COMMITTEE ON WORLD FOOD SECURITY
Rome, 31 May - 3 June 1999
ASSESSMENT OF THE WORLD FOOD SECURITY SITUATION
1. World Food Summit undertakings and guidance provided by the CFS at past sessions call for improvements in the annual assessment document that would give "a more consistent and up-to-date picture of undernutrition in the developing countries, including access to food, local food, and measures of poverty, and should provide a quantitative basis for Summit follow-up analysis." (Report of the 24th Session of the Committee on World Food Security, CL 115/11, November 1998).
2. To respond to this guidance, this document presents a new structure for annual world food security assessments so as to address the following four objectives:
3. Improvements in databases and assessment methodologies are under development through the Food Insecurity and Vulnerability Information and Mapping Systems (FIVIMS) initiative. This document introduces a new method for presenting range estimates and trajectories for use in future monitoring of the number of undernourished in the developing world, and suggests a more systematic approach for monitoring underlying structural conditions through time. It also summarises recent developments with regard to the global food security indicators that have been traditionally used by CFS to monitor the current situation. The Committee's views on the appropriateness of the proposed structure are sought.
4. The method used by FAO to estimate the number of undernourished people1 in the developing world has already been described in other widely-circulated documents at the time of the World Food Summit itself and, more recently, when issuing the first update of the Summit estimate in June 1998 (See, for example, the Sixth World Food Survey, FAO, 1996 and Information Note on Estimation of the Number of Undernourished, CFS: 98/Inf. 8).
5. In most countries direct and up-to-date measurements of amounts of food actually consumed are not readily available. Therefore FAO has developed a method for estimating the prevalence of undernourishment in a country that combines information on food availability and population with information about access to food (inequality in the distribution of food among households). The aggregate of these estimates for 98 developing countries for the period 1990-92 gave the estimate of about 840 million undernourished people in the developing world that was used at the time of the World Food Summit, based on information available at the time.
6. Some of the data used by FAO to produce its estimates are, of necessity, quite old and reliable only within a certain range. This is particularly the case for data that measure the inequality of food consumption among people, and for data on foods that form an important part of the diet, but for which production is not always recorded accurately. In many countries, population data also is not accurate. For these reasons, the estimate of the number of undernourished people in the world in any given year can be considered accurate only within a certain range representing the likely margin of error surrounding it (about plus or minus 5%). Later revisions are unavoidable, since countries frequently report retroactive revisions for important data series such as population and production.
7. Using the concept of a range instead of a point to represent the number of undernourished, the original estimate for the reference period 1990-92, published at the time of the Summit, has been transformed from a point estimate of 840 million into a range estimate of 800 - 880 million. Graph 1 shows this range, together with a revised point estimate generated in 1998 for that same reference period. For purposes of comparison the estimated ranges and revised point estimates for two earlier periods (1969-71 and 1979-81) are also shown.
8. Looking ahead to the year 2015, the graph also shows how a trajectory starting in 1996 would need to move through time in order for the target of 400 - 440 million undernourished to be achieved. Projections contained in FAO's 1995 publication, World Agriculture: towards 2010, which are also shown, suggest that, under a business as usual scenario, without effective implementation of the Summit Plan of Action, the number is likely to decline at a much slower rate.
9. Graph 1 shows that between the periods 1990-92 and 1994-96 there was in fact a slight increase rather than a decline in the point estimates. While the magnitude of this change is not large enough to suggest any structural change in the underlying trends, it does confirm the concerns that led to the convening of the Summit. In short, aggregate performance up to the time of the Summit gave little ground for optimism and, indeed, much cause for concern.
10. Regular monitoring of global market conditions for staple foods and other traded commodities and of effects of changes in prices and foreign exchange rates on cereal import bills, provides the basis for current assessments and warnings of unfavourable conditions that may be developing regionally or globally. Food and agricultural emergencies caused by natural or man-made disasters are also a regular occurrence that create food insecurity for affected people almost continuously in one part of the world or another.
1. FAO estimates of undernourishment are subject to uncertainty, due to imperfect information available and are at intervals revised retrospectively. Range estimates are therefore more appropriate to illustrate knowledge and projections about undernourishment. The estimated range for the past, projected and target trajectories in Graph 1 is based on a range of 5 percent above and below the past, projected and target numbers considered by the WFS. Within these range trajectories, the most recently elaborated point estimates are shown.
2. The _ 5% range is consistent with the observed margin within which these estimates have varied over the years, as new and better data has become available. As work continues to validate national estimates and reduce the margin of error in the FAO estimates, it is hoped that greater accuracy can be introduced in the monitoring and assessment of this important outcome indicator.
3. Under the FIVIMS initiative, work has begun to validate and improve the FAO estimates. Of particular importance in this context is the systematic use of food consumption distribution data derivable from existing household income and expenditure surveys. At the same time, a number of countries have indicated their intention to use other methods suitable to their own circumstances to obtain national estimates of the number of undernourished. As these national estimates become available, the results obtained will be compared with the FAO estimates, and reasons for any differences explained.
4. A supplementary estimate for the number of undernourished people in developed countries and countries in transition will also be generated and monitored in subsequent years, in order to provide full coverage of progress being made toward the Summit goal.
11. Table 1 reports the results of assessments of general market conditions and the incidence of food and agricultural emergencies undertaken by FAO's Global Information and Early Warning system (GIEWS) with reference to the 1998-99 marketing year.
12. Table 1 shows seven individual food security indicators that have been in use since the mid-1970s. These indicators, while confined to cereals, shed light on the present and future global food situation due to the weight of cereals in the global food basket. The analysis of the changes in 1998/99 compared with the previous year, and their implications, are discussed below.
13. The first indicator provides information about global stocks in relation to probable magnitude of market demand. In general, the FAO Secretariat considers that a 17-18 percent ratio of end-of-season cereal stocks to trend utilisation for the coming marketing year, as the minimum necessary to safeguard world food security. This consists of a 12 percent working stock and a 5-6 percent reserve stock element. This year, in spite of an anticipated build-up in carryovers held by the major exporters, global cereal stocks are forecast to be reduced slightly from their opening levels, as smaller stocks are expected to be held by other countries that are important in world markets, in particular China and the Russian Federation. The ratio of global end-of-season cereal stocks in 1998/99 to trend utilisation in 1999/2000 is forecast to decline slightly to 17.2 percent. This is well below the average of the first half of the 1990s, but higher than in the mid-nineties, and is within the 17-18 percent range which the FAO Secretariat considers as the minimum necessary to safeguard world food security.
14. The second indicator measures the ability of the five major grain exporting countries to meet the demand for wheat and coarse grains imports. It relates the sum of their production, imports and opening stocks to the sum of their domestic utilisation plus exports. This ratio is expected to improve slightly in 1998/99 due largely to bumper crops in the EC and the United States, especially for coarse grains, and larger average opening stock levels than during the past few years. The current indicator of 1.17 is also about the same as the average ratio estimated for the 1991/92-95/96 period.
15. The third set of indicators presents the ratio of the volume of closing cereal stocks held by the major exporters of wheat, coarse grain and rice to the total disappearance of these cereals (domestic consumption plus exports). Based on current forecasts, these indicators, except for rice, reflect some improvement in the global grain supply situation in 1998/99 compared to the previous year and the averages during the earlier period. Among the major rice exporting countries, particularly China and Vietnam, are expected to draw down their stocks due mostly to smaller and/or stagnant production in 1998.
16. Indicator four, measuring changes in cereal production among the major cereal importing countries of China, India and the CIS against the trend and the preceding year, shows a negative output growth in 1998/99. There was a significant reduction in grain output in the CIS in 1998, with less pronounced shortfalls in China and India compared to the previous season. Variations in annual cereal output of this group of countries have traditionally had a significant influence on the size of world trade in cereals. However, this influence may be on the decline, since cereals imports by this group have fallen from 20-25 million tonnes as recently as in the mid-1990s (about 10 percent of world imports) to 10-11 million tonnes in 1998/99, or 5 percent of global trade.
Table 1 - CHANGES IN FOOD SECURITY INDICATORS
|1.Ratio of World Cereals Stocks to World Cereal Consumption Trends||18.3||16.2||17.6||17.2|
|2. Ratio of Five Major Grain Exporters4 Supplies to Requirements||1.16||1.12||1.15||1.17|
|3. Closing Stocks as %age of Total Disappearance of Major Cereal Exporters|
|Annual Trend Growth Rate||Percentage Change from Previous Year|
|4. Changes in Cereal Production in China, India and CIS||1.66||5.40||3.12||-6.02|
|5. Changes in Cereal Production in LIFD Countries||2.31||7.30||-2.15||0.88|
|6. Changes in Cereal Production in LIFDCs less China and India||1.88||8.71||-6.01||4.82|
|Percentage Change from Previous Year|
|7. Export Price Movements7 (Annual Averages)||Wheat (July/June) Maize (July/June) Rice (Jan./Dec.)8||-16.3 -15.0 -12.7||-21.3 -17.0 -8.5||-16.9 -17.0 0.5|
17. Changes in aggregate cereal production of the Low Income Food Deficit Countries (LIFDCs), shown by indicator five, provide another way of measuring food security in a group of countries considered vulnerable to fluctuations in supplies. However, given that production in China and India heavily influences the aggregate output for LIFDCs, indicator six excludes these two countries. Thus, while indicator five shows only a slight improvement in 1998, indicator six, i.e. changes in aggregate cereal production of the LIFDSs, excluding China and India, points to a substantial aggregate gain for the remaining countries. The large difference in the two indicators for 1998 reveals the heavy weight of the world's two largest cereal producers, which both harvested smaller crops. It should be noted, however, that the fall in production in several countries is measured in relation to record and/or bumper crops which had been registered during the previous season.
18. A comparison of export prices for the major cereals is provided by indicator seven. Except for rice, international cereal prices continued to trend downward as a result of weak global import demand and large carryover stocks, especially among the major grain exporting countries.
19. In general, the seven indicators confirm an overall improvement in the cereal supply situation in 1998/99 compared to the preceding year. Of some concern, however, has been the continuing decline in grain prices which have already led to lower-plantings, especially among the major exporting countries of the EC, Canada and the United States. As these countries hold the bulk of the world's surplus supplies of cereals, next season's inventories could be drawn down even if import demand remains unchanged.
20. Although cereal production in developing countries in 1998 is estimated to have improved marginally from the previous year, the number of countries facing food emergencies as of mid-March 1999 stands at 38 compared to 37 towards the end of 1998, mainly due to adverse weather and civil strife. Locations of particular concern include:
21. FAO's forecast of cereal food aid shipments in 1998/99 (July/June) now points to a total of 8.5 million tonnes, a 2.7 million-ton rise over those estimated for the previous year. The greater availability of grain supplies among the major donor countries combined with higher food aid requirements, particularly from Asia, Central America and the CIS, is expected to account for this impressive boost in global cereal food aid shipments.
22. Cereal import bills are expected to continue to decline in 1998/99 (July/June) for the third consecutive year, thus returning a level similar to that prevailing in the years just before the high import bills of 1995/96. The most pronounced reduction is forecast for the developing countries whose import bill is expected to decline by 16 percent from US27 000 million in 1997/98 to US$ 22 800 million in 1998/99. This is due to a combination of reduced import volume, lower international prices and larger food aid shipments. For the Least Developed and Net Food-Importing Developing Countries, the group covered by the Uruguay Round Ministerial Decision, the forecast for the cereal import bill in 1998/99 is for a decrease of 19 percent, amounting to about US$ 1 500 million. In the case of these developing countries, the smaller cereal import cost has more to do with reduced imports due to a recovery in domestic output as well as lower import prices than with larger food aid shipments.
23. Monitoring trends in structural factors that underlie under-nourishment provides a basis for assessing the extent to which this condition is, or is not, changing fundamentally over time. Various methods for doing this are under investigation9. The method reported here involves clustering countries according to prevalence of under-nourishment, and monitoring indicators for key food security outcomes, namely, food availability and access to food, and key vulnerability factors related to demographic, economic, environmental, political and social conditions.
24. This method was introduced for the first time in last year's assessment document. Several improvements have been made in its application over the past several months. A statistical technique has been introduced for clustering countries according to prevalence of under-nourishment, and a more systematic analysis of potentially relevant indicators has been carried out.
25. Graph 2 shows the 98 countries in the developing world, clustered according to the share of the undernourished in their national populations as estimated at the time of the Summit. It also shows the number of undernourished for each cluster. Whereas, not surprisingly, the largest absolute number of undernourished people are found in Asia, this graph clearly reveals that the most widespread problems are found in the low-income food-deficit countries of Africa, where the share of undernourished in the national population often exceeds twenty-five percent, and in twenty of them reaches nearly 40% or more.
26. Chronic food insecurity, i.e. under-nourishment, is the direct result of inadequate levels of food consumption through time (see box). However, the actual level of food intake is not easy to monitor.
27. The underlying conditions that create food insecurity for specific locations or population groups are large in number, and vary in relative importance from one location to another, and from one group to another. Nevertheless, these underlying conditions are more easily monitored than changes in food intake. Furthermore, even without very precise information, it is possible to make some qualitative statements about the way in which changes in these conditions are likely to be affecting, positively or negatively, the food security status of vulnerable people and countries. Such information is valuable not only for monitoring progress but also for indicating what kind of policy and programme interventions are most needed in different situations.
Table 2: Indicators representing Food Security Status and Vulnerability Factors, 1990-92 - 1996
|Indicators of desired food security outcomes||Indicators of demographic conditions||Indicators of environmental conditions||
Indicators of economic conditions
|Indicators of political conditions||Indicators of social conditions|
|Food availability||Food access||Urban/rural population distribution||Natural resource base||Structure of national economy||Level of development in rural economy||Performance of food economy||Incidence of civil conflict||Child nutritional status|
|Daily energy supply per caput
|GNP per caput
|Share of rural population
|Per capita arable land
|Share of agriculture in
|Proportion of roads that
|Yields per hectare (major
|Proportion of countries that
experienced an emergency situation
|Under 5 mortality rate
(deaths per thousand live births)
|Countries grouped by prevalence of undernourishment||1990-92||1996||1990-92||1996||1990-92||1996||1990-92||1996||1990-92||1996||1990-92||1996||1990-92||1996||1992-96||1990||1996|
|Group 1: >50% undernourished||1786||1806||169||116||71.0||68.1||0.30||0.25||41.6||38.8||13||16||8||9||56%||251||225|
|Group 2: 38-50% undernourished||2179||2243||402||427||69.9||67.3||0.25||0.22||29.7||31.5||17||18||12||15||36%||153||156|
|Group 3: 27-37% undernourished||2162||2149||312||298||69.8||65.8||0.21||0.20||29.0||30.3||22||22||12||12||29%||178||130|
|Group 4: 17-26% undernourished||2337||2387||538||685||56.0||53.4||0.18||0.16||24.0||21.9||29||33||24||26||21%||124||98|
|Group 5: 8-16% undernourished||2711||2855||827||1101||44.1||41.5||0.14||0.13||17.6||18.0||39||38||28||31||18%||52||51|
|Group 6: < 8% undernourished||3023||3124||2884||3690||31.8||28.4||0.24||0.24||15.8||16.0||75||67||22||26||8%||69||45|
28. Outcomes sought by the World Food Summit, as by all other recent international conferences on economic and social issues, relate essentially to quality of life in a human rights context, that is, they establish goals that would lead to human wellbeing for all persons on this planet. Two major goals established by the Summit as relevant for reducing the number of undernourished and achieving food security for all are: availability of adequate amounts of safe and nutritious food, and physical, social and economic access by all people at all times to the food they need to meet their dietary needs and food preferences for an active and healthy life. Other social goals such as poverty eradication, good health, universal education, gender equality, and so forth, also figure importantly in the Summit commitments because of their close relationship to food security. Due attention will be given to appropriate indicators of social goals in developing the overall monitoring system to be implemented under the FIVIMS initiative. However, as the focus of the annual CFS assessment report is on the world food security situation, only the two directly related outcomes - food availability and access - are proposed for coverage. At the national level, the outcomes retained for monitoring in this report are food availability and food access, represented by per caput dietary energy supply for food availability, and GNP per caput for food access.
29. As noted above, a large number of vulnerability factors may be associated with underlying structural conditions in each of five main categories. In this section, a brief description of various factors pertaining to each main category is provided, together with an indication of the relevant vulnerability factor(s) selected for coverage in this document.
30. These indicators have been selected with regard to the logical coherence and internal consistency among them. Work is underway to create more composite indicators that would help to place each individual indicator in a more meaningful socio-economic context. Pending completion of this work, this document presents a first set of results based on monitoring each of the selected indicators individually. These are shown in Table 2.
31. Demographic conditions create vulnerability when size of population exceeds the carrying capacity of a particular area, and there is limited opportunity for out-migration or for development of physical, social and economic infrastructure so as to provide more productive alternatives to the dominant livelihood systems in the area. High share of rural population in the total may indicate the presence of this kind of vulnerability, particularly if it occurs together with a low level of economic development in rural areas. A high proportion of dependent persons within a family, community, locality or nation also increases the risk of under-nourishment for these persons. Urban/rural population distribution is the demographic vulnerability factor retained for monitoring in this report, represented by share of rural population in the total.
32. Environmental conditions can create chronic vulnerability in several ways. People living in areas where the natural resource base is poor or deteriorating often have limited opportunities for earning their livelihood. Their situation is worsened if acts of man lead to pollution and environmental degradation. Variable climatic and geophysical conditions and biological threats create additional risk. Availability of arable land per caput usually declines with economic development, as more and more land is dedicated to non-agricultural use, and high-technology, high-yielding agricultural practices are introduced on the remainder. However, if availability of arable land per caput is declining solely as a consequence of population growth, without compensating improvements in productivity or in the performance of the national economy, the result is likely to be increasing levels of under-nourishment. Natural resource base is the environmental vulnerability factor retained for monitoring in this report, represented by arable land per caput.
33. Economic conditions can be monitored and assessed at various levels - national, sectoral, or zonal. The structure and performance of the national economy and its components can affect the food security situation of an entire nation through the performance of food markets. These are in turn affected by factors such as food prices, interest rates, inflation rates, labour market conditions, foreign exchange rates, and trade balances. Economic conditions can also create vulnerability and food insecurity if assets and incomes are distributed inequitably among the population, or if public and private sector investment is inadequate or skewed. The degree to which an economy is or is not diversified will often determine whether or not employment and income-generating opportunities exist that would provide sufficient purchasing power to meet basic food needs to all segments of the population. Also, the level of development and the dynamism of economic activities in rural areas often have a strong influence on the level of under-nourishment nationwide. Structure of national economy, level of development in rural economy, and performance of food economy are the economic vulnerability factors retained for monitoring in this report, represented by share of agriculture in GDP, proportion of roads that are paved, and yields per hectare for major cereals.
34. Political conditions can affect food security positively or negatively. Political structures that encourage people's participation tend to reduce vulnerability. But, the presence of civil conflict is a vulnerability factor which can restrict employment and market opportunities, and may lead to loss of assets, destruction of social and physical infrastructure, and even displacement from their homes for affected households. Some of the highest levels of under-nourishment are found among people living in areas affected by chronic conflict. Incidence of civil conflict is the political vulnerability factor retained for monitoring in this report, represented by proportion of countries experiencing emergency situations.
35. Social conditions, including both the state of social services and prevailing social attitudes, have a very important influence on vulnerability and food insecurity. Where people have access to social infrastructure such as primary education, health care centres and extension services, serious under-nourishment is less likely to be found. People's traditional attitudes influence the kinds of food they eat, the way in which available resources and food are distributed, and the kinds of hygiene, food preparation and caring practices that they most commonly follow. Traditional knowledge offers possibilities for finding innovative solutions to local problems but traditional attitudes may also create a stumbling block in certain circumstances. A social vulnerability factor that combines access to social services with social attitudes regarding appropriate care and feeding practices, is retained for monitoring in this report. It is child nutritional status, represented by under 5 mortality rate.
36. As expected, in general, the table shows that the lower the level of per caput DES and per caput GNP, the higher the prevalence of under-nourishment. One exception is that Group 2, with 38-50% undernourished, appears to be slightly better off than Group 3, with 27-37% undernourished. However, the differences between these two groups are slight, and it may be that a few large countries within one or the other of the groups are pulling the group average in one direction or the other. Overall, looking at the two indicators together, conditions in Groups 1, 2 and 3 appear to have worsened during the five years since the 1990-92 reference period for estimating the number of undernourished in the developing world, a finding consistent with the slight increase in the number estimated for 1995 compared to the reference period.
37. An attempt is underway to isolate the most important vulnerability factors associated with structural conditions that affect the prevalence of under-nourishment in each of the six classes of countries shown in Graph Two. Although more work is still needed to refine the approach, and apply it at sub-national as well as national level, Table 2 already indicates that, on average, consistent relationships can be observed between the performance of key vulnerability factors and the level of under-nourishment in each of the six clusters of countries depicted in Graph 2.
38. In brief, among countries one would expect the prevalence of under-nourishment to decline as the magnitude of vulnerability represented by each of the selected factors declines. Thus, prevalence of under-nourishment should decline with decreases in:
and with increases or improvements in:
And these are precisely the relationships that one does find, with the exception of the relatively higher per caput availability of arable land in Group 6 (where the large cultivable land area in one or two large countries probably influences the result), and the relatively higher under-five mortality rate for Group 3 in the reference period.
39. Similarly, one would expect that, if the vulnerability factors moved in the above-mentioned directions over time, then the prevalence of under-nourishment should decline. For the period between 1990-92 and 1996 (the latest year for which reliable data are available), all indicators for all country groups moved in positive directions, except share of agriculture in GDP which increased in groups 2, 3, 5 and 6, and proportion of roads paved, which remained stagnant for Group 3. This suggests that recession in the non-agricultural sector for a number of countries may have contributed to the slight increase in the number of undernourished that was observed during the five year period just preceding the Summit.
40. These conclusions are based on averages for each country group. The usefulness of this approach for individual countries is still under investigation. Current thinking is that, within each country grouping, there are probably subgroups of countries for which one or a few underlying factors dominate. Furthermore, it seems likely that composite indicators may more completely capture the interaction of relevant vulnerability factors at country level than the monitoring of individual factors one-by-one. These ideas will be further developed by a technical subgroup of the IAWG-FIVIMS led by the FAO secretariat and the Geographic Information System (GIS) - Consortium of the Consultative Group on International Agricultural Research (CGIAR) during the coming year.
41. For specific groups of people, the nature of vulnerability differs according to the kind of food insecurity involved. These are of three kinds:
42. Vulnerability for the first and second types of food insecurity is present when structural conditions exist that create a risk of chronic food insecurity for the people affected by these conditions. Only for the third type of food insecurity does vulnerability derive from risk that is attributable to predictable, but irregular, variability in underlying conditions.
43. Selection and interpretation of key indicators to assess the vulnerability of different population groups is largely determined by the procedure that is used to define the vulnerable groups.
44. Work is currently underway to ascertain which approaches are most commonly used across countries, with the intention of eventually deriving a classification system that can be used for monitoring these groups at global level. This work is not yet sufficiently advanced to be used as a basis for this document. However, an Information Note will be tabled (CFS: 99/Inf. 6) that reports on results of work in progress.
1 See box for definitions of this and related terms.
2 Council document CL 116/2, "Current World Food Situation", April 1999, covers trends and outlook for production and marketing of food staples, as well as other agricultural commodities, and gives up-to-date information on food emergencies, food aid, and cereal import bills.
4 Argentina, Australia, Canada, EC and the United States.
5 China, Pakistan, Thailand, United States and Viet Nam.
6 For 1998/99 wheat and maize prices, eight months averages only (July/February). Changes are calculated by comparing the first eight months of the season with the corresponding period in 1997/98.
7 Wheat: U.S. no. 2 Hard Winter; Maize: U.S. no. 2 Yellow; Rice: Thai Broken (A1 Super)
8 Rice prices are based on the calendar year of the first year shown.
9 A first attempt to monitor access to food was made prior to the World Food Summit, when the aggregate household food security index (AHFSI) was introduced. This index gave a rank order for the household food security status of developing countries, but it did not indicate whether the absolute situation was improving or worsening time within a particular country. Since the publication of actual estimates of the number of undernourished at national level, the AHFSI is no longer being maintained.
10 This indicator, strictly speaking, is not appropriate for monitoring the situation at national level, since it refers to the situation within a group of countries. Also, some emergency situations not related to conflict are included in the database. This indicator will be replaced by a more appropriate one in future reports.