The Food Problem: Theory and Policy

Author(s): Amartya Sen and Amaryta K. Sen

Source: Third World Quarterly , Jul., 1982, Vol. 4, No. 3 (Jul., 1982), pp. 447-459
Published by: Taylor & Francis, Ltd.

Stable URL: https://www.jstor.org/stable/3990688

REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/3990688?seq=1&cid=pdf-
reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Taylor & Francis, Ltd. is collaborating with JSTOR to digitize, preserve and extend access to
Third World Quarterly

This content downloaded from

154.59.124.38 on Fri, 04 Nov 2022 21:09:03 UTC
All use subject to https://about.jstor.org/terms
 THE FOOD PROBLEM: THEORY AND POLICY*

Amartya Sen

The various approaches to the food problem that can be found in the literature
fall broadly into two groups. One group emphasises the natural sciences and
engineering, and relates the food problem to technological issues of various
kinds. The other group concentrates on social issues, including political
economy, and sees the food problem primarily in social terms. At the risk of
oversimplification, the two classes of approach may be called 'nature-focused'
and 'society-focused', respectively. These are not, of course, pure, unmixed
categories; the classification reflects the relative emphasis that is placed on the
different factors. It is really a question of focus rather than of coverage.
It is fair to say that the nature-focused view has been traditionally the
dominant one. Scepticism about the contribution that the social sciences can
make to the food problem is not new. In fact, the antiquity of the nature-focused
view of the food problem is brought out by an interesting conversation in Plato's
Statesman, one of his celebrated dialogues involving Socrates. The so-called
Stranger says the following:

We come lastly to the getting of food and of all the substances the parts of which are
capable of combining with the parts of the body to promote its health. This will make a
seventh class and call it 'nourishment' unless we can find some better name for it.
Provision of it is rightly to be assigned to the arts of farming, hunting, gymnastics,
medicine, or butchering rather than to political science.'

To this Young Socrates replies: 'Of course.' It is with that 'of course' that much of
this lecture will be concerned. The role of political economy and politics in the
nature of the food problem is the subject matter of this article.

Malthusian Pessimism

Robert Malthus's Essay on the Principle of Population, published in 1798, falls

fairly clearly on the nature-focused side as far as his analysis of the origin of the
food problem is concerned, though he also happened to think rather little of the
ability of the natural sciences to expand the possibilites of food production.
Malthus saw the food problem in terms of the growth of food supply falling
behind the expansion of population, and saw both these growths as being
primarily determined by nature. This is, of course, the context in which Malthus
* This article is based on the Special University Lecture delivered on 23 February 1982, at Imperial
College of Science and Technology, University of London.

I Plato, Statesman, 288e in the 1578 edition by Henri Estienne (Stephanus); English translation in E
Hamilton and H Cairns (eds), Plato: The Collected Dialogues, Princeton University Press: 10th
printing, 1980, p 1057.

July 1982 Volume 4 No. 3

This content downloaded from

154.59.124.38 on Fri, 04 Nov 2022 21:09:03 UTC
All use subject to https://about.jstor.org/terms
448 THIRD WORLD QUARTERLY

used his famous argument about food production growing in arithmetic

progression, while population grew in geometric progression, soon overtaking
the former. Mathus's particular fascination with the AP and the GP is not really
crucial to his argument, and reflects - I suppose - an attempt to get profound
insights from elementary mathematics - a tendency not altogether unknown in
modern economics as well. What is important is his belief that: (1) the crucial
variable is the ratio of aggregate food supply to population, and (2) there is a
natural tendency for that ratio to fall, leading to starvation and other calamities
until the ratio goes up again as a result of increased mortality. I shall refer to the
former (i.e., the concentration on the ratio of aggregate food supply to
population) as the 'Malthusian focus', and the latter (i.e., the natural tendency
for that variable to fall) as 'Malthusian pessimism'.
While Malthus did suggest checking population growth through 'moral
restraint' to combat what he called 'the excessive and irregular gratification of
the human passions',2 his scepticism of the actual possibility of achieving such a
'moral' solution is also abundantly clear. Malthusian pessimism reflects his view
of a natural conflict, but more importantly, the Malthusian focus itself represents
a far-reaching abstraction from various social influences on hunger, starvation
and mortality, making the food problem turn on the ratio of two physical
magnitudes.
Malthusian pessimism has not been well vindicated by history. His fears have
proved to be not merely ill-founded but fundamentally misconceived, as the
enormity of technical progress and the vast expansion of food production - far in
excess of the growth of population which has itself been very rapid - have shown
in the span of nearly two centuries since the publication of Malthus's well-known
Essay. However, Malthusian pessimism has survived this bit of empirical failure,
as indeed economic theories often do (witness the performance of so-called
'monetarism' and its nine lives). But there is a belief - backed by some quite
respectable reasoning - that while Malthus's fears did not come true in the first
two hundred years, the stage is now set for his pessimistic predictions to be
realised. The natural constraints on food production, it is argued, will now begin
to bite in a way they have not in the past, and signs of decadence, it is alleged, can
already be seen around us.
Indeed, this pessimism has been extended from food supply to other influences
on the standard of living. Various studies commissioned by the Club of Rome
and other organisations have tried to estimate the potentials of future economic
growth in comparison with the growth of population.3 Various types of crises
with different timings have been read into the results that have emerged from
such studies. While some models have not predicted disaster, others have, and

2 Robert Malthus, 'A Summary View of the Principle of Population,' 1830 (revised version of his
piece for the Encyclopaedia Britannica, 1824); reprinted in T Malthus, J Huxley and F Osborn,
Three Essays on Population, Calcutta: Oxford & IBH, 1969, p59.
3 See particularly D H Meadows, D L Meadows, J Randers, and W W Behrens III, The Limits to
Growth, Washington DC: Potomac Associates, 1972; J W Forrester, WorldDynamics, Cambridge,
Mass.: Wright-Allen, 1971; M D Mesarovic and E Pestel, Mankind at the TurningPoint, New York:
Dutton, 1974; A Peccei, The Human Quality, Oxford: Pergamon, 1977.

This content downloaded from

154.59.124.38 on Fri, 04 Nov 2022 21:09:03 UTC
All use subject to https://about.jstor.org/terms
 THE FOOD PROBLEM: THEORY AND POLICY 449

some have indeed sounded screeching alarms, e.g.: 'Basically the prediction is
that mankind has perhaps 40 or 50 years left ... The human race will be wiped out
- mostly or completely - by the year 2100.'4
The assumptions underlying these pessimistic models have received a great
deal of scrutiny recently, and there is little doubt that the shrill announcements of
disaster and doom are not easy to justify in terms of rigorous economic
reasoning. There are a great many arbitrary assumptions in the calculation and
the results happen to be quite sensitive to the precise values assumed.5 Further,
gloomier assumptions have often been fed into the models than could be justified
in terms of the available evidence. The more recent studies give far less
pessimistic pictures. These studies include a variety of methodology used
respectively in the United Nations World Model,6 the so-called Latin American
World Model,7 the extensive study by Interfutures,8 the Global 2000 report
commissioned by Jimmy Carter,9 and even the later study done for the Club of
Rome itself, called MOIRA (Model of International Relations in Agriculture). 10
They also leave much more room for policy response. It does not look as if the
human race is about to be wiped out. (At least not for these reasons, though I
believe it might be easy to underestimate our capacity and propensity to blow
ourselves - literally - out of existence through devices precisely aimed to do that.)
Malthusian pessimism has typically been presented so forcefully that the
impression has often been created that the world population has already been
growing faster than the world food supply. That is most certainly not the case -
indeed there has been a steady increase in the amount of food output per head
even in recent decades. In fact, except for a part of Africa there is no substantial
region in the world in which the food supply trend has fallen behind population
growth. Some countries with well-known food problems, e.g., India, have moved
from being consistent big importers of food to being typically self-sufficient and
often more than that. While the future of the world food supply cannot be
predicted with any confidence, there is really little reason to expect that food
supply will presently start falling behind the growth of population.
I would now like to argue that the real problem is not Malthusian pessimism,
but what - for want of a better term - I would like to call Malthusian optimism. If
the Malthusian focus is retained, concentrating on the ratio of food supply to

4 T H Nelson, Computor Lib, Chicago: Hugo's Book Service, 1974; quoted in Council on
Environmental Quality and the Department of State, The Global 2000 Report to the President,
Harmondsworth: Penguin Books, 1982, p 612.
See, for example, H S D Cole, (ed), Models of Doom, New York: Universe, 1973; Interfutures,
Facing the Future, Paris: OECD, 1979.
6 W Leontief, et. al., The Future of the World Economy, New York: Oxford University Press, 1977.
A 0 Herrera et. al., Catastrophe or New Society? a Latin American world model, Ottawa:
International Development Research Centre, 1976; Handbook of the Latin American WorldModel,
Paris: UNESCO, 1977.
8 Interfutures, Facing the Future, 1979.
9 Council on Environmental Quality and the Department of State, The Global Report to the
President, 1982.
' H Linnemann, MOIRA: a model of international relations in agriculture, Amsterdam: North-
Holland, 1981.

This content downloaded from

154.59.124.38 on Fri, 04 Nov 2022 21:09:03 UTC
All use subject to https://about.jstor.org/terms
450 THIRD WORLD QUARTERLY

population and ignoring everything else, and that ratio is seen to be rising rather
than falling, a sense of optimism about the food problem can be generated that
would be both unjustified and dangerous. It is this question of Malthusian
optimism that I would like to take up next.

Maithusian Optimism

If one were, to start with, worried about food supply falling behind population,
and then were to find out that the converse happened to be the case - with food
supply outrunning population - one might well end up being quite unworried,
even smug. But that optimism, based on the Malthusian focus, can be quite
unfounded if the Malthusian method of analysis is itself wrong. Is the causation
of starvation and malnutrition best seen in terms of the relation between the
physical magnitudes of food supply and population? Are famines caused by the
decline of food availability per head? Are the Malthusian categories the right
ones to use in studying the food problem?
I have argued elsewhere - in a book called Poverty and Famines" - that food
availability per head is a very poor indicator of starvation. Major famines have
taken place without any significant reduction in the ratio of food to population,
and indeed some famines have occurred during years of peak food availability.
Furthermore, the Malthusian focus has contributed to some of these famines not
being anticipated, thereby causing a great many more deaths. The focus on food
per head and Malthusian optimism have literally killed millions.
The problem may be illustrated with an example. The Bengal Famine of 1943,
which killed about three million people,'2 was arguably the largest famine of this
century, though there are also other claimants to that distinction. In terms of
food availability per head, 1943 was not an exceptionally bad year, and indeed
just two years earlier in 1941 the availability of food per person in Bengal had
been a great deal less. One reason for the lack of government efforts to save the
population from the famine was the absence of any recognition by the
government that such a famine was developing. There were other reasons of
course, including the deeply unsympathetic nature of the British government
then in power in India (of this the nationalist critics made a great deal, with good
reason), and the priorities of fighting the war (of this the apologists of the regime
made a great deal, again with good reason). But in addition to these general
problems, the Malthusian focus on food output per head powerfully contributed
to the absence of public policy to counter the famine.
The British Indian government went on calculating food availability in
aggregate terms and saw no real reason for alarm.'3 Indeed, even when the
famine erupted in Bengal with people dying in the streets, the government
evidently had some difficulty in believing what was happening. The govern-
ment's data about food availability in Bengal was fairly accurate, but its theory of

" Poverty and Famines: an essay on entitlement and deprivation, Oxford: Clarendon Press, 1981.
12 See Poverty and Famines, Appendix D.
13 See Poverty and Famines, chapter 6.

This content downloaded from

154.59.124.38 on Fri, 04 Nov 2022 21:09:03 UTC
All use subject to https://about.jstor.org/terms
 THE FOOD PROBLEM: THEORY AND POLICY 451

starvation - based on the Malthusian focus - was totally wrong. Given the choice
between theory and facts, the government stuck to its theory - a phenomenon
that is not unknown to us in Britain today. When eventually the government did
concede the existence of a large famine in Bengal, more than a million had
already died, and it was difficult to turn the forces of mortality back. The
epidemic diseases including malaria and other fevers, cholera, smallpox,
dysentry, etc., were already in full swing - as they typically are in these famines -
and could not be made to stop before millions more were to perish. Interestingly
enough, when the government did eventually concede the existence of the famine,
it still stuck to its theory including the Malthusian focus, and simply revised the
facts, by claiming the existence of an unobserved decline in the stocks of food
carried over from previous years. There was no direct evidence whatever to back
this claim, and there was little indirect reason to expect such a decline.14 The
Malthusian focus proved hardier than the contrary hard facts.
But why did food availability per head prove to be such a bad indicator of
starvation possibilities? What is really wrong with the Malthusian focus? Does
not commonsense suggest that starvation must be closely related to the
availability per head of the commodity food, the absence of which leads to
starvation? I don't think commonsense suggests any such thing, nor does
economic reasoning, but it is an issue that requires a closer examination, and this
is what I take up next.

Ownership and Entitlement

If we live in a society in which food is equally distributed among all the members
of the society, quite clearly it will be the case that the command over food that
each person has will be simply given by the aggregate food availability per head.
We don't live in such a society, and indeed there is no such society. In every
society that exists, the amount of food that a person or a family can command is
governed by one set of rules or another, combined with the contingent
circumstances in which that person or that family happens to be placed vis-ai-vis
those rules. For example, in a private ownership market economy, how much
food a person can command will depend on (1) what he owns, and (2) what he can
get in exchange for what he owns either through trade, or through production, or
some combination of the two. Obviously, in such an economy a person may
suddenly face starvation either because his ownership bundle collapses (e.g.,
through alienation of land to the money lenders), or because the 'exchange
entitlement' of his ownership (i.e., the command of what he owns) collapses (e.g.,
through his becoming unemployed and not being able to sell his labour power, or
through a decline in his terms of trade vis-a'-vis food).
A person starves when he cannot establish his entitlement to the food that he
needs. This is not directly related to the aggregate food availability per head in
the area, and in so far as aggregate availability has any effect at all, it must work

14 See Poverty and Famines, pp. 62-3, 78-83.

This content downloaded from

154.59.124.38 on Fri, 04 Nov 2022 21:09:03 UTC
All use subject to https://about.jstor.org/terms
452 THIRD WORLD QUARTERLY

through some variable or other that affects the person's legal entitlement to food.
Such links, of course, do exist. A general shortage of food could raise the price of
food vis-a'-vis other goods and make it more difficult for a person to buy food in
the market. But that is only one influence among many, and it is quite possible for
food prices to rise even without any decline in food availability per head, e.g., due
to an increase in the demand from others competing for the same volume - or
indeed an increased volume - of food. This is basically what happened in the
Bengal Famine of 1943, with a war-based boom leading to a food price inflation
related primarily to an expanded demand."5
Furthermore, starvation can take place with little or no rise in food prices, and
may be caused by factors such as loss of employment, or a decline in the output of
goods that people make and sell to buy food, or a fall in the relative price of these
goods. In the big Ethiopian famine of 1973, centred on the province of Wollo,
hundreds of thousands died without any sustained rise in food prices.'6 In the
1974 famine in Bangladesh, while food prices did rise, the mortality of rural
labourers - the most affected occupation group - was also directly linked with the
loss of employment. The employment loss was due to the floods which destroyed
the work opportunities immediately. As it happens this occurred in a year of peak
food availability, and the famine was, in fact, over before the food output fell
after the appropriate gestation lag, many months after the floods.'7
If starvation is seen in terms of a failure of entitlement rather than within the
Malthusian focus of food supply per head, much of the mysteries of modern
famines disappear. Rather than concentrating on the crude variable of food
output per head, which is just one influence among many affecting the
entitlement of different groups to food, the focus of analysis has to be on the
ownership patterns of different classes and occupation groups and on the
exchange possibilities - through production and trade - that these groups face.
The forces leading to famines affect different occupation groups quite
differently, and famine analysis has to be sensitive to these differences rather than
submerging all this in an allegedly homogeneous story of aggregate food supply
per head affecting everyone's food consumption. The world isn't like that.
It is important to see the divisive nature of famines and starvation. There
probably has never been a famine in which every group has suffered. There are, of
course, many stories to the contrary, but they don't stand up to examination. The
authoritative Encyclopaedia Britannica, in its vintage Eleventh Edition, does
support the view - often asserted - that the Indian famine of 1344-5 was one such
famine in which everyone suffered. The Encyclopaedia states firmly that even 'the
Mogul emperor was unable to obtain the necessaries for his household'. This
remarkable story does not bear scrutiny. This is not merely because the Mogul
empire was not founded until 1526, that is, nearly two centuries after the famine

15 See Poverty and Famines, chapter 6, and also my 'Starvation and Exchange Entitlements: a general
approach and its application to the Great Bengal Famine,' Cambridge Journal of Economics (1)
1977.
6 See Poverty and Famines, chapter 7.
1 See Poverty andFamines, chapter 9, See also M Alamgir, Famine in South Asia -political economy of
mass starvation in Bangladesh, Cambridge Mass.: Oelgeschlager, Gunn and Hain, 1980.

This content downloaded from

154.59.124.38 on Fri, 04 Nov 2022 21:09:03 UTC
All use subject to https://about.jstor.org/terms
 THE FOOD PROBLEM: THEORY AND POLICY 453

in question! The Tughiak king who did rule much of India then - Mohammad
Bin Tughlak, to be exact - not only had no difficulties with his own household
necessities, he also succeeded in organising one of the most extensive famine
relief programmes in history."8
As a matter of fact, Robert Malthus himself had the occasion to comment on
the different impacts of food shortage on different sections of the society. In his
Investigation of the Cause of the High Price of Provisions, published in 1800 - two
years after his essay on population - Malthus discussed the question of
distribution between the rich and the poor. While the shortages typically affect
the poor, Malthus considered the possibility that redistributive relief through the
Poor Laws pursued too vigorously could lead to a situation in which 'all the
people would starve together' - a prospect which Malthus evidently relished
rather little compared with the usual unequal sharing of the shortage. He went on
to reassure the presumably nervous reader that 'there is no kind of fear, that any
such tragic event should ever happen in any country.'"9
Despite being interested in the distribution between the poor and the rich,
Malthus did not pursue the distributional question sufficiently to recognise the
limitation of his focus on aggregate food availability per head. In fact, famines
divide the poor themselves just as they divide the poor from the rich, and
depending on the position of different occupation groups vis-a'-vis entitlement
relations, groups that are normally equally poor end up in quite different boats in
a famine situation. I shall have to come back to this question again presently
when I go into policy issues, but here I just note that Malthus's categories are
much too broad and crude. This crudeness also prevents him from seeing even
when food availability per head is unchanged or increased.

Policy: Anticipation and Relief

While understanding the causes of starvation and famines is of some interest on

its own, its main interest must lie ultimately in the policy implications that can be
derived from this understanding. Food policy does, of course, have different
facets, which require different types of analysis. At the risk of oversimplification,
the policy issues connected with starvation and famines may be divided into three
broad categories, namely (1) anticipation, (2) relief, and (3) prevention.
I have already discussed how and why the Malthusian focus on food
availability per head can lead to failures of anticipation, delaying preventive and
curative action. The entitlement approach points towards the variables that have
to be watched to anticipate any growing food crisis whether or not it is
accompanied by a decline in food availability per person. While the Malthusian
focus is aggregative, the entitlement focus - by its very nature - is group specific.
The ownership positions as well as exchange entitlement possibilities of different
occupation groups can differ radically from each other, and correspondingly
Ix See Poverty and Famines, p 43.
'9 T R Malthus, An Investigation of the Cause of the Present High Price of Provisions, London, 1800, p
18.

This content downloaded from

154.59.124.38 on Fri, 04 Nov 2022 21:09:03 UTC
All use subject to https://about.jstor.org/terms
454 THIRD WORLD QUARTERLY

economic changes of one kind may affect one group while changes of a different
kind will affect quite a different group. People who normally share the same level
of poverty may be torn asunder by the dynamics of economic change.
For example, in the Bengal famine of 1943, the rural wage-labourers were
among the hardest-hit group and provided the largest number of famine victims,
whereas peasants and share-croppers, who are typically not much richer, were
among the least hit. A rise in the price of foodgrains vis-ai-vis rural wages ruined
the rural labourers, but the peasants and share-croppers were to a great extent
shielded from the effects of price change by the simple fact that their incomes
typically take the form of a quantity of foodgrains itself. Similarly, urban
labourers, particularly in Calcutta, who are normally not very much better off
than the rural labourers were helped both by the war boom and by a system of
rationing to survive the famine in good shape, while the unprotected rural
labourers took the full impact of the rise in food prices - without a boom in
incomes and without rationing - and went to the wall.20 While the focus on food
supply per head would not have - and indeed did not - give any clue to the
coming doom of vast sections of the rural population, a more discriminating
analysis focusing on entitlements could have led to early identification and
possibly even vital anticipation before the event, providing scope for remedial
public policy.
The approach of entitlements also provides guidance regarding relief of
famines should it occur or threaten to occur. Moving food into famine areas will
not in itself do much to cure starvation, since what needs to be created is food
entitlement and not just food availability. Indeed, people have perished in
famines in sight of much food in shops. This was widely noted in the Bengal
famine of 1943. There was no run on food in the markets in Dessie, the capital of
Wollo in Ethiopia, during the famine in Wollo in 1973. During the Irish famines
in the 1840s, this was also a noted feature, and in late 1846 when people were
dying in vast numbers from starvation, the local Relief Inspector, Major Parker,
sent the following report from Skibbereen on 21 December: 'On Saturday,
notwithstanding all this distress, there was a market plentifully supplied with
meat, bread, fish, in short everything.'2' Since famines reflect a collapse of
entitlement, famine relief has to take the form of generating entitlements through
other channels.
There are many obvious ways of generating entitlements. The simplest, of
course, is to provide free food in relief centres. It is an expensive form of relief
since it is difficult to discriminate between famine victims and others who might
not be averse to having some free food. A somewhat more organised form of
relief is to set up work centres and pay food in exchange for employment. One
side benefit of this is to get some useful things done through the work of the relief
centre workers, e.g., digging wells or canals, but also the necessity of work can
make, if the system is well organised, the abuse of the facilities that much more
20 See Poverty and Famines, chapter 6, and Appendix B.
21 See C Woodham-Smith, The Great Hunger. Ireland 1845-9, London: Hamish Hamilton, 1962; New
English Library edition, 1975. p 159.

This content downloaded from

154.59.124.38 on Fri, 04 Nov 2022 21:09:03 UTC
All use subject to https://about.jstor.org/terms
 THE FOOD PROBLEM: THEORY AND POLICY 455

difficult. When a famine is caused by the loss of employment of rural labourers -

as many famines in India or Bangladesh have been - creating alternative
employment is an obviously sensible response. There is little doubt that the
extensive floods in West Bengal in October 1978, in which about 3.5 million
people lost their livelihood would have led to quite a large famine but for quick
and effective relief operations. In fact, as a result of public policy there was no
real famine, even though the incidence of malnutrition certainly did go up.22
Both direct food relief as well as food-for-work programmes require adequate
stocks of food in the public distribution system. Sometimes such reliefs cannot be
organised because that condition is not met. For example, even when the
existence of the Bengal famine of 1943 was belatedly acknowledged by the
government, the famine was not officially 'declared', since such a declaration
would have created the obligation on the part of the government to offer relief in
requisite quantities, as specified by the Famine Code of 1883. Sir T Rutherford,
the Governor of Bengal, wrote to the Viceroy: 'The Famine Code has not been
applied as we simply have not the food to give the prescribed ration.'23 The relief
operations in the Bangladesh famine of 1974 were similarly delayed because of
the lowness of the stock of food-grains in the public distribution system. That
problem was, incidentally, vastly worsened by the decision of the United States
government to cut off food aid to Bangladesh just at the peak of the famine
because of Bangladesh's temporary refusal to accept the US demand that exports
of Bangladeshi jute to Cuba be stopped. US food aid was resumed only when
Bangladesh cancelled further exports of jute to Cuba, by which time the famine
was largely over.24
The importance of having a sizeable stock of food in the public distribution
system in famine-prone economies cannot be overemphasised. The most
effective way of immediately creating entitlements for a famine-affected
population is through public relief, and this is certainly a great deal easier if the
relief can directly take the form of food distribution. The question, however, is
worth asking whether the second-best policy in the absence of enough food in the
public distribution system is to do nothing, or whether cash relief, say, through
work programmes can help. It should be obvious that it could help if the
additional purchasing power would lead to more food being drawn into the
famine affected area. Very rarely would this condition not be fulfilled. While
transport difficulties are often emphasised and can in some relatively rare cases
be decisive, they very often are not so. Indeed, some of the cases that are typically
cited as examples of food not moving into the famine area because of transport
failure have turned out - on closer examination - to be not such cases at all, the
non-movement being primarily due to the absence of effective demand for food

22 See my 'Family and Food: sex-bias in poverty,' in P Bardhan and T N Srinivasan, (eds), Rural
Poverty in South Asia, (forthcoming) to be published by Columbia University Press.
23 See N Mansergh, (ed), The Transfer of Power 1942-7, vol. IV, London: HMSO, 1973, p363,
Document No. 158.
24 See D F McHenry and K Bird, 'Food Bungle in Bangladesh, Foreign Policy, 27 (Summer) 1977; R
Sobhan, 'Politics of Food and Famine in Bangladesh,' Economic and Political Weekly, (14) 1979.

This content downloaded from

154.59.124.38 on Fri, 04 Nov 2022 21:09:03 UTC
All use subject to https://about.jstor.org/terms
456 THIRD WORLD QUARTERLY

in the famine area.25 Cash reliefs can indeed help in creating this effective
demand, i.e., in translating human needs into market demand to which traders
will respond.
There is an interesting 'classical' issue here that may be worth clearing up. The
authority of Adam Smith has often been quoted by non-interventionists in
famine situations. The role of market mechanism in moving food to distress areas
has been emphasised in that context, e.g., by the Governor of Bombay in 1812 in
justifying his refusal to intervene in a famine that was developing in Gujerat,
referring to 'the digression of the celebrated author of the Wealth of Nations
concerning Corn-trade' as being 'irresistibly applicable to every state of society
where merchants or dealers in grain may be established'.26 Such non-
interventionist policy has almost always led to disastrous results, since private
merchants and traders will not move food to famine victims when their needs are
not translated into money demands. Indeed, frequently food does move out of
famine areas when the loss of entitlement is more powerful than the decline - if
any - of food supply, and such food 'counter-movement' has been observed in
famines as diverse as the Irish famines of the 1840s, the Ethiopian famine in
Wollo of 1973, and the Bangladesh famine of 1974.
The failure of the market mechanism to help famine victims has puzzled non-
interventionist public servants. During the Orissa famine of 1865-6, Commis-
sioner Ravenshaw expressed astonishment at the lack of food movement into the
famine area despite what he understood to be 'the ordinary rules of political
economy', under which 'the urgent demand for grain ought to have created a
supply from other and more favoured parts.'27 Of course, political economy says
no such thing, since there is no incentive for the traders to move food to famine
victims as they lack purchasing power. (Poor Adam Smith must have grown very
used to turning in his grave.)
So far, so bad. But when cash relief is given to famine victims, that does, of
course, provide the missing link, creating effective demand for food. Even in the
absence of food in the public distribution system, it is most plausible that food
will move in response to the increased market demand. It is most unlikely that
doing nothing could be even the second best policy - to be chosen when there is
an inadequate public stock of food - since adding purchasing power itself
contributes to the entitlement of the famine victims and would make the food
supply respond. The case for public sector trading would also have to be
considered in that context, especially if private trade has elements of monopoly
or oligopoly.

Policy: Prevention and Securing

The entitlement approach also suggests lines of analysis for preventing famines

25 See Poverty and Famines, chapters 7 and 10.

26 See S Ambirajan, Classical Political Economy and British Policy in India, Cambridge University
Press, 1978, p 71.
27 See Ambirajan, p 76. On this general question, see also S Rashid, 'The Policy of laissez-faire during
Scarcities,' Economic Journal, (90) 1980, and my Poverty andFamines, chapter 10.

This content downloaded from

154.59.124.38 on Fri, 04 Nov 2022 21:09:03 UTC
All use subject to https://about.jstor.org/terms
 THE FOOD PROBLEM: THEORY AND POLICY 457

and avoiding starvation. It is a question not just of providing relief but of creating
circumstances in which failures of entitlements will not occur. The reason why
starvation is typically absent in the richer countries is not just the high level of the
average income or average wealth in these countries but the arrangements for
social security that provide a minimum entitlement to everyone. Even in a rich
country widespread unemployment would lead to starvation and possibly even
to a famine but for social security in the form of unemployment insurance,
supplementary benefits, etcetera.

Can a poor country afford to provide security to everyone in these ways? It is

common to argue that it cannot since this would prove much too expensive. I
doubt this. I am not arguing that social security of this kind would be cheap; but
then large costs might well be justified if the benefits are great, and what is at
stake is the most elementary issue of life and death. Some poor countries have, of
course, done much in this direction. China is an example, even though data
problems make it difficult to assess its exact achievements, and there are big
disputes that are currently going on within China as to what was or was not
achieved. Sri Lanka is a country which has tried a system of free rice distribution,
treating it as a general right of all those below a certain level of income. The costs
have been substantial but evidently not unbearable.

It is sometimes argued that by spending money in providing free rice to all

those who need it, Sri Lanka has starved capital formation and reduced its
economic growth. While Sri Lanka's growth rate has not been exceptionally low
- no lower than the average performance of South Asia - it has certainly not been
very high either. On the other hand, Sri Lanka's concern with social services and
social security - the rice policy is only one part of this - has certainly contributed
to the high expectation of life in that country, which comes close to European
standards, and this is quite astounding for a country as poor as Sri Lanka. There
is obviously some trade-off here. It is interesting to ask how long would it have
taken Sri Lanka to reach its present level of longevity at birth if instead of trying
to do it through social services it did it through faster economic growth,
following the pattern of other developing countries. There is obviously much
uncertainty in calculations of this kind, and it is necessary to try out alternative
assumptions regarding the effect of transferring social service expenditures to
capital formation of the usual kind.

Sri Lanka's longevity level is, of course, very much higher than what would
correspond to its income per head in cross-country comparisons. What is
calculated first is the income level at which Sri Lanka would have achieved its
present level of longevity in that international fit. Then it is calculated, with
alternative assumptions, how many years it would take Sri Lanka to raise its
income per head from what it is now to what it would need to be for it to achieve
the same level of longevity as it already actually has if Sri Lanka were just another
country in the cross-country international comparison. The answer depends on
the exact effects of transferring its social service expenditure to straightforward
capital formation, and it turns out that the answer lies somewhere between 58 to

This content downloaded from

154.59.124.38 on Fri, 04 Nov 2022 21:09:03 UTC
All use subject to https://about.jstor.org/terms
458 THIRD WORLD QUARTERLY

152 years" - a very long haul indeed no matter which figure we choose. To
replicate what Sri Lanka has got from its social service-oriented public policy
through the more traditional means of capital formation and growth would take
a terribly long time. Paying direct attention to people's needs rather than doing it
through economic growth is not such an expensive strategy after all in terms of
human life. Social security may be expensive, but the rewards are high too.
How easy it might be for other countries to emulate Sri Lanka's social service-
oriented policy is a difficult question to answer. The government of Sri Lanka
itself has recently tended to underplay these remakable aspects of its public
policy. Certainly, the fact that Sri Lanka has so far avoided getting into heavy
military expenditure and has indeed shunned so-called 'defence' makes it rather
unique among the developing countries - indeed in the world. It might look as if
this makes Sri Lanka's achievements hard to emulate, but I believe Sri Lanka's
experience also helps to bring out the real social costs and sacrifices induced by
defence expenditure.
I turn now to aspects of prevention other than social security. Failures of
entitlement - as I have already discussed - might originate either in a fall in
ownership bundles (e.g., through the alienation of land, or loss of grazing
grounds for animals), or in a fall in the 'exchange entitlement' (e.g., through
unemployment, or worsening of terms of trade).29 While everyone might be
ultimately vulnerable to these fluctuations, some groups are a great deal more
vulnerable than others. In fact, two groups stand out in terms of special
vulnerability judging from empirical studies of modern famines. They are
respectively landless rural labourers, especially in South Asia, and pastoralist
nomads, especially in sub-Saharan Africa.
Landless rural labourers have to live by selling the only substantial thing they
own, namely, their own labour power. If they can't sell it, i.e., not find
employment, they must starve. If they do find employment but food prices rise to
make the market command of their wages decline sharply, then again they may
have to start starving. Even their normal wage rates earned over periods of high
agricultural activity typically leave them no option but to starve in less busy
seasons - often called 'the hungry months'. While vagaries of weather have
always produced uncertainty in the lives of the agriculturalist, the early stages of
capitalist development in agriculture make the rural labourers particularly
vulnerable. Historically, as the landless rural proletariat emerges in a peasant
economy, the security of ownership of land - no matter how small -
characteristic of a peasant economy is lost, and at the same time social security
arrangements, which are characteristic of more advanced capitalist economies,
are yet to come. This produces exceptional vulnerability for the class of rural
landless labourers in this intermediate period, and, indeed, famine after famine
have brought home how terribly vulnerable this class actully happens to be.
The pastoralist in sub-Saharan Africa has also been made more vulnerable by
28 See my 'Public Action and the Quality of Life in Developing Countries,' Oxford Bulletin of
Economics and Statistics, (43) 1981, pp. 301-6.
29 See also K Griffin, International Inequality and National Poverty, London: Macmillan, 1978.

This content downloaded from

154.59.124.38 on Fri, 04 Nov 2022 21:09:03 UTC
All use subject to https://about.jstor.org/terms
 THE FOOD PROBLEM: THEORY AND POLICY 459

economic development, in this case by the development of capitalist agriculture,

encroaching on grazing land - often the very best grazing land, e.g., in the Awash
river valley in Ethiopia. This has made the pastoralist even more prone to disaster
in drought situations.30 But on top of that the dependence of the pastoralist on
the market mechanism adds sharply to his vulnerability. The pastoralists
typically sell animal products to buy cheaper calories in the form of foodgrains,
which in the Sahel and in the Horn of Africa accounts for about half the calories
they consume. A rise in the price of foodgrains vis-a-vis animal products can ruin
the pastoralist. And, indeed, it did just that in the Ethiopian famines of 1973-4.
The drought reduced both the agricultural output and the animal stock, but the
economic impoverishment led to a general shift in demand away from richer
food, namely, animal products, to relatively less expensive food, namely,
foodgrains. There was as a consequence a sharp fall in the terms of trade of
animal products vis-a-vis foodgrains, making it hard for the pastoralist to get
enough to eat.3'
In the prevention of starvation and famines, particular importance has to be
attached to monitoring the position of specially vulnerable groups. The
entitlement approach helps not merely to identify which groups are vulnerable,
but also precisely why they are vulnerable. Food security cannot be built without
public policy being geared to countering the sources of these different types of
vulnerability.

Food and Society

I began with the contrast between nature-focused and society-focused approaches

to the food problem. I have tried to show why food is a social problem and it is not
just a question or raising food production vis-a-vis population. It is not my
intention to deny, in any way, the importance of technological issues in the
production of food, or those related to its storage and transportation - they are
indeed important. But, ultimately, the food problem is not concerned just with the
availability of food but with the disposition of food. That involves economics,
politics and even law.32 Starvation and malnutrition relate ultimately to
ownership and exchange in addition to production possibilities.33
I believe Young Socrates was much too rash in agreeing so readily that the
problem of nourishment did not belong to political economy and to political
science. It does, in fact, belong to political economy and to political science. There
is, indeed, no such thing as an apolitical food problem.
30 See L Bondestam, 'People and Capitalism in North-East Lowlands of Ethiopia,' Journal of Modern
African Studies, (12) 1974; G Flood, 'Nomadism and Its Future: the Afar,' RoyalAnthropological
Institute News (RAIN), (6) 1975; also my Poverty and Famines, chapters 7 and 8.
3' See Poverty and Famines, pp. 104-1 1.
32 See my 'Ingredients of Famine Analysis: availability and entitlements,' Quarterly Journal of
Economics, (95) 1981.
33 In this article I have not gone into the question of distribution of food within the family. There is
much evidence for an age bias (against the children) and a sex bias (against women) in the division
of food within the family in developing countries and even in some developed ones in periods of
economic crises. On this question see my 'Family and Food: Sex-Bias in Poverty,' referred to
earlier. This issue adds a further social dimension to the food problem.

This content downloaded from

154.59.124.38 on Fri, 04 Nov 2022 21:09:03 UTC
All use subject to https://about.jstor.org/terms