 1957-62: Badly-Hit Provinces
50
45
40
per 1,000
35
30
25
20
15
10
1
2
3
ningxia
4
gansu
sichuan
qinghai
5
henan
6
[2] Gender and Famine:
• Proportionately greater impact on the
male population
• In seven provinces losing population
1957-61 725,000 fewer females and 366,000
fewer females
• In the rest of China both males and female
populations rose by 1.5 million; slight
male edge
[3] Migration as disaster relief:
• Migration spreads disease
• But reduces pressure on scarce resources
• Lifeboat ethics
• Compare Great Irish Famine
• But: USSR in 1932-33
Henan
50
40
30
20
10
0
1957
-10
1958
1959
1960
1961
1962
1963
1964
-20
Anhui
80
60
40
20
0
1957
-20
-40
-60
-80
1958
1959
1960
1961
1962
1963
1964
Guizhou
60
50
40
30
20
10
0
1957
1958
1959
1960
1961
1962
1963
1964
-10
-20
-30
Gansu
50
40
30
20
10
0
1957
-10
-20
-30
-40
1958
1959
1960
1961
1962
1963
1964
Guangdong
20
15
10
5
0
1957
-5
1958
1959
1960
1961
1962
1963
1964
-10
-15
Jilin
50
40
30
20
10
0
1957
-10
1958
1959
1960
1961
1962
1963
1964
Heilongjiang
70
60
50
40
30
20
10
0
-101957
-20
-30
-40
1958
1959
1960
1961
1962
1963
1964
Shaanxi
20
15
10
5
0
1957
-5
-10
1958
1959
1960
1961
1962
1963
1964
[4] What do people die of during famines?
Usually literal starvation accounts for only small
fraction of deaths
• Nutrition-related causes (diarrhoea, dysentery)
• Causes resulting from societal breakdown
(typhus, malaria)
• ‘Modern’ famines: World War II
• USSR 1932-3
• Leningrad versus Bengal
• What about China?
Table 5. Main Causes of Death in Ireland in 1840 and in
Yunnan Province, China in 1940-4
Cause
Ireland, 1840 Yunnan, 1940-4
Smallpox
4.35
6.73
Dysentery/diarrhea
1.04
14.09
Cholera
0.19
11.97
Fever (incl. Typhoid)
12.69
12.08
Other infectious (incl. measles, scarlet fever)
12.36
6.66
5.00
7.25
Coronary, respiratory
15.23
12.66
Digestive
11.44
5.54
Infirmity, old age
19.08
6.16
3.32
2.32
Other and unspecified
15.25
14..44
Total
100.0
100.0
Convulsions
Total violent and sudden (incl. external)
Source: Mokyr and Ó Gráda, ‘What do people died of’, Table 1; Chen,
Population in modern China’, Tables 25 and 26
Rice occupied up to nine-tenths of cultivated
area
Daily rice consumption about 1,500 kcals daily
per adult male equivalent [over 3,000 kcals in
pre-famine Ireland]
Bengal even more dependent on rice than
Ireland on potatoes in the 1840s
Peasant cultivators more dominant in Bengal
than Ireland; still, considerable inequality
Bengal 1943-44: locus classicus for Sen’s reorientation of famine studies away from a
Malthusian to a distributionist perspective.
Sen found producers and grain merchants
had converted a 'moderate short-fall in
production... into an exceptional short-fall in
market release'.
The famine was due in large part to
'speculative withdrawal and panic purchase
of rice stocks... encouraged by administrative
chaos'.
Which ‘Truth’ to Believe?
•
•
•
•
London/Leo Amery (Secretary of State)
Sir John Herbert (Governor of Bengal)
Linlithgow (Viceroy I)/Wavell (Viceroy II)
Leonard Pinnell
• H.S. Suhrawardy/Muslim League
• Syamaprasad Mookerjee (Hindu Mahasabha)
• The Communist Party/People’s War
• The Report on Bengal [1945]
Issue of FOOD SUPPLY still contested.
FIC claimed Bengal contained enough food
in 1943 to feed everybody. Even though
‘total supply, including the carry-over, was
probably smaller in 1943 than in any of the
preceding 15 years’;
still, likely ‘absolute deficiency of supply [was]
of the order of 3 weeks’ requirements’.
Mark Tauger [2004] highlights brown spot
disease [Helminthosporium oryzae]
• Affected west Bengal only. District-level
statistics corroborative
• Overall impact unclear: botanists versus
others
Pinnell: ‘One was inclined to suspect that the
reports might be exaggerated by speculators, but
there was no doubt that there was a crop disease’
Agricultural statistics fallible
 Substitute evidence?
• Contemporary correspondence, memoranda
• Evidence on hoarding
• Winners and Losers
Mid-1943, Herbert (Calcutta) to Linlithgow
(Delhi):
Hitherto I have studiously avoided overstating the
case and I have faithfully reported any day-to-day
alleviation of the situation: I am now in some doubt
as to whether I have not erred in the direction of
understatement.
 …the essential fact remains that we cannot keep
Bengal fed (certainly we cannot assume the
responsibility of rationing in Calcutta and elsewhere)
unless we get food grains from outside
Hoarding
Sen/FIC blame speculation/panic hoarding.
Also position of post-April 1943 Bengali
government and Leo Amery.
This explanation
• suited ruling Muslim League
• rejected need to import emergency
supplies
The poor always, at first at least,
blame speculators.
In times of crisis, mass opinion both educated
and uneducated, likes to picture a small collection
of scapegoats, a few enemies of society who should
be ‘hanged on the lamp posts’. This is a
comforting view for society in general to take
when the faults of society are shared by the
majority of its members [Leonard Pinnell,
Bengal, 1943]
[1] ‘PRUDENTIAL’ HOARDING: correctly
recognizes that there is going to be a
shortage. Invisible hand rules.
[2] ‘EXCESSIVE’ HOARDING: makes a
wrong judgment and overestimates
shortage. The speculator loses. Evil
compensated by ‘glut’ when price falls
below what it would have been.
[1] ‘PRUDENTIAL’ HOARDING:
correctly recognizes that there is going to
be a shortage. Invisible hand rules.
[2] ‘EXCESSIVE’ HOARDING: makes a
wrong judgment and overestimates
shortage. The speculator loses. Evil
compensated by ‘glut’ when price falls
below what it would have been.
[2]  price plummets in due course
BUT
• Fall from 30 rupees per maund in late
August 1943 to 20 rupees a month later
mainly a ‘mirage’.
• In reality market prices fell to ceiling
levels only in December 1943.
•Plausible interpretation is that small
producers held back and consumed
what they held back over time.
•Moreover, the real price during the first
half of 1944 was still higher than before
the crisis.
Calcutta Rice Prices, July 1942-Dec. 1943 (Rupees/maund)
45
40
35
Black
Market
30
Controlled
25
Market
Price
20
15
10
5
0
Jul-42
Sep-42
Oct-42
Dec-42
Feb-43
Mar-43
May-43
Jul-43
Aug-43
Oct-43
Nov-43
However,
• Across Bengal, rice discovered in hoards
represented only small fraction of annual
supply. Failure of ‘drive’ left poor with sense
of foreboding calamity, because actual
shortage much worse than they had been
lulled into believing
• Ashok Mitra: raided warehouses as ICS
official in east Bengal. Found little grain, but
‘just to create an atmosphere against hoarding’
arrested owner and walked him handcuffed
around village before locking him up.
And:
• In Calcutta/Howrah drive against hoarders
produced measly results. Had it been
otherwise, ministers would have proclaimed
this ‘from rooftops’
• Official ‘food drive’ located only 100,000 tons
of rice held in hoards of 400 maunds or more
throughout Bengal, and only 43,000 tons of
rice and paddy requisitioned.
Pinnell memo:
I let it be known…that I was willing to
guarantee a reward of 4 annas a maund on any
concealed stock successfully disclosed to be of
10,000 maunds or more in one place: but I
never got information on a single case. It
is true that profits to be made by concealment
were large, but a minimum of Rs. 2,500 is not
a small sum for a common informer.
Pinnell, cont.
My conclusion then was, and still is, that the
idea that a few people sitting on a large
quantity of concealed stocks was a facile myth.
I do however believe that probably an
enormous number of people were each sitting
on quantities which were individually small
but collectively ruinous in a year of shortage.
In the vast majority of cases, fear and not
greed was the motive.
WINNERS AND LOSERS:
Short-run context suggests 2-sector specific
factors model
[1] Pure ‘entitlements famine’ 
• ↑ in PF -> ↑ in w, though less than ↑ in PF
• ↑ in PF -> ↑ in AgLF and QF
[2] ‘FAD’ famine 
• ↓ in QF due to a downward shift in PFnctn
• Relative ↑ in PF ; ↓ in AgLF, QF, w, w/PF,
and rent.
[1] and [2] imply loss to wage-earners and
food consumers.
[1] predicts producers gain, [2] predicts they
don’t.
CONTEMPORARY SURVEYS
MAHALANOBIS [1946]: pioneering, much
cited: n=16,000
 landless suffered most. But lots of
landholders became labourers
DAS [1949]: anthropologist’s survey of
Calcutta destitutes
One in five ‘had enough land to maintain
themselves throughout the year’ before
1943
LAND TRANSFERS
Macro:
16.4 m. landholders in 1940; 2.7 m. sales in
wake of famine
Micro [Mukerjee, 1947]:
54 of 168 families in village in Faridpur [E.
Bengal] alienated all or part of holding in
1943, mostly due to ‘scarcity and food
purchase’.
Amery to Linlithgow, 5 August 1943:
Famine in Greece has been, I imagine, even
worse than in Bengal and one of the most
urgent needs of the immediate future will be
the shipping of food into Greece to help the
insurgents, of whom something like 50,000 are
under arms today and playing a really
important role in the whole war effort.
CONCLUSION:
Accounts of both Chinese and Indian famines
pay insufficient attention to exogenous
harvest shocks
In China, politics mattered
• But history and geography also mattered
In India the blame is put mainly on speculators
• Also public inaction and procrastination
• But ‘Mars’ mattered more than ‘Malthus’