National Academies Press: OpenBook

Food Aid Projections for the Decade of the 1990s (1988)

Chapter: Workshop Proceedings

« Previous: Executive Summary
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 9
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 10
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 11
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 12
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 13
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 14
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 15
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 16
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 17
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 18
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 19
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 20
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 21
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 22
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 23
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 24
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 25
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 26
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 27
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 28
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 29
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 30
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 31
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 32
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 33
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 34
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 35
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 36
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 37
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 38
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 39
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 40
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 41
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 42
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 43
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 44
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 45
Suggested Citation:"Workshop Proceedings." National Research Council. 1988. Food Aid Projections for the Decade of the 1990s. Washington, DC: The National Academies Press. doi: 10.17226/1418.
×
Page 46

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Food Aid Projections for the 199Os: Workshop Proceedings INTRODUCTION At the request of the Agency for International Development's Bureau for Food for Peace and Private and Voluntary Organizations, the Board on Science and Technology for Inter- national Development (BOSTID) of the National Research Council (NRC) arranged for an NRC-appointed pane} and a group of experts to convene for two days of discussions concern- ing projections of needs for food aid in the decade 1990-2000. The objective of the meeting was to examine the projections of food commodity trade and, either directly or by deduction, food aid needs of developing countries, relying upon work of six principal groups engaged in food commodity analysis the United Nations Food and Agriculture Organization (FAO) Commodities and liade Division (and others), the World Bank International Commodity Markets Division, the U.S. Department of Agriculture (USDA) Economics Research Service Commodity Bade & Analysis Branch (and others), the International Food Policy Research Institute (IFPRI), the Iowa State University Center for Agricultural Research and Develop- ment (CARD), and the International Institute of Applied Systems Analysis (~lASA). The projections of these groups were discussed from the perspective of a number of specialists engaged in parallel types of analysis of future global economic, regional econo-political, demographic, and climatic impact, scientific and technological research impact, and risk forecasting. DEFINITIONS There are two basic approaches to defining food aid needs: ~ "Supply-stabilization~ food aid- to stabilize market prices in the recipient country by making up the shortfall ("need gap") between production and consumption. Such aid is thus "demand-based" or "demand-driven". Most typically it is delivered into the local governmental or commercial food distribution system; seldom is there an attempt to target local populations beyond existing policies. Food stabilization aid of this type helps developing countries with economic growth, through saving foreign exchange, assisting in generating local income, and insuring against domestic economic instability. It can help 9

to cushion their vulnerability with respect to world commodity markets, though food aid volumes may decline when prices increase, as many donors make allocations in money terms. · "Bunger-responsiven food aid aid sufficient to alleviate hunger through making up nutritional deficiencies. This estimate of food needs attempts to identify populations at nutritional risk. The goal is to provide food to affected populations, such as women and children, suffering disproportionate deprivation, attempting to alleviate both chronic hunger and acute periodic deprivation. This goal leads to estimates of need necessarily higher than the minimum caloric shortfall for any specified hungry populations since some substitution effects occur among this population. Currently, relatively little regular food aid is based on this hunger-responsive type of estimate. Projections for "supply stabilization" and "hunger-responsive" food aid do not include soft credit programs of the European Community or the United States, such as the Guar- anteed Market Supply (GMS) or Export Enhancement Program (EEP) entitlements, but they do include sales under Title ~ of Public Law 480, as well as U.S. and other donor nation grant aid. Past emergency aid, because of its stabilizing eject, as well as its humanitarian motivation, is used for calculating the trade proportion of food aid in some estimates (not those of IFPRI). In calendar year 1987 alone, export subsidies on wheat and flour provided by the Euro- pean Economic Community and the United States to developing countries amount to about $1.6 billion, or more than half the value of all food aid recently provided by all countries in the Organization for Economic Cooperation and Development (OECD). Although these programs also provide importers an opportunity to use their foreign exchange for other uses, perhaps for development-oriented investments or for debt repayments, they are not appropriately defined as food aid. METHODOLOGY . Projecting food commodity trade and aid has consistently been attempted since the Second World War, thanks, in part, to the creation of the FAO, and the expansion of trade and development concerns in industrial countries. Mostly, estimates were based on immediate demand and supply availability rather than on detailed need estimates. However, with improvements in the global system of reporting production and export and import statistics, and, more recently, with the ability to store and process large amounts of information on computers, sophistication in using agricultural statistics for forecasting has improved. But the complex interdependence among the variables requires models that account for many relationships. There are, therefore, no entirely satisfactory methods of predicting food trade and food aid needs, even over periods as short as five or ten years, let alone the fifty-year time horizon set by Brown University's Hunger Program (Kates et al., 1988~. The most widely used studies on food aid (cereal) needs or requirements are those prepared by the FAO and USDA. These are principally made for the coming year, or at most, the year after. The methods that are used to project future trade, and derivatively, aid needs, all Involve a complex set of estimates and assumptions. These include estimates of relevant forces affecting individual countries, for example, the expected rate of general economic growth (which reflects purchasing power), the growth of population (which together with income is the major determinant of demand for food), and the amount of cereal grain production. Assumptions leading to these estimates, along with decisions as to which countries to include in estimates (such as China and India, whose size tends to swamp 10

the data on international trends arising from smaller countries) account for the primary differences among the main forecasts. here is no general agreement about the validity of various assumptions. Most economists agree that a general equilibrium model, in which as many as possible of the variables of relevant countries' economies are allowed to interact together to assess the impact of varying any or each of them, including the insertion of food aid, would be able to give the most fully considered estimates. However, these general equilibrium models are so complex that none currently exists that is agreed to be especially useful In projecting food aid needs. The International Institute for Applied Systems Analysis QUASAR, in Laxenburg, Austria, a center for international cooperation in developing this type of modeling, among other types of research, has creates} one basic world food model. Since suitable general equilibrium models are not available, many econorn~sts use partial equilibrium models, in which a number of the variables are assumed to remain constant, or to grow at an assumed fixed (positive or negative) rate, while other variables are projected. Partial equilibrium analysis is widely considered to be a valid means to assess the impact of limited changes in the dependent variables. However, the longer the period and the greater the changes projected, the less confidence there is in the results. For an economic factor as important to some countries as food aid, it is likely that supplying food aid over a number of years will create economic changes that will alter econorn~c variables assumed to remain constant. In the extreme this could mean that food aid supplied over the first half of a projected period would cause economic growth rates to rise, hence incomes would rise, local food prices would rise without large increases in supply, and therefore the numbers of nutritionally disadvantaged would increase. After some lag, however, higher food prices could stimulate both income growth among rural producers and large increases in production, as occurred in China in the early 1980s, for example. The implicit expectation is that eventually the effects of food aid could actually result in a decline in its need, an agricultural growth would be adequate to cover growing domestic demand. The estimates presented at the workshop, with the exception of those prepared by the IFPRI, were derived as rather simple projecl;ions or logical deductions. They were put forward within the framework of fairly robust estimates of LDC food imports in the 1990s, generated by models of international commodity trade in cereals and national staple foods consumption behavior. Developing the estimates of hunger-responsive and price-stabilizing food aid amounts involved a recognition of "normative elements in forecasting. That is, they included those elements which, from prior experience, are taken to be normal or reasonable amounts. In addition to the needs projected from import gaps or nutritional requirements, there are food aid needs arising from exceptional circumstances. These may result from political, economic or "exogenous shock" factors (such an drought, civilstrife, or other unforeseen occurrences). Actual food aid deliveries, as opposed to needs, may be expected to be lower than the workshop projections of food aid needs, if we assume no changes in donor country policies. This reflects constraints on the ability of donors to provide the estimated amounts. As Ascher (1978) points out, "The introduction of 'normative' forecasting, which is the use of projections to systematically explore and select goals and alternatives, is the most prominent indication of the effort to transform the forecast from isolated information into a decision-making process in its own right." Aside from the key assumptions of a normative nature regarding target gods for food aid, other assumptions regarding econorn~c behavior were involved in the estimates derived at the workshop. 11

A second point about the estimates is that they were based on expert judgment. A con- siderable degree of consensus among the experts encouraged a certain degree of confidence in the average figures derived, and the boundaries around them. The ultimate test of esti- mates which forecast future food aid flows and levels of food aid needs lies in their accuracy. In the 1990s, actual deliveries of food aid will occur. Whether these satisfy either of the two "needs" supply-stabilization for poor developing countries (the FAO Tow-income food deficit countries), and hunger-responsive food aid will be difficult to assess, certainly more difficult than whether forecasts of actual flows are accurate or not. Both needs estimates require, in turn, projections of the demographic composition and amounts of food needed to eliminate the caloric deficiency of hungry people in the affected countries a subject of widely varying estimates and observations of uneven reliability, depending on the criteria adopted. FOOD COMMODITY TRADE AND AID FORECASTING - In introductory remarks, Owen Cylke, Assistant Administrator for FVA, poir~ted out that food aid is one of the most widely shared instruments of international cooperation among executive branch agencies of the U.S. government. With the impending change in administration it is timely to Took at future requirements. The concern is for efficient food systems and food security in the Third World, not just for food aid. The following is a synopsis of the papers contributed by the six principal organizations. The synopsis is derived from the synthesis evaluation paper presented to the meeting by Bruno Larue of the University of Guelph, and from the ensuing discussions of participants. Dr. Larue pointed out that it is always easier to judge forecasting models than to create them. Forecasts are only as good as the available data, and, because they require many compromises between what should ideally be included and what is practical to include, they are inherently limited. There are also many uncertainties, arising from such causes as the weather for example, Hurricane Gilbert in Jamaica and floods in Banglaclesh and unexpected market demands arising from political changes, such as the coup d'etat in Haiti and other so-called exogenous factors. Most of the "models" reviewed at the workshop are designed to predict future prices and tonnages in the export-import trade market. Except for the IFPR! process employed by Hannan Ezekiel, they are not explicitly designed to predict food aid requirements. Three models, the Food and Agricultural Policy Research Institute (FAPRI) model developed by Iowa State University and the University of Missouri, the Static World Policy Simulation Modeling (SWOPSIM) of the U.S. Department of Agriculture, and the World Bank model, predict export-import trends in commodities, mainly cereal grains, not food aid. The FAPR! model is based primarily on predictions of trade to and from the United States, while the Bank and others treat U.S. policies in detail in their models because of their importance to world food trade and aid. The IFPRI effort, which does project food aid needs, is based on the aggregate impor- tation of food by developing countries, translated into cereal equivalents.* The import gap is then calculated as the difference between total domestic use of the major food crops and production. Food and feed use projections are obtained from estimates of per capita GNP, population growth, and income elasticities of demand. Commercial import projections are subtracted from the import gap. These are based on growth rates of per capita income *Using cereal equivalents for oilseeds, roots, and tubers assumes that foods are equally interchangeable; since the bulk of food aid is in the form of cereals, there is no great error introduced by this assumption. 12

applied to base period commercial imports as a measure of the ability to pay for food. The balance is the projected food aid needed for food price stabilization. Some issues arise from these assumptions: there are problems in looking only at the demand side, since if you know how much food aid you can expect to receive, it can affect your demand for commercial imports of food and other goods. Theoretically, food aid demand should not be limiting, since it is a free good. On the other hand, modeling the supply side is difficult, since you are modeling political and economic behavior. For one thing, it would have to be based on assumed relations specifying political propensity to tie food exports to economic aid or new loans in order to reduce undesirably large domestic stocks. However, given the data available over the short period in developing countries, the existing LDC demand side at projected prices and availabilities remains the normal basis for estimates. This "need", as distinct from food aid policy implementation, is affected by many exogenous factors, and it is very difficult to predict or control. OUTPUT OF THE MODELS All models show better economic situations for developing countries in the high-growth scenario; Tow growth has a negative effect. The FAPRI and World Bank models are surpris- ingly close on production forecasts. FAPRI predicts a substantial increase in commodity prices, whereas the World Bank and SWOPSIM models predict a decrease in prices in real terms. Net trade prediction figures for the period 1990-1995 show 10-15 percent growth in exports from the producing countries, and 1~17 percent growth in LDC imports. Figures for the Centrally Planned Economies (Figures 1 and 2) show a swing from net exporters to net importers. Corn trade shows an increase in exports by 30-40 percent, with imports by developing countries increasing 36 percent in response to rising incomes and demand for animal products. Growth in soybean trade will be much smaller than wheat or corn trade. The import gap of demand exceeding domestic supply should get wider by 2000. An interesting implication is that high growth, stimulated through resource transfer to developing countries from industrial countries, has a greater economic impact than freer trade resulting from removal of restrictions on commodity trade. Comparison of the IFPR! and FAPRI models yields different results, but not sur- prisingly since they have different assumptions and countries included. The import gap estimates of these two models (all figures in million metric tonnes EMMA of cereal equiva- lents) are: High Income E. Asia Asia (excluding China & India) Latin America 21 (excluding Argentina) FAPRI 16 30 IFPRI 19.2 22.6 13.2 The proportion of food aid needed to close the import gap is anticipated to become larger as import needs rise faster than foreign exchange earnings. One reason for this is the expectation that the growth impact of the "green revolution" package of technologies and support to farmers, a factor in high food production growth in the 1970s and 1980s, is largely over and will not continue into the next decade. High-yielding varieties produced by recent research are more likely to lead to greater stability in crop yield, rather than much higher yields. There is, however, a prospect for greater gains from a marketing and distribution revolution. It is increasingly possible to deliver food across national and international 13

Wheat 100 80 60 40 20 o 70 oh O 60 11 o oh J - 50 40 30 20 10 o 10 5 o -5 -10 -15 ~ . . . . .~;~;~;~;~;~;~;. ,, 'of, :~:~:~:~:~:~:~:e , ..... ~ ' ': ', :~:~:~:~:~:~:~: it 2:,.:.:.:~.~ ·:~:~:~:~:~:~:. / 'aim .".'.2.2. .. 2 ',.2. :2:::::: ~ ::. .~.:.; ;.. :::::::-: :~:~:~:~:~:~:: ·:-:- - - . .: :-:-:-:-:-:-:-: :::::::: ::: ·:-:-:- -::: :-:-:-:-:-:-:-: ...2............ :~:~:~:~:~:~:. ·:-:-:- - . .: ........ ·:-:-:-:-:-:-:- :-:-:-:-:-:-:-; ·:-:-:-:-:-:-: :::: :-:-:- .~...:... .~ ... ['2'2-L ; 20 FAPRI World Bank 1 5 o 10 _ If,,, .-.-.-.-...... ............ ·:-:-:-:-:-:-: :-:-:-:-:-:-:- , .~ ~ _ :.:-.:.:.:.:-.: ....... ·:-:-:-:-:-:-: :-:-:-:-:-:-:- _ - -5 _ Coarse Gralns .-...-.-...... :-:- - - .:: .............. :::.::: :-:-:-:-::::: I:::::: :::: :-:-: :~:~:~:~::: ·:-:- - .:: ::: :-:-:-: :-:-:-::-:-:- ::: :. .: Soymeal ....... '.2.~.:': ·:-:-:-:-:-::: :::: :-.-:- :-:-:- -: :-: ·:-:-:-:-:-::: :-:-:-:-:-:-:- ::::::::-: ·:-:-:-:-:-::: :~:~:~:~:~:~:~: :::::: :-:-:- ·:-:- - -: :-: :-:-:-:-:-:-::: ::::::::::::::: ·:-:-:-:-:-:-:- :~:~:~:~:~: :~: ~- . ,,,, ,,~ i,., Hi' . / . . ,." ,, -10 70 :-:-:-:-:-:-:- :::--::::: a:-:-:-:-: :-:- - - .:: ·:-:-:-:-:-:-- .............. :-:-:-:-:-:-:- ............... ,.:~:~:~:~:~:~:. :~:~:~:~:~:~:~: , .. - . - ........ :~:~:~:~:~:::~. :::::----::.:: ·:-:-:-:-:-:-:- :~:~:~:~:~:~:~: . - -. -.- ~ :-:-:-:-:-:-: :1 LL (D 50 I z LL to 40 30 20 10 o 25 20 15 10 5 -10 -15 -20 L :~:~:~:~:~:~: ·:-:-:-:-:-:-: :-:-:-:-:-:-:- :::: :-:-: ·:-:-:-:-::::: :-:-:-:-:-:-:- .:~:~:~:~:~:~: t:::::::::::::: l:-:-:-:-:-:-:- I-:-:-:-:-:-:-: I:~:~:~:~:~:~:. [::::::::: k:-:-:-:-:-:-: ' ~ . c.:~:~:.:.:. .: I ~. [:~:~:~:~:~:~::: c ....... t.2 2. c~ ·: 0r ~ -5 L ~ ~ , .. . 1 1 t.:~:~:.:. ;. ;~ t: :~:~:~:~:~:~:~ ~...2.. ~ h:-:- - - .: :-4 :~:~:~:.:.:.. :: k:~:~:~:~:~:~:~:d k::::::: :-:-; :~:~:~:.:. . :~:: :~:~:~:~:~:~:::d ;~:~:~:~:~:~:~:~: ::::::::d :~:~:~:~:~:. .:~:~:.:. . :~:3 :~:~:~:~:~:~:::4 E | ~ FAPRI 1 ~ | ~ World Bank e 1 ___L; 1-----.-.-.-.-.1 t::::::::::::::] I:-:-:-:-:-:---] 1-:-:-:-:-:-:::] 1:-:-:-:-:-:-:-3 1::::::::::::::3 1-:-:-:-:-:-:-:4 I:-:-:-:-:-:-:-3 I: ::-. :: : : : ::- :i 7 , 1 t.:;::::~:~:~: i:~:~:.:: a:~: F- - i:-:- - - :-:-: [-:-:-:-:-:-.:: t::: :-:-:-:- i:~:~:.:: :~:~: 1-:-:-:-:-: ::: t:~:~:~:~:~:~:~. 1::::::::::: :: 1-:-:-:. .: :-, p-:-:-:-:-:-::: .~:-:-:-:- L EXPORTS IMPORTS IMPORTS EXPORTS IMPORTS IMPORTS DCS LDCs CPEs DCs LDCs CPEs FIGURE 1 Projected 1995 Net Trade. SOURCE: Adapted from Larue and Meilke, 1988. FIGURE 2 Percentage Change in Net Trade 1990 to 1995. SOURCE: Adapted from Meilke and Larue, 1988. 14

regions to feed people. More efficient means of marketing food, therefore, would have a positive impact on food needs, even without increases in the rates of production growth to keep up with population and economic growth. PRESENTATION OF THE IFPRI FIGURES Hannan Ezekiel, International Food Policy Research Institute Dr. Ezekiel indicated that the IFPRI method estimates demand-based food aid needs. It does not deal with the gap between need and the ability to pay. It assumes that prices are kept reasonable through food aid, and do not cause a reduction of demand on the part of the poor. One methodological problem is to distinguish between demand and nutritional need; adequate data for estimating nutritional deficiencies satisfactorily are not available. Another problem—access of poor to the food is beyond the scope of the present IFPRI study. Increasing the incomes of the poor (or the average incomes of the country as a whole) to the levels at which the poor are able to buy the food they require is a gigantic problem, one far beyond the solution of food aid by itself. Food import gaps cannot always be filled commercially because of countries' foreign exchange shortages; food aid fills this unfilled gap. There is no such thing as food aid "needs" per se, however. "Needs" derive from assumptions about stable per capita food availability and commercial import capacity. Ultimately they are determined by donor policies, which result in suitable criteria being framed for the purpose. Seeing food aid as a tool for development helps to set these criteria properly. In addition, there is an important need for equity which, because of the nature of food, must be estimated independently, and addressing it by, say, feeding the urban poor, must be done without leading to neglect of agriculture. If used properly, food aid should increase commercial demand faster because it supports good policies, thereby increasing incomes. However, if it supports bad policies, decreasing incomes may lead to more need for food aid to avoid starvation. Thus future food aid needs are in part a product of its effectiveness in earlier periods. In every case, food aid must be seen in a specific context, or else it is only a mechanical response and is not likely to be used as an important development resource. Specific IFPRI projections for market stabilizing program food aid are given in tables in the paper presented (Appendix A; Ezekiel, 1988b.) These IFPRI estimates are based on trends. Income trends combined with a population growth factor are used to estimate demand, which, in turn, is the basis for import gap estimates. In principle, food aid should fluctuate in response to production in recipient countries, though the volume of food aid exports may lag behind production changes, based on food stock adjustments taking place first (Johnson, 1979~. Upper and lower limits for need can be projected around a central trend, assuming average variability. Data are aggregated over a large number of countries from FAO supply utilization accounts, without modification. The period 1972-1983 is used to estimate trends because using long base period (the 23 years from 1961-1983) to make projections misses significant recent alterations in the trends in some countries, both positive and negative. The future is likely to be more like the immediate past 10-12 years than like the last 23. 7 PRESENTATION OF FAO FIGURES Richard Perkins, Commodities and Made Division Dr. Perkins pointed out that food aid is already included in historic trade data, and 15

therefore appears in the estimates of countries' future import needs, which assume that supplies of donated food will continue. Using country import statements may give rise to anomalies, he noted. Factors that cause this may be upward revisions of production estimates without revisions in income or population estimates. This could lead, for example, to Ethiopia being seen as requiring no food aid, and/or, as in the Sudan, refugees not being counted since they are absorbed into the population without distinction for their special food aid needs. Needs estimates based on national averages do not account for distribution inequities and are therefore very inaccurate. The FAO gathers a great deal of data, and much of it is used in studies of future world food trends. It has not, however, made projections for food aid requirements recently. The most recent FAO food aid requirements estimates are based on production, demand, and trade Projections for the 1985-1990 neric~c] mind m?~.tlL`~mn.t.ic~.~1 meanly similar to ~ ~ r ~ O _ ~ A ~ ~~ _~~ V _ ~ ~ T ~ ~ ~ ~ / ~ ~ _ ~ ~ ~ · . ~ ~ ~ . . ~ . ~ the Ub~A/~UL Strain, Unseen, and Livestock' model and plann~ng-type models with a fairly optimistic production and trade scenario. For the 1985 estimates of food aid requirements, a revised approach was adopted to include separate components reflecting balance of payments support and project food aid (for example, for Food for Work and food security reserves) plus an emergency component based on shortfalls in cereal production. The total requirement was estimated at 20.2 MMT of cereals. The projection models are usually neutral with respect to income distribution and food price policies. They are also neutral about targeting populations or leakage limits. Another approach adopted by FAO has been to base food aid needs on nutritional considerations—to increase food intake to "acceptable levels." The guidelines for food aid adopted by FAO in 1981 give priority to low income food deficit countries. Out of 12 MMT of food aid in 1987, 10 MMT went to low income food deficit countries, an allocation trend to be welcomed. The definition of a nutritionally "minimal level of intake" was established by the Fifth World Food Survey of 1985, based on the distribution of food energy intake measured through recent household consumption surveys. Very conservative low limits were set of 1.2-1.4 BMR units per capita (corresponding to 145~1610 KCal/day in India and 1550-1720 KCal/day in Egypt) as minimum maintenance requirements for energy, with no allowance for movement or work energy for adults. It then calculated the minimum additional amounts of energy (as cereal equivalents) required to bring the average daily intake up to these minimum levels for the people in developing countries which at the time were below this level. These estimates, without allowing for leakage or targeting difficulties, amount to 8-14 MMT respectively for reaching the 1.2-1.4 BMR limits. The energy distribution curves used in the survey in 1979-1981 implicitly included in their established base the approximately 9 MMT of food aid actually provided during that period. Therefore the 8-14 MMT plus any further increase required for the portion of food aid that "leaks" (that is, fails to reach the targeted malnourished population) should be added to the existing food aid levels in any projection of extra needs of that period. As a basis for extrapolating such figures, by the year 2000 FAO projections for the absolute number of malnourished people in LDCs indicate an increase over 1979-1981 figures of roughly 10 percent. (While the fraction of the worId's population receiving lower than a minimum calorie need will drop, the absolute numbers are not expected to do so.) A further 10 percent allowance for leakage would give a total estimate of 20.6-27.8 MMT additional food required.* *The calculations are: 9 MMT + ~ MMT + 10% of 17 = 18.7, + 10% of 18.7 = 20.6 MMT; 9 MMT + 14 MMT + 10% of 23 = 25.3, + 10% of 25.3 = 27.8 MMT. 16

Such figures are based, however, on the assumptions that 1) all countries would be able to purchase their net import requirements commercially by 2000; and 2) real world market prices equal to those in 1983-1985 would prevail. It may be noted, however, that a rather large decline is projected in the agricultural trade surplus of the developing countries as a whole, which could undermine such assumptions. PRESENTATION OF WORLD BANK FIGURES Ronald Duncan, International Commodity Markets Division Dr. Duncan indicated that World Bank figures are not for food aid, but projections of _ 1 ~ · te · ~ ~ ~ ~ ~ . ~ ~ ~ ~ · . · ~ . lV-U~IVll ~t 6~O mu my ~~. iiie shorts DanK proJectlons see snort-run price increases through 1989, as the last year of a 3-year cycle, then falling prices in 1990-1991. In the Tong run, to 2000, they see prices in real terms declining as a continuation of the post-Second World War trend. Growth in LDCs' per capita GNP will increase food demand. Their ability to meet this demand will depend on policy responses within the LDCs to restraints on production, and on import policies. Areas of particular concern for food imports are sub-Saharan Africa, Bangladesh, and the small southeast Asian countries (excluding China and India). The World Bank focuses on LDCs in its projections. Other world bodies are taking care of projections for the industrial countries. Major countries within the World Bank's area of interest, however, and which the World Bank is assisting with policy reforms, include countries other than those included above, such as India, Indonesia, Thailand and Argentina, and also China. Macroeconomic assumptions indicate a 1989-1990 slowdown in industrial country eco- nomic growth, and then a rather optimistic period 1990-200~during which econorn~c growth in OECD countries should be greater than 3 percent, with real interest rates falling, boosting investments compared with 197~1985 a period when investments fell as real interest rates rose. LDCs' growth averages are predicted to be 4.5 percent, but this includes China, with 6-7 percent, and India with 4.5 percent, as well as African and Latin American growth, which is much lower. The Centrally Planned Economies (excluding China) are projected at a somewhat pessimistic 2.3 percent. If the CPEs grow faster, the grain trade market could be much larger. The outlook for grains trade is the subject of some dispute within the World Bank. When growth in grain yields slows, some tend to assume that more investment in irrigation will lead to higher rice yields. However, with increasing incomes in LDCs, there will be a further shift in relative demand toward coarse grains and from rice to wheat (Japan's market for grains in 1955 was 95 percent rice, but fell to less than 70 percent in 1985~. Bennet's Law suggests the inevitability of the growth in demand for coarse grains for feeding livestock. There should thus be less alarm at falling rice consumption. Africa is different, showing increased rice consumption with rising incomes. There will therefore be a projected annual import growth in wheat and coarse grains to both LDCs and CPEs of 5-8 percent. The actual rate of growth will, it is believed, depend on the rate of implementation of price/policy reform. The emphasis on the economic environment is justified because, as in the case of demand for rice. food consumption r~fl~rt~ ~~`r~rn1 f~f~t.~r~ for inet~n'*= a'7he;~;~ .. ~~;_A~ HAT ~~ ~ ~ ~ ~ ~ ~ ~ ~ At. ~ ~ ~ _ ~ ' ~ ' _ ~ ~ A I _ _ ~ _ ~ ~ _ I ~ 1 _ _ _ 1 ~ __¢ ~ ~ ^_~ w_^ ~ ^~^ ^~—US—~ ~ ~—~1~ ~1 ~11 PI 1~. ~1~1_ _ ~ · 1 · · 1 1 . ~ · ~ · · . ~ ~ nag ~r Is possible to remove suns~ct~es without economic chaos was shown in one province in China where the price was tripled, and consumers were given a lump sum to assist their adjustment to the new prices. There was no upheaval; a 25 percent decline in rice purchases reflected mainly an end to feeding cheap rice to chickens. 17

There is a major problem in using aggregated commodity trade figures because of the huge influence of the large countries such as India and China. PRESENTATION OF THE FAPRI FIGURES W. Meyers, Center for Agriculture and Rural Development, Iowa State University Dr. Meyers observed that the meeting was an important step in bringing together the commodity trade modelers with those making food aid projections to compare the output of the models and understand why differences might occur. Key assumptions used in the FAPRI model are macroeconomic, based on Wharton/Chase econometrics and Project Link (the University of Pennsylvania macroeconomic modeling project) data rather than World Bank projections. Its forecast is for an average GNP growth rate of less than 4.5 percent. The baseline used in the paper circulated to participants was that presented at the Buenos Aires XX International Conference of Agricultural Economists in August 1988, but prepared before the drought in the United States, and does not include changes in U.S. policies in response to drought. The model's major conclusions were that U.S. acreage planted, as opposed to idled through set-aside programs, will increase towards the minimum limit of 40 million acres conservation reserve mandated in the 1985 Food Security Act, leaving the United States with reduced flexibility to respond to future stresses. The effect of the drought on prices has driven up estimates for 1988-1989, but these should return to the trend as Tong as policies remain the same. A slight increase in prices for 1986-1996 is foreseen, returning to resume the long-term decline thereafter (Figure 3~. Cereal stocks will also be drawn down to the pre-1985-1986 level by the drought, and there will be much less criticism of the reserve policy than there was in 1987 (Figure 4~. FAPRT trade trends project growth of net imports by LDCs, less by the Centrally Planned Economies (CPEs) and industrial countries. There will be greater pressure on foreign exchange in developing countries. This will reflect factors such as the growing impact of debt servicing. Real income growth in Africa (ranging from 0-3 percent per year) is expected to result in a widening trade gap in wheat and coarse grains (Figures 5 and 6~. A gap requiring ~ ~ ~ v ~ cat imports is broadly typical of all developing country areas. Even though Asian economic growth will stay at about a constant 5 percent per year, well above African growth, the higher African growth in its import gap probably reflects the Tow base starting point for African predictions and higher marginal propensity to import food as incomes rise. Though the improved economic trends are encouraging, in many countries growth in production cannot keep pace with growth in demand from population and income increases. While countries in east Asia with high growth can cover their increased grain imports using foreign exchange generated by exports, many other LDCs have heavy debt burdens and foreign exchange constraints that will limit their ability to increase imports substantially. Importers' costs will also depend heavily on whether the local currency is likely to appreciate or depreciate with respect to the U.S. dollar, and it is the same countries with foreign exchange shortages whose currencies are likely to depreciate, causing the costs of imported commodities to rise. The region under the greatest pressure, in terms of potential reductions in per capita consumption of commodities, is Africa, followed by the Middle East and Latin America, excluding Argentina. The Asian region is likely to do better, because production growth is 18

500 450 400 O 350 - ire To ~ a. \ / \ / \ Soybean 1 \ / \ -1 \ I ' 300 / l\\ 250 200 150 100 I \"\ \ ~ _ ~ | ~ Wheat \ ''` 1/: \\~~~~\ '` // ~ ~ ~ Corn g /\\ / 'A \ _ ~x _' 50 71/72 76/77 81/82 86/87 91/92 YEAR FIGURE 3 Major Cereals: U.S. Gulf Port Real Prices FAPRI Model 0.29 0.23 O 0.21 0.19 0.17 0.15 0.13 0.11 1~ j L\ . _\\ ~ J i ~ I ~ ~ ~ / I I HI I \ / ~ I / t1 ~ ~ ma/ 1 1 I/ 1 71/72 76/77 81/82 86/87 YEAR FIGURE 4 World Cereal Levels: Stocks-to-Use Ratio FAPRI Model 19 96/97 / The at / / \\ / 1 \ \ I \ \ I \ 1 \ 1 \ 1 \ 11 \ 1 l l Feed Grains - 1 1 91/91 96/97

70 A o E IL z o 65 60 55 50 40 35 - - Domestic Use - Production _ _— 76/77 78/79 80/81 82/83 84/85 86/87 88/89 90/91 92/93 94/95 96/97 YEAR FIGURE 5 Africa and Middle East Wheat 65 60 z o E A o FIGURE 6 Africa and Middle East Coarse Grains 55 50 45 40 35 30 / x\' - Domestic Use - - Production _— _ _ _ _ - _ _ - 251 1 1 1 1 1 1 1 1 1 1 76/77 78/79 80/81 82/83 84/85 86/87 88/89 90/91 92/93 94/95 96/97 YEAR expected to keep pace with population growth rate. Although imports are increasing in all developing regions, they are insufficient to offset the slower rates of growth in production. The most desirable solution to these problems would be to increase productivity. Another important measure would be to resolve the debt problem. Food assistance programs are 20

a short-term measure, not a solution to static or declining per capita consumption levels. Targeted export subsidies can have a similar effect, provided they are based on need rather than on geopolitical or strategy considerations. There needs to be a shift in priorities. Poorer Third World countries must have more assistance to develop their entire infrastructure (education, health, transport, marketing, and distribution systems) as well as agriculture and industry if the growing import gap projected is to be avoided. Only through Tong-term sustainable growth will food security problems be solved. When the Third World countries can produce and sell products for which they have a comparative advantage, they will command the resources necessary to feed their populations. PRESENTATION OF USDA FIGURES R. Nightingale and R. TrostIe, Economic Research Service ~ O Dr. Nightingale indicated that IFPRI and ERS short-term food aid projections use similar data sources and assumptions, but that food need in the ERS analysis is based on maintenance of per capita consumption (the status quo) and on nutrition requirements rather than income-based economic demand. However, the ERS method (FNA) has not been extended to longer periods because the analyses are directed to Congressional and National Security Council requirements for frequent short-term reporting. Short-range projections have been undertaken since 1975, and the basic methodology has been stable since 1981. Since then, methods have been introduced to assess cereal needs to maintain stocks or to meet explicit national food stock objectives. Methods have also been developed to project debt repayment, a factor entering into estimation of commercial import capacity. The FNA analysis is at the national aggregate level and does not address internal distribution issues. The Economic Research Service uses a number of models to project commodity trade. These include the Static World Policy Simulation Model (SWOPSIM) and the domestically focused Food and Agriculture Policy Simulation Mode} (FAPSIM). Both are comparative static equilibrium rather than dynamic projection models. Dr. Jostle summarized the model assumptions and projections: population growth will slow, incomes will rise; debt servicing will constrain consumption, but Tower agricultural prices will be an offsetting stimulus to consumption. LDC agricultural production averaged a 2.9-percent compound annual growth rate between 1950 and 1987, but is expected to decline in the 1990s to 2.4 percent. Countries approaching 100 percent self-sufficiency in food will slow production rather than export food at Tow prices. Debt servicing will restrain the ability to import food commercially. There are no radical technical production breakthroughs on the shelf that will raise yields sharply during the next decade. An exception may be animal growth hormones, coupled with improvement in livestock management, but this will not cause a large effect in overall estimates. LDC area planted to crops has expanded a modest 0.7 percent per year. Much of the growth has been on poor land. Area expansion will slow unless prices increase significantly. Grain prices are projected to follow a downward trend of about 2 percent per year, roughly offsetting increases in productivity. Grain and oilseed output has shown a 2.7-percent growth, 1.9 percent due to yield increase, 0.8 to area planted. Slower growth is forecast, with cereal yield increases declining to 1.7 percent per year. Consumption is likely to increase faster, and self-sufficiency will decline. LDC net cereal imports have been increasing by 8 percent per year, but will probably slow to 5-6 percent per year. The net agricultural trade surplus of LDCs fell from $15 billion in the rn~-1970s to zero in recent years. Some countries 21

TABLE 2 IIASA/BLS Projections of the Incidence of Hunger Year 1990 2000 As Percentage of Population in LDCe 17 In Millions of People: In All LDCs In Middle-Income Countries In Low- to Middle-Income Countries In Low-Income Countries 11 470 400 30 60 310 are doing well; others very badly. Since 1975, food aid shipments have risen 3.7 percent per year. Foreign exchange will be a major factor affecting future growth of commercial imports. In 69 countries, food imports use 10 percent or more of available foreign exchange. PRESENTATION OF IIASA FIGURES K. Frohberg, OECD, Paris Dr. Frohberg indicated that the IIASA mode! is a general equilibrium model, while the others are partial equilibrium models. The IIASA Basic Linked System (BES) model looks at income and demand aspects, explicitly including what actually happens. It is based on the system of 20 national models and 14 regional models. It is deterministic, and while one must resist the temptation to use it as a forecasting tool, it is used to simulate the impact of policies over a 20-year perspective, including 9 agricultural commodities, and 1 nonagricultural commodity. The model generates the following information: 1) at the international level, prices, volumes traded, produced, consumed; 2) at the national level, population, production, human consumption, feed use, intermediate consumption, net trade volumes, stocks, prices for producers and consumers, and input use: labor, capital, land, fertilizer. Table 2 gives the projections for the incidence of hunger 1990-2000. Table 3 shows the share of global demand among the Industrial Market Economies (IME), developing countries (LDCs), China, and the Countries for Mutual Economic As- sistance (CMEA). Table 4 gives the results of a reference run estimating the numbers of hungry people in selected countries for which the BES models have a common structure. Tables 5 and 6 show the projections from the BES models of different scenarios of the impact on hunger. 22

TABLE 3 Share of Global Demand in 1980 and 2000 (percent) IME LD Ce China CMEA 1980 2000 1980 2000 1980 2000 1980 2000 Wheat 23 21 28 38 12 11 33 27 Rice 6 5 56 66 36 28 1 1 Coarse Grain 42 44 20 27 10 9 24 18 Beef, Mutton, Lamb 49 43 29 36 6 6 17 15 Dairy Products 44 39 23 32 1 2 30 26 Other Animalsa 41 38 20 28 19 22 50 9 Protein Food 50 48 15 21 12 13 9 6 Other Foodb 22 18 42 52 16 14 15 11 Fibres 35 26 30 31 14 18 27 30 bPork, poultry, eggs, fish. Oils and fats, sugar, vegetables, fruits, nonalcoholic beverages. TABLE 4 Numbers of People Hungry (in Millions), Reference Run 1980 2000 Argentina Brazil Mexico Egypt Kenya Nigeria India Indonesia Pakistan Thailand Turkey 1 12 3 3 3 1 6 25 o 219 156 21 0 9 6 8 1 Note: The low year-2000 figures for Egypt and Indonesia derive from a continuation of the trend of high percent per capita per annum economic growth. (See the text of Frohberg's paper for the explanation.) 23

TABLE 5 Impact of Some Scenarios on Hunger (Changes in Percent from Reference Run) 50 MMT Wheat as "Manna" (Gift) -3.0 Reduced Meat Consumption in IMEs -1.4 " Green" Scenario of Reduced Use of Fertilizer and Pesticides in OECD Countries Drought in the North Drought in the South +9.0 +8.6 +8.2 TABLE 6 Alternative Aid Scenarios for All LDCs (Millions of People Hungry, Percentage Change) Reference Run A-Capa A-Bopb 1990 2000 1990 2000 1990 2000 All LDCs 470 400 -13 -32 -24 -32 Poorest LDCs 345 305 -15 -40 -29 -40 b Aid added to capital investment. Aid support of balance of payments. MACROECONOMIC CONTEXT COMMENTARY Lawrence Klein, Economics Department, University of Pennsylvania Dr. Klein noted that the Project Link macroeconomic model, which the University of Pennsylvania Department of Economics has developed to project changes in the global economic picture as a result of different variable assumptions, uses the same assumptions as the IlASA BES model. He observed that it is difficult to think of assumptions for the next 10 years. Only one year ago, Tong-range projections of food aid needs would have been low. But was the drought of 1988 a one-year blip, like the stock market crash of October 1987? Will the growth in world economies continue on the same trend line as before? It is possible to be horrendously wrong in making these important predictions. Oil prices were high, are now low, and will again be high. The stock market is back to its pre-crash level in some countries (not the U.S.~. Droughts are a major feature of food forecasting and led to depletion of grain stocks in 1972-3. India suffered a major drought in 1987, the U.S. in 1988, and there is the possibility of more frequent climatic impact events. He made a suggestion with regard to methodology: employ stochastic simulations using better input from the climatologists; this would be better than a purely random draw. Stochastic simulation allows realism based on past errors, by running a large number of variations which can be checked against reality. Weather and climate may be particularly important variables to check. (Figures 7 and 8 give examples of the graphic output of stochastic model random runs of models of carbon dioxide production and atmospheric levels, from the 1983 NRC report on Climatic Change.) Regarding the assumptions of growth, he thought that World Bank projections of 24

A) 8 - ~4 So O CM rat O O Cal fir LO N O CM (quo} lo SU°!ll!q) NOISSIW~ NOSHVO O0 \~ co ~ a, ~ v o X o o a) ~ _ ~ a) ~ O . JO ~ a' ° . v, ~ a' a) cIS ~ a) =._ ~ _ ·— V ~ o ~ o ~ _C ° ° V 04 ,=._ o Do O be o,= m ~ `= ~ ~ O ~ ~ O of ~ ~ . O ~ ~ ~ O 111 ~ ° ~ o - _ O .0 L~ ~ ~ ~ X ao V . V ,~ U2 · O COM _ I I I ~ L~ ~ O O O U~ O O U O ~) CN C\l CM ~ co c~ (Wdd) NOIlVUlNEONOO OIH9HdSOWlV 25 , ~ . CO ~ o ~ ~ V ~ ~ C5) a) . _ ~ ° ~ ~ V ° ~ V° ~ ._ ~ ~ ~ ~ o V.= s.= o ~ ~ ~ V~; ._ ~ ~l C.) tn V2 ~ o o =) ~ V, o ~ ~ V ·= XV_ o ~ ~ . - - 5 ~ ~ o i_ o 04 ~ ~= o~

around 3 percent were probably good; LDC growth rates of 4.5 percent are on the high side (including China: global LDC figures are raised or lowered by +0.5 percent and LDC Asia figures by +1.0 percent if China is included or not included). Estimates of CPE's growth of 2.3-2.8 percent are on the Tow side. Recent discussions with Soviet economists elicited their projections of around 4 percent, which may be on the high side. However, if China can achieve growth rates of 6-8 percent per year, the Soviet Union through its own reforms may come closer to that level of growth than 2.8. Inflation estimates are very low: in 1987 OECD countries' rates of inflation averaged 5.5 percent; LDC inflation rates were much higher, and 4 percent projected for 1990-2000 may be Tow. Regarding the FAPRI projections, if one were to measure the impact of the Reagan Administration on the economy for the period 1981-1988, performance would be much worse than for the period from 1982-1988 it depends where you start. He suggested testing whether different time periods might account for different values in the projections. The FAPR! 1981-1986 figures show an annual negative growth rate of-0.23 percent for LDCs, because of the impact of oil prices on African countries and oil-importing countries of the Middle East 1981 is a misleading year from which to start projections. Regarding LDC growth projections, Asia's average is 5-6 percent (6 with China in or 5 with China out), and Hong Kong, Singapore, Taiwan, and Korea are all in the 7-10 percent range; with the other Asian countries, growth is lower. Projections for sub-Saharan Africa are very depressing, showing negative per capita growth because population is growing faster than GDP. In the period 1945-1970s the OECD countries' economies grew at over 4 percent. Bade in the 1960s often grew at over 10 percent. The question is whether the 1990s will be like the 1970s, or will recovery in growth lead to a recurrence of the 1960s' conditions, with a high growth scenario and higher inflation? He indicated that he felt that the conditions responsible for slowing growth in the 1970s were no longer operating, and therefore higher growth rates might be more likely in the 1990s. In the high growth scenario, interest rates may be higher than World Bank projections, because high growth and inflation drive interest rates up. Interest rates depend on savings and rebuilding capital stock, such as re-equipping industry. However, the impact of modern electronics and other new technologies in industry is not major capital re-equipment, but largely tearing down old industries. LDCs must not focus only on primary agricultural products; they must diversify. If there is a productive tendency to increase the value added to primary products, there may not be such a decline in their commodity prices, since less will be available for export. The gradual increase in prices in 1987-88 was partially due to the weakness of the dollar, and there may be less overproduction of commodities than anticipated. Dr. Kates observed that in recent discussions with Soviet scientists, they felt that the Soviet Union may be entering a rising phase of the Kondratieff* cycle, and the United States may have bottomed-out in its falling phase. He wondered whether there might be some hidden surprises trend extrapolation should be sensitive to surprises, since their likelihood becomes quite large over a decade. There may be a recurrence of the energy crisis in the l990s. Dr. Klein mentioned several surprise possibilities: debt renunciation; greater harvest failures; nuclear winter (the long-term lowering of global temperature by 3-5 degrees); the greenhouse effect- all might occur in "big change" scenarios. His own positive surprise *The Kondratieff cycle, named for Russian economist Nikolai D. Kondratieff, is a business cycle with a periodicity of between 50 and 60 years. 26

would be disarmament and development—the transfer of capital from defense to develop- ment. The UN, many governments, and many academics, were enthusiastic about it, but the U.S. government, although supporting both disarmament and development, regards them as separate policies, and is against the linkage of the two. Regarding the elect of debt renunciation, the impact on money markets should not show up directly in food. It would show up in the cost of money, maybe adding 2-3 percentage points to the interest rates, and driving the U.S. prime rate up to 10 percent for the Ton ger term. U.S. trade and OECD exports to LDCs have suffered because of debt. U.S. exports to LDCs were 40 percent of total exports, but now are only 30 percent. U.S. and LDCs have gained less from the debt than other countries. Many LDCs showed negative growth in food imports and consumption. Those with minerals were not as badly off as those dependent on cash crops, such as cocoa and coffee, the prices for which are now Tow. If disarmament could make available a $50 billion development fund, LDC growth rates could be increased by 0.5 annual percentage points, and even more in Africa and the Middle East. Regarding the potential impact of the green revolution, discussions indicated that in the short run (the coming decade) biotechnology might be more likely to reduce crop losses than increase yields, since yields are multigenic (determined by a number of genes) while resistance to disease can be influenced by single genes. Dr. Duncan reported that the recent Buenos Aires XX International Agricultural Economics Conference discussions saw no big technology improvement until after 2000. The green revolution was not a surprise: it was foreseen and achieved operating within known technologies. However, the impact of Chinese cultural and economic reforms on the supply side was not foreseen. An NAS panel visiting China in 1979 estimated annual agricultural growth at 3-4 percent it actually shot up to 10 percent. Dr. Ezekiel felt that risk had been increased by greater variability in climatic and economic conditions, and we have to learn how to manage variability better, including financial and economic variability. Dr. Jostle thought it unlikely that the Soviet Union could emulate the major structural changes that allowed the growth of Chinese productivity in agriculture. Though there may be production surprises as prices increase, these are unlikely to be a major factor in commodity trade in the 1990s. DEMOGRAPHIC CONSIDERATIONS COMMENTARY Robert Chen, Alan Shawn Feinstein World Hunger Program Dr. Chen pointed out that UN projections of African population growth have tended to underestimate range of uncertainty significantly, even for projections a few years into the future (Figure 9~. Moreover, UN assumptions about the demographic transition in Africa are clearly suspect, since the assumed decline in fertility does not follow the actual pattern observed in most other regions of the world (Figure 10~. Indeed fertility declines have not yet begun in many African countries. Substantial variations in projections of future African population growth remain, especially beyond the year 2000 (Figure 11~. Much of this growth will take place in cities; UN figures show an African urban population of more than 400 million by the year 2000 (Figure 12~. Notably, these projections clo not allow for unpredictable factors like AIDS, or for interactions with economic, social, or technological trends. For example, migration is assumed to dwindle to zero in the future, even though the level of migration itself is likely to be a function of food needs, among other things. Thus, the assumption that population can be treated exogenously in those models could prove to be dangerous. 27

1 000 co - 9oo - O~ 800 O 700 IL C) 600 o CC 500 _ 400 _ 1955 Projections for the Year 2000 High Variant Medium Variant Low Variant I (36) (27) (22) (20) (32) Projections for the Year 1985 (16) High Variant. ' __ ~ ,.` Medium Variant Low Variant &~ (17) (12) (22) 1 1 1 ~ 1 1 1960 1965 1970 1975 ASSESSMENT YEAR 1980 1985 1990 FIGURE 9 Trends in U.N. Projections for the African Population as Assessed Since 1963 The composition of the diet in terms of the proportion of food calories consumed as animal products can have an important effect on the quantity of primary agricultural products needed. Using FAO/WHO/UNU standards for caloric needs and computing food supply from FAO data there is presently enough food in the world to feed some 6 billion people if almost all foods are consumed directly (Figure 133. However, as diets increase in the utilization of animal products, the need for primary agricultural products also increases. For a diet consisting of about 30 percent animal products flower that that currently consumed in much of the industrial worId), the need for primary agricultural products doubles, rising significantly above what is presently produced in the world. This increase due to dietary composition is much larger than the potential differences in projected demand due to different population growth projections. The World Hunger Program has also made a back-of-the-envelope estimate of the number of hungry people there will be in the world by the turn of the century. Over the past several decades, the proportion of hungry in the developing world as defined by the FAO standard has declined from about 34 percent in 1948-1950 to 17 percent in 1978-1980. However, the rate of decline appears to have levelled off in the past decade, with only about a 1 percent decline in the subsequent five-year period. Projecting this current slow rate of decline into the future, and applying it to the UN medium variant population projections, we find that the actual numbers of hungry could still rise during the next decade, even if the proportion of hungry declines. The FAO estimates of the numbers of hungry do not include China. According to recent Chinese statistics, roughly 7 percent of the population, or 70 million people, fall below the poverty line. The World Hunger Program uses half this number (35 million) in its estimates of the numbers of hungry by FAO standards. 28

_ 7 a) in ._ 53 - LL 111 ILL at: o 5 4 3 L 2 1 1 1 1 1 1 1 ,1 1 1 1 1 1 1 1 <~ coo (~ ~0 8 ~ ~^ -age W~ O Africa · Latin America · North America O East Asia - ~- ~ South Asia Europe ~ Oceania _ o Soviet Union by FIVE-YEAR PERIOD FIGURE 10 Estimated and Projected Fertility Rates, World Regions, 1950-2025 (UN Medium Variant) NUTRITIONAL CONTEXT COMMENTARY Per Pinstrup-Andersen, Food and Nutrition Policy Program, Cornell University Dr. Pinstrup-Andersen observed that nutritional status is influenced by food and health, both of which are influenced in turn by prices and incomes. Food aid must influence nutrition or health. It is not a question of total food quantity per se, nor average supply and demand per se, but rather how food prices are deterrr~ined and why incomes of populations at risk are insufficient to afford them. One cannot simply convert MMT of cereal into calories and divide by the population; nutrition is not adequately described by average calories or tons ~ . O : grain. The assumption is that: food aid need (to keep prices level) = demand - production/supply at that price level. There are two components of need: a market-demonstrated response to need, and a lack of response from those who cannot express need through the market. Program food aid can have an important impact on nutrition by lowering prices, or at least keeping them level; that is, there is an impact beyond nutrition projects. Program food aid to support subsidy schemes and project food aid to make up the rest of the deficiencies do not add up to meeting nutritional needs. If food aid is really intended to eliminate malnutrition, we cannot just dump cereals on the market, since this only meets effective demand and lowers prices to the point that producers are hurt. Combining the goals of stabilizing prices and meeting nutritional needs would be the 29

Q ~ ~ N\~< \ ._ ° ~ s ° ~ ° ~ ~ z z z m m m :~: ~ ~ ~ ~ ~ ~ t ~ ~ t t ~ ~ ~ - o CO o C~ o o CM o N o CO o CM CM o CM o O O O o C~ o cn o 00 o r~ o o C~ o U: - - ._ s~ U: U: - o ._ - :- o a) v a) ._} o ~ _ CD =~ L 1 1 1 ~ ~ ~ ·~ ° o o o o o ° ° o o o ~ o ~ o ~ - N CNJ (SUO!ll!U1) NOIlblndOd ~ 30

most effective use of food aid from a nutritional point of view. How could food aid be used to increase the incomes of the poor, and what are the final impacts on incomes and prices? One suggestion would be to direct (target) food aid to the malnourished through a system of food stamps or ration cards. Assuming 50 percent efficiency (a net addition in food consumption equal to one-half of the food aid received) the way the program would work is that if the targeted individual is projected to require an additional 300 calories a day, he or she would be given food stamps or a ration allocation of 600 calories-equivalent. The total food deficit is estimated at 15 MMT/year (current FAO figures), which at 50 percent substitution is equivalent to 30 MMT/year cereal equivalent as food aid entitlement to the poor. The poor would actually use 15 MMT and the other 15 MMT would serve to keep prices stable in the local market. The IFPRI estimate is that 37 MMT will be needed just to keep prices stable, using current commercial mechanisms of nontargeted food aid. Therefore if we are willing to entertain imaginative solutions, it would be possible to use targeted food aid as a resource and to maintain prices (or substitute for imports) and at the same time eliminate malnutrition. An additional benefit would be to provide an outlet for more food aid grain, which may be important for producers. Dr. Rogers commented that (as far as meeting nutritional needs is concerned) demand does not work, and distribution is the main issue. We need to look at actual behavior of income categories to project income/price situations to predict what would have to happen to raise poor peoples' intake. FAO figures of 1.2-1.4 BMR assuming household uniformity and distribution efficiency are simply insufficient. In a recent keynote address at the Smithsonian "Closing the Food Gap" Symposium, Prof. Nevin Scrimshaw of the Massachusetts Institute of Technology (MIT) indicated that recent MIT studies of dietary amino-acid requirements by Vernon Young et al. had indicated that there was a need for an upward revision of protein requirements. There is also need to Took at the consumption behavior for the variety of the diet and calorie:protein ratio rather than just calories, and requirements are going to be much higher than the 1,200, 1,400 or even 1,700 calories per person per day, when normal movement and work are included and "everybody knows over 2000 are needed". - rip v^_~ Dr. Ezekiel expressed delight with the way in which Dr. Andersen used the figures from the IFPR] model to project needs to eliminate malnutrition, and also that his proposed scheme would eliminate the cost of transaction. He felt that the employment/asset creation aspect of supplying the additional 50 percent would furnish an important development resource, and have a substantial multiplier effect in the economy. Dr. Duncan observed that food aid is a paternalistic and inefficient instrument for meeting needs, and financial assistance should be provided. If food aid could be monetized, it would be possible to shift all needs to the market, otherwise, the transaction costs of administering programs is too high. However, Dr. Ezekiel and Dr. Rogers pointed out that this does not change the need for movement of food, or avoid the costs of moving it, but only transfers costs to the commercial sector. Dr. Riley pointed out that most food aid is provided under Title IT of the Foreign Assistance Act loans to countries to purchase food, and since the loans are tied to purchases in the United States, the purchases mean mainly wheat, which does not meet the needs for rice, maize or sorghum in rural areas of many poorest countries. Dr. Andersen suggested that there is no reason why his suggested scheme should not be handled entirely by the private sector at both wholesale and retail levels, and the commercial sector could find triangular arrangements for matching country needs with supplies of commodities on the world market. 31

G ~ ~ G — 0 in 0 ~ I I I t ~ me, I I . 8 8 8 ' 'A 8 8 8 ° co Cat ~ ~ 'a ~ ~ Cat (Suo!ll!w) NOll~lndOd Nv9~n (Suo!ll!w) NOll~ndOd NV9Hn ~ CD - - a' 1 1 o o o o o o Cat Cat ~ (Suo!ll!w) NOll~llndOd NVOdn to Cat to Cat to to c~ o 8 ~ C~ o - - o ~ o _ o ~ ~ o i .. ,...... , . - 1 . . ... :: L O O o O O o C~ C~ ~ (SUO!II!W) NOll~lndOd NVEUn a) ._ ~n o o ~o CC ._ V rQ o ._ - P~ CO ~ ~ ~_ ~ ~_ O ~ a' . ~ ------------. ................. b~ \\\\\\\ V--2~ 2-"2.~ "--2~ CO ( W.2~ ~ rO il: ~ Z _~ io O O O O O C ~ ~ O O O O O O O O O O O O ~ ~ ~ C~ ~ (Suo!ll!~) NOll~lndOd NV8dn (Suo!ll!~) NOll~ndOd NVOU 32

z 1 0000 LL ~ 8000 a LL a 6000 G CO Oh O 4000 ~ 2000 Ad o - - _ _ Full Butt O _ 1980 1 990 Improved Diet Basic Diet World Food Supply - 2000 2010 2020 YEAR 2030 2040 2050 FIGURE 13 Projected Primary Food Supply/Demand for the World, 1980-2050. SOURCE: Kates et al., 1988, p. 31. (The "basic" diet assumes that all primary food production is humans at FAO/SHO/UNU- recommended caloric levels, whereas the "improved" and diets assume that 10$~ and gown clef faced calories re~nenti~r~l~r caroms from ~nim:~1 nr~Allrte primary foods.) consumed directly by I "full-but-healthy" , ---~--~ ~, -a A-. =.~ is are derived from TECHNICAL RESEARCH AND DEVELOPMENT CONTEXT Donald Plucknett, Senior Scientific Adviser, World Bank Dr. Plucknett observed that the announcement of the death of the green revolution was premature. It is very much in operation, and must be sustained. It is still largely based on dwarf high-yielding varieties of rice and wheat responsible for the early successes, mainly in Asia, Mexico and Brazil. Grain production continues to climb, but the rate of increase will not climb as quickly to 2 and 3 times the levels of current yields. Yields will show slow incremental gains of 2-3 percent per year based on yield stability, disease resistance and tolerance to stress. There remains a gap between yields on-farm and on the experiment station, though the gap is narrowing in some countries. The maximum yields currently possible are in the 11-14 MT per crop per ha range (Figure 14~. Anything that will give a boost to increase rice yields beyond the current experiment station 11-14 MT per ha per crop must be accomplished through increasing the biomass production of the rice crop. Wheat yields may be increased through increased yield stability, dependent on multigenic resistance factors. Substitutions of wheat for rice will be difficult in Asian countries; wheatiands are drier and more difficult to manage, so the potential for increasing wheat production is poor except through increased application of irrigation. There is more hope through genetic improvement of some other crops, such as sorghums, which could show a significant expansion in the 1990s analogous to rice and wheat in the 1970s and 1980s. Improvements in shortening life cycles of some crops could provide greater security of crops in areas with uncertain rainfall; millets are now available that mature in 7~80 days, and cowpeas mature in 60~70 days. The models agree that for the 1990s, yield trends will continue upwards, and exogenous risks could be reduced through research input. One important difference over earlier times is the unprecedented level of research collaboration around the world, with access to research 33

15 6 I 1Q 9 5 o - // / Farm Experiment Station 11-14 MT/CROP/HA ~~ —LESS—- -—~ MORE OPTIMAL CONDITIONS FIGURE 14 Yield Potential of Cereals data through the CGIAR global agricultural research system. The response time has been shortened between undertaking the research to solve specific problems and being able to implement solutions in the field. The biggest research problem is understanding the biological limits of theoretical pro- duction of the major crops, so that the limits to land productivity can be understood. The biggest animal disease problem is malnutrition; when stable food supplies are available with adequate amounts to feed animals, improving the genetic potential of livestock will pay off. An important point to remember is the limits to recycling nutrients in farming systems; sooner or later we have to replace mined nutrients, and this means supplying fertilizer. Improving crop productivity will be increasingly difficult as populations increase and higher yields need more complex inputs, which in turn brings non-biological problems of organization and management. It took India until 1950 to harvest the first 50 MMT national crop, until 1973 for the first 100 MMT, and until 1985 for a 150 MMT crop. A population of 1.4 billion Indians in 2015 will require an additional 50 MMT every 8-10 years to achieve the 400 MMT that will be necessary in 2015. This will need an amount of fertilizer to be produced at the rate of one railroad carload every 20 seconds, and the railroad capacity to deliver it to where it is needed, as well as the irrigation to sustain productivity without causing salinity. Research will be needed to breed crop adaptation to a broader range of environmental conditions. Sustained research will be necessary to support continuous productivity in face of the constant threat of environmental degradation. Some countries, such as Haiti and Nepal, may be already past the environmental point of no return. Priority for future stabilization of food production through food aid is likely to be in Asia rather than Africa, because of greater man/land ratio and much higher numbers of hungry people. CLIMATIC IMPACT CONTEXT Norman J. Rosenberg, Climate Resources Program, Resources for the Future Dr. Rosenberg indicated that the climatic context has two questions: Are climate forecasts adequately incorporated into projection models? Are projections improving to the point they can be of any use? 34

The modern availability of statistical techniques, computers, and modelling can be employed to I=k econom~c/climatic/ecosystem information to models of plant growth. Forecasts are improving in the short range through information from satellites, and more sensitive global climatic models. In the medium scale, accuracy decays beyond ~3 days. Microscale forecasting is expected to improve dramatically because of better instruments for detecting phenomena like w~ndshear at airports. Improvements in seasonal forecasts are much more difficult because of the absence of real understanding about global climatic systems, and their extraordinary complexity. One recent improvement was the discovery of connections between global phenomena such as the E] Nino Southern Oscillation, drought in the southwest United States, and the timing and duration of monsoons in the Indian Ocean. Predictions of temperature and precipitation in the next growing season are only possible with at best 60 percent confidence, which is of little use to farmers. CLIMATE CHANGE The main factor implicated In possible climatic change ~ the detected increase in carbon dioxide (CO2) concentrations in the atmosphere. At Moana Loa over the period 1958-1988, CO2 increased from 250 ppm to 350 ppm. Observations made in Antarctica over the same period show similar increases. The "greenhouse effects as the result of the increased CO2 concentration, as well as other radiation-interactive gases such as methane, nitrous oxide, tropospheric ozone, and chlorinated fluorocarbons (CFCs). These gases are transparent to most of the solar radiation reaching earth, but are partially opaque to the infrared radiation emitted by the earth's surfaces and the gases of the atmosphere itself. The aggregate effect ~ to warm the lower layers of the atmosphere while cooling the stratosphere. Greatest warming will occur in the high latitudes because of increased absorption of solar radiation as the highly reflective ice melts there. Warming in the tropics will be moderate. Accordingly, the temperature gradient between the poles and tropics which drives the weather systems will diminish in strength. Weather systems are therefore likely to change. The impact of this change is still obscure. Three modelling groups the Geophysical Fluid Dynamics Laboratory (GFDL), the Goddard Institute of Space Studies (GISS), and the National Center for Atmospheric Research (NCAR) have modelled the impact of doubling atmospheric CO2 from its pre-industrial level (about 280 ppm) or the equivalent effect due to all of the greenhouse gases. The models disagree profoundly with respect to the geographic distribution of chances in temperature. Drecinitatic~n and coil mni~til~r~ {Fi~llr-. 15,16, and 17). ~ _.1 __ · 1 . · — O ~ ~ ¢~ I_ a_ ~~ ~^ ^~—user V 40~. —~ boll moisture content in summer is predicted by the GDFL model to decrease throughout all of North America. The other models show less drastic changes. From an examination of these figures one gains an idea of the current "state of the art" in modeling the possible impacts of the greenhouse effect. The models do not yet provide information on mter- seasonal variability as it might be affected by greenhouse warming nor on changes in sunshine, windiness, relative humidity and other climatic factors at least as important temperature in determining crop and forest growth and yield. DROUGHT IMPACT Over the period 1905-1986, yields of wheat in the Great Plains increased as a result of improved varieties and technology, but this increase was interrupted a number of times by drought. Droughts were especially severe in the 1930s, 1950s, and mid-1970s. Wheat 35

GFDL it' ~0 {4>D--\ NCAR ~ x 4~J FIGURE 15 Temperature (June, July, August). The distribution of surface air temperature change (°C) for a doubling of atmospheric carbon dioxide concentration for June, July, and August simulated by the global climatic models of GFDL (left), GISS (center), and NCAR (right). Stipple indicates temperature increases greater than 4°C. GFDL q~7 o: go ~? OWEN GISS ~1~ FIGURE 16 Precipitation (June, July, August). The distribution of precipitation rate change (mm/day) for a doubling of atmospheric carbon dioxide concentration for June, July, and August simulated by the global climatic models of GFDL (left), GISS (center), and NCAR (right). Stipple indicates a decrease in precipitation rate. 36

GFDL I... . l . ~ \ - 1~1.~1~1141.1.\ ~2'gIl./ o ~7 GISS _~N KG O \ ~7 oN NCAR of FIGURE 17 Soil Water Content (June, July, August). The distribution of soil water change (cm) for a doubling of atmospheric carbon dioxide concentration for June, July, and August simulated by the global climatic models of GFDL (left), GISS (center), and NCAR (right). Stipple indicates a decrease in soil water. yields have been generally high and stable since the m~d-1970s in the Great Plains region because the weather has been generally good until, of course, spring and summer of 1988. Some have attributed the 1988 drought to the long-awaited appearance of the greenhouse effect. CO2 concentration has been increasing throughout the 20th century. How then do we explain the generally benign weather of the last decade? Hot years, cold years, wet years, and drought years are all well known in the Great Plains. One need not invoke the greenhouse effect to explain the most recent drought. Jones and Wigley in the United Kingdom and Hansen and Lebedeff in the United States have shown an increase in global mean temperature of about 0.5°C since 1900. Their records show that the warming trend was interrupted by cooling for about 20 years between 1940-1960. Since 1960, warming has resumed. Karl et al. (1988) have computed the average temperature for the 48 contiguous states during the same period. They observe a warming trend until the mid-1930s. From then on, temperatures have decreased not only until 1960 but, for all intents and purposes, to this time, although there has been a slight upswing since the mid-1970s. Although the Karl et al. data show that the mean temperature in the 48 states is now about 0.4°C warmer than it was in 1905, it is 0.5°C cooler than it was in 1935. The impact of increased temperature and consequent changes in climate could change the comparative agricultural advantage many countries now enjoy. Canada and Japan and parts of the Soviet Union might benefit with yields increasing in a warmer world. Other countries such as the United States might suffer losses particularly if the scenarios of reduced precipitation come to pass. However, some adjustments are possible, for example, changes in planting dates, changes in varieties, introduction of new crops. Most agricultural scientists hold that there are many tactics and strategies available to help agriculture adapt to the 37

kinds of temperature change that are predicted by the global climate models. Major changes in precipitation might be more difficult to deal with. However, the certainty assignable to model predictions of regional precipitation changes is much Tower than that assignable to predictions of temperature change. Another factor to consider is that plants will respond to the higher concentration of CO2 In the air with increased rates of photosynthesis and with improved water use efficiency, that is photosynthesis per unit of water consumed. These responses, often termed the "CO2 fertilization effect" should act to moderate detrimental climate changes and perhaps augment the effects of beneficial climate changes. Dr. Hutchinson discussed the ability to predict emergency food aid requirements, and concluded that in the 1990s our ability to predict drought or famine will not improve enough to make any effective difference to the farmer. He pointed out that we must be able to monitor better, and though remote sensing is very useful for observing changes over time in soil moisture and crop acreage, there is a need to improve ground truth with better data. We really do not know people's needs, market situations, patterns of drought and famine, and the ability of people to respond to them. The 1974/1975 famine-affected population in the Sahel should not have survived, but they did. What did they do? What institutional responses are needed? Drought will continue, and will be incorporated in some models, so that there should be plans for drought in national planning, as we are admonished in biblical tradition. . AFRICAN REGIONAL CONTEXT Christopher DeIgado, IFPRI Dr. DeIgado observed that world food price projections can be useful to the policy debates concerning the particular problems of Africa. He underlined the African situation: on the demand side it is useful to distinguish three types of commodities rice and wheat, other cereals, and non-cereals. The main factors influencing demand are income and population growth in the urban and rural areas. He agreed that world commodity prices will decline. The composition of cereal import needs of LDCs are driven by GNP per capita. There is a shift in commodity composition which is not linked to price, and usually not much discussion of non-cereal food aid, which is relatively import ant in some parts of Africa. There axe five issues touched on by the conference discussion that are especially relevant to food aid in Africa: 1. The projected continued decline in world food prices. This is very important, because it is acting in opposition to the strategy for increasing African food production through price incentives to farmers. 2. The continued rising share of income in urban areas. This is more important in determining the commodity composition of demand than any other factor; the latter is a prime determinant of the demand for imported wheat and rice there has been a shift in commodity consumption in West Africa since the 1960s, according to FAO figures, where rice and wheat consumption rose by 16 kg per head while sorghum fell by 23 kg per head; 3. IMP structural adjustment policies are working in Africa, but the debt situation is grim and demand on foreign exchange for debt servicing is a major constraint on growth; this will tend to reduce the ability to import food commercially; 4. Hard times ahead suggest that political stability will be a major issue: food aid is an important tool for reducing pressure on governments; 38

PRICES Rice price = Rice import parity price Sorghum import parity price Sorghum price Rice export parity price Sorguhum export parity price ~ \ Rice supply \ - / ~ ___ ~~ ~ ——————, I /~ \ / Sorghum 1~1 1 ~ 1 1\1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ~ 1\1 1 1 1/1 1 supply / \ \\ / curve beck / \ V weather I ~ \ A: Rice demand Sorghum t I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I demand I _ QUANTITI ES FIGURE 18 Commodity Domestic Supply and Demand in Africa 5. The non-cereal food aid picture is confused by subsidies to farmers in the industrial countries: milk is cheaper in the Sahel than it is in Europe, where there is a large surplus; the impact of this on production incentives in Africa will be the hot political issue in food aid in the l990s. Food aid in Africa in the 1980s is very controversial, because it is linked to structural adjustment, price policies, devaluation, and the commodity demand for rice and wheat. The commodity composition in consumption is not synchronous with production and leads to very important equilibrium issues. Some donors want African governments to resist using food aid. However, rice and wheat play an important role in assuring the food security of the poor under increasing urbanization in Africa, particularly West Africa, which is particularly ill-suited to produce these crops locally. Food aid is thus very important in stabilizing prices of these commodities and keeping governments in power. The main problem is that governments have little control over relative prices among cereals because of the large gap between export and import parity prices for major cereals (due to high internal transport and marketing costs see Figure 18~. However, food aid can play a role in helping governments to stabilize relative cereals prices. Political stability requires rice and wheat imports which governments cannot buy, hence the need for food aid. Biangular arrangements offer the potential for even relatively rich countries to use food aid productively, for example, for Zimbabwe to import wheat and export maize to countries like Mozambique. Structural adjustment to get agriculture moving will increase the benefit/cost ratio of domestic production, but prices are beyond the ability of governments to control. Therefore, the only way to get benefits to farmers is to decrease unit production costs. This requires a massive increase in capital investment. Public investment in infrastructure and institutions can be a major catalyst to help price incentives encourage the massive private investment (by farmers) to do the job. Delgado suggested food aid could be used to allow farmers not to plant for one season and instead use their labor for capital improvement. In the ensuing discussion, the dilemma was underlined in which African countries typically base food security policies on increasing food prices. Governments cannot bank on a long-term downward trend in prices, yet this is what has happened over the past 39

100 years. They may be right, since it is possible that the agricultural policy dilemmas of overproduction in the industrial countries may not continue. Demand for food marches on at an average 3 percent increase per year, while supply fluctuates. In the Tong run, a change will occur to the benefit of producers when livestock demand increases in the LDCs, as incomes rise, perhaps in 20 years. This implies that the concentration on coarse grains is correct, since in addition to human consumption they will be required to feed livestock, and must therefore be produced in Africa on a large and growing scale. ESTIMATING FOOD AID The Chairman pointed out that earlier projections of needs were made by a few people on the basis of very primitive figures and guesswork. Now complex methodologies are available for processing data, and we have 25 years experience; what have we been able to learn, and can we forecast food aid with any degree of accuracy? Projections are complicated when they are made for longer than short run: 1990 projections should be reliable, 1995 fairly good, 2000 less reliable. Complications arise from regional breakdowns for different commodity demand, exogenous shocks, and other factors. They are based on demand-driven analysis, projections of trade numbers, and supply-side forecasting. The latter—a supply-side forecast is relatively simple: FAt=FAt_n b (where Food Aid, FA, at time t, is estimated from a previous food aid time t minus n years ago—times a weighting factor b reflecting institutional inertia in the budget. Other factors such as price effect and growth forces, either high or low, might also be added to this equation.) The AID representatives observed that they not only need numbers, but must be able to explain the numbers to Congress. How do the numbers relate to commercial markets, non-commercial supplies and other subsidy programs? Are we looking for a single figure for food aid, or trends? There is now a very large export subsidy program, the Export Enhancement Program (EEP), amounting in 1987 to $2 billion. How do we account for a need that is partially satisfied by these concessional but non-aid programs? Since actual food aid allocations are made in a highly political environment which is largely divorced from "need criteria", and since most t50 percent or more] food aid in fact ~ , replaces commercial imports (even though by international convention it is supposed to be "in addition" to commercial food trade), would it not be valid to first estimate how much food will likely flow through commeTciai trade channels (as the BES/~lASA, World Bank and FAPR! models do), and then separately estimate how many people would still be hungry or severely malnourished given these food trade levels, and finally derive the food aid need estimate from these trade and malnourished projections? There followed a discussion concerning the ability to model food aid estimates. There is a methodological dilemma inherent in the validity of the modeling process. On the one hand it is argued that a sound methodological approach to medium-term food aid estimates (5-10 years) must be based on a general equilibrium model. (A general equilibrium model requires that the interrelationships among the many variables of the entire economy can only be described validly by including all the variables together in the analysis.) The IlASA model described by Dr. Frohberg is such a general equilibrium model, but because of the complexity of the economy and data constraints, it is still fairly primitive and it may be some time before an adequate general equilibrium mode] is available to assess food aid 40

1 needs and their impact, if supplied, on the economies into which they are inserted. On the other hand, many economists argue that it is not necessary to include all the variables simultaneously, and that a partial equilibrium model, in which the outcome of changing key factors such as world cereal prices, or supplying food aid at varying levels to a country or region, can be modeled independently of other factors (such as GNP growth rates) that can essentially be treated as constants. The IFPRI, World Bank and USDA estimates assume that this partial equilibrium modeling is valid. As there was no way to resolve this methodological dilemma in the time available for discussion, the modelers agreed that the only way to arrive at estimates was to assume that many of the factors could be held constant and see what food aid projections would result. Dr. Kates also requested the modelers to indicate both high and low levels of food aid that would surprise them if they were the actual levels in 1995 or 2000. In exchanges that followed, various approximations and extrapolations from trade projections using assumptions about the future proportion of food aid, allowed various participants to propose food aid need estimates as starting points for refinement, but not as conclusions of their research. Estimates USDA Projection to 2000 56 MMT represents annual nutritional needs of 69 countries (1988/89 level of 37.5 MMT at compounded 3.8 percent growth to 2000) 29.4 MMT status quo current annual amount at 3.8 percent growth to 2000 21.1 MMT annual food aid shipments 1976-85 trend extrapolated to 2000 (excluding export subsidies) 17.6 MMT lean year trend extrapolated to 2000. 56 MMT is the high side surprise limit; 21.1 the Tow side. World Bank Projection By the year 2000, 229.3 MMT is the projected net industrial country commodity export to LDCs. In 1988, 10 percent of industrial country commodity exports to LDCs are food aid; if this level is maintained to 2000, the Bank estimate of food aid will thus be 23 MMT. IlASA/:B[S Estimate Projection of the mode} to 2000 shows 165 MMT net imports annually by LDCs (excluding China). Food aid or other subsidized imports estimates for 2000 are around 30 MMT/year. "Food gap" (nutritional) deficit is estimated at 50 MMT, calculated from 400 million hungry people. IFPR] Estimate Based on 85 Tow-income LDCs, excluding China, India, Nigeria, and Brazil, and at 2 levels of per capita GNP cutoff: 41

Program Food Aid Requirements in MMT Cereal Equivalents (CE) per year By 1990 By 2000 By 1995 All countries 37 74 55 Countries with < $800 average GNP/head 19 39 29 FAPR] Estimate Net average imports by LDCs (MMT CE/yr) 1990 2000 1995 World Bank estimate 103 177 129 FAPRI 95 117 IIASA/BLS — 165+11 Net exports from industrial countries World Bank 134 229 171 FAPRI 129 151 World Bank shortfall 31 52 42 FAPRI shortfall 36 34 Nutritional Need Estimates Dr. Pinstrup-Andersen projected three levels (A, B. and C) with slightly different assumptions, based on Dr. Ezekiel's IFPRI figures, as shown in Table 7. FAO Estimate The 1979 projection of food aid need in 1985 was 17-18 MMT, reasonably close to the actual estimate of need in 1985 of 22 MMT. Actual food aid provided was 10 MMT. In 1988, the actual food aid supplied is projected to approximate 9 MMT. The unofficial estimate of need for 1990 is 18-20 MMT, and 8-10 MMT is likely to be a "politically acceptable estimate" of actual amount to be supplied. Food aid approximates 10 percent of total overseas development assistance (ODA). OECD countries are unlikely to increase the total ODA level in 2000. FAO estimates of net cereal import needs of all countries in 2000 are 115 MMT, and the gross needs (including China and India) 160 MMT. These figures do not include an estimated additional 10 percent for dairy, fish and fruit components. For the Tow-income food deficit countries, based on requests and percentage of food aid in imports 1984-86 (with very little for feed), food aid needs are 19 MMT minimum and 38 MMT maximum. The average is thus about 30 MMT. The lower end of the range assumes an optimistic domestic cereal production scenario for developing countries. POINTS OF CONSENSUS REACHED BY THE WORKSHOP 1. There will be an increase in trade of cereals with LDCs, especially imports. 2. There will be, on average, a decline in real-term commodity prices. 3. The shortfall between domestic production and requirements will increase. 4. More ambitious domestic policies on nutrition using food targeting would increase the estimates of food aid needed. 5. African needs have the widest boundaries, showing the greatest uncertainty. Ten years is very short to expect dramatic change in growth rates in agricultural production. 42

TABLE 7 Pinstrup-Andersen Estimate of Nutritional Food Aid Needs A B C Estimated numbers of food deficient people 400 (in millions) Assumed calorie deficiency/person/day 500 400 300 300 300 Numbers and calorie' expressed as 15 18 15 MMT grain per year Substitution (percent)3 50 50 67 Food aid total (from need/~ubetitution)4 30 36 45 Net addition to food supply5 15 18 30 1Two alternative scenarios 2Additional grain consumption by calorie deficient population to eliminate deficits 3Percent of food aid used for replacing current consumption. Two alternative scenarios. 4Amount of food aid targeted on deficient households needed to eliminate deficiencies. 5Since the net addition to market supply is less than what is projected to be needed to keep prices stable (IFPRI estimates 37 MMT), no price-depressing effect will occur. These often require investment in economic infrastructure, as well an development of agricul- tural technology. It takes 20 years to introduce a new improved crop variety widely enough to farmers to make an impact on national production. If production grows, however, income grows (the rule of thumb is: one job created in agriculture creates five jobs in the general economy) and demand grows, both production and demand changes, affecting "need". Asia may be a larger arena of growth in need by the year 2000. How much food aid is possible? Just to achieve a minimum estimate of 19 MMT by 2000 requires doubling current levels, and this would mean a 10 percent per year increase, whereas current budget plans do not anticipate such a rise. A growing "needs" gap, especially in Africa, is possible, therefore, although not certain, given the uncertainty already noted. 6. Exogenous shocks create the need for systematic projection planning under uncer- tainty. Food aid projections are not good at including exogenous shocks.Negative exogenous shocks include: . Climatic effects - increased variability increased UV affecting photosynthesis change in ozone layer volcanism, and "volcanic winter" effect occurrence of worst weather events together in the eastern and western or northern and southern hemispheres. Economic effects - debt renunciation 43

recurrence of rapidly rising energy costs recession due to industrial countries' policies. Positive exogenous shocks include: Climatic effects climatic improvements in short-growing areas of the USSR, Canada, etc. · . Economic effects diversion of armament expenditures to development funding. O Technology effects unexpected research breakthroughs in plant breeding, production, storage or marketing technologies. There is little likely impact of research on productivity during the next decade, because of the length of time it takes for research to be translated into technology available widely to farmers. Research is able to respond more quickly to outbreaks of disease than previously, because of increased sophistication in both biotechnology and communications among re- searchers. Short-term research contributions may be anticipated in improving postharvest technology. The only way of handling exogenous shocks is to recalculate the estimates to take into account their projected effects, because including exogenous shocks in forecasts would make modelling impossible because of the number of variables. The only way they can be handled is analogous to the disclaimer included in insurance policies: "in the event of exogenous shocks, the estimates have to be revised". This is therefore a potentially fruitful area for research, including the stochastic modeling suggested by Dr. Klein. 7. Numbers of hungry people. Dr. Kates pointed out that the estimates of hungry people vary by a factor of 6, from Dr. Duncan's conservative 100 million to Dr. Pinstrup-Andersen's 400-500 million, with FAO's "Food-Poor Population" of 350 million close to the mean. This is partly as a result of differences in definition of "hunger", partly because of the difficulty in obtaining reliable figures. In fact, there has been remarkable progress in reducing the percentage of hungry in the population, from close to 35 percent in 1950 to 17 percent in 1988 (Figure 19~. There is some evidence that the rate of improvement slowed over the past 5-6 years, and the curve may not continue downward. It needs to be underscored that the program to reduce hunger is stagnating and must be improved over the next decade to overcome a serious hiatus. Dr. Rogers pointed out the need to revise the definition of hunger, to take into account the different thresholds of need for women, children and growth and activity. Dr. Perkins noted that in 1988 10 MMT is the total amount of food aid estimated to be delivered to Tow-income food deficit countries, while 20 MMT is the minimum actual level of need. Doubling the present level is the maximum that could be realistically expected by 2000 from the standpoint of political acceptance, since it involves a 10 percent annual increment. Dr. Perkins indicated the actual commitments of food aid in 1967 4.7 MMT committed by the food aid convention 1980 7.6 MMT 2000 16 MMT projected. The range of figures derives from a number of different sources. Dr. Andersen suggested that the most useful references include the FAO Fifth World Food Survey (1985) and the Agriculture: toward 2000 (1987 revision), the Kates et al. Hunger Report 1988, and the Sub-Committee on Nutrition report (ACC/SCN, 1987~. It was agreed there is need to revise upward the nutritional need thresholds to take 44

- - 30 z I 20 at LL to 10 _ —_ _, - ~. _ _ _ —_ 1 950 1 960 1 970 1 980 1 990 2000 ,` YEAR FIGURE 19 Change in Percentage of Hungry People with Time into account the special needs of women and children, and to stress that the demand- based estimates of need do not represent nutritional needs estimates these must be more precisely targetted by country and by needy groups within countries. Nutritional needs must be met by methods that go beyond food aid, through entitlement, food for work, food stamp programs, or other mechanisms. REFERENCES AND BIBLIOGRAPHY ACC/SCN. 1987. First Report on the World Nutrition Status. Food and Agriculture Organization of the United Nations, Rome. Ascher, W. 1978. Foreca~tir~g: An Appraisal for Policy-Maker and Planners. The Johns Hopkins University Press, Baltimore. Brown, L. 1988. '`The Changing World Food Prospect: the Nineties and Beyond." Washington, D.C.: Worldwatch Institute, Worldwatch Paper 85. Ezekiel, H. 1988a. "Medium-term Estimates of Demand-based Food Aid Needs and their Variability: A Note on Methodology.n International Food Policy Research Institute, Washington D.C. 9pp. Ezekiel, H. 1988b. "Medium-term Estimates of Demand-based Food Aid Needs and their Variability.n International Food Policy Research Institute, Washington D.C. 13pp. FAO. undated. aFAO Agricultural Commodity Projections to 1ggo.n FAO Economic and Social Develop- ment Paper No. 62, pp. 201-212. FAO. 1984. "Assessing Food Aid Requirements: A Revised Approach." FAO Economic and Social Development Paper No. 39. FAO. 1985. "The Fifth World Food Survey." Rome. FAO. 1987. "Agriculture: Toward 2000." Conference of FAO, Twenty-fourth Session, Rome, 7-26 November 1987. Document C 87/27, Revised Version 1987. Food and Agriculture Organization of the United Nations, Rome. FAPRI. 1988. "FAPRI Ten-Year International Agricultural Outlook. Summary and Tables," and "FAPRI Staff Report #1-88.~ Ames, Iowa: Center for Agricultural and Rural Development, Iowa State University. Frohberg, K. 1988. "Food Aid Requirements of Developing Countries.n Organization for Economic Cooperation and Development, Paris, France. Xerox, llpp. Hanrahan, C.E. 1988. "The Effectiveness of Food Aid: Implications of Changes in Farm, Food Aid, and Trade Legislation." Proceedings of a CRS Workshop held on April 25, 1988. Washington, D.C.: Congressional Research Service, The Library of Congress, Document 88-493 ENR, 155pp. Johnson, D. Gale. 1979. PA Liberal View." Chapter 9 in Raymond F. Hopkins and Donald J. Puchala, "The Global Political Economy of Food." 1979. The University of Wisconsin Press. 45

Johnson, S.R., W.H. Meyers, P. Westhoff and A. Womack. 1988. "Agricultural Market Outlook and Sensitivity to Macroeconomic, Productivity and Policy Changes." Paper prepared for the Twentieth International Conference of Agricultural Economists, Buenos Aires, Argentina, August 23-31, 1988. Karl, T.R., H.F. Diaz and G. Kukla. 1988. "Urbanization: Its Detection and Effect in the United States Climate Record. Journal of Climate. (in press). Kates, R.W., R.S. Chen, T.E. Downing, J.X. Kasperson, E. Messer and S.R. Millman. 1988. The Hunger Report: 1988 The Alan Shawn Feinstein World Hunger Program, Brown University, Providence, Rhode Island, 45pp. Larue, B. and K.D. Meilke. 1988. A Di~cwsiorz of Lor~g-Term Agr~c?tltural Commodity Forecasts and Food Aid Needs. Department of Agricultural Economics and Business, University of Guelph, Ontario, Canada. Xerox 15pp. Meyers, W.H., S. Devadoss, and B. Angel. 1988. "Commodity Market Outlook and Trade Implications Indicated by the FAPRI Analysis." Center for Agricultural and Rural Development, Iowa State University, Ames. Xerox, 34pp. Meilke, K.D. and B. Larue. 1988. "A Critique of Agricultural Trade Models. Paper prepared for the Twentieth International Conference of Agricultural Economists, Buenos Aires, Argentina, August 23-31, 1988. Mitchell, D.O. 1988. "Outlook for Grains and Soybeans to 2000." Paper prepared for the Twentieth International Conference of Agricultural Economists, Buenos Aires, Argentina, August 23-31, 1988. Washington, D.C.: International Commodity Markets Division, The World Bank. Xerox, 21pp. Parikh, K.S. and W. Tims. 1986. 'From Hunger Amidst Abundance to Abundance Without Hunger." (An overview of the policy findings of the Food and Agriculture Program of the International Institute for Applied Systems Analysis). Laxenburg, Austria: International Institute for Applied Systems Analysis. Executive Report 13. 37pp. Paulino, L.A. 1986. "Food in the Third World: Past Trends and Projections to 2000.~ Washington, D.C.: International Food Policy Research Institute, Research Report No. 52. Perkins, R.J. 1988. "Statement for the NRC Workshop on Food Aid Requirements and Food Trade in the 1990s.~ Rome: U.N. Food and Agriculture Organization, Commodities and Trade Division. Xerox, 9pp. Pinstrup-Andersen, P. 1988. "Changing Patterns of Consumption Underlying Changes in Trade and Agricultural Development. (Publication from the December 1986 Trade Consortium Meeting in Mexico, pp 69-90.) Washington, D.C.: U.S. Department of Agriculture. Pinstrup-Andersen, P. 1988. Prevalence of Malnutrition and Dietary Deficiencies as they Relate to Health and Performance Problems (Paper Prepared for the 1988 World Food Conference, Des Moines, Iowa, June 5-9, 1988. Ithaca, New York: Cornell Food and Nutrition Policy Program, Division of Nutritional Sciences, Cornell University. Xerox, 30pp. Roningen, V.O., P.M. Dixit and R. Seeley. (undated) "Agricultural Outlook for the Year 2000: Some Alter- natives~ Washington, D.C. U.S. Department of Agriculture, Economic Research Service. Unpublished draft. Xerox, 10pp. Rosenberg, N.J. 1988. "Global Climate Change Holds Problems and Uncertainties for Agricultures, in Ann Tutwiler, (ed.) U.S. Agriculture in a Global Setting: Art Agenda for the Future. Washington, D.C.: Resources for the Future. Chapter 13. Rosenberg, N.J. 1988. "Some Highlights of a Resources for the Future Workshop: Controlling and Adapting to Greenhouse Warming (held at the National Academy of Sciences, Washington, D.C., June 14-15, 1988.) Xerox l9pp. Sanderson, F.H. 1984. World Food Prospects to the Year 2000.~ Washington, D.C.: Resources for the Future, Reprint 219. Sarma, J.S. and P. Yeung. 1985. "Livestock Products in the Third World: Past Trends and Projections to 1990 and 2000~. Washington, D.C.: International Food Policy Research Institute, Research report No. 49. Schlesinger, M.E. and J.F.B. Mitchell. 1985. "Model Projections of the Equilibrium Climatic Response to Increased Carbon Dioxide," in M.C. MacCracken and F.M. Luther, eds., Projecting the Climatic Effects of Increa~r~g Carbon Dioxide. DOE/ER-0237. Washington, D.C.: U.S. Department of Energy, Carbon Dioxide Research Division. Trostle, R.G. 1988. "Food Aid Needs During the 1990s." Washington, D.C.: U.S. Department of Agriculture, Commodity Economics Division, Economic Research Service. Xerox, 10pp. USDA. 1988. "World Food Needs and Availabilities, 1988/89: Summer." Washington, D.C.: Economic Research Service, U.S. Department of Agriculture. Young, V., D. Bier, and P. Pellett. 1989. A Theoretical Basis for increasing current estimates of amino acid requirements in adult man, with experimental support. American Journal of Clinical Nutrition 50 :80-92. 46

Next: Appendix A: Medium Term Estimates of Demand-Based Food Aid Requirements and Their Variability »
Food Aid Projections for the Decade of the 1990s Get This Book
×
Buy Paperback | $50.00
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

This book presents the results of a meeting held by the Board on Science and Technology for International Development. At this meeting, six groups of modelers presented their best estimates of the food aid/food commodity trade picture during the period 1991-2000. These estimates are based on each modeler's own database and own assumptions about trends in global policy, climate, population, and economics.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!