The Bali Declaration of the Non-Aligned Movement and Other Developing Countries defined food security as "access to food for a healthy life by all people at all times" (NAM, 1994). It recognized that, in spite of a substantial increase in the world's food output, the number of people suffering from hunger and malnutrition has increased during the last decade in many developing countries. The Bali Declaration reaffirmed that "food security should be a fundamental goal of development policy as well as a measure of its success". It called attention to the diverse causes and nature of food insecurity in different localities and countries as well as to the complex policy and institutional issues that have to be dealt with at global, national and sub-national levels in order to assure adequate access to food by all people at all times.1
In order to promote a useful policy-oriented discussion it is necessary to break down the concept of food security into components or criteria which render the concept meaningful. The five dimensions listed below can be applied in evaluating food systems at diverse levels ranging from households and commu-nities to nations and groups of nations.
1. A food system offering security for its participants should have the capacity to produce, store, import or otherwise acquire sufficient food to meet the needs of all its members at all times.
2. It should provide maximum autonomy and self-determination (without implying autarchy), thus reducing vulnerability to market fluctuations and other social and political pressures.
3. It should be reliable, so that seasonal, cyclical and other variations in the access to food are minimal.
4. A secure food system should be equitable, meaning, as a minimum, dependable access to adequate food for all individuals and groups both now and in the future.
5. Finally, it should be socially and environmentally sustainable so that the ecological systems on which all societies and food production depend are protected and enhanced over time.
Some analysts have compressed these five criteria into three: food availability, stability and access (FAO, 1996a).2 This is logically defensible as these three headings implicitly include all five criteria. De-emphasizing questions of autonomy, equity and long-term ecological sustainability, however, may mean that these crucial issues receive less attention than they should in international debates and recommendations dealing with food security issues. Equally troublesome is the fact that in discussions in international fora on food security little attention is paid to the issue of who will implement their recommendations and of how they are to be implemented.
There are several other rather worrying aspects in recent documents and agendas treating food security issues by international organizations that are in large measure dominated by the North. Two underlying assumptions are particularly controversial. These are that the adoption of neoliberal economic policies would nearly always contribute to greater food security, and that a country's food security can be rather adequately indicated by aggregate food availability per capita.
Neoliberal economic policies: There seems to be a rather unquestioning acceptance by most international organizations of the Washington consensus that gives primacy to "market forces", "free trade" and "privatization" in development strategies. These policies are assumed to be necessary conditions for assuring sufficient food production, adequate access to food by the poor and also for good governance, although it is recognized that they are not sufficient ones. There is little in the history of the now rich industrial countries, or of the relatively successful developing ones, to suggest that this has been the course that these states actually followed. Indeed, a historical review of the development strategies and especially of the food and agricultural policies followed by the "developed countries" that belong to the OECD is instructive in this respect. Western European countries, Japan and the United States all have highly "subsidized" agricultural sectors, as well as an array of institutions and policies designed to protect poor food producers and consumers. Producers of basic foods everywhere face a plethora of special problems that require skilful and purposeful state interventions of a kind often antithetical to neoliberal dogmas. The same is true of providing adequate access to adequate food for those who lack it.
Over-reliance on quantitative indicators of food security: The principal quantitative indicator used for international comparisons of food security is estimated average daily per capita food availability at national, regional and global levels. This is understandable as these are data FAO generates and that it considers roughly comparable over the last three decades, although estimates for most developing countries are admittedly very crude with wide margins of error. The assumption that national average per capita food supplies is a good indicator of a country's food security, however, is frequently not justified.
Obtaining comparable quantitative estimates of trends in undernutrition as an indicator of the absence of reliable access to adequate food and of equity in its distribution among different social groups is much more difficult.3 Quantitative indicators of the autonomy of food systems and their long-term ecological and social sustainability are also necessarily extremely partial because of the qualitative nature of these concepts. In fact, good quantitative indications of these aspects of food security may not be possible. Nonetheless, neglecting these crucial dimensions of food security in discussions of the issues can be highly misleading. One has only to recall the wide divergences in undernutrition and of acute hunger in many countries that apparently have more than sufficient food, and the virtual absence of serious hunger in a few others with very tight food supplies, to realize how misleading an indicator national level per capita food availability can be. Needless to say, it can be an even more misleading indicator of food security at regional and global levels. Moreover, it may leave the mistaken impression that the principal issue is a race between population growth and food production.
The documentation prepared for the global initiative to focus on world food security matters-- the 1996 World Food Summit -- suggests that the leading root cause of chronic inadequate access to sufficient food for individuals and households is poverty, although it emphasizes that natural or human-made disasters also often contribute to serious hunger. This is not very helpful because blaming food insecurity primarily on poverty is something of a tautology.4 Poverty lines (the income levels below which households are considered to be living in poverty) in developing countries are commonly determined by estimating the income required for a family or an individual to enjoy a low cost adequate diet together with a few other basic necessities. In low-income countries, food usually accounts for most of the estimated consumption by the poor in terms of market values.
The elimination of poverty is crucial for improved food security by definition. How this might be accomplished in different contexts is a key issue. But an analysis of how poverty can be reduced in both rural and urban areas seems largely to be ignored in international discussions of food security issues (e.g. FAO, 1996b). It is suggested here that to explicitly address poverty issues primarily in terms of food security could contribute to additional insights and sharpened recommendations.
There is no point in this introduction in attempting to qualify or second guess FAO's or the World Bank's estimates of the extent of hunger and undernutrition at national, regional and global levels. Whether well over or somewhat less than one billion people are chronically undernourished in developing countries will not change very much the issues that have to be faced. The same is true of whether the situation is getting slightly better or worse than it was in the 1970s, or whether the numbers suffering acute food insecurity associated with natural catastrophes, wars, economic embargoes and other disasters is less than 100 million or over 200 million. The estimates are necessarily very rough and the concepts used in making them tend to be extremely controversial. The fact remains that hunger in a world of plenty remains morally, socially and politically unacceptable whether it affects a large or small portion of any country's population.5
A little reflection suggests that the broad concept of food security outlined above is practically the same as that of the role of the food sector in sustainable development. Social sustainability implies meeting people's needs on a continuous basis, together with a widely shared perception of at least minimally acceptable equity among diverse social groups and sufficient autonomy for all social actors to participate meaningfully in establishing the rules regulating their societies. Even if by some miracle these conditions could be attained, food systems could not be sustained unless the natural ecosystems on which they ultimately depend were protected adequately to provide future generations with at least equal opportunities as those enjoyed by present ones to improve their livelihoods.
The world has ample food. Global food production has grown faster than population since the 1950s. Global food production could have grown much more rapidly if the poor had enjoyed access to sufficient resources to produce, or incomes to purchase, all the food they needed.6 There is adequate scope for economically and sustainably increasing per capita food supplies significantly in the foreseeable future. In spite of these favourable conditions, about one fifth of humanity remains underfed. In some countries and sub-national regions this proportion is far higher, while in others it is much less. This gap between reality and what is clearly possible is the central problem to be addressed by the international community.
The rest of this paper discusses broad food security problems which are of special interest for developing countries. It reviews food security issues in the light of the five criteria mentioned earlier. It then discusses several issues requiring international reform and co-operation. It highlights the potential for South-South co-operation, and, finally, in the concluding section it summarizes the key principles and policy objectives which emerge from the preceding discussion and analysis.
II.
1 It needs to be appreciated that access to adequate food is essential for good nutrition but it is not in itself sufficient. Household nutritional security implies depends also on the capacity to utilize food in a way that meets nutritional needs. This is affected by infectious and parasitic diseases, poor sanitation, inadequate food preparation and eating habits, and many other factors. Access to health care by the poor is therefore essential, as is sound information about the causes of poor nutrition and how they can be remedied by poor households within the means available to them. Improvements in the nutritional and health education of women is crucial in this respect, as in most societies they prepare household meals and are responsible for the care and feeding of their young children. (UNICEF, 1990).
2 FAO's initial documentation for the World Food Summit was revised after drafts were circulated for discussion and comment. Subsequently, revised documents were produced, including re-ordered technical papers. The references cited here for the most part refer to the pre-Summit initial drafts, but where substantive changes were introduced in the final documents this is noted in the text or in footnotes.
3 FAO uses three quantitative indicators related to food security: calories available per capita (cal/cap), the Aggregate Household Food Security Index (AHFSI) and the percentage of undernourished in the total population (UNNUR) (FAO-Tech 7, 1996 and Tech-11, 1996). It admits that reliable data for estimating these indicators, and especially the latter two, are simply unavailable for most developing countries (FAO-Tech 9, 1996).
4 Of course, in some situations the poor have enjoyed relative food security. This becomes increasingly difficult where customary institutions governing food production and its distribution are being disrupted by commercialization and monetization before alternative sources of livelihood become available for those negatively affected.
5 For the purposes of designing some kinds of corrective policies and programmes, however, it would be useful to have more reliable data than are usually now available. These include the numbers, locations and characteristics of the hungry, the nature of the food insecurity and increasing malnutrition they are facing, the processes generating it and the contexts in which it occurs. Such analyses have to be made at local and sub-national levels in order for meaningful generalizations to emerge nationally and internationally about the problem.
6 To the extent that the rural poor have access to sufficient resources for increasing their food production, this would tend to improve their livelihoods from self-provisioning while at the same time making more food available for sale or barter, helping to keep prices for consumers reasonably low. Greater incomes for the urban poor and the rural landless, however, would increase effective demand for food and tend to push up food prices in some circumstances. These contradictory tendencies could result in a modest overall increase in food prices in the future.