This study is the latest forward assessment by FAO of possible future developments in world food, nutrition and agriculture, including the crops, livestock, forestry and fisheries sectors. It is the product of a multidisciplinary exercise, involving most of the technical units and disciplines present in FAO, as well as specialists from outside FAO. It continues the tradition of FAO's periodical perspective studies for global agriculture, the latest of which was published in 1995 (Alexandratos, 1995). Earlier editions were Alexandratos (1988), FAO (1981a) and FAO (1970). An interim, less complete version of the present study was published in April 2000. Comments received on the interim report helped shape the study in its present form.
The projections were carried out in considerable detail, covering about 140 countries and 32 crop and livestock commodities (see Appendix 1). For nearly all the developing countries, the main factors contributing to the growth of agricultural production were identified and analysed separately. Sources of productivity growth, such as higher crop yields and livestock carcass weights, were distinguished from other growth sources, such as the area of cultivated land and the sizes of livestock herds. Special attention was given to land, which was broken down into five classes for rainfed agriculture and a sixth for irrigated agriculture. This level of detail proved both necessary and advantageous in identifying the main issues likely to emerge for world agriculture over the next 30 years. Specifically, it helped to spot local production and resource constraints, to gauge country-specific requirements for food imports and to assess progress and failure in the fight against hunger and undernourishment. The high degree of detail was also necessary for integrating the expertise of FAO specialists from various disciplines, as the analysis drew heavily on the judgement of in-house experts (see Appendix 2 for a summary account of the methodology). Owing to space constraints however, the results are mainly presented at the regional level and for selective alternative country groups, which of course, masks wide intercountry differences.
Another important feature of this report is that its approach is«positive» rather than«normative». This means that its assumptions and projections reflect the most likely future but not necessarily the most desirable one. For example, the report finds that the goal of the 1996 World Food Summit - to halve the number of chronically undernourished people by no later than 2015 - is unlikely to be accomplished, even though this would be highly desirable. Similarly, the report finds that agriculture will probably continue to expand into wetlands and rainforests, even though this is undoubtedly undesirable. Therefore, the prospective developments presented here are not goals of an FAO strategy but they can provide a basis for action, to cope both with existing problems that are likely to persist and with new ones that may emerge. It should also be stressed that these projections are certainly not trend extrapolations. Instead, they incorporate a multitude of assumptions about the future and often represent significant deviations from past trends (see Appendix 2). To give an impression of how well previous projections compared with actual outcomes, in Chapters 3 and 4, actual outcomes for some major aggregate variables (e.g. developing country cereal production and net imports) for the latest year for which historical data are available are compared with those implied for that same year by the 1988/90-2010 trajectories projected for these variables in 1992-93 for the 1995 study.
A long-term assessment of world food, nutrition and agriculture could deal with a great number of issues, the relevance of which depends on the reader's interest in a particular country, region or topic. As a global study, however, this report had to be selective in the issues it addresses. The main focus is on how the world will feed itself in the future and what the need to produce more food means for the natural resource base. The base year for the study is the three-year average for 1997/99 and projections are made for the years 2015 and 2030. The choice of 2015 allows an assessment of whether or not the goal of the 1996 World Food Summit (WFS) - to halve the (1990/92) number of chronically undernourished people in developing countries no later than 2015 - is likely to be reached. Extending the horizon to 2030 creates a sufficiently long period for the analysis of issues concerning the world's resource base; in other words, the world's ability to cope with further degradation of agricultural land, desertification, deforestation, global warming and water scarcity, as well as increasing demographic pressure. Naturally, the degree of uncertainty increases as the time horizon is extended, so the results envisaged for 2030 should be interpreted more cautiously than those for 2015.
The analysis is, inter alia, based on the long-term developments expected by other organizations. The population projections, for instance, reflect the latest assessment (2000 Assessment, Medium Variant) available from the United Nations (UN, 2001a), while those for incomes are largely based on the latest projections of gross domestic product (GDP) from the World Bank. Most of the agricultural data are from FAO's database (FAOSTAT), as of June 2001. Because these assumptions critically shape the projected outcomes, it is important to note that they can change significantly, even over the short term. For example, the historical data and the projections for the growth of population and GDP used in the 1995 study (Alexandratos, 1995), have since been revised in many countries, often to a significant extent. World population in the 1995 study, for instance, was projected at 7.2 billion for 2010, whereas the current UN projections peg the figure at 6.8 billion, i.e. about 400 million less. Similarly, it is now projected that sub-Saharan Africa's population will reach 780 million by 2010, compared with 915 million in the 1995 study. Also, projections for GDP growth (largely based on the latest projections from the World Bank) are different now from what they were in the 1995 study. Finally, FAO's historical data for food production, demand and per capita consumption were often drastically revised for the entire time series as more up-to-date information became available (for an example, see Box 2.2).
In the chapters that follow, particularly Chapters 2 to 4, reference is made on a number of occasions to conceptual and practical problems that arise in the process of making the projections. Many of these problems result from the approach followed in this study, i.e. the fact that the analyses are conducted in considerable country and commodity detail and that all the projections are subjected to a process of inspection, evaluation and iterative adjustment of the numbers by country and discipline specialists. This means that more data problems are encountered than if the analyses were conducted at a more aggregate level. It also means that an alternative scenario cannot be derived without repeating a good part of the whole cycle of inspection-evaluation-adjustment. In Appendix 2 some of these problems and the way they are or are not dealt with are highlighted, so that that the reader can better form an opinion as to how to view and evaluate the statements in this study.
The report is structured as follows. The remainder of this chapter gives an overview of the main findings of the report. The first part of the report (Chapters 2 to 7) presents the main perspective outcomes for food, nutrition and agricultural production and trade, as well as for forestry and fisheries. Chapter 2 presents and discusses expected developments in food demand and consumption and the implications for food security and the incidence of undernourishment. Chapter 3 discusses expected developments in demand, production and trade of total agriculture (crops and livestock) and for the major commodities. Chapter 4 presents the underlying agronomic projections for growth in crop production, including expected developments in crop yields, and resource (land, water) and input (fertilizer, farm power) use in the developing countries. It includes revised and updated estimates of land with rainfed crop production potential, estimates of irrigation potential and possible expansion of irrigated areas and, for the first time, an estimate of the volume of fresh water withdrawals needed to sustain irrigated agriculture in developing countries as projected in this report. Chapter 5 deals with projections of livestock production and related issues. Chapters 6 and 7 present the current state of and plausible developments in world forestry and fisheries.
The second part of the report (Chapters 8 to 10) continues with a discussion of the main issues raised by these developments. Issues of poverty (including international targets for its reduction), nutrition-development interactions, the role of agriculture in the development of the rural and overall economy, and macroeconomic policies and agriculture are the subjects of Chapter 8. Chapter 9 deals with trade policy issues, focusing on lessons from the implementation of the Uruguay Round Agreement on Agriculture (URAA), the possible implications of further liberalization and issues relating to further reforms during the follow-up negotiations in the context of the broader round of multilateral trade negotiations launched by the World Trade Organization (WTO) at its Fourth Ministerial Conference (Doha, November 2001). Chapter 10 then places agricultural trade in the broader context of globalization and discusses the possible effects of globalization on trade and on the concentration and location of the food processing industry.
The final part of the report (Chapters 11 to 13) deals with perspective issues concerning agriculture and natural resources (land, water, air and the genetic base). Chapter 11 examines the potential of existing and incipient agricultural technologies (including modern biotechnology) to underpin further growth in production while conserving resources and minimizing adverse effects on the environment, and the needed directions of agricultural research in the future. Chapter 12 presents an assessment of the environmental implications of agricultural production based, to the extent possible, on the quantitative projections presented in the preceding chapters, and examines options for putting agriculture on a more sustainable path. Finally, Chapter 13 deals with the interactions between climate change and agriculture, the role of agriculture as source of greenhouse gases but also as a carbon sink of potentially growing importance in the context of the Kyoto Protocol, and the broad impact of climate change on agriculture and food security.
Historical developments. The 2001 FAO assessment of food insecurity in the world (FAO, 2001a) estimates the incidence of undernourishment in the developing countries at 777 million persons in 1997/99 (17 percent of their population). The estimate was 815 million (20 percent of the population) for the three-year average 1990/92, the base year used by the 1996 WFS in setting the target of halving the number of the undernourished in the developing countries by 2015 at the latest. Obviously, the decline between 1990/92 and 1997/99 has been much less than required for attaining the target of halving undernourishment by 2015 (see further discussion in Box 2.5).
In a longer historical perspective of the last four decades, considerable progress has been made in raising the average world food consumption (measured in kcal/person/day), a variable that is a close correlate of the incidence of undernourishment. The world average kcal/person/day grew by 19 percent since the mid-1960s to 2 800 kcal. What is more important, the gains in the world average reflect predominantly those of the developing countries whose average grew by 31 percent, given that the industrial countries and the transition economies had already reached fairly high levels of per capita consumption in the mid-1960s. This progress in the aggregate of the developing countries has been decisively influenced by the significant gains made by the most populous among them. Of the seven countries with a population of over 100 million (China, Indonesia, Brazil, India, Pakistan, Nigeria and Bangladesh), only Bangladesh remains at very low levels of per capita food consumption. A significant number of countries however failed to participate in this general thrust towards improved national average food consumption levels and, by implication, towards reduced incidence of undernourishment. There are currently still 30 countries with per capita food consumption under 2200 kcal, most of them in sub-Saharan Africa.
Population, incomes and poverty in the future. The latest assessment of world population prospects by the UN (UN, 2001a) indicates that there is in prospect a rather drastic slowdown in world demographic growth. In the time horizon of this study, the world population of 5.9 billion of our base year (the three-year average 1997/99) will grow to 7.2 billion in 2015, 8.3 billion in 2030 and 9.3 billion in 2050. The growth rate of world population, which had peaked in the second half of the 1960s at 2.04 percent p.a. and had fallen to 1.35 percent p.a. by the second half of the 1990s, is projected to fall further to 1.1 percent by 2015, to 0.8 percent by 2030 and to 0.5 percent by 2050.
This deceleration in demographic growth and the gradual saturation in per capita food consumption for parts of the world population are important factors that will contribute to slow the growth of food demand and, at the world level, also of production. Despite the drastic fall in the growth rates of both population and aggregate demand and production, the absolute annual increments continue to be large. For population, 79 million persons are currently added to world population every year. The number will not have decreased much by 2015. Even by 2030, annual additions will still be 67 million. Practically all of the increases in world population will be in the developing countries. Within the developing countries themselves there will be increasing differentiation. East Asia will be reaching a growth of under 0.5 percent p.a. towards the later years of the projection period. At the other extreme, sub-Saharan Africa's population will still be growing at 2.1 percent p.a. in 2025-30. For demand and production, the world cereal totals must increase by almost another billion tonnes by 2030, from the current level of 1.9 billion tonnes, an increase almost equal to that of the period since the mid-1960s (see Chapter 3, Table 3.3).
The growth of incomes is the other major determinant of the growth of food demand and of changes in food security and nutrition. The outlook for income growth is mixed. The latest World Bank assessment for the period 2000-15 foresees higher growth rates in per capita GDP than in the 1990s for all regions and country groups (particularly the reversal of declines in the transition economies) with the exception of East Asia. However, for several countries that have low food consumption levels and significant incidence of undernourishment, the economic growth rates that may be achieved are likely to fall short of what would be required for significant poverty reductions. In particular, there is great contrast as regards the prospects of the two regions with high relative concentrations of poverty and food insecurity, South Asia and sub-Saharan Africa. In the former region, a continuation of the relatively high GDP growth holds promise of positive impact on poverty alleviation and increases in per capita food consumption. However, progress in sub-Saharan Africa may be very limited and far from sufficient to make a significant dent on poverty and food insecurity.
The World Bank also projects possible developments in the incidence of poverty - the percentage of the population and numbers of persons below the US$1/day poverty line. This is of particular importance for the evaluation of possible reductions in undernourishment, as poverty is a close correlate of food deprivation and insecurity. The Bank assessment concludes that the proportion (not the absolute numbers) of the population living in poverty in the developing countries as a whole may fall from the 32 percent it was in 1990 to 13.2 percent in 2015. This fall, if it materialized, would meet the target of halving the proportion in poverty between 1990 and 2015. It is recalled that the target of halving the proportion of the poor is one of the International Development Goals of the Millennium Declaration adopted by the UN General Assembly (see Chapter 8 for details). However, the absolute numbers in poverty in the developing countries (a measure more directly relevant for the target of reducing undernourishment - see below) are not halved, as they are projected to decline from 1.27 billion in 1990 to 0.75 billion in 2015.
Prospects for food and nutrition. The projections of food demand for the different commodities suggest that the per capita food consumption (kcal/person/ day) will grow significantly. The world average will be approaching 3000 kcal in 2015 and exceeding 3000 kcal by 2030. These changes in the world averages will reflect above all the rising consumption of the developing countries, whose average will have risen from the 2680 kcal in 1997/99 to 2850 kcal in 2015 and close to 3000 kcal in 2030. These gains notwithstanding, there will still be several countries in which per capita food consumption will not increase to levels that would imply significant reductions in the numbers undernourished from the very high levels they have at present. In 2015 there could still be 6 percent of world population (462 million) living in countries with very low levels of food consumption (under 2200 kcal). At the regional level, sub-Saharan Africa would still have in 2015 medium-low levels of per capita food consumption, 2360 kcal/person/day, and even less if Nigeria is excluded from the regional total.
These gains in average consumption mean that the great majority of people will be better fed and the incidence of undernourishment should decline. But will it decline sufficiently to achieve the objectives set by the international community? The 1996 WFS set the target that the numbers undernourished (not just the proportion of the population in that condition) should be reduced by half by 2015 at the latest. Improved nutrition, in addition to being a human right and a final objective of development in its own right, is also an essential precondition for societies to make progress towards overall economic and social development within a reasonable time span. This is because undernourished persons are impeded by their very condition (undernourishment) from fully contributing to, and profiting from, the economic activities that are part and parcel of the development process. There is sufficient empirical evidence (reviewed in Chapter 8) establishing how persons in such condition have smaller earnings and fewer opportunities in life than others. Removing the causes of undernutrition is a prime area for public policy interventions (e.g. through public health, sanitation and feeding programmes for pregnant women and children) since economic growth, although a necessary condition, is rarely sufficient by itself to achieve this goal within a reasonable time span.
The implication of the projected higher levels of average national food consumption per person is that the proportion of the population undernourished in the developing countries as a whole could decline from the 17 percent in 1997/99 to 11 percent in 2015 and to 6 percent in 2030. All regions would experience declines in these percentages and, by 2030, all of them, except sub-Saharan Africa, should have incidence in the range of 4 to 6 percent of the population, compared with a range of 9 to 24 percent in 1997/99. Sub-Saharan Africa could still have 15 percent of its population undernourished in 2030, down from 34 percent in 1997/99.
Because of population growth, declines in the relative incidence of undernourishment do not necessarily translate into commensurate declines in the absolute numbers (which is of relevance to the WFS target). Notwithstanding the slowdown in their demographic growth, the developing countries' population will still grow from 4555 million in 1997/99 to 5804 million in 2015 and to 6840 million in 2030. Therefore, the numbers undernourished will decline only modestly: from the 776 million in 1997/99 to 610 million in 2015 and to 440 million in 2030.1 If these projections came true, it would mean that we might have to wait until 2030 before the numbers undernourished are reduced to nearly the target set for 2015 by the WFS, i.e. one half of the 815 million estimated for 1990/92.
Can faster progress than projected here be made? Empirical evidence suggests that in the countries with high dependence on agriculture, assigning priority to the development of food production holds promise of overcoming the constraint to better nutrition represented by unfavourable overall economic growth prospects. Several countries, mainly in sub-Saharan Africa (Nigeria, Ghana, Chad, Burkina Faso, the Gambia, Mali, Benin and Mauritania) have at times achieved in the past quantum jumps in their food consumption per capita (over 20 percent) over periods comparable in length to the first subperiod of our projections (17 years or less), at a time when national average per capita incomes were not growing or outright falling. The common characteristic of these experiences has been rapid growth in the production of staple foods (cereals, roots and tubers).
Several empirical studies (discussed in Chapter 8) document the mechanism of how agricultural growth can be a potent factor in initiating a process of development that favours poverty reduction and improved food consumption in rural areas in countries with high dependence on agriculture and poverty profiles that are predominantly rural. The key conditions for this to happen are (i) that agricultural growth be somehow initiated, e.g. through policies that develop and diffuse affordable productivity-raising innovations such as improved seeds; and (ii) that the distribution of the ownership of land or other productive assets of agriculture not be too unequal so that benefits from higher agricultural production are widely spread and do not accrue predominantly to large landowners (who would spend most of their additional income on things other than locally produced goods and services).
If these conditions are met, a virtuous cycle of causation can set in: the initial increases of agricultural production, in addition to providing more food for the producers themselves and for others, create incomes that are spent locally and create demand for produce and services from the non-farm rural sector, thus generating incomes there. These, in turn, feed back into increased demand for food and more goods and services from the rural non-farm sector itself. This«pump-priming» role of agriculture is seen as necessary because the rural non-farm sector produces goods and services that are in the general category of non-tradables. This means that in that particular context and stage of development their production can increase only in response to local demand, otherwise it would not increase. The services that can be produced locally are non-tradable almost by definition, even if most goods may be inherently tradable but become non-tradable in the absence of adequate transport and marketing infrastructure.
Returning to the projections, the slow pace of progress in reducing the absolute numbers undernourished notwithstanding, the considerable overall improvement implied by the projected numbers should not be downplayed. It is no mean achievement that in the developing countries the numbers well-fed (i.e. not classified as undernourished according to the criteria used here) could increase from 3.8 billion in 1997/99 (83 percent of their population) to 5.2 billion in 2015 (89 percent of the population) and to 6.4 billion (94 percent) in 2030.
With the reduction of the relative incidence of undernourishment, the problem will tend to become more tractable through policy interventions, both national and international. This is the consequence of the prospect that the number of persons involved (and possibly needing to be targeted by policies) will be smaller, as will be the number of countries with high incidence (see Table 2.3). In addition, many more countries than at present will have relatively low incidence of undernourishment. For example, 16 countries had less than 5 percent undernourishment in 1997/99. Their number will increase to 26 in 2015 and 48 in 2030. By this latter year, such countries (including most of the largest developing countries) will account for three-quarters of the population of the developing countries and for 178 million undernourished out of the total of 440 million (in 1997/99: 8 million out of 776 million). Thus a growing part of the undernourished will be in countries with low relative incidence. For this reason, it will be easier than at present for them to respond to the problem though national policy interventions.
In parallel, the number of countries with high incidence of undernourishment (over 25 percent) and most in need of international policy interventions will be reduced considerably: from 35 in 1997/99 to 22 in 2015 and to only five in 2030. None of them will be in the most populous class (over 100 million population in 1997/99). They will account for an ever-declining proportion of the undernourished, 72 million out of the 440 million in 2030 (1997/99: 250 million out of the 776 million).
Total agriculture. The bulk of the increases in world consumption of crop and livestock products has been originating in the developing countries. With slower population growth and the gradual attainment of medium to high food consumption levels in several countries, the growth rate of the demand for food (and at the world level, also of production) will be lower than in the past. These are positive factors from the standpoint of human welfare. On the negative side, there is the prospect that in several countries with severe incidence of undernourishment, the growth of the demand for food will be well below what would be required for significant improvement in their food security. How much lower the growth of aggregate global demand will be in relation to the past depends on a number of factors. Foremost among them are the relative share, in world totals and other characteristics, of the countries that have already attained medium to high levels of consumption per capita, say over 2700 kcal/person/day. Several developing countries (29 of those covered individually in this study) belong to this class, including some of the most populous ones (China, Indonesia, Brazil, Mexico, Nigeria, Egypt, the Islamic Republic of Iran and Turkey). They have one half of the population of the developing countries and account for two-thirds of their aggregate demand.
In the period from the mid-1960s to 1997/99, this group of 29 developing countries made spectacular progress in raising per capita consumption, from an average of 2075 kcal to 3030 kcal (see Table 3.2 in Chapter 3). China's performance carried a large weight in these developments. The group's population growth rate was 1.8 percent p.a. and that of its aggregate demand for all uses was 4.2 percent p.a. In the projections, commodity by commodity, the implied per capita consumption for this group in terms of kcal/person/day rises to 3155 kcal in 2015 and to 3275 kcal in 2030, i.e. to levels not much below those of the industrial countries today. The implied annual growth rate for the whole period 1997/99-2030 (in terms of kcal per capita) is only 0.25 percent. In parallel, the group's population growth rate is projected to fall to 0.9 percent p.a. The net result is that a drastic deceleration of the growth of aggregate demand is in prospect for this group of countries, from 4.2 percent p.a. in the preceding three decades to 1.7 percent p.a. in the period to 2030.
Given the large weight of this group of countries in the totals of the developing countries and the world, their drastic deceleration is reflected in all the aggregates. Thus, the growth rate of the developing countries as a whole declines from 3.7 percent p.a. in the preceding three decades to 2.0 percent p.a. in the period to 2030. This happens despite the prospect that the demand of the other developing countries (those below 200 kcal in 1997/99 which have the other half of the population) will not decelerate much (less than the decline in their population growth rate) and that their per capita food consumption would rise from 2315 kcal in 1997/99 to 2740 kcal in 2030. At the world level, the impact of the deceleration in the developing countries is muted (the deceleration in aggregate demand, from 2.2 percent p.a. to 1.5 percent p.a., is not very different from that of world population, see Table 3.1). This reflects the fact that in the past the world growth rate was depressed because of the collapse of consumption and production in the transition economies. The cessation (and eventual reversal) of this effect in the future offsets in part the deceleration in world totals caused by the slowdown in the developing countries.
At the world level, production equals consumption, so the preceding discussion about global demand growth prospects applies also to that of global production. For the individual countries and country groups, however, the two growth rates differ depending on movements in their net agricultural trade positions. In general, the growth rates of production in the developing regions have been below those of demand, and as a result their imports have been growing faster than their agricultural exports. These trends led to a gradual erosion of their traditional surplus in agricultural trade. In fact, the developing countries have turned in recent years from net agricultural exporters to net importers. This trend continues in the projections. The net imports of the developing countries as a whole of the main commodities in which they are deficit, mainly cereals and livestock products, will continue to rise fairly rapidly. In parallel, their net trade surplus on account of their traditional exports (e.g. tropical beverages, bananas, sugar and vegetable oils) will either rise less rapidly than their net imports of cereals and livestock products or outright decline. This does not mean that the exporting developing countries will not be expanding their net exports. It rather reflects the fact that several large developing countries are turning into growing net importers of products that are exported mainly by other (exporting) developing countries, e.g. vegetable oils, sugar and natural rubber.
Cereals. The preceding discussion about future slowdown in the growth of demand applies in varying degrees to individual commodities.
The deceleration of the world cereal sector has been taking place for some time. It will continue in the future, but the difference between past and future growth rates will not be as pronounced as in other sectors, particularly in livestock and oilcrops (see below). Cereals will continue to be by far the most important source (in terms of calories) of total food consumption. Food use of cereals has kept increasing, albeit at a decelerating rate. In the developing countries, the per capita average food use is now 173 kg, providing 56 percent of total calories, compared with 141 kg and 61 percent in the mid-1960s. The level of around 173 kg has been nearly constant since the mid-1980s. We project that it will remain around that level over the projection period. Cereals will continue to supply some 50 percent of the food consumption (in terms of kcal/person/day) in the developing countries, which is projected to reach nearly 3000 kcal/person/day in 2030. Within the cereals group, per capita food consumption of rice will tend to stabilize at about present levels and will decline somewhat in the longer term, reflecting developments in, mainly, the East Asia region. In contrast, food consumption of wheat will continue to grow in per capita terms and, in the developing countries, such growth will be associated with growing wheat imports. Increases in the demand and trade of coarse grains will be increasingly driven by their use as animal feed in the developing countries.
As noted earlier, world consumption and production of cereals are projected to increase by almost another billion tonnes by 2030, from the 1.89 billion tonnes of 1997/99. Of this increment, just over one half will be for feed, and about 42 percent for food, with the balance going to other uses (seed, industrial non-food use and waste). Feed use, and within it that of the developing countries, will revert to being the most dynamic element driving the world cereal economy, as it will account for an ever-growing share in aggregate demand for cereals. It had lost this role in the decade to the mid-1990s following events and policies that had depressed feed use of cereals in two major consuming regions, the transition economies and the European Union (EU) (see Chapter 3, Sections 3.2 and 3.3).
The dependence of the developing countries on imports of cereals (wheat and coarse grains) will continue to grow, notwithstanding lower growth of demand compared with the past. This follows from the prospect that in the post-green revolution period, and in the face of growing resource scarcities, particularly of irrigation, developing countries' potential to increase production is also more limited compared with the past. Their net cereal imports are projected to rise from 103 million tonnes in 1997/99 (and the forecast 110 million tonnes for the current trade year July 2001/June 2002) to 190 million tonnes in 2015 and to 265 million tonnes in 2030. These numbers imply a resumption of the growth of the world cereal trade after a period of near stagnation. The latter was mainly the result of the virtual disappearance of net cereal imports of the transition economies in the 1990s as well as the slowdown of the economies and oil export earnings in many countries, particularly in the major importing region Near East/North Africa.
These factors that depressed export markets available to the traditional exporters of cereals will be less limiting in the future, but they will not disappear entirely. Not only are the transition economies unlikely to revert to being the large net importers they were in the pre-reform period, but in the longer term they have the potential of transforming themselves into net exporters of cereals. We have made an allowance for this eventuality by estimating their net exports in 2015 and 2030 at 10 and 25 million tonnes, respectively. The traditional cereal exporters in the industrial world (United States, Canada, the EU and Australia) are expected to increase their net exports from the 144 million tonnes in 1997/99 to 224 million tonnes in 2015 and 286 million tonnes in 2030 (see Table 3.8).
The question is often raised whether these traditional exporting countries have sufficient production potential to continue generating an ever-growing export surplus. Concern with adverse environmental impacts of intensive agriculture is among the reasons why this question is raised. The answer depends, inter alia, on how much more they must produce over how many years. Production growth requirements are derived by adding the projected net exports to the projections of their own domestic demand, including demand for cereals to produce livestock products. The result is that these countries are required to increase their collective production from the 629 million tonnes of 1997/99 to 758 million tonnes in 2015 and 871 million tonnes in 2030, an increment of 242 million tonnes over the entire period, of which about 80 million tonnes would be wheat and the balance largely coarse grains. The annual growth rate is 1.1 percent in the period to 2015 and 0.9 percent in the subsequent 15 years, an average of 1.0 percent p.a. for the entire 32-year projection period. This is lower than the average growth rate of 1.6 percent p.a. of the past 32 years (1967-99), although the historical growth rate has fluctuated widely, mostly as a function of the ups and downs of export demand, associated policy changes and occasional weather shocks. The overall lesson of the historical experience seems to be that the production system has so far had the capability of responding flexibly to meet increases in demand within reasonable limits. This is probably also valid for the future.
Livestock. The world food economy is being increasingly driven by the shift of diets towards livestock products. In the developing countries, consumption of meat has been growing at 5-6 percent p.a. and that of milk and dairy products at 3.4-3.8 percent p.a. in the last few decades. However, much of the growth has been taking place in a relatively small number of countries, including some of the most populous ones, especially China and Brazil. Including these two countries, per capita meat consumption in the developing countries went from 11.4 kg in the mid-1970s to 25.5 kg in 1997/99. Excluding them, it went from 11 kg to only 15.5 kg. Including or excluding China in the totals of the developing countries and the world makes a significant difference for the aggregate growth rates of meat, although not of milk and dairy products, given the small weight of these products in China's food consumption. However, many developing countries and whole regions, where the need to increase protein consumption is the greatest, have not been participating in the buoyancy of the world meat sector. In this category are the regions of sub-Saharan Africa (with very low per capita consumption reflecting quasi permanent economic stagnation) and the Near East/North Africa where the rapid progress of the period to the late 1980s (oil boom) was interrupted and slightly reversed in the subsequent years, helped by the collapse of consumption in Iraq. Similar considerations apply to the developments in the per capita consumption of milk and dairy products.
The world meat economy has been characterized by the rapid growth of the poultry sector. Its share in world meat production increased from 13 percent in the mid-1960s to 28 percent currently, while per capita consumption increased more than threefold over the same period. In more recent years, the world meat trade has been expanding rapidly. This expansion reflected, among other things, significant moves towards meat trade liberalization, including in the context of regional trade agreements. Some drastic changes occurred as to the sources of imports and destination of exports. Japan became the world's largest net importer (it increased net imports fourfold since the mid-1980s), followed by the countries of the former Soviet Union (mainly the Russian Federation). Australia and New Zealand (together) continue to be the world's largest exporters while, in the last decade or so, the United States turned from a large net importer of meat to a large net exporter, mostly on the growing strength of its poultry sector.
Concerning the future, the forces that made for the rapid growth of the meat sector in the past are expected to weaken considerably. The lower population growth is an important factor, as is the natural deceleration of growth following the attainment of fairly high consumption levels in the few major countries that dominated past increases. For example, if China's growth in the 1990s of about 2.6 kg/person/year (leading to the 45 kg of 1997/99)2 were to continue for much longer, the country would soon surpass the per capita consumption of the industrial countries. Similar considerations apply to other countries, such as Brazil. Therefore, rather drastic deceleration in at least these countries and, given their large weight, also in the global aggregates, is to be expected.
Other countries may do better in the future than in the past, but their weight in the totals is not sufficient to halt the expected deceleration in the broad aggregates. India, with its very low per capita meat consumption (4.5 kg in 1997/99) and a population rivalling that of China, could be thought of as a potential growth pole of the world meat economy. This is not likely to be, as its per capita consumption would not exceed 10 kg even by 2030, and this under rather optimistic assumptions. At least this is what a number of studies indicate. The result is that the growth rate of world meat demand and production could grow at rates much below those of the past, 1.7 percent p.a. in the period to 2030, down from 2.9 percent in the preceding 30 years. The deceleration in the developing countries would be more drastic, because of the above-mentioned large-country effect, following the projected slower growth of aggregate consumption in China, and to a smaller extent in Brazil. Remove these two countries from the developing countries aggregate and there is only a modest reduction in the growth of their aggregate demand for meat, reflecting essentially the lower population growth rate.
Unlike meat, we project higher growth in the world milk and dairy sector than in the recent past because of the cessation of declines and some recovery in the transition economies. Excluding these latter countries, world demand should grow at rates somewhat below those of the past but, given lower population growth, per capita consumption would grow faster than in the past.
The structural change that characterized the evolution of the meat sector in the past will probably continue, although at attenuated rates. Poultry will continue to increase its share in world meat output and the meat trade will continue to expand. The trend for the developing countries to become growing net importers of meat is set to continue. This is another important component of the broader trend for the developing countries to turn from net exporters to growing net importers of food and agricultural products. Imports of poultry meat will likely dominate the picture of growing dependence on imported meat. Trade in dairy products will also likely recover with the net imports of the developing countries resuming growth after a period of stagnation from the mid-1980s onwards.
As noted earlier, the world feed use of cereals had slowed considerably in the recent past. It had grown less fast than the aggregate production of the livestock sector, although not in the developing countries. The main reasons were the collapse of the livestock sector in the transition economies, high policy prices in the EU up to the early 1990s, the shift of livestock production to poultry, and more general productivity increases creating more output from a given input of cereals. Growth in the world feed use of cereals will resume thanks to the cessation of the downward pressure on world cereals feed use exerted by events in the transition economies and their eventual reversal; the turnaround of the EU to growing feed use of cereals; the increasing weight of the developing countries in world livestock; and the attenuation of structural change towards poultry.
Oilcrops and products. This category of food products with a high calorie content has played an important role in the increases of food consumption in developing countries. Just over one out of every five calories added to their food consumption in the period since the mid-1970s originated in this group of products. In the projections, this trend continues and indeed intensifies: 45 out of every 100 additional calories in the period to 2030 may come from these products. This projection reflects above all the prospect of only modest further growth in the direct food consumption of staples (cereals, roots and tubers, etc.) in the majority of the developing countries in favour of non-staples such as vegetable oils which still have significant scope for further consumption increases.
On the demand side, the major driving force of the world oilcrops economy has been the growth of food demand in the developing countries, with China, India and a few other major countries representing the bulk of such growth. Additional significant demand growth has been in the non-food industrial uses of oils and in the use of oilmeals for the livestock sector. The growth of aggregate world demand and production (in oil equivalent) would continue to be well above that of total agriculture, although it would be much lower than in the past, 2.5 percent p.a. in the next three decades compared with 4.0 percent p.a. in the preceding three. This deceleration will reflect the factors that were discussed earlier in relation to other commodities, i.e. slower population growth, more and more countries achieving medium-high levels of consumption and, of course, persistence of low incomes in many countries limiting their effective demand.
On the production side, the trend has been for four oilcrops (oil palm, soybeans, sunflower seed and rapeseed) and a small number of countries to provide much of the increase in world output (Table 3.16). With the lower demand growth in the future and changes in policies (e.g. limits to subsidized production), the pace of structural change in favour of some of these crops could be less pronounced in the future. The sector accounted for a good part of cultivated land expansion in the past and in the industrial countries it made up for part of the declines of the area under cereals. The projections of land use in the developing countries indicate that oilcrops will continue to account for a good part of future expansion of harvested area in the developing countries. Being predominantly rainfed crops, the expansion of their production depends on area expansion (rather than yield growth) by more than is the case of other crops, such as cereals.
The rapid growth of demand of the developing countries was accompanied by the emergence of several of them as major importers of oils and oilseeds. If we exclude the five major net exporters among the developing countries (Malaysia, Indonesia, the Philippines, Brazil and Argentina), the others increased their net imports of oils and oilseeds (in oil equivalent) from 1 to 17 million tonnes between 1974/76 and 1997/99. In parallel, however, the five major exporters increased their net exports from 4 to 21 million tonnes, so that the net trade balance of all the developing countries increased slightly (see Table 3.20). In the future, these trends are likely to continue and the net trade balance of the developing countries would not change much, despite the foreseen further rapid growth of exports from the main exporter developing countries. The developing countries have so far been net exporters of oil meals, which have enabled them to maintain a positive, although declining, net trade balance in value terms in their combined trade of oilseeds, oils and meals. With the development of their livestock sector, the prospect is that their net exports of oil meals could turn into net imports. This is yet another dimension of the above-mentioned trend for the developing countries to turn into net importers of agricultural products.
Roots, tubers and plantains. These products (mainly cassava, sweet potatoes, potatoes, yams, taro and plantains) represent the mainstay of diets in several countries, many of which are characterized by low overall food consumption levels and food insecurity. The great majority of these countries are in sub-Saharan Africa, with some countries of the region (Ghana, Rwanda and Uganda) deriving 50 percent or more of total food consumption (in terms of calories) from these foods. In general, high dependence on these foods is mostly characteristic of countries that combine suitable agro-ecological conditions for their production and low incomes. With the exception of potatoes, diet diversification away from these products occurs when incomes and overall food consumption levels improve. In a number of countries (e.g. Nigeria, Ghana and Benin), quantum jumps in the production of these products were instrumental in raising food consumption from low or very low levels.
The evolution over time shows declining per capita food consumption of these products for the developing countries and the world as a whole up to about the late 1980s, followed by some recovery in the 1990s. These developments reflected two main factors: (i) the rapid decline in food consumption of sweet potatoes in China (from 94 kg in the mid-1970s to 40 kg at present), only in part counterbalanced by a parallel rise in that of potatoes, in both China and the rest of the developing countries, and (ii) the rapid rise of food consumption of all products in this category in a few countries, such as Nigeria, Ghana and Peru.
Significant quantities of roots are used as feed, including potatoes (mainly in the transition countries and China), sweet potatoes (mainly China) and cassava (mainly Brazil and the EU). The EU's feed consumption of cassava (all imported) amounts to some 10 million tonnes (fresh cassava equivalent). This is less than half the peak it had reached in the early 1990s, when EU policy prices for cereals were high and rendered them uncompetitive in feed use, leading to a process of substitution of cassava and other imported feedstuffs for cereals. The reversal is mostly the result of the policy reforms in the EU which lowered the policy prices for cereals after 1993 and re-established their competitiveness. In Thailand, the main supplier of cassava to the EU, cassava production and exports followed closely the developments in the EU. The rapid expansion of cassava production for export in Thailand is thought to have been a prime cause of land expansion and deforestation, followed by land degradation in certain areas of the country. This link provides a good example of how the effects of policies (or policy distortions such as the high support prices in the EU) in one part of the world can exert significant impacts on production, land use and the environment in distant countries.
The food products in this category will continue to play an important role in sustaining food consumption levels in the many countries that have a high dependence on them and low food consumption levels overall. The main factor responsible for the decline in the average of the developing countries (precipitous decline of sweet potato food consumption in China) will be much weaker, as the scope for further declines is much more limited than in the past. In parallel, the two factors that made for increases in the average, the increase in demand for potatoes when incomes rise and the potential offered by productivity increases in the other roots (cassava and yams), will continue to operate. It will be possible for more countries in sub-Saharan Africa to replicate the experiences of countries such as Nigeria, Ghana, Benin and Malawi, and increase their food consumption. Thus, the recent upturn in per capita consumption of the developing countries is projected to continue.
The main export commodities of the developing countries. The agriculture and often the overall economy, as well as the incidence of poverty and food security of several developing countries, depend to a high degree on the production of one or a few agricultural commodities destined principally for export, e.g. tropical beverages, bananas, sugar, oilseeds and natural rubber. In such cases the overall economies, and often the welfare of the poor are subject to changing conditions in the world markets, i.e. the rate of expansion of such markets and the prices these commodities fetch. For some commodities, the rate of expansion of world consumption has been too slow. For other commodities, such as sugar, protectionism in the main traditional import markets of the industrial countries has been a prime factor in restricting the growth of exports. In the face of such constraints, competition among exporters to capture a market share has resulted in declining and widely fluctuating export prices.
This has been particularly marked for coffee in recent years. The industrial countries account for two-thirds of world coffee consumption and their consumption per capita has been nearly constant for two decades. Given their low population growth, aggregate demand has been growing very slowly while production has kept increasing, including from recent entrants in world markets such as Viet Nam. The result has been that prices have precipitated and this has worsened poverty in the countries where significant parts of the rural population depend on coffee for a living.
The importing developing countries have been increasingly providing market outlets for the exports of other developing countries. As noted, this has been the case for commodities such as sugar, vegetable oils and oilseeds, and natural rubber. It has been much less so for coffee and, to a lesser extent, cocoa. For these latter commodities the growth of export demand over the projection period will be slow, as it will continue to be dominated by consumption trends in the industrial countries. In contrast, higher growth is foreseen in exports of sugar, vegetable oils and oilseeds, and natural rubber, fuelled by demand generated in the importing developing countries where the scope for expansion of consumption is still considerable. The policies of the industrial countries severely restricted imports of commodities such as sugar in the past (including the turning of the EU from net importer to net exporter). These policies will be less restrictive in the future, following policy reforms already agreed (e.g. the Uruguay Round limits on subsidized exports, the North American Free Trade Agreement [NAFTA] eventually leading to tariff-free access of Mexican sugar to the United States, the EU's Everything but Arms Initiative [EBA]) and the new ones that may come in the future.
As noted in the preceding section, the traditional agricultural trade surplus in the balance of payments of the developing countries has been diminishing over time and turned into a net deficit in recent years. This gradual erosion has reflected the many factors discussed above that influenced demand and supply and associated imports and exports of the individual commodities in the different countries. Among these factors are the agricultural and trade policies of the main players in world markets, foremost among them the Organisation for Economic Co-operation and Development (OECD) countries. Most OECD countries have traditionally protected their agriculture sectors heavily, partly through policies granting domestic support and partly, and closely related to the former, through trade policies, such as tariffs, quotas and export subsidies.
The impact of these policies on the trade performance and on the welfare of the developing countries has varied widely. There have been clear losers among the countries exporting products competing with those of the OECD (e.g. Argentina for cereals and livestock products, Brazil and Cuba for sugar), and clear gainers among the countries that had preferential access to the protected markets. Possible gainers (from lower prices and plentiful supplies of cereals in world markets) are to be found among the consumers of the food import-dependent developing countries, including those receiving food aid, but often at the expense of farmers in these same countries. More often, the situation is mixed, e.g. countries in North Africa benefiting from low-priced cereal imports but harmed by barriers to their exports of fresh vegetables.
These support and protection policies affected above all the trade performance (changed market shares) as well as consumers (paying higher prices) and the taxpayers (paying the subsidies) of the OECD countries themselves. For this reason, most studies that examined the possible effects of agricultural trade liberalization conclude that the lion's share of gains in welfare would accrue to the high-income countries, and certainly to some developing countries exporters of competing products, e.g. cereals, livestock, sugar and vegetables. However, some developing countries could be harmed, such as those that enjoy preferential access in protected markets or those that have few agricultural exports but import much of their food.
The late 1980s and the 1990s witnessed intensified efforts to discipline policies that distorted trade. The resulting Uruguay Round Agreement on Agriculture (URAA) enshrined a certain measure of discipline. It mandated reductions in border protection, trade-distorting domestic support and export subsidies. These reductions, however, still left the countries that made heavy use of them in the pre-URAA period, i.e. primarily most OECD countries, with considerable scope for continuing them, albeit at lower levels than before. For countries that made little use of domestic support and export subsidies, overwhelmingly the developing ones, the URAA meant that they were left with very little scope for using such policies in the future, generally within the limits of the de minimis clause. In practice, the URAA legitimized and in a sense froze the divide between high-protecting countries and the rest. However, there were some compensations. Developing countries were not required to reduce and could increase, or introduce for the first time, support aimed at their agricultural development, e.g. investment or input subsidies. In addition, the widely diffused practice of binding tariffs at levels well above those effectively applied in the pre-URAA period afforded significant scope for increasing border protection through higher tariffs in the future. These possibilities notwithstanding, some analysts argue that the URAA may have«institutionalized» the production and trade-distorting policies of the OECD countries without addressing the fundamental concerns of developing countries.
Continuing negotiations on agriculture to liberalize trade further are under way. They began in March 2000 and were later subsumed in the broader round of multilateral trade negotiations launched by the WTO at its Fourth Ministerial Conference (Doha, November 2001). In these negotiations the potential exists for the concerns of the developing countries to be addressed more effectively than in the past under the provision for special and differential treatment of the developing countries to reflect their development needs. At the same time, however, further liberalization will tend to erode the gains enjoyed by several developing countries of preferential non-reciprocal access to the protected markets of the major OECD countries (though generally to the benefit of other competing developing countries). Four non-reciprocal preferential arrangements are of particular relevance: the Generalized System of Preferences (GSP) under the WTO; the African, Caribbean and Pacific Group of States (ACP)-EU Cotonou Agreement; the US Trade and Development Act of 2000; and the EBA to provide duty-free and quota-free market access to the EU for the products of the least developed countries (LDCs).
While the importance of«classical» border measures (such as tariffs and quotas) gradually diminishes, the prominence in trade of the safety and quality standards increases. The latter concerns mainly the WTO Agreement on the Application of Sanitary and Phytosanitary Measures (SPS), but also the quality attributes covered under the Agreement on Technical Barriers to Trade (TBT) and the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS). Recent food safety scares and the advent of genetically modified products have added to the influence of these standards in trade. Their application will be increasingly coming under scrutiny in order to minimize the risk that they will be used inappropriately to protect domestic producers from foreign competition.
The importance of standards in trade has been greatly boosted by the advance of globalization in food and agriculture. What was once a set of national markets linked by raw material trade from land-rich to land-scarce countries is gradually becoming a loosely integrated global market with movements of capital, raw and semi-processed goods, final products and consumer retail services. Intrafirm and intra-industry trade is increasing in importance so that the food trade is assuming certain characteristics of the trade in manufactures. Trade policy with respect to food and agriculture is gradually shifting beyond concentration on issues pertaining to primary farming to encompass issues and interests of the whole food chain, including food processing, marketing and distribution.
The thrust towards a globalized food and agriculture economy is seen as offering opportunities for developing countries to improve the performance of their agricultural and food sectors. This is part of the wider argument that, generally speaking, policies that favour openness of the economy boost economic growth. This has knock-on effects on reducing the numbers below the poverty line, although not necessarily on reducing the income gap between the rich and the poor or, for that matter, between rich and poor countries. These are not arguments in favour of«big bang» liberalization, whether of trade or capital flows. Empirical evidence suggests that openness and outward-oriented policies are not per se guarantors of success. More important are the companion policies on the domestic front that facilitate the integration into global markets. These are policies that provide appropriate transition periods towards freer trade; help adapt new, external technologies to the domestic environment; or provide competition policy settings and design contracts that also allow small-scale agriculture to thrive within the operations of transnational corporations (TNCs).
As an important element of globalization in food and agriculture, experience suggests that TNCs can make an important contribution as vehicles of capital, skills, technologies, access to both domestic and export marketing channels, and creation of linkages to the rural economy, for example through contract farming. There are, however, snags such as excessive concentration of market power (and its eventual abuse) in the hands of a few large, sometimes vertically integrated, enterprises operating in many countries. These should be mitigated by appropriate policies, not by closing the economy to the broader benefits and, of course, domestic enterprises are not devoid of monopolistic elements; on the contrary, competition from new entrants, even if foreign, can be a welcome boost to competition.
1 Numbers refer to the population of the countries for which estimates of undernourishment were made.
2 It is thought that these very large increases appearing in the FAO food balance sheet data and the discrepancies from independent consumption data result mainly from an apparent overestimation of China's meat production (see Chapter 3).