© FAO, 1996
1.1 According to United Nations projections (medium variant), the world's population will increase by 72 percent between 1995 and 2050. It is hoped that by that time food deficits will be reduced, per caput food consumption in countries suffering from shortages will increase and the diets of populations will be diversified in order to eliminate specific deficiencies. All these changes will weigh heavily on food production systems, natural resources and the environment.
1.2 The main question is whether the required improvements in food production and available natural resources will be enough to cope with this population growth in a sustainable manner until 2050, when the worlds population is projected to stabilize. The distribution of natural resources needed for agricultural production does not correspond to the geographical distribution of population, and migration does not necessarily compensate for this difference in distribution, which can cause additional difficulty. This issue is relevant at the local, national, regional and international levels.
1.3 This paper focuses on the relative importance in the trends of three factors: the demographic factors determining food energy requirements,1 the closing of existing gaps in energy requirements and the diversification of dietary patterns to meet nutritional needs better. From this perspective, a population is seen not only as a sum of people with demographic or socio-economic characteristics, but as a collection of individuals with differing nutritional requirements because of variables such as gender, height, age, degree of physical activity and dietary pattern.
1.4 This paper, therefore, uses mainly minimum energy requirements as a basis to estimate the plant-derived energy needed to meet them. These energy requirements are evaluated for 2050 according to several demographic and nutritional scenarios and their interplay.
1.5 Obviously, the relative importance of the three demographic and nutritional factors of change (identified above) depends on the assumptions adopted. Changing assumptions would modify the results. The United Nations population projections present three highly different scenarios which show the importance of demographic factors. In contrast, nutritional assumptions adopted in terms of closing the gap in energy requirements and diversifying diets are conservative: the food energy level expected to be reached by developing countries in 2050 would be that projected by FAO for East Asia for 2010 and the level of diversification of diet adopted does not differ markedly from that projected by FAO for the whole world in 2010.
1.6 Such scenarios make it easier to assess the consequences of different assumptions for population and nutritional changes in 2050 and have been adopted for the following reasons.
1.7 These results will give an idea of the magnitude of population challenges for food production to meet nutritional requirements by 2050, without consideration of the calendar during that period. Consequently, the paper expands the population projections of the United Nations to include the nutrition field.
1.8 It should be clear that the paper cannot be seen as an extension of population projections into the field of economics.
1.9 It should also be taken into account that focusing on sustainable development over a period roughly corresponding to two generations highlights the importance of human capital development and policies that cover more than one generation.
1.10 Besides using the common distinctions between developed and developing countries (i.e. socio-economic) and by continent (i.e. geographic), it is also useful to attempt to classify countries according to the main sources of energy in the diets of their populations. The three modes of classification should be seen as complementary.
1.11 In determining the number of dietary pattern classes, there is a trade-off between the precision that comes with defining a large number of classes and the stability of the classification over the long term. Of course, by choosing the stability option as far as possible, and thus reducing the number of classes, some anomalies can occur, which will be discussed where appropriate.
1.12 Owing to the small number of classes adopted here, there will probably be few changes in diet class. Even though some countries will have to import cereals to augment their food supplies and may have to buy wheat in spite of their diet being structured differently, these imports would have to be on a very large scale at the national level to change a countrys diet class.
1.13 The following important assumption is adopted implicitly: populations that consume mainly roots or tubers (cassava, yams, taro, etc.) will not change this pattern radically. The assumption can be explained by three observations: the countries concerned can intensify cultivation of those crops without damaging soils; they still have considerable land reserves for rain-fed agriculture; and, as a result of their great poverty, they might not be able to import enough cereals to generate a change in their dietary class for a number of decades.
1.14 The predictability of the various aspects of the food/population question varies considerably.
1.15 Of these aspects, population changes are probably the least difficult to project. Soon after the Second World War the United Nations was able to project world population for the year 2000, and it has produced subsequent projections without any substantial changes in the results (Table 1).
United Nations population projections for the year 2000
|Year of reference||Date of publication||Projected population
for the year 2000
1.16 Because of internal and international migration flows, often caused by economic changes, the observation made on the projections is better verified for larger geographical areas than for smaller ones.
1.17 It is difficult to project changes in farming practices such as technological changes, the development of new cultivars, the selection of new animal species, the success of research on adapting farming to the environment or the environment to farming (irrigation, etc.) and the dissemination of expertise and innovations.
1.18 Adopting a time frame of 50 years, or even 20 years, presents difficulties because of the unpredictable nature of certain exogenous variables such as: public investments in infrastructures that may influence the development of agriculture; the supply of agricultural inputs; the regulation of agricultural markets; and the effectiveness of structural adjustment policies that affect factors such as farmers income or debt. But, by focusing on relatively robust population projections and nutritional aspects, these difficulties, to a certain extent, can be avoided.
1.19 The projections in this paper draw from a series of projections produced by international organizations.
1.20 The three United Nations scenarios for population projections up to 2050 have been examined to estimate the relative or absolute impact of certain demographic factors on energy requirements and of dietary changes on food demand. Even stopping at this level, it is apparent that the magnitude of the challenges for food production is unprecedented and raises fundamental questions regarding possible improvements in productivity, sustainability and environmental costs.
1.21 FAO projections for food and agriculture until 2010 have served as a framework for this study. Thus this document draws on World agriculture: towards 2010 (WAT2010) (FAO, 1995a) for its country-level results and its assumptions for trends up to the year 2025.
1.22 The procedures for evaluating energy requirements have been elaborated by FAO.
1.23 The combined use of these sources has permitted a long-term projection of energy requirements and of the demographic and nutritional factors that determine these requirements.
1.24 The scenarios presented in this paper imply that meeting challenges of the magnitude described will require:
1.25 In Chapter 2, the historical balance between population and food production is examined and the countries are classified according to level of development, continent and dietary patterns. The factors affecting the balance between energy requirements and food supplies are discussed and some lessons are drawn from the observed trends.
1.26 In the third chapter, some future scenarios are examined in order to illustrate the relative impact on energy requirements of population growth and other demographic factors such as urbanization. The importance of closing existing gaps in energy requirements is then shown, as well as the additional requirements made necessary by changing dietary habits. Finally, the aggregate impact of the previously mentioned factors is discussed and some conclusions are drawn. It should be noted that each factor is presented as a multiplying coefficient of plant-derived energy. The aggregate effects are obtained by multiplying each specific coefficient.
2.1 The first consideration is an examination of population trends that may have influenced the energy requirements of populations and thus affected the food supply needed to satisfy those requirements. Changes in food supply are described below. Finally, the paper examines whether food supply has covered these requirements.
2.2 Population changes that have affected food supplies since the Second World War can be summarized as follows:
On the basis of these remarks, population movements can therefore be expected to increase in volume and diversify in destination. Appropriate national and international policies will be necessary to ensure that the impact of migration is positive. In this respect, policies designed purely to stem migration flows, whether internally or internationally, cannot be successful and will often be counterproductive without effective development policies in sending areas.
2.3 In addition to the obvious role played by population growth itself, all the structural changes mentioned above have influenced the energy requirements of populations to varying degrees.
2.4 The increase in life expectancy contributes to the increase in population of all ages. Furthermore, improving the diet of children contributes to increasing their size and weight, thus increasing their average energy requirements as adults.
2.5 The decline in fertility has two opposite effects. On the one hand, it tends to reduce the average energy requirements, because the requirements of pregnant and nursing women are somewhat higher than those of other women of the same age. On the other hand, it tends to increase energy requirements through a reduction of the proportion of children, whose requirements are less than those of adults. On balance the latter effect dominates, but the net effect is a weak one.
2.6 Physical activity increases requirements, and levels of physical activity tend to be greater in rural areas than in urban ones. Inversely, urban populations change their diet and adopt eating habits that generally require more plant energy (Calories).
2.7 This document describes the specific effects of these demographic factors on energy requirements.
2.8 Since 1971, FAO has been developing an integrated and computerized system for compiling and maintaining, in the form of supply/utilization accounts (SUAs), current agricultural statistics for 300 primary food, agricultural and fishery commodities and 310 processed products derived from those commodities for about 200 countries and territories with data series from 1961 to 1990.
2.9 The total quantity of foodstuffs produced in a country added to the total quantity imported and adjusted for any change in stocks that may have occurred since the beginning of the reference period gives the supply available during that period. On the utilization side, a distinction is made between quantities exported, fed to livestock, used for seed, put to industrial and other non-food uses or lost during storage and transportation and those food supplies available for human consumption at the retail level, i.e. the form is noted in which the food leaves or otherwise enters the food supply for consumption (FAO, 1993a).
2.10 It is important to note that the quantity of food available indicates the quantity of food reaching the consumer but not necessarily the amounts of food actually consumed. Amounts consumed may be lower than the quantity shown because of the losses of edible food and nutrients in the household during storage, preparation and cooking (which affect vitamin and mineral content more than they affect energy, protein or fat), plate-waste, quantities fed to domestic animals and pets or food thrown away (FAO, 1993a).
2.11 FAO has thus evaluated food supply for 1992 at an average (after losses) of 2 718 Calories per person per day, of which 2 290 Calories were from plant products and 428 Calories from livestock products.
2.12 Globally, food supplies have more than doubled in the last 40 years. This has resulted in global food supplies increasing faster than the population, which has led to a substantial increase in average per caput food supplies in Calories. The data available show that between 1962 and 1991 average daily per caput food supplies increased by more than 15 percent, but these global averages conceal important regional variations.
2.13 In developing countries the increase in per caput food supplies was substantial, rising from almost 1 990 Calories in 1962 to 2 500 Calories in 1991, while in the same period the total population almost doubled, growing from 2.2 billion to more than 4.2 billion people. At the same time food supplies in developed countries rose from 3 000 Calories in 1962 to a maximum of about 3 300 in 1982, then diminished to about 3 150 Calories in 1991. The increase was especially significant in Asia, which fully exploited the advantages of the green revolution, and in Latin America, which benefited greatly from technological progress in the form of hybrid varieties of maize.
2.14 Three categories of countries are not included in these trends. The European countries, first of all, reduced their supplies between 1982 and 1992, while North American countries greatly increased theirs. African countries, especially those whose populations consume cassava, yams or taro, underwent a reduction in food supplies during the same period. This African evolution needs to be seen particularly as a consequence of the failure to carry out a green revolution on that continent. It should be noted that during the same period (1982 to 1992), those populations that obtain most of their energy requirements from maize also underwent a decrease in their supplies.
2.15 Only half the cereals produced are destined for human consumption: 48 percent in 1969-1971 and 50 percent in 1988-1990 (FAO, 1995a).
2.16 Most of the cereals put to other uses are destined to feed livestock. Just over 20 percent of the worlds cereal production was used to feed livestock in 1988-1990 (15 percent in 1969-1971). Developing countries used a little less than 20 percent for livestock in 1988-1990 (a little more than 10 percent in 1969-1971) (FAO, 1995a). The rest is used for seed reserves (seed requirements are estimated at around 5 percent) (James and Schofield, 1990) or is lost between harvesting and retailing. (The proportions are difficult to evaluate except by looking at the leftover balance.)
2.17 No one knows exactly the harvest losses between production and retailing. There are existing case-studies that focus on losses in yields caused by pests, but it is difficult to generalize (FAO, 1993b). The FAO evaluation of food supplies takes into account various losses. The loss of food during storage may be considerable. Important losses have been observed in Latin America (SOLAGRAL, 1995). Some authors estimate losses at 10 to 20 percent. Others estimate losses to be as high as one-third of the quantities produced (Erlich and Erlich, 1991). At any rate, losses of 10 to 15 percent in commercial warehouses are not infrequent (James and Schofield, 1990) (see Paragraph 3.61).
2.18 To compare national requirements and per caput food supplies, a percentage representing food losses occurring between retailing and domestic consumption (at the stage of preparation or consumption) must be added to the evaluation of requirements. A figure of 5 to 10 percent is frequently quoted (James and Schofield, 1990). These losses are greater in developed countries than in developing countries, and in the latter losses are greater in urban areas than in rural ones.
2.19 Losses from household security stocks, usually made in rural areas by the farmers themselves, must also be considered. Rural societies often experience crop failures. The stocks are exposed to deterioration or destruction by pests, mould, etc. Losses are greater when reserves are larger and when they are made for periods exceeding one year. Stocks are small in developed countries and larger in developing countries. In the latter, they are larger in rural areas than in urban areas. Stocks are smaller in rural societies with higher living standards.
2.20 Available figures cannot be used as statistics because they are so imprecise and specific to certain cases. However, in the absence of more precise figures, losses resulting from stock constitution and the losses sustained between retailing and domestic consumption are estimated to amount to between 10 and 40 percent of the total food supplies of a family (Uvin, 1995).
2.21 The evaluations of per caput food supplies cannot be considered evaluations of energy requirements. Nutritionists have evaluated human energy requirements. A manual for planners and nutritionists details current knowledge and suggests methods for evaluation at the country level (James and Schofield, 1990).2
2.22 It should be recalled that there is a difference between energy requirements and food demand. All people consume food to satisfy their requirements for energy and nutrients, which vary according to age, sex, height, weight, etc. However, the demand for food to satisfy these requirements varies according to food supplies as well as such factors as consumers tastes, income and relative prices. A population increase and a change in its structure, notably by age and sex, will lead to changes in requirements, especially in energy, according to the parameters previously mentioned, but these requirements may be satisfied by a great variety of combinations of food products.
2.23 What are the factors that most affect energy requirements? With a constant population, height and weight account for 49 percent, the age structure for 35 percent and urbanization for 15 percent in the national average allocation of energy (James and Schofield, 1990). The main factors to consider are the following:
2.24 Retrospective evaluation of energy requirements shows that they vary greatly in different countries. Those of developed countries are greater than those of developing countries.
2.25 The energy requirements of North American countries average almost 2 400 Calories per person per day, which is a little more than the requirements for European populations. At the other end of the scale, the lowest energy requirements are those of African populations (less than 2 150 Calories), which are slightly below those of Asian or Latin American populations (almost 2 150 Calories) (see Figure 1). For a definition of dietary classes, see Paragraphs 2.42 to 2.48.
2.26 In general, populations whose diets are based on rice, maize, wheat, millets or cassava (Figure 2, Classes 1, 2, 3, 5 and 6, respectively), for the most part in developing countries, have energy requirements that are almost 10 percent lower than those of developed countries, where the diet is richer in animal products.
2.27 The energy requirements of populations have generally increased since 1970 (after having decreased during the previous decade), reaching a level of 2 220 Calories per caput per day. The requirements in developed countries have been increasing more rapidly since 1970 after slow growth during the preceding decade. The requirements in developing countries have increased even more rapidly than in developed countries. This is because the energy requirements of Asian countries, especially those consuming mainly rice, and of Latin American countries have increased very rapidly, much more so than those of developed countries (Figure 1).
2.28 The trends in Africa are different (Figure 1). Average per caput energy requirements have decreased very slightly since 1960 because of the slight reduction in the requirements of populations consuming mainly millets or sorghum and of populations consuming cassava, yams, taro or plantain (Figure 2, Classes 5 and 6, respectively). This is also the case for the energy requirements of populations consuming maize. The changes observed in average energy requirements are mainly caused by changes in population structures by age.
2.29 Coverage of human energy requirements, i.e. the ratio of food energy supplies to the value of requirements, improved rapidly during the 1960s, but this progression slowed considerably during the 1970s. The average rate of coverage even diminished during the 1980s. This does not mean that the food situation of developing countries worsened during the same period. It is in the developed countries that consumption has diminished in relation to constant energy requirements. Even though progress is now slower, the food situation in developing countries has, on average, improved.
2.30 Food supplies in developed countries increasingly surpass their energy requirements. With a rate of coverage already exceeding 1.35 in 1980 and approaching 1.5 in 1990, the populations of North America seem assured of almost total security of their food supply, even in the advent of massive losses at the production level or before or after retailing. Their supplies now exceed their requirements by almost 50 percent (Figure 3). European countries, on the contrary, are reducing their supplies. The rate of coverage of their requirements has gone from 1.4 in 1980 to 1.36 in 1990 (Figure 3). Given the moderate losses that probably occur between the production and domestic consumption of food products, the food situation undoubtedly remains at a surplus level for most of the population.
CHANGES IN ENERGY REQUIREMENTS FROM 1960 TO 1990, BY CONTINENT
CHANGES IN ENERGY REQUIREMENTS FROM 1960 TO 1990, BY CLASS OF DIET
2.31 The coverage situation in developing countries has greatly improved from a deficit of 6 percent in 1962 to a surplus of 17 percent in 1990.
2.32 The improvement was especially noticeable in Asia. The rate of coverage of energy requirements went from a little more than 0.9 in 1962 to slightly less than 1.2 in 1990, a rapid progression observed during the entire period from 1962 to 1990. In Latin American countries, where the food situation is not as favourable on average as in Asian countries, the coverage rate regressed slightly between 1980 and 1990 after a marked improvement between 1962 (1.10) and 1980 (1.25). Further details and an explanation of how these figures should be interpreted will be given below.
2.33 The figures confirm the alarming situation in African countries. As supplies have been insufficient to meet these countries needs since 1962 and have never exceeded their requirements by more than 8 percent (coverage rates below 1.06), the food situation is obviously inadequate in most of Africa (Figure 3). The situation is no doubt more serious in some African countries than the average figures indicate. North African countries, where wheat is an important part of the diet, are able to import grain to satisfy their needs. Conversely, because North Africa is part of the average, the situation has to be worse than average in parts of sub-Saharan Africa.
CHANGES IN COVERAGE OF ENERGY REQUIREMENTS BY FOOD SUPPLIES FROM 1962 TO 1990, BY CONTINENT
2.34 Despite an improvement in the ratio between supply and requirements for populations that obtain most of their food energy from millets and sorghum, their food supplies did not cover their needs, even in 1990 (coverage rates below 1.00) (Figure 4, Class 5).
2.35 Even allowing for the poor quality of data, the trends for populations that consume mainly roots or tubers, cassava, yams or taro differ from those of other diet classes. Their food situation is deteriorating and their food supplies were below their requirements in 1980 and in 1990 (Figure 4, Class 6).
2.36 The food situation of sub-Saharan African populations is completely different from that of rice-growing populations. From 1962 to 1990 the coverage rates for Classes 5 and 6 went from 0.89 and 1.02, respectively, to 1.00 and 0.98, while that of Class 1 (rice) rose from 0.88 to 1.18 (Figure 4). Taking into account food losses after retailing, which are usually considerable in poor countries, the situation of these populations is often critical.
CHANGES IN COVERAGE OF ENERGY REQUIREMENTS BY FOOD SUPPLIES FROM 1962 TO 1990, BY CLASS OF DIET
2.37 In the developing countries, the classes include a large number of countries (with the exception of Class 5, which is small). They thus represent averages and mask local situations. It should be stressed that Africa is not the only continent facing serious food shortages at the national level. Countries such as Ethiopia, Mozambique, Central African Republic, Sierra Leone and Somalia (FAO, 1995b) are effectively in this situation, but so are some Asian countries such as Afghanistan or even Mongolia and Nepal. World Health Organization (WHO) bulletins show that other countries such as Bangladesh, Cambodia and Myanmar also suffer from chronic undernourishment (Erlich and Erlich, 1991). There also exist chronic deficits at the subnational level in India and China. This is why chronic undernutrition continues to be an important problem in East Asia and South Asia (FAO, 1992).
2.38 Large numbers of people still suffer from chronic undernourishment. Figures show that 918 million people suffered from undernourishment in 1969-1971, 906 million in 1979-1981 and 841 million in 1988-1990 (FAO/WHO, 1992).
2.39 The unequal distribution of food supplies occurs at both the micro and macro levels. Until now, no mention has been made of the widespread inequality of peoples access to food which occurs regardless of the degree of coverage for the entire population. Women do not always have the same access to food as men, which not only affects their health, but also the future development and growth of their children. It is also known that childrens diets can be up to 20 or 30 percent inferior to their needs (FAO, 1987) and that, in the case of food shortages, men sometimes have preferential access to food. It is also true that economic inequalities can generate unequal distribution of food supplies.
2.40 The approach developed here has assumed so far that food is distributed to individuals exactly according to their requirements. This is rarely the case. A fairer distribution of food supplies would probably eliminate most undernourishment. But just as accepted losses between retailing and domestic consumption require a compensatory increase of food supplies, when there are inequalities in distribution food supplies should be substantially larger than needs if there is to be enough food to satisfy the energy requirements of those who come last in the distribution chain (see Paragraph 3.62). Increasing food supplies can facilitate improved distribution only if accompanied by appropriate policies (Paragraph 3.58). The trickle-down effect does not occur automatically.
2.41 At the intercountry level, the differences are increasing. The combination of the macro and micro differences in distribution explains the persistence of the high number of undernourished people in spite of the overall improvement. Current population and poverty trends exacerbate this situation in many developing countries.
2.42 The major eating patterns of the world and their socio-cultural dimensions have been classified using the same national information that was used for the analysis of the principal components of daily per caput consumption. Six classes have been defined using an ascending hierarchical classification of 151 countries for which FAO had information. Another 32 countries have been excluded because of insufficient information. However, since these 32 countries together represent only 0.4 percent of the worlds population, their exclusion has minimal impact.
2.43 The six classifications include 16, 25, 25, 27, 5 and 21 countries, respectively (Table 2). The inclusion of a category with only five countries may be questioned. But even in a typology programmed with five classes instead of six, the small category would remain because, as will be seen, its characteristics are very important. As explained in the following paragraph, a typology reduced to five classes would group Classes 1 and 2 together, although they are composed of populations consuming mainly rice and maize, respectively.
2.44 To simplify this typology, each class has been named according to the food product that best characterizes its diet. Thus, Class 1 is rice, Class 2 is maize, Class 3 is wheat, etc. But this simplification should not be taken literally. Each class contains countries with similar diet structures according to the FAO food balance sheets. Here it is important to keep in mind that this classification process considers only the energy aspects of each element of the diet in order to compare the structures. It does not consider qualitative aspects such as proteins from fish origin or glucides from wheat. Because each food is recognized only for its energy value in the diet structure, a country will be grouped with a specific class if it has a similar diet structure, even if it differs in an important food. The few anomalies that have emerged in the classification process are of considerable interest, such as the case of Japan, which is explained below in some detail. The differences between some structures can be relatively small. For example, the structure of the Class 1 diet (consumers of rice) is closer to that of Class 2 (consumers of maize) than to any other class. So, if only five classes were defined, countries consuming mainly rice would be grouped with those consuming mainly maize. The diet of this combined group is substantially different from the diet of those countries that consume mainly wheat, because a diet of wheat also includes other sources of energy. On the contrary, if more than six classes were programmed, it would mean separating some countries, such as Japan, into independent classes. With the division of six classes, Japan is grouped with the countries that consume mainly maize, despite the fact that its population does not eat maize. This can be explained by three facts: Japan is not classified within the group of high-revenue countries (Class 4) because its level of energy intake is lower than that of high-revenue countries; its consumption of meat is low; and its consumption of rice is low compared with that of other countries of the rice class. Thus the consumption of plant products other than rice places Japan in a different category from the countries that consume mainly rice (Class 1), and the low consumption of meat, especially beef or mutton, places Japan in a different category from the countries that consume mainly wheat (Class 3). A similar explanation could be given for Malaysia.
ASCENDING HIERARCHICAL CLASSIFICATION OF COUNTRIES ACCORDING TO AVERAGE NATIONAL CONSUMPTION FOR EACH COMMODITY
Milk, meat, wheat
Cassava, yams, taro, plantains
Ethiopia (including Eritrea)
Central African Republic
1 Based on three-year average consumption for 1989-1991, in Calories.
2 See Paragraph 2.44 for an explanation of the appearance of these countries in this group.
Note: 32 countries not classified because of insufficient data: Bahamas, Barbados, Belize, Brunei, Cape Verde, Comoros, Cyprus, Djibouti, Dutch West Indies, Fiji, French Polynesia, the Gambia, Guadeloupe, Guinea-Bissau, Guyana, Hong Kong, Iceland, Kuwait, Maldives, Malta, Martinique, Mongolia, New Caledonia, Papua New Guinea, Réunion, Samoa, Solomon Islands, Suriname, Swaziland, United Arab Emirates, Vanuatu, Zambia.
2.45 This typology distinguishes six major types of diets (Table 3).
2.46 The classes in this typology closely correspond to the principal plants of the different civilizations in the world, rice, wheat, maize, millets (to which can be added sorghum) and cassava (to which can be added yams and taro). The effects of diet diversification can be observed only in developed countries which have introduced many livestock products into their diets (Class 4).
2.47 Three developing countries are in Class 4 alongside developed ones. However, these developing countries, Saudi Arabia, Israel and Lebanon, are on or near the Mediterranean rim and are considered high-revenue countries by the World Bank. Their economic situations have enabled their populations to diversify their diets (World Bank, 1994).
2.48 This classification is not very different from that obtained using earlier data. Similar results would probably have been obtained if even older data had been taken into account. The differences would probably have concerned the three countries mentioned above and some developed countries, such as France, which until recently consumed much greater proportions of cereals. There is therefore a certain stability of diets in developing countries, although some changes in dietary patterns can be observed as countries develop.
2.49 Classifying countries according to dietary patterns also organizes them according to important demographic characteristics regarding the energy requirements of their populations. These relations are associative, not causal. In addition, these classifications correspond to different food situations (see Table 3):
MAIN CHARACTERISTICS DIFFERENTIATING THE DIETARY CLASSES (ratio to average (1.00))
2.50 Even though improvements in the coverage of energy requirements in developing countries may have been remarkable, they have been too slow. As mentioned earlier, the increase in per caput daily food supplies in developing countries has been substantial, increasing from almost 1 990 Calories in 1962 to more than 2 500 Calories in 1991, which exceeds these countries requirements (2 160 Calories in 1990). In many countries these gains have not benefited the poor.
2.51 However, the losses of food that occur between retail sale and household consumption have played an important role in increasing the gap between supply and needs. Such losses can be considerable, especially in the case of cereals or vegetables. Losses of roots and tubers are low because they are consumed in countries with relatively small markets (so there are few transport and storage losses) and because they can be left in the ground until they are needed without too much damage occurring. The losses are greater when households are forced to make reserves as a guarantee against crop failure or catastrophes caused by weather, flooding or supply problems. The longer these reserves are kept, the more vulnerable they are. Average annual losses of 10 percent would increase the average food supplies needed to 2 380 Calories, which is not far below the available food supplies of developing countries in 1988-1990 (2 470 Calories).
2.52 The inequality of food distribution within countries also increases the gap between average requirements and the food supplies needed by populations. It has already been pointed out that in countries with the greatest inequalities in distribution, food supplies per caput would have to be 20 to 30 percent above the average requirements to overcome malnutrition (FAO, 1995a). However, such an increase does not automatically solve the problem of distribution, which is why policies in this area are crucial.
2.53 It is understandable, therefore, that the food supplies of countries that consume mainly rice (about 2 520 Calories) should be inadequate and that, since these countries represent the majority of the worlds population, most people suffering from malnutrition in the world live in these countries. But the most acute food shortages are in those countries that consume mainly cassava, yams or taro. Their food supplies, which in 1990 were 2 090 Calories, do not equal their average food energy requirements. This class includes many food-insecure countries which will probably account for the majority of people suffering from undernutrition in 2010 (FAO, 1995a). They are also countries with rapid population growth.
2.54 It is not within the scope of this study to define the factors that determine increases in food supplies. In fact, not much is known about these factors. But with the above results, past conditions that have led to increased food production can be recognized. They are partly demographic since population growth almost automatically leads to an increase in total food energy requirements. They are also partly economic.
2.55 Since the Second World War, those populations that produce their own food and find employment in and earn their living from agriculture have steadily declined in relation to the total population, but they still remain in the majority. In general, the poorest populations are still to be found in rural areas (World Bank, 1990). It is observed that population growth is greatest among the low food consumption countries. This underscores the importance of studying the interactions between increasing requirements and increasing demand. The damage caused to policy formulation by an insufficient understanding of these linkages should be appreciated. Still, general economic growth and a decline in poverty which leads to an increased demand for food commodities seem to be the main factors for ensuring success in meeting increasing energy requirements in many developing countries. Policies addressing rapid population growth can also make an important contribution to these factors.
2.56 A considerable increase in food supplies made possible by increased productivity and, to a lesser degree, by an increase in cultivated area has been observed. According to FAO (1995a), 69 percent of the increase in plant production between 1970 and 1990 was due to improved yields and 31 percent to increases in cultivated area.
2.57 The increase in productivity was mainly for wheat (2.8 percent per year), rice (2.3 percent per year) and, to a lesser degree, maize (1.8 percent per year) and sorghum (1.5 percent per year). Barley, millet and cassava yields only progressed by 1 percent per year (FAO,1995a). Intensification was also made possible by irrigation, which, apart from directly increasing yields by allowing the use of high-yielding varieties (HYVs) of cereals (hybrids, etc.), has increased the number of harvests. In developing countries (China not included), 123 million hectares of arable land were irrigated in 1988-1990, of which 35 million hectares were arid or very arid land (FAO, 1995a). Although no detailed studies have been made on the subject, the substitution of high-yielding crops for low-yielding ones and changing plant products also added to the increase in food production.
2.58 Technological changes have made food products less expensive, which has allowed increased consumption by humans. Developing countries have also started to feed their livestock food initially intended for human consumption because of lower costs. Increased demand has generated technological innovations, which in turn have become cheaper to use. The increased concentration of populations has probably stimulated production as a result of mass consumption, although this hypothesis has not yet been proved.
2.59 In spite of the increase, production in developing countries has not been enough to satisfy needs. Apart from a few large rice-producing countries, China in particular, developing countries have had to import cereals. These imports have been especially large since prices were abnormally low as a result of farm subsidies given by some countries.
2.60 The long-term trend has been a decline in food prices, partly because, at a global level, there is probably no obstacle to food production increasing to meet demand. The global decline in per caput cereal production during the last ten years can only be explained by a decline in the production of the major exporters.
2.61 Two incompatible policy objectives are being pursued. The first is to reduce poverty and help the poor have access to food, which leads to increasing demand. The second is to maintain prices at a level sufficiently attractive to encourage production on the part of the big cereal exporters, which leads to land being left fallow and curbs increases in available per caput food supplies.
2.62 Increased food production and increased agricultural productivity are the principal means of reducing poverty and improving the food situation in developing countries. The main cause of chronic undernourishment has been the inability to reduce poverty in these countries. The potential for increasing production, e.g. in Africa, is still considerable, but it requires appropriate and effective policies. These policies need to improve the situation of rural women, who are burdened with the major role of food production but who, in the absence of men, frequently have to reduce the number of tasks or the time allocated to each, which can result, inter alia, in land degradation.
2.63 A great many developing countries have, to varying degrees, increased their grain imports over the last few decades. The major grain exporters have easily met the increase in demand. Support given to agriculture in exporting countries partly explains the increase in imports by developing countries. However, this support is beginning to ebb.
2.64 As gaps between national production and energy requirements have increased, poor countries have become increasingly dependent on cereal imports. But the size of the imports depends on the solvency of the country: the impact of population size behind these imports becomes apparent only if the countries are solvent.
2.65 Beyond a certain level of economic progress which may initially encourage a decline in fertility, cereal imports increase as fertility and population growth decline. Essentially, it has been observed that the more countries prosper, the more fertility declines and the more cereal imports increase.
2.66 This is not the case when fertility is high. The situation of countries in sub-Saharan Africa is different for two reasons. First, imports are smaller when population growth is rapid, probably because of the poor solvency of countries with high fertility rates. Second, the higher the population pressure in relation to farming area in countries consuming cassava, yams or taro, the smaller the cereal imports per caput. This reveals how extremely insecure food supplies are in these countries (Collomb, 1988, 1989). They cannot import enough cereals to feed their populations.
2.67 An observation specific to Arab countries is that the higher the fertility rate, or population growth, the higher the cereal imports (and the slopes of the regression curves are steep) (Collomb, 1988, 1989). Oil revenues and tourism probably make such imports possible.
2.68 The main lesson learned is that poverty is the major economic factor behind the failure to improve access to food, a situation exacerbated by rapid population growth.
2.69 There has been a general improvement in coverage of energy requirements by food supplies in developing countries. The most illustrative case is that of Asia, with an increasing rate of coverage over the period 1962-1992. More recently, the rate of improvement has slowed in Latin America.
2.70 Africa is the exception to these positive trends, with no improvement in coverage of energy requirements during the period 1962-1992. There was an actual decrease in coverage in Class 6 during the period 1970-1980.
2.71 These trends reveal the consequence of the green revolution in Asia and to a lesser degree in Latin America. The absence of such a green revolution in Africa is clearly apparent. These results will have an impact on future scenarios.
2.72 A related factor is probably the insufficient development of human capital. Health and the fulfilling of energy requirements of populations are obviously the preliminary conditions for rural and agricultural development. Humankind is far from ensuring health for all by the year 2000, as set out in 1978 at the International Conference on Primary Health Care in Alma Ata, USSR. High illiteracy and lack of information on new techniques or innovations hinder development. The positive effect of training on production and productivity has been proved, but the effects on gross domestic product vary greatly from one country to the next. The higher the initial level of education, the more productive are the investments in education. It would seem that sub-Saharan Africa is an exception to this rule, probably because of the lack of appropriate infrastructure and institutions (World Bank, 1990).
2.73 Thus ensuring that countries suffering from food deficits are able to make progress in their population and development policies in a sustainable manner, a condition for social order in the world, presents a great challenge for agricultural production. It also presents a great challenge for the general development of countries where demographic factors, isolated from or combined with the effects of dietary patterns, play an important role, as will be shown in the rest of this document. These considerations highlight the time-scale necessary for development, which could amount to two generations.
1 In the title of the paper, the term food requirements has been used. Strictly speaking, it should be replaced by the term food demand. However, in view of the economic meaning of the term demand, the expression food requirements was preferred. It covers the nutrient requirements, which include energy, protein requirements and micronutrients.
2 See also WFS companion paper 5, Food security and nutrition.