u pIf) 0: ; ff t; ofY : n;f ffE: 0 000 dfE A; .:1

> || 00 it*, # X ; S0 v 'R ' - X

r ¢x.7';t J0 s''tr6 Xen7X, : > eit'] '"II1';.F- '.01~~~~~ -| :: f ::;.KI n R ,. 0g ,t ,..0 f: 0|

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Published by the Consultative Group onInteratonlAgricultundl Research, COIAR

Scretaritat. The World Dan1k. 1818 H St.,

NW., Washington. D.C., 20433, UnitedSae,October 1994.

Sir John CrawfordMemorial Lectures

1985 Robert S. McNamara, United States1986 Bukar Shaib, Nigeria1987 Amartya Sen, India1988 Helen Hughes, Australia1989 Jacques Diouf, Senegal1990 M. S. Swaminathan, India1991 CGIAR Twentieth Anniversary Commemoration1992 Enrique V. Iglesias, Uruguay1993 James Gustave Speth, United States1994 Alex F. McCalla, Canada

The Sir John Crawford Memorial Lecture has beensponsored by the Australian government since 1985 inhonor of the distinguished Australian civil servant, educa-tor and agriculturalist who was one of the founders of theConsultative Group on International Agricultural Research(CGIAR). Sir John (1910 -1984) was the first chairman ofthe CGIAR's Technical Advisory Committee.

Agriculture and Food Needs to 2025:Why We Should Be Concerned

Alex F. McCallaChair, Technical Advisory Committee, CGIARDirector, Agriculture and Natural Resources

Department, The World Bank

Introduction

Everyone agrees that the world's population willexceed 8 billion people by 2025, an increase of over 2.5billion in the next thirty years. Everyone agrees thatmost of the increase will occur in developing countrycities-urban population is expected to rise from 1billion in 1985 to 4 billion by 2025. Most everyoneagrees that world food supplies will have to more thandouble by 2025, because of increases in income andurbanization in addition to population growth.

Given this widespread agreement on the needs ordemand side of the equation and its magnitude-thegreatest numerical growth in human numbers in his-tory and a required magnitude of increased food pro-duction never before achieved-why is there so littleagreement on the ease or difficulty of generating the

1

supply to meet that demand? The spectrum of viewsranges from the one extreme, 'there is no problem," tothe other, "the imminent arrival of the Malthusian night-mare, unless effective population control is implementedimmediately." By far the predominance of views is to-ward the "no problem" end, and can only be character-ized as bordering on complacency.

Therefore, the puzzle that this lecture identifies is,how can intelligent students of the international foodeconomy agree so closely on the demand side and dis-agree so wildly on the capacity of the world to providethe supply to meet that demand? The cacophony ofviews is muddying the waters and, in my view, retard-ing needed attention to this critical issue.

This lecture has a modest objective. It is to criti-cally appraise the competing viewpoints and to showthat, regardless of which view you prefer, the produc-tivity improvement challenge facing world agriculturein the next thirty years is enormous. Twenty twenty-five is just thirty years away. From initiation to imple-mentation in farmers' fields, agricultural research takesten to twenty years to have an impact. Twenty yearsfrom now there will be at least 1.8 billion more peoplein the world to feed. Research and technology devel-opment to contribute to the needed production muststart today. Everyday spent on further debate aboutwhether "Malthus must wait" or "Malthus is finally right"is "fiddling while Rome burns."

Specifically, I shall do five things. First, I willreview briefly the past history of "food crises" debates.Second, I will quickly summarize the demand side uponwhich most people agree. Third, I will summarize four

2

different viewpoints on the supply side of the worldfood equation, from the "no problem" view, as exempli-fied by Donald 0. Mitchell and Merlinda D. Ingco intheir paper entitled "The World Food Outlook," to theominous predictions of Lester R. Brown and Hal Kanein their book, Full House: Reassessing the Earth'sPopulation Carrying Capacity. Fourth, I will criticallyappraise the consequences of each of these scenariosfor future agricultural development and technologygeneration needs. Finally, I will focus on the conse-quences of not recognizing the urgency of the produc-tivitv challenge.

Past Debates

The sufficiency of future food supplies has been arecurrent question in the international debate over mostof the post-World War II period. The debate is mostfrequently driven by supply side considerations. SinceThomas Malthus wrote his "Essay on the Principle ofPopulation as It Affects the Future Improvement ofSociety" in 1798, the debate has focused on the racebetween supply (seen to grow linearly) and population(seen to grow exponentially). New lands, new technol-ogy, and capital investment in irrigation have delayedthe "Malthusian cross" (i.e. when population growthrates exceed the rate of food supply increases) for mostof the world, but the debate, for how long?, has ragedfor years.

Immediately after World War II, there were con-cerns about imminent food shortages. These quicklygave way to food production surges and rising stocks

3

in the 1950s and early 1960s. Two bad monsoons inSouth Asia during 1965-66 led to resurgent concernsabout imminent famine. William and Paul Paddockwrote a best seller in this period called Famine-1975!,which predicted imminent famine by 1975. In the late1960s and early 1970s, Malthus was kept at bay againby expanded output. The years 1972 to 1974 saw acoincidence of events-production shortfalls in severallocations simultaneously and rapid demand expansion,particularly from the Soviet Union, which caused agri-cultural prices to skyrocket. Grain prices tripled overan eight-month period. Global food shortages werepredicted. Then U.S. Secretary of Agriculture, EarlButz, exhorted farmers to 'plant fencerow to fencerow.'

Surpluses rebuilt in the early 1980s. The UnitedStates instituted its most comprehensive and expen-sive supply control program, Payment in Kind (PIK), in1983 as stocks soared. The 1988 drought brought abrief return of the issue of possible shortages, but-concerns about excess supplies soon cooled the de-bate, at least in developed countries. Mitchell andIngco, in their recent paper, extensively reviewed pre-vious chapters of this debate and concluded that tech-nological pessimists have always been wrong.

Currently, the food production versus populationgrowth issue is the subject of some debate, but theissue is not viewed as critical, even though there arewidely divergent views on what the next twenty to thirtyyears will hold regarding the world's capacity to in-crease production to feed more than 8 billion people.

I will return to four of these widely divergent viewsin a moment, but first let me quickly review the de-mand side that most people agree on.

4

World Food Needs to 2025-The Convergent View

World population will double in the next forty years.By 2025-thirty-one years from now-the median vari-ant of projections by the United Nations suggests aglobal population of 8.5 billion people. A larger shareof that population will live in developing countries. In1985, 75 percent of the world's population lived indeveloping countries. By 2025, more than 83 percentwill live there. At present, approximately 31 percent ofthe population of developing countries live in cities,although there are strong regional differences. By 2025,it is estimated that 57 percent of the population indeveloping countries will live in cities. The number ofpeople living in cities will quadruple from 1 billion to4 billion. Regionally, the population in Asia will nearlydouble to over 4 billion, while that in Sub-SaharanAfrica will more than triple from 420 million in 1985 tonearly 1.3 billion by 2025 (Table 1). The number ofmalnourished will rise from the current level of 750million to over 1 billion.

In addition to population growth, income growthalso increases the demand for food. Even with modestincome growth in developing countries, the demand forfood in 2025 will be more than double current levels ofproduction. Further, urbanization, in conjunction withincome growth, will cause the character of diets to shiftaway from roots and tubers and lower quality staplegrains to higher quality cereals, such as rice and wheat,livestock products, and vegetables. [See Mitchell andIngco, 1993, Chapter V, for an excellent review.] Withmassive urbanization will come increased need formarkets and basic infrastructure as well as for urbanoriented food security policies.

5

Table 1. Global Population and Distribution Patterns

Developed Developing Sub-Saharan Latin Asia and West Asia-Year World Countries Countries Africa America the Pacific North Africa

POPULATION (millions)

1960 3,019 964 2,055 209 218 1,505 1231985 4,855 1,210 3,645 421 404 2,575 2452010 7,191 1,365 5,826 916 631 3,810 4692025 8,467 1,422 7,045 1,296 761 4,379 609

DISTRIBUTION (percentage)

1960 100.0 31.9 68.1 6.9 7.2 49.8 4.11985 100.0 24.9 75.1 8.7 8.3 53.0 5.02010 100.0 19.0 81.0 12.7 8.8 53.0 6.52025 100.0 16.8 83.2 15.3 9.0 51.7 7.2

Source: CGIAR, Technical Advisory Committee. 1990. 'A Possible Expansion of the CGIAR." Paper AGR/TAC:IAR/90/24. TAC Secretariat, Food and Agriculture Organization of the United Nations, Rome, Italy.

Most observers also agree that there will be wideregional differences in the severity of hunger and mal-nutrition. All agree that Sub-Saharan Africa and SouthAsia will face particularly difficult problems. Let usrecall that by 2025 these two regions alone will ap-proach a population of 3 billion people.

The Supply Side-The Divergent Views

Perceptions of the capacity of the world to meetthe above challenges vary widely. On the optimisticside are analysts who use global projection models basedon past trends, which basically conclude that on aglobal basis the world can feed itself until at least 2010.The clearest and most comprehensive presentation ofthis view is by Mitchell and Ingco. At the opposite endof the spectrum is the most recent Worldwatch Insti-tute study by Brown and Kane. Reading Mitchell andIngco and then Brown and Kane one after the other isa study in contrasts which makes one wonder whetherthey are talking about the same planet.

Between these two poles are two other views. OneI call the conventional scenario which argues that thechallenge is serious indeed, requiring developing coun-tries to increase significantly their capacity to feedthemselves and in a sustainable fashion. The percep-tion is that it can be done; but, if current investmentsin agricultural development and productivity improve-ment are not maintained or increased, the world willspin toward the Brown-Kane model. A fourth wild-card scenario is a hypothesis presented by IanCarruthers called "Going, Going, Gone! Tropical Agri-culture as We Knew It," which argues that developingcountries will not be able to meet their growing urbancereal demands and that the developed countries mustfill the gap with greatly expanded trade.

7

I begin by presenting the conventional view andthen turn to the other three scenarios.

Scenario 1: The Conventional View

The challenge facing world agriculture is enor-mous. World food production has to more than double.Until the middle of the twentieth century, expansionof cultivated area roughly kept pace with populationgrowth. In the last forty years, the doubling of cerealoutput came from three sources-area expansion,increased intensity of land use (mainly through ex-panded irrigation), and yield increases. While irri-gated area more than doubled from 1950 to 1980, itsrate of growth has since slowed substantially as hasarea expansion in rainfed areas. The current view isthat the next doubling of food production must comeprimarily from increased productivity (i.e. yield). Al-ready increasing productivity in many developing coun-tries is putting stress on the natural resource base-in some countries as much land is lost to erosion andsalinization as is brought into production through ir-rigation or area expansion.

Therefore, the difficult challenge facing world ag-riculture is to double production on the same landbase while maintaining or, hopefully, improving thenatural resource base. These are the twin challengesof creating environmentally-sustainable production sys-tems-productivity improvement and improved man-agement of natural resources.

The aggregate challenge is staggering enough: but,when we begin to disaggregate food demand, the task

8

is more complicated. As noted, rising incomes andurbanization shift the composition of food demand.Consumers demand more diverse and higher-qualitydiets and need foods that can be transported and stored.While yields of some cereals, such as wheat and rice,have doubled in the last thirty years, yields of mostother developing country crops-such as maize, cas-sava, sorghum, millet, beans, and edible legumes-haveshown less rapid increases. To double again wheatand rice yields and more than double yields of otherbasic food products will be problematic without in-creased research and development efforts. While bio-technology holds the promise of significant geneticimprovements, that promise is becoming reality muchmore slowly than earlier forecasts suggested.

This scenario implicitly views the food supply prob-lem as basically a nationalistic one (i.e. countries areresponsible for their own food security). This is gen-erally translated to mean responsible for their own foodproduction. Trade enters the scenario in a limitedway. If food demands double, grain consumption-ofwheat, rice, and maize-will increase from 1.9 billionmetric tons to 3.8 billion metric tons. Trade is nowaround 200 million metric tons, or approximately 10percent of the supply, and is not likely to grow as apercentage. If developing countries are to grow theirown food, and if population increases 2 percent peryear, then their food production must rise by 2 percentper year.

Scenario 2: The Optimists

Analysts have been projecting world food supplyand demand balances for decades. In their simplest

9

form, rates of population growth are added to rates ofincome growth, modified by the income elasticity of thedemand for food, to project a rate of growth in foodneeds (i.e. demand). This rate is then compared torates of growth in productivity (i.e. production), usuallymade up of an estimate of new land availability plusprojected yield increases. These models, therefore, arebasically projections of two compounding growth rates.Any deviation between these rates either leads to foodgaps or surpluses and the difference increases thefurther the projection. If the model has endogenousprices, then real prices either rise or fall. Over a twenty-five year time horizon a one-tenth of 1 percent differ-ence leads to substantial divergence. In reality, ofcourse, food gaps or food surpluses do not occur be-cause prices in the marketplace equilibrate quantitysupplied to quantity demanded. Thus, the strong fo-cus on the direction of real prices of food over futureperiods.

One such model deserves our attention here.Mitchell and Ingco have produced a substantial andcontroversial paper in 'The World Food Outlook." Afterreviewing past predictions of global food shortages overthe last several decades, Mitchell and Ingco concludedthat the world has really done quite well. Using threeindicators-real food prices, calories available to con-sumers, and per capita food production-they concludedthat, overall, the world was better fed in 1990 than in1960. Real food prices, except for a blip from 1972 to1974, have continued their century-long decline. 'Percapita calorie supplies in developing countries rose by27 percent from 1961-63 to 1987-89" (Mitchell andIngco 1993, p. 20) and overall per capita food supplies

10

"increased at a steady pace since 1961-63" (p. 23).

Their basic question is, what can we expect to2010? Their model is based on two critical assump-tions. The first is that the global population growthrate will decline from 1.74 percent in 1994 to 1.4 per-cent in 2010. The second is that world grain produc-tion will grow at 2 percent per annum from now until2010. The result of their baseline simulation model isthat global food production increases will more thankeep pace with increases in demand. Food imports bydeveloping countries will increase by more than 4 per-cent per annum, doubling imports by 2010; but thesewill easily be provided by expanded exports from devel-oped countries and reduced net imports by formerlycentrally planned economies.

Among the study's conclusions are:

1. "The simulation results strongly suggest that theoutlook for the world food situation is good, de-spite regional problems" (p. 151).

2. "It should become increasingly easy to meet theworld's demand for grain if past trends in produc-tion and consumption continue" (p. 175).

3. "The most important conclusion to come from ouranalysis is that the world food system has manyoptions to meet future demand" (p. 175).

4. Mitchell and Ingco's final paragraph:

The world food situation has improveddramatically during the past thirty yearsand the prospects are very good that the

11

twenty-year period from 1990 to 2010will see further gains. However, thesegains depend on continued increases infood production along the trends of thepast. This will not occur automatically.rather it will require continued invest-ments in research to increase crop yieldsand in other factors of production. Ifpast crop yield trends continue and ifpopulation growth rates slow as pro-jected, then the gains in the world foodsituation seen during the past thirty yearsshould continue. If Malthus is ultimatelyto be correct in his warning that popu-lation will outstrip food production, thenat least we can say: 'Malthus must wait"(p. 232). [Emphasis added.]

Other studies-such as "Agriculture: Towards2010" by the Food and Agriculture Organization of theUnited Nations (FAO) and an International Food PolicyResearch Institute (IFPRI) paper by Mark W. Rosegrantand Mercedita Agcaoili entitled "Global and RegionalFood Demand, Supply, and Trade Prospects to 2010"-reach similar, though not identical, conclusions. TheFAO study uses the same population growth rate for2010 (i.e. 1.4 percent), but a slightly lower rate of glo-bal production increase (i.e. 1.8 percent). The studyconcludes that per capita calorie supplies will rise andthe absolute numbers of people suffering chronic un-dernutrition will decline. Production increases for grainsare projected to be 2.2 percent per annum, made up ofa 1.4 percent per annum increase in yield and a 0.8percent per annum increase in area harvested.

12

Rosegrant and Agcaoili use an IFPRI simulationmodel to project to 2010. Aggregate simulation re-sults suggest declining or constant real prices of majorfood commodities, which suggest optimism for futureaggregate food supplies. As with the Mitchell andIngco and the FAO studies, the IFPRI study also pointsto potential regional problems, particularly in Sub-Saharan Africa, but in general it is upbeat.

Their conclusion is:

If governments and the internationalcommunity maintain (or renew) theircommitment to agricultural growththrough policy reform and sustained,cost-effective investment in agriculturalresearch, extension, irrigation and wa-ter development, human capital, andrural infrastructure, there will be nooverwhelming pressure on aggregateworld food supplies from rising popula-tion and incomes. Projected per capitaavailability of food will increase and realworld food prices will be stable or de-clining for key food crops. However,these aggregate price trends concealemerging problems at the regional andcountry level, which show that there willcontinue to be problems in getting foodto those who need it most (Rosegrantand Agcaoili 1994, p. 40-41).

Thus, these models project that growth in globalproduction will keep pace with global demand. Infact, they argue that production could grow faster than

13

2 percent if land currently held out of production indeveloped countries returns to production. The con-clusion is that supply will continue to press on de-mand leading to a continuation of the decline in realgrain prices which has persisted with few exceptionsfor the last 100 years. They conclude that there willbe no global world food problem as aggregate supplywill be equal to or greater than aggregate demand atconstant or lower real prices. Finally, none of thestudies sees resource degradation as a critical issue.In fact, the Mitchell and Ingco study suggests less landwould be needed.

They do, however, admit that there will be pocketsof problems such as in Sub-Saharan Africa and SouthAsia, particularly where there will be problems of mal-nutrition. They identify this as a problem of access tofood, which is a poverty problem not a food problem.If pressed, the supporters of these models will admitthat a 1.8 percent to 2 percent output growth assump-tion is critical, and that there will, therefore, be a rolefor research and technology development; but they donot see global food supplies as a crisis in the nextdecade or two. They are generally silent about thelonger-term.

Scenario 3: The Pessimists

At the opposite end of the spectrum we find Brownand Kane in Full House: Reassessing the Earth's Popu-lation Carrying Capacity. This book is in stark con-trast to the Mitchell and Ingco analysis. The basicpremise is that the 1990s mark the beginning of a newera where it will be much more difficult to expand foodoutput.

14

Many knew that this time would eventu-ally come, that at some point the limits ofthe earth's natural systems, the cumula-tive effects of environmental degradationon cropland productivity, and the shrink-ing backlog of yield-raising technologieswould slow the record growth in food pro-duction in recent decades. But becauseno one knew exactly when or how thiswould happen, the food prospect waswidely debated. Now we can see that sev-eral constraints are emerging simulta-neously to slow the growth in food pro-duction (Brown and Kane 1994, p. 22).

The "facts" according to Brown and Kane are dif-ferent from Mitchell and Ingco. Brown and Kane saygrain production expanded at 3 percent per year from1950 to 1984, but the rate of growth dropped to scarcely1 percent annually during the period 1984-93. RecallMitchell and Ingco projected a continuation of the 2percent per year growth in production that occurred inthe 1980s. Further, Brown and Kane argue that pro-duction of fish has reached its biological limit and thecarrying capacity of rangelands has been exceeded,requiring future food needs to be met by only the crop-land food system, whereas before it was met by allthree-fish, livestock, and crops.

Therefore, Brown and Kane argue future supplytrends will be subject to six new constraints:

1. The shrinking backlog of unused agriculturaltechnology.

2. The growing human demands that are pressingagainst the limits of what fisheries and rangelandscan contribute to increase food needs.

15

3. The demands for water that are pressing againsthydrologic limits.

4. The declining response of crops in many countriesto additional fertilizer application.

5. The substantial losses of cropland to industrial-ization and urbanization.

6. The 'social disintegration, often fed by rapid popu-lation growth and environmental degradation [thatlis undermining many national governments andtheir efforts to expand food production" (p. 24).

On this last point, Brown and Kane cite exten-sively a chilling article by Robert Kaplan entitled "TheComing Anarchy."

F1uX House presents quantitative information toback-up these basic propositions. The food productionincreases have slowed perceptibly in the last ten yearsand may slow even more in the future. Per capitagrain production has fallen from a peak of 346 kilo-grams per capita in 1984 to 303 kilograms per capitain 1993. World grain stock, as a percentage of produc-tion, is at an all time low. Relatively little land iscurrently being held out of production in the UnitedStates and the European Union, and what is out is oflow productivity. Bringing all this land back into pro-duction would "expand the world grain area by only1.6 percent, not half enough to get it back to the his-torical high reached in 1981" (p. 99).

Further, China is losing nearly 1 million hectaresor 1 percent of its cropland per year to industrializa-tion. Brown and Kane predict that China will follow a

16

similar path to Japan, South Korea, and Taiwan, wheretheir combined grain areas decreased from 8 millionhectares to 4 million hectares from 1950 to 1990.Thus China, in their scenario, will experience a 66million metric ton reduction in grain production from1990 to 2030 and an increase of 210 million metrictons in imports-more than total world trade in the1990s.

These facts plus others-declining fertilizer use;a falling off in yield increases in recent years in manycountries (world grain yields increased 2.3 percentper year from 1950 to 1984, but only 1 percent peryear from 1984 to 1993); declining investments inagricultural research; and increasing environmentalpressures-lead Brown and Kane to conclude that theworld is close to exceeding its carrying capacity. Theiranalysis suggests that by 2030 world grain importneeds will 'exceed exportable supplies by 526 milliontons, an amount approaching the current grainconsumption in the United States and China com-bined" (p. 188). Their bottom line is that the growingimbalance between food and people can only be re-dressed by frontally attacking the population issue.In sum, if Brown and Kane were to paraphrase Mitchelland Ingco's concluding sentence, it could read "Malthusis here."

Scenario 4: The Developed CountriesFill the Gap

The most radical scenario is one put forward byCarruthers, a professor at Wye College, in 'Going,Going, Gone! Tropical Agriculture as We Knew It."

17

Carruthers' view is that our traditional model of devel-oped countries (i.e. rich) supplying the world withmanufactured goods and financial services while thedeveloping countries (i.e. poor) provide primary prod-ucts-such as food, natural resource products, andminerals-is not sustainable. Carruthers' view is thatin the long-run developing countries will produce manu-factured goods and trade them for food from developedcountries. His argument in simplified terms runs asfollows:

1. Carruthers is convinced that the tropics are inca-pable of producing enough basic foodstuffs for bur-geoning cities in the developing world-where popu-lation is estimated to be 4 billion by 2025-in thelong-run. The fragile tropical and subtropicalenvironments will be lucky to support the remain-ing 50 percent that still subsist from the land.

2. The trend has already started; it is developed coun-tries-the United States, Canada, Europe, andAustralia-which export food to developing coun-tries and increasingly import labor intensivemanufactured goods.

3. Production increase potentials are greater in thetemperate zone because of better technology andsignificant areas of land held out of production.Therefore, developed countries can provide in-creased supplies through trade.

4. If the scenario occurs with developing countriesexporting manufactured goods (i.e. from laborabundance), urbanites in developing countries willhave enough income to import basic foods (i.e.grains).

18

The implications of this scenario are enormous. Ifthe additional 3 billion urban dwellers are to be fed bytrade, exports of grain will have to increase 4 timesfrom 200 million metric tons to 800 million metric tons,assuming minimum consumption of 200 kilograms percapita, in next thirty years. This is-physically, bio-logically, and economically-a huge task. If trade doesnot expand this rapidly, the impact on food prices couldbe substantial, causing greater increases in malnutri-tion in poor countries. Carruthers' paper contains nonumbers, so it may be that the physical magnitude ofthe increases in production in developed countriessuggested was not fully comprehended. The UnitedStates currently provides about half of world grainexports. To maintain that share, US grain productionwould have to triple by 2030.

Some Comparisons of the Scenarios

The four scenarios presented look at the same'facts" and reach vastly different conclusions. The rea-sons for the differences, despite all the rhetoric, residein four projection parameters:

1. The rate of increase in biological cereal yields tobe expected over the next fifteen to thirty years.

2. The amount of new land to be added to or lostfrom agricultural production.

3. The amount of land subject to increased intensi-fication primarily through irrigation.

19

4. The impact of environmental degradation on foodproduction capacity.

Mitchell and Ingco assume a continuation of therate of increase in production of the last several de-cades. Ninety percent of that increase resulted fromyield increases. In their terms, yield is output per unitof land, which includes the impact of both biologicalyield increases and intensification. Presumably boththe rate of growth in biological yield and irrigation isassumed to continue at the same rate as during 1960-90. Therefore, the assumption with respect to increasedland area appears to be close to zero. They minimizeany significant negative impact on production of re-source degradation.

On the other extreme, Brown and Kane argue thatbiological yield growth has slowed to about 1 percentper year in the last decade and may decline further.Herein lies the major difference-a 1 percent differencein a compound growth rate over thirty years makes anenormous difference at the end of the projection pe-riod. Further, Brown and Kane argue that land lostfrom agricultural production, coupled with increasedurban competition for water, will lead to a projecteddecline in irrigated acreage. Environmental degrada-tion will further constrain production increases.Carruthers makes no explicit presumptions about anyof these parameters, but must implicitly be assuminglow yield growth in developing countries, very high ratesof yield growth plus expanded land area in developedcountries, and environmental constraints mainly in thetropics and subtropics.

The conventional scenario argues that biologicalyields must increase to about 2 percent per annum to

20

replace the contributions made by area expansion andintensification in the last three decades. These yieldincreases must be accomplished without degrading theenvironment further.

All scenarios recognize the need for sustained orincreased investments in research and technology de-velopment.

One must be somewhat cautious in assuming thatfuture biological yield increases at past rates will beeasy. Research by the International Rice Research Institute(IRRI) and the International Center for the Improve-ment of Maize and Wheat (CIMMYr) found significantslowing in the rate of yield increases of rice and wheatunder experimental conditions. Nor should we be blaseabout area or irrigation expansion. While the area ofpotentially useable arable land seems large, its poten-tial for production has been seriously questioned in arecent study by Gershon Feder and Andrew Keck en-titled "Increasing Competition for Land and Water Re-sources: A Global Perspective."

The trade implications of the four scenarios arealso widely different. Carruthers appears to be arguingthat, given population increases in developing coun-tries, exports from developed countries would need toincrease by 400 percent to possibly 800 million metrictons by 2025. Brown and Kane have export require-ments which appear to exceed 700 million metric tons.The Mitchell-Ingco model sees a doubling of developingcountry imports by 2010 and, presumably, if the mod-els were projected further, developing country importscould double again by 2025. This would imply a tri-pling of cereal trade. Finally, the conventional scenariowould imply a doubling of cereal trade.

21

The first three scenarios raise two critical issues.First, the capacity of the developed countries and,possibly, the formerly centrally planned economies toachieve the required rate of increase, particularly givenenvironmental concerns and resource limits. Second,the physical capacity of developing country infrastruc-ture to handle the volume of trade projected. All ofthese models-Mitchell-Ingco, Brown-Kane, andCarruthers-are very cavalier in assuming that thesetwo barriers can be overcome.

Of course, no one knows who will be right. Pro-jections thirty years ahead, particularly those by econo-mists, are invariably wrong. This is partly because ofquestionable assumptions, limited models, and poorinformation, but also because a dynamic world economyis self-adjusting since it does not tolerate disequilibrium.easily.

The Consequences for the Future

While my own views tend to be more consistentwith the conventional view than any others, this is notcrucially important. Regardless of who is correct, theproductivity-food production challenge for the globe isvery substantial. Given the agreement on the demandside, all scenarios 'require" at least a 2 percent ormore per year increase in global food production.However, each scenario would have a different distri-bution of required relative increases. At one extreme,Carruthers places almost the entire burden of produc-tion increases on developed countries, and seems toimply a rate of increase approaching 4 percent peryear-a rate never before accomplished. Mitchell andIngco clearly imply a larger rate of increase in devel-

22

oped countries. The conventional view places more of theburden on developing countries and implies that virtuallyall of these increases must come from biological yieldincreases. Brown and Kane are skeptical about it hap-pening at all. Under all scenarios biological yield in-creases accomplished over the last thirty years must beat least maintained or, better yet, increased.

Several other points need to be made:

1. The global requirements for production systems tobe non-degrading to the environment (i.e. sustain-able production systems) increases an already enor-mous research and development challenge. Fewsystems have sustained increases of over 2 per-cent per year, and these have often been at theexpense of resource degradation.

2. Sources of increased rainfed land are limited, andthe rate of increase of irrigated land has slowedconsiderably because of rising costs and the threatof long-term salinization. Therefore, productionincreases must come from yield increases. Howdifficult will it be to get 250 bushels per acre ofcorn or to increase the average irrigated rice yieldsin developing countries from 3.5 tons per hectareto 7 tons per hectare? Doubling sorghum yieldsin the Sahel from 500 kilograms to 1 metric tonper hectare is not going to help much in meetingglobal food security needs no matter how impor-tant it is to the Sahel.

3. The mix of crops will need to change to producemore tradable surpluses which are transportableand storable. Further, the increased foreign ex-change earnings required by developing countries

23

for imports require a much more open tradingsystem than we now have, even given the advancesmade during the Uruguay Round of negotiationsof the General Agreement on Tariffs and Trade(GATI).

4. Even increasing production does not solve themalnutrition problem which will surely grow. It isa problem of access and income.

Further, we must recognize that the agriculturalproductivity issue is not just an issue of food suppliesor even of biological food security. Let me make threequick, but very important, points in this regard. First,in the poorest countries of the world, the agriculturalsector remains the most important, both in terms ofemployment and income generation. Increased produc-tivity in subsistence and smallholder agriculture is apowerful engine of labor intensive growth, income im-provement, and better access to food. It is a majorcontributor to poverty alleviation and equity improve-ment.

Second, more of the poorest of the poor and themalnourished currently live in low-potential areas thanin high-potential areas, and rural numbers far exceedurban numbers so far (Pinstrup-Andersen and Pandya-Lorch 1994b). Thus, improvement in productivity inagriculture in both low- and high-potential areas hasthe multiple impact of increasing production, reducingpoverty, reducing malnutrition, and generating growth,thereby improving food security broadly defined. Agri-cultural development is not just increasing cereal yields.Further, for the growing number of urban poor, ever-declining real food prices are a positive contribution toreducing malnutrition and poverty.

24

Third, increased yield per unit of land, particu-larly biological yield increases, reduces pressure onfragile environments. Feder and Keck argue that .... ev-ery 0.1 percent of yield increase in the period 2010 to2025 'substitutes' for about 25 million hectares ofrainfed cropland" (p. 22). Further, given that agricul-tural production systems are dominant users of thearable landscape, attention to environmental issues inthe development of sustainable production systems isan indispensable component of any successful futurestrategy.

Concluding Comments

The frightening part of this story to me is that,while the challenge just outlined is, in my view, criticaland immediate, funds to support agricultural develop-ment and productivity improvement are being reducedin developed countries, and aid agencies and interna-tional development institutions are reducing the shareof resources going toward agriculture. This trend ismade worse by the overall decline in development as-sistance. Even the interest of developing country gov-ernments in agricultural development appears to be insteep decline. There is at least a twenty-year lag be-tween initiating strategic research and significant in-creases in production in farmers' fields. Twenty yearsfrom now there will be 2 billion more people to feedand most of them will be in developing country cities.To not recognize the challenge and increase efforts isbad enough, but it is much worse to allow existingresearch capacity to erode.

25

Explanations for the apparent neglect of a criticalproblem abound: the short time span of attention ofpoliticians; perceptions of over-production and surplusesin rich countries; protectionist domestic agriculturalpolicies, which reduce incentive prices in developingcountries; aid fatigue; fiscal crises in countries of theOrganisation for Economic Co-operation and Develop-ment, the end of the Cold War, which reduced theurgency for development assistance; and on and on.Regardless of the reason, the consequences of notaddressing these issues now will clearly have seriousfuture consequences. Unfortunately, extremely pessi-mistic or optimistic scenarios, both of which must bequestioned, detract us from serious debate on thiscritical issue. The efforts of IFPRI through their 2020Vision for Food, Agriculture, and the Environment are,therefore, to be applauded as they seek a broadly ac-cepted consensus on the chaflenges ahead. A recentIFPRI paper by Peter Hazell entitled "Prospects for aWell-Fed World" begins to move us on that direction.More than anything, the global community needs abalanced and reasonable analysis upon which to basecritical future decisions.

26

Bibliography

Brown, Lester R., and Hal Kane. 1994. Full House:Reassessing the Earth's Population Carrying Ca-pacity. The Worldwatch Environmental AlertSeries. New York: W.W. Norton and Company.

Carruthers, Ian. 1993. "Going, Going, Gone! Tropi-cal Agriculture as We Knew It." Tropical Agri-culture Association Newsletter (United Kingdom)13(3): 1-5.

Crosson, Pierre, and Jock R. Anderson. 1992. "Resourcesand Global Food Prospects." Technical PaperNumber 184. World Bank, Washington, D.C.

Feder, Gershon, and Andrew Keck. 1994. 'IncreasingCompetition for Land and Water Resources: AGlobal Perspective." Paper presented at work-shop, Social Science Methods in AgriculturalSystems: Coping with Increasing Resource Com-petition in Asia, held May 2-4, in Chiang Mai,Thailand. Agriculture and Natural ResourcesDepartment, World Bank, Washington, D.C.

Food and Agriculture Organization of the United Na-tions. 1993. "Agriculture: Towards 2010." Con-ference Paper C-93/24. Rome, Italy.

Hazell, Peter. 1994. 'Prospects for a Well-Fed World."Paper presented for roundtable meeting, Popu-lation and Food in the Early Twenty-First Cen-tury: Meeting Future Food Needs of an Increas-ing World Population, held February 14-16.International Food Policy Research Institute,Washington, D.C.

27

Kaplan, Robert D. 1994. "The Coming Anarchy." TheAtlantic Monthly 273(2):44-76.

Mitchell, Donald O., and Merlinda D. Ingco. 1993."The World Food Outlook." Draft paper. Inter-national Economics Department, World Bank,Washington, DC.

Paddock, William, and Paul Paddock. 1976. Famine-1975! Boston: Little, Brown and Company.

Pinstrup-Andersen, Per, and Rajul Pandya-Lorch.1994a. "Alleviating Poverty, Intensifying Agri-culture, and Effectively Managing Natural Re-sources." Food, Agriculture, and the Environ-ment Discussion Paper Number 1. InternationalFood Policy Research Institute, Washington, DC.

1994b. "Poverty, Agricultural Intensification,and the Environment." Paper presented at theTenth Annual General Meeting of the PakistanSociety of Development Economists, held April2-5, in Islamabad. International Food PolicyResearch Institute, Washington, DC.

Plucknett, Donald L. 1993. "Science and AgriculturalTransformation." IFPRI Lecture Series Number1. Washington, DC: International Food PolicyResearch Institute.

Rosegrant, Mark W., and Mercedita Agcaoili. 1994."Global and Regional Food Demand, Supply, andTrade Prospects to 2010." Paper presented atroundtable meeting, Population and Food in theEarly Twenty-First Century: Meeting FutureFood Needs of an Increasing World Population,held February 14-16. International Food PolicyResearch Institute, Washington, DC.

28

Alex F. McCalla is Chair of the Technical AdvisoryCommittee (TAC) of the CGIAR and Director of theAgriculture and Natural Resources Department of theWorld Bank. Dr. McCalla is also Professor Emeritus ofAgricultural Economics at the University of California,Davis.

Dr. McCalla's service at the University of Califor-nia, Davis spanned three decades and encompassed anumber of positions, including Dean of the College ofAgricultural and Environmental Sciences, AssociateDirector of the California Agricultural Experiment Sta-tion, and Founding Dean of the Graduate School ofAdministration. He was also Professor of AgriculturalEconomics and Professor of Management.

Dr. McCalla was Study Director of the first reviewof the CGIAR. He was appointed as a Member of TACin 1984, and has served as TAC Chair since 1988.

Dr. McCalla received a Bachelor of Science degreein Agriculture and a Master of Arts degree in Econom-ics from the University of Alberta in Canada and aDoctor of Philosophy degree in Agricultural Economicsfrom the University of Minnesota. His awards includea Ford Foundation Travel and Study Award to reviewworld food policy research on a global basis and beingnaamed a Fellow of the American Agricultural Econom-ics Association.

® Printed on Recycled Paper

29