Poverty and Famines 
by Amartya Sen.
Oxford, 257 pp., £8.95, October 1981, 0 19 828426 8
Show More
Show More

In our hearts, most of us are Malthusians. We associate mass starvation with too many people chasing too little food. There are too many people because they reproduce themselves too fast, in ‘geometric progression’, as Malthus put it. There is too little food because the supply of food cannot keep pace with the pressure of population. From time to time, there are famines, natural disasters like drought, flood or blight, which reduce the supply of food sharply, causing widespread death. Amartya Sen’s view is different. ‘Starvation,’ he says in the opening paragraph of this book, ‘is a characteristic of some people not having enough food to eat. It is not a characteristic of there being not enough food to eat. While the latter is a cause of the former, it is but one of many possible causes. Whether and how starvation relates to food supply is a matter for empirical investigation.’ Sen’s empirical investigation leads him to the startling discovery that in several major recent famines, which killed vast numbers of people, the supply of food did not decline or declined only slightly.

It is worth getting a debating point out of the way. A famine means food shortage leading to widespread death from starvation. The air of paradox about Sen’s discovery arises in part from his unpicking the meaning of ‘famine’ and setting part of the meaning in opposition to the rest. The statement ‘mass starvation often occurs without any reduction in food supply’ is startling enough, but it is slightly less startling than the statement ‘famines often occur without any reduction in food supply,’ because reduction in food supply is built into the concept of famine. In other words, Sen’s discovery really amounts to saying that many recent episodes of mass starvation, like the Great Bengal Famine, should not properly be called famines. I shall not pursue this line of argument any further. Mass starvation is just as horrible whether it is called ‘famine’ or by some other name. I shall, like Sen, use the word ‘famine’ to mean episodes of widespread death from starvation.

In Sen’s scheme, starvation has nothing directly to do with food supply: it is a matter of whether people can acquire command over food. Command over food, or for that matter any scarce good, depends upon the prevailing ‘entitlement system’, the set of rules governing the acquisition, use and transfer of property rights.

Consider a private ownership market economy. I own this loaf of bread. Why is this ownership accepted? Because I got it by exchange through paying some money I owned. Why is my ownership of that money accepted? Because I got it by selling a bamboo umbrella owned by me. Why is my ownership of the bamboo umbrella accepted? Because I made it with my own labour using some bamboo from my land. Why is my ownership of the land accepted? Because I inherited it from my father. Why is his ownership of that land accepted? And so on. Each link in the chain of entitlement relations ‘legitimises’ one set of ownerships by reference to another, or to some basic entitlement in the form of enjoying the fruits of one’s own labour.

Thus the entitlement system is related to an individual’s command over goods in two ways. First, it heavily influences an individual’s initial endowment through the rules governing inheritance. Secondly, it defines the possibilities open to him of transforming his initial endowment into goods he wants through the processes of production, trade, employment, state support etc. Sen calls these possibilities the exchange entitlement mapping (the E-mapping). The E-mapping depends inter alia on whether he can find employment and if so at what wage rate, what he can earn by selling his non-labour assets, what he can produce with his own labour power and the resources he can buy and manage, the cost of the resources he buys, the value of the products he sells, the social security benefits he is entitled to, the taxes he pays, and so on. The E-mapping, in combination with his initial endowment, determines an individual’s exchange entitlement, the set of alternative combinations of goods that an individual can obtain by making full use of all possibilities that are legally open to him. Famines develop when the exchange entitlements of large numbers of people collapse and no longer contain enough food to ensure their survival. This could happen because of loss of endowments and assets. For example, epidemic disease may reduce the fitness to work of agricultural workers (loss of labour endowment); farmers may be forced to sell land or livestock in order to survive, thereby reducing their ability to produce and to survive in the next period (loss of non-labour endowment). But famines can develop even with unchanged asset ownership because of unfavourable movements in the E-mapping. Landless labourers may become unemployed or find that food prices have risen sharply while their money wages are unchanged; cash crop farmers, craftsmen and fishermen may find that the price of their product has declined relative to the price of food. There is no necessary connection between these changes in the distribution of purchasing power and the volume of aggregate food output. Even if food output is constant, a section of the population could be wiped out because they suffer an ‘entitlement failure’ and lose command over food.

Analysing exchange entitlements ultimately requires a detailed examination of the social, economic and political structure of the society in question and of its ‘modes of production’. Entitlement relations obviously differ as between capitalist and socialist systems and between market and non-market economics. But most economies are mixed systems and an individual’s prospects depend upon the details of his position in terms of production relations. For example, a sharecropper and a landless labourer have similar ownership positions in that they do not own land: the difference between them arises in the way they can use the only resource they own – their labour power. The landless labourer is employed in exchange for a money wage while the sharecropper is employed in exchange for ownership of a part of the product of the farm. If food prices are stable and the labourer is employed at a fixed money wage, he is in some ways better placed than the sharecropper who bears the production risk. On the other hand, if food prices rise sharply, the sharecropper is better placed since he receives his payment directly in food while the labourer has to buy his food in the market. The sharecropper, like the owner-cultivator, has a direct entitlement to some food; the labourer’s command over food, like that of a fisherman or a barber, is a trade entitlement.

Having set up his analytical framework, Sen conducts a detailed and scholarly investigation into some recent famines: the Great Bengal Famine of 1943, the Ethiopian famines of 1973/74, the Sahelian famines of the early 1970s and the Bangladesh famine. In each case, he is concerned not merely with investigating the role played by decline in food availability but with exploring the causes of movements in exchange entitlements. These chapters make fascinating reading and some of their highlights are listed below.

1. The hypothesis that famines are caused by food availability decline (FAD) is fairly decisively rejected.’ The Great Bengal Famine of 1943 killed three million people, but food output in that year was only slightly below that of 1942 and substantially higher, both in total and per head, than in 1941, which was certainly not a famine year. The Bangladesh famine of 1974, too, did not involve a decline in food availability. Indeed, the three districts worst affected by the famine, Rangpur, Mymensingh and Sylhet, seem to have been relatively better supplied with food than in the preceding year! The Ethiopian case is more complex, In 1973, food production did decline significantly in the province of Wollo, but not in Ethiopia as a whole. Sen argues convincingly that the famine in Wollo does not support the FAD case since the natural economic unit is Ethiopia, not Wollo. This would not be so if there had been a transport constraint preventing food from reaching Wollo: Sen shows that the ‘transport constraint’ hypothesis does not stand up. Food could have reached Wollo: it did not reach there because the population of Wollo lacked purchasing power. The Ethiopian famine in Harerghe in 1974 and the Sahelian famines of the 1970s did involve reductions in aggregate food output, but, again, less than one might have expected.

2. The macroeconomic influences leading up to the various famines differed considerably. The Bengal famine was clearly a boom famine. The rise in food prices originated in the expansion of monetary demand related to British government war expenditures in the region. The Ethiopian famine of 1973 was a slump famine. The drastic fall of food output in Wollo reduced the purchasing power of the farmers there and caused derived destitution among the suppliers of other goods and services in the region. The Bangladesh famine showed a mixture of boom and slump characteristics. There was an overall inflationary tendency which fuelled the rise in the price of rice, but the floods of June-September 1974 also sharply reduced the demand for labour, leading to unemployment and a decline in money wages. (But Sen is not able to penetrate fully the mystery of why money wages fell more sharply in the districts worst hit by the famine.)

3. The groups which suffered most varied considerably and so did the causes of entitlement failure. In cases where food output did not fall, those with direct entitlement to food – peasants and sharecroppers – escaped relatively lightly. So also did those with guaranteed trade entitlements: for example, the urban dwellers of Calcutta in 1943, who were protected by a subsidised rationing scheme as a matter of government policy. On the other hand, groups with market-dependent trade entitlements proved to be extremely vulnerable. In Bengal, when food prices rose, the money wages of landless labourers, which were ‘ruled by custom’, remained stationary. In some areas, as a result, the wage in terms of rice fell by two-thirds in a matter of a few months. Similarly, fishermen suffered because of a decline in the demand for fish and a severe fall in the price of fish relative to rice. (Another reason for destitution among fishermen was that the Government confiscated their boats in order to prevent the Japanese, who were expected to invade, from using them. In the event, the Japanese did not turn up for other reasons.) The collapse of the exchange entitlement of landless labourers was seen again in the Bangladesh famine, where floods reduced the demand for labour and money wages at the same time as food prices were rising because of an anticipated reduction in food output in the following year.

In cases where food output fell, as in Wollo, there was, of course, acute distress among farmers. An interesting and complex case is presented by the entitlement failures of the pastoralists in Ethiopia and the Sahel who rely on buying cheaper calories from grain by selling animals. They suffered not merely because the drought killed large numbers of animals (an ‘endowment failure’ since their wealth is held in the form of animals and a ‘direct entitlement failure’ since animals are a source of food), but also because the relative price of animals fell sharply in terms of food grain (‘trade entitlement failure’). The reason for the latter phenomenon was that dissaving by herdsmen through the sale of animals was accompanied by a shift of demand from animals to foodgrains, i.e. from ‘superior’ to ‘inferior’ foods, as incomes fell.

4. Food movements through the market, aided and abetted by private speculation, were sometimes beneficial and sometimes perverse. In a boom famine, free food movement can help, since food is attracted into the famine region. Governments have sometimes misguidedly prevented this, as in the Bengal famine. On the other hand, in a slump famine, free food movement can make matters worse. The lack of purchasing power in the famine region could mean that food is actually exported from the famine region. This phenomenon was observed in Wollo and Bangladesh. (It was also observed in the Great Irish Famine of the 1840s.)

To my mind, there is no doubt that Sen has changed the focus of famine analysis in the right direction – away from mere juggling of numbers of per caput availability to the underlying factors governing people’s command over food. The problem with his ‘entitlement approach’ is that it risks being too comprehensive. To be sure, as Sen points out, the entitlement approach to famines excludes explanations in terms of ‘extra-entitlement transfers’ – for example, looting or brigandage – or ‘choice characteristics’, such as voluntary starvation as a form of religious or political protest. Nevertheless, it needs to be sharpened further if it is to be really useful. ‘Exchange entitlement’ in Sen’s scheme refers not to rights as such but to the effective value of the rights. A large number of different factors therefore combine to yield exchange entitlements and it is their relative importance that needs to be established. If famines are to be prevented, is it more important to change initial endowments (through redistribution of wealth, including land) or to change the E-mapping? The E-mapping itself is a product of many different influences. One of the crucial issues is whether the most efficacious way of improving exchange entitlements through the E-mapping is to establish a universal consumer right against the state for a subsistence ration of food. (The state could honour this right by measures such as free distribution of food, provision of social security benefits of adequate real value or guarantees of employment at specified minimum real wages.) An affirmative answer to this question would imply that famines are not primarily to be attributed to the many other features of economic policy and administrative performance which affect the E-mapping and, therefore, exchange entitlements. Yet these other features of economic policy may be very important in causing famines. People may starve because the Government restricts non-consumer rights connected wth labour mobility, freedom of trade etc, thereby reducing incomes or increasing food prices. People may starve, on the other hand, because the Government does not interfere sufficiently with private rights in order to correct market failures. People may starve because the Government creates disincentives to food production, indirectly reducing their command over food. People may starve because of bureaucratic inefficiencies in storing, transporting and distributing food.

The considerations outlined above become even more relevant if the entitlement approach is applied, as Sen surely intends it to be, to wider, more long-term problems such as the redressal of poverty. How should countries go about improving the standard of living of the poorer sections of their populations? To answer that the exchange entitlements of the poor must be improved is only a starting-point. What means should be adopted to achieve this end? The evidence yielded by thirty years of economic development does not point to any simple answers. It used to be thought that fast growth of GNP was the best way to reduce poverty. The experience of Brazil and Mexico is not at all encouraging: they have both had rapid economic growth but this has almost exclusively benefited the richer income groups. Countries such as Sri Lanka, China and Cuba, on the other hand, have made substantial progress in reducing poverty in spite of the slow growth of GNP. In these countries, policy has been characterised by the establishment of rights to minimum quantities of food, minimum wages, guaranteed employment, and so on. Does this suggest that the establishment of such rights is the key to poverty reduction? Not necessarily. The Gang of Four – South Korea, Taiwan, Hong Kong, Singapore – have achieved both poverty reduction and fast growth without the establishment of such rights or guarantees. Indeed, some people would claim that the economic success of these countries was due, in part, to their not having established such rights. A superficial view may suggest that the Gang of Four followed extreme laissez-fairist policies. Recent studies have cast doubt on this. Governments in these countries have been quite active in promoting labour-intensive industrialisation. But they have certainly avoided setting up the usual trappings of the welfare state. Lest we jump to the conclusion that poverty reduction requires a competitive, capitalist economy, we have the counter-example of Yugoslavia, which has also had an excellent record in poverty reduction and economic growth. The evidence thus opens up some challenging questions about economic development. Success and failure in poverty reduction seem to have cut right across differences in political and economic systems. Making useful and substantive generalisations about how to improve exchange entitlements would be the natural direction in which Sen’s entitlement approach should be extended.

Notwithstanding these comments, there can be no doubt that this is an important work on an important subject by an economist of the highest powers. It combines rigorous analysis and meticulous research with a broad vision and a deep humanity. It is written in an easy, extraordinarily lucid prose style.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN


Please include name, address, and a telephone number.

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences