Widespread hunger in the world is primarily related to poverty. It is not principally connected with food production at all. Indeed, over the course of the last quarter of a century, the prices of the principal staple foods (such as rice, wheat etc) have fallen by much more than half in 'real' terms. If there is more demand for food, in the present state of world technology and availability of resources, the production will correspondingly increase. The demand for food is restrained mainly by lack of income. And the same factor explains the large number of people who are hungry across the world. Given their income levels, they are not able to buy enough food, and as a consequence these people (including their family members) live with hunger.

But it is not adequate to look only at incomes. There is need to look also at the political circumstances that allow famine and hunger. If the survival of a government is threatened by the prevalence of hunger, the government has an incentive to deal with the situation. Incomes can be expanded both by policies that raise overall income and also by redistributive policies which provide employment, and thus tackle one of the principal reasons for hunger (to wit, unemployment in a country without an adequate social security system). In democratic countries, even very poor ones, the survival of the ruling government would be threatened by famine, since elections are not easy to win after famines; nor is it easy to withstand criticism of opposition parties and newspapers. That is why famine does not occur in democratic countries. Unfortunately, there are a great many countries in the world which do not yet have democratic systems.

Indeed, as a country like Zimbabwe ceases to be a functioning democracy, its earlier ability to avoid famines in very adverse food situations (for which Zimbabwe had an excellent record in the 1970s and 1980s) becomes weakened. A more authoritarian Zimbabwe is now facing considerable danger of famine. Alas, hunger in the non-acute form of endemic under-nourishment often turns out to be not particularly politically explosive. Even democratic governments can survive with a good deal of regular under-nourishment. For example, while famines have been eliminated in democratic India (they disappeared immediately in 1947, with Independence and multi-party elections), there is a remarkable continuation of endemic under-nourishment in a non-acute form. Deprivation of this kind can reduce life expectancy, increase the rate of morbidity, and even lead to under-development of mental capacities of children. If the political parties do not succeed in making endemic hunger into a politically active issue, hunger in this non-acute form can go on even in democratic countries.

What should rich countries do, and is trade liberalisation the answer?

The rich countries can do a great deal to reduce hunger in the world. First, the displacement of democracies in poor countries, particularly in Africa, often occurred during the Cold War with the connivance of the great powers. Whenever a military strongman displaced a democratic government, the new military dictatorship tended to get support from the Soviet Union (if the new military rulers were pro-Soviet) or from the United States and its allies (if the new rulers were anti-Soviet and pro-West). So there is culpability on the part of the dominant powers in the world, given past history, and there is some responsibility now for rich countries to help facilitate the expansion of democratic governance in the world.

Second, hunger is related to low income and often to unemployment. Poverty could be very substantially reduced if the richer countries were more welcoming to imports from poorer countries, rather than shutting them out by tariff barriers and other exclusions. Fairer trade can reduce poverty in the poor countries (as the recent Oxfam report Rigged Rules, Double Standards discusses in detail). Third, there is a need for a global alliance not just to combat terrorism in the world, but also for positive goals, such as combating illiteracy and reducing preventable illnesses that so disrupt economic and social lives in the poorer countries. Trade liberalisation on the part of the richer countries could certainly make a difference to employment and income prospects of poorer countries. The situation is a little more complex in the case of liberalisation of the poorer countries. Even those countries which have greatly benefited from the expansion of world trade (such as South Korea or China) often went through a phase of protecting industries before vigorous expansion of exports and trade. So, trade liberalisation is partly an answer, but the economic steps involved have to be carefully assessed: the policies cannot be driven by simple slogans.

What is the solution?

There is no 'magic bullet' to deal with the entrenched problem of hunger in the world. It requires political leadership in encouraging democratic governments in the world, including support for multi-party elections, open public discussions, elimination of press censorship, and also economic support for independent news media and rapid dissemination of information and analysis. It also requires visionary economic policies which both encourage trade (especially allowing exports from poorer countries into the markets of the rich), but also reforms (involving patent laws, technology transfer etc.) to dramatically reduce deprivation in the poorer countries. The problem of hunger has to be seen as being embedded in larger issues of global poverty and deprivation.

Countries of the South increasingly seek food self-sufficiency. Could this solve the problem of hunger and starvation?

Food self-sufficiency is a peculiarly obtuse way of thinking about food security. There is no particular problem, even without self-sufficiency, in achieving nutritional security through the elimination of poverty (so that people can buy food) and through the availability of food in the world market (so that countries can import food if there is not an adequate stock at home). The two problems get confused, because many countries which are desperately poor also happen to earn most of their income from food production. This is the case, for example, for many countries in Africa. But if these countries were able to produce a good deal of income (for example through diversification of production, including industrialisation), they can become free of hunger even without producing all the food that is needed for domestic consumption. The focus has to be on income and entitlement, and the ability to command food rather than on any fetishist concern about food self-sufficiency.

There are situations in which self-sufficiency is important, such as during wars. At one stage in the Second World War, there was a real danger of Britain not being able to get enough food into the country. But that is a very peculiar situation, and we are not in one like that now, nor are we likely to be in the near future. The real issue is whether a country can provide enough food for its citizens - either from domestic production or imports or both - and that is a very different issue from self-sufficiency. We have to look at ways and means of eliminating poverty, and to undertake the economic, social and political processes that can achieve that.

Amartya Sen, who won the Nobel Prize in Economics in 1998, is Master of Trinity College, Cambridge. This is a longer version of an article, expanded by the author, that appeared in Le Monde