Ifpri global water outlook to 2025 averting an impending crisis

REPORTFOOD POLICY

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE

sustainable options for ending hunger and poverty

GLOBAL WATER OUTLOOK TO 2025Averting an Impending CrisisMark W. Rosegrant, Ximing Cai, and Sarah A.Cline

THE INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE (IFPRI)

IFPRI was established in 1975 to identify and analyze national and international strategies and policies for

meeting the food needs of the developing world on a sustainable basis, with particular emphasis on low-income

countries and poor people; to make the results of its research available to all those in a position to use them; and

to help strengthen institutions conducting research and applying research results in developing countries.

While IFPRI’s research is geared to the precise objective of contributing to the reduction of hunger and

malnutrition, the factors involved are many and wide-ranging, requiring analysis of underlying processes and

extending beyond a narrowly defined food sector.The Institute’s research programs reflect worldwide

collaboration with governments and private and public institutions interested in increasing food production and

improving the equity of its distribution.

A 2020 VISION FOR FOOD,AGRICULTURE,AND THE ENVIRONMENT is an initiative of IFPRI launched in 1993 in

collaboration with partners around the world.The 2020 Vision Initiative seeks to develop and promote a shared

vision for how to meet the world’s food needs while reducing poverty and protecting the environment and seeks

to generate information and encourage debate to influence action by all relevant parties.

The 2020 Vision Initiative gratefully acknowledges support for its general activities from the following donors:

Canadian International Development Agency (CIDA), Danish International Development Agency (DANIDA),

Swedish International Development Cooperation Agency (SIDA), and Swiss Agency for Development

Cooperation (SDC).

THE INTERNATIONAL WATER MANAGEMENT INSTITUTE (IWMI)

IWMI’s mission is: improving water and land resources management for food, livelihoods, and nature. IWMI

works to: (1) identify the larger issues related to water management and food security that need to be

understood and addressed by governments and policymakers; (2) develop, test, and promote management

practices and tools that can be used by governments and institutions to manage water and land resources more

effectively and address water scarcity issues; (3) clarify the link between poverty and access to water and help

governments and the research community better understand the specific water-related problems of poor

people; and (4) help developing countries build their research capacities to deal with water scarcity and related

food security issues.

IWMI’s agenda is to put research into action through five research themes: Integrated Water Resources

Management for Agriculture, Sustainable Smallholder Land and Water Management, Sustainable Groundwater

Management,Water Resource Institutions and Policies, and Water, Health, and Environment.

The Institute leads three international programs that bring together the expertise of a number of Future

Harvest Centers and other partners, including national research systems and NGOs: the Comprehensive

Assessment of Water Management in Agriculture; the SIMA Malaria and Agriculture initiative; and the CGIAR

Challenge Program on Water and Food.

FUTURE HARVEST

IFPRI and IWMI are two of 16 food and environmental research organizations known as the Future Harvest

Centers.The centers, located around the world, conduct research in partnership with farmers, scientists, and

policymakers to help alleviate poverty and increase food security while protecting the natural resource base.The

Future Harvest Centers are principally funded by governments, private foundations, and regional and

international organizations, most of which are members of the Consultative Group on International Agricultural

Research (CGIAR).

Global Water Outlook to 2025Averting an Impending Crisis

A 2020 Vision for Food, Agriculture, and the Environment InitiativeInternational Food Policy Research Institute

Washington, D.C., U.S.A.

International Water Management InstituteColombo, Sri Lanka

September 2002

Mark W. RosegrantXiming Cai

Sarah A. Cline

The views expressed in this document are those of the authors and are not necessarily endorsed by or representative of thecosponsoring or supporting organizations.

Copyright © 2002 International Food Policy Research Institute. All rights reserved. Sections of this report may be reproducedwithout the express permission of but with acknowledgment to the International Food Policy Research Institute.

ISBN 0-89629-646-6

ContentsPreface v

Introduction 1

A Thirsty World 2

Alternative Futures for Water 4

Consequences of Key Policy Changes 17

Implications for the Future 22

Notes 25

Box:The IMPACT-WATER Model 26

Preface

For some time, experts have argued about the Earth’s capacity to support ever larger human popu-lations. Can the Earth produce enough food to feed 8 billion people? 10 billion? It now appearsthat one of the main factors limiting future food production will be water. This scarce resource is

facing heavy and unsustainable demand from users of all kinds, and farmers increasingly have to competefor water with urban residents and industries. Environmental uses of water, which may be key toensuring the sustainability of the Earth’s water supply in the long run, often get short shrift.

Based on a global model of supply and demand for food and water, this report shows that if currentwater policies continue, farmers will indeed find it difficult to meet the world’s food needs. Hardest hitwill be the world’s poorest people. The results from the model used in this report also show the conse-quences of changing the course of water policy. Further inattention to water-related investments andpolicies will produce a severe water crisis, which will lead in turn to a food crisis. A commitment tosustainable use of water, through appropriate policies and investments, however, will lead to a morewater- and food-secure world. Water may be a scarce resource, but humans have developed many waysof using it more efficiently—that is, getting more from each unit of water. But water-saving policies,practices, and technologies are of no help if they are not used. Inappropriate incentives and institutionsoften hinder effective use of water. This report spells out the future results of our current choices.

Many more details about the scenarios described in this report are available in a book called WorldWater and Food to 2025: Dealing with Scarcity, by Mark W. Rosegrant, Ximing Cai, and Sarah A. Cline, alsoavailable from IFPRI.

v

GLO

BA

L W

AT

ER O

UT

LOO

K T

O 2

025

Demand for the world’s increasingly scarce water supply is rising rapidly, challenging its

availability for food production and putting global food security at risk. Agriculture,

upon which a burgeoning population depends for food, is competing with industrial,

household, and environmental uses for this scarce water supply. Even as demand for water

by all users grows, groundwater is being depleted, other water ecosystems are becoming

polluted and degraded, and developing new sources of water is getting more costly.

Introduction

Will there be enough water to grow food forthe almost 8 billion people expected to populatethe Earth by 2025? It is impossible to answerthat question without an understanding of theevolving relationship between water availabilityand food production.This understanding willallow decisionmakers to look squarely at theconsequences of the choices they make tobalance water supply and demand among all usersin the years to come.

To spell out these consequences, we havedeveloped a global model (see box on page 26) ofwater and food supply and demand to addressthe following questions: How are water avail-ability and water demand likely to evolve by theyear 2025? What impact will various water poli-cies and investments have on water availability forthe environment, domestic and industrial uses,and irrigation? What steps can policymakers taketo ensure a sustainable use of water that meetsthe world’s food needs?

GLO

BA

L W

AT

ER O

UT

LOO

K T

O 2

025

1

Water development underpins food security, people’s livelihoods, industrial growth, and

environmental sustainability throughout the world. In 1995 the world withdrew

3,906 cubic kilometers (km3) of water for these purposes (Figure 1). By 2025 water with-

drawal for most uses (domestic, industrial, and livestock) is projected to increase by at least

50 percent. This will severely limit irrigation water withdrawal, which will increase by only 4

percent, constraining food production in turn.

A Thirsty World

About 250 million hectares areirrigated worldwide today, nearly fivetimes more than at the beginning ofthe 20th century. Irrigation has helpedboost agricultural yields and outputsand stabilize food production andprices. But growth in population andincome will only increase the demandfor irrigation water to meet foodproduction requirements (Figure 2).Although the achievements of irriga-tion have been impressive, in manyregions poor irrigation managementhas markedly lowered groundwatertables, damaged soils, and reducedwater quality.

Water is also essential fordrinking and household uses and forindustrial production. Access to safedrinking water and sanitation is criticalto maintain health, particularly for chil-dren. But more than 1 billion peopleacross the globe lack enough safewater to meet minimum levels ofhealth and income. Although thedomestic and industrial sectors use farless water than agriculture, the growthin water consumption in these sectorshas been rapid. Globally, withdrawalsfor domestic and industrial usesquadrupled between 1950 and 1995,

5,000

4,000

3,000

2,000

1,000

0

Figure 1 Total water withdrawal by region, 1995 and 2025

Asi

a

Wor

ld

Lati

n A

mer

ica

Sub-

Saha

ran

Afr

ica

Wes

t Asi

a/

Nor

th A

fric

a

SOURCE: Authors' estimates and IMPACT-WATER projections, June 2002.NOTE: Projections for 2025 are for the business as usual scenerio.

Dev

elop

ed

coun

trie

s

19952025

Cub

ic k

ilom

eter

s

Dev

elop

ing

coun

trie

s

1,600

1,400

1,200

1,000

800

600

400

200

0

Figure 2 Total irrigation water consumption by region, 1995 and 2025

Asi

a

Wor

ld

Lati

n A

mer

ica

Sub-

Saha

ran

Afr

ica

Wes

t Asi

a/

Nor

th A

fric

a


Dev

elop

ed

coun

trie

s

19952025

Cub

ic k

ilom

eter

s

Dev

elop

ing

coun

trie

s

GLO

BA

LW

AT

ERO

UT

LOO

K T

O 2

025

2

compared with agricultural uses, for which with-drawals slightly more than doubled.1

Water is integrally linked to the health of theenvironment.Water is vital to the survival ofecosystems and the plants and animals that live inthem, and in turn ecosystems help to regulate thequantity and quality of water. Wetlands retainwater during high rainfall, release it during dry

periods, and purify it of many contaminants.Forests reduce erosion and sedimentation ofrivers and recharge groundwater. The importanceof reserving water for environmental purposes hasonly recently been recognized: during the 20thcentury, more than half of the world’s wetlandswere lost.2

GLO

BA

LW

AT

ERO

UT

LOO

K T

O 2

025

3

Business As Usual ScenarioIn the business as usual scenario current trends inwater and food policy, management, and investmentremain as they are. International donors andnational governments, complacent about agricul-ture and irrigation, cut their investments in thesesectors. Governments and water users implementinstitutional and management reforms in a limitedand piecemeal fashion.These conditions leave theworld ill prepared to meet major challenges to thewater and food sectors.

Over the coming decades the area of landdevoted to cultivating food crops will grow slowlyin most of the world because of urbanization, soildegradation, and slow growth in irrigation invest-ment, and because a high proportion of arable landis already cultivated. Moreover, steady or decliningreal prices for cereals will make it unprofitable forfarmers to expand harvested area. As a result,greater food production will depend primarily onincreases in yield. Yet growth in crop yields will alsodiminish because of falling public investment in agri-cultural research and rural infrastructure.Moreover, many of the actions that produced yieldgains in recent decades, such as increasing thedensity of crop planting, introducing strains that aremore responsive to fertilizer, and improvingmanagement practices, cannot easily be repeated.

In the water sector, the management of riverbasin and irrigation water will become more effi-cient, but slowly. Governments will continue totransfer management of irrigation systems tofarmer organizations and water-user associations.Such transfers will increase water efficiency whenthey are built upon existing patterns of coopera-tion and backed by a supportive policy and legalenvironment. But these conditions are oftenlacking.

In some regions farmers will adopt more effi-cient irrigation practices. Economic incentives toinduce more efficient water management,however, will still face political opposition fromthose concerned about the impact of higherwater prices on farmers’ income and fromentrenched interests that benefit from existingsystems of allocating water.Water managementwill also improve slowly in rainfed agriculture as aresult of small advances in water harvesting,better on-farm management techniques, and thedevelopment of crop varieties with shortergrowing seasons.

Public investment in expanding irrigationsystems and reservoir storage will decline as thefinancial, environmental, and social costs of buildingnew irrigation systems escalate and the prices ofcereals and other irrigated crops drop.

The future of water and food is highly uncertain. Some of this uncertainty is due to rela-

tively uncontrollable factors such as weather. But other critical factors can be influ-

enced by the choices made collectively by the world’s people. These factors include income

and population growth, investment in water infrastructure, allocation of water to various

uses, reform in water management, and technological changes in agriculture. Policy deci-

sions—and the actions of billions of individuals—determine these fundamental, long-term

drivers of water and food supply and demand.

To show the very different outcomes that policy choices produce, we present three alter-

native futures for global water and food, followed by an assessment of specific policy options.3

Alternative Futures for Water

4

GLO

BA

L W

AT

ER O

UT

LOO

K T

O 2

025

5

Nevertheless, where benefits outweigh costs, manygovernments will construct dams, and reservoirwater for irrigation will increase moderately.

With slow growth in irrigation from surfacewater, farmers will expand pumping from ground-water, which is subject to low prices and littleregulation. Regions that currently pump ground-water faster than aquifers can recharge, such as thewestern United States, northern China, northernand western India, Egypt, and West Asia and NorthAfrica, will continue to do so.

The cost of supplying water to domestic andindustrial users will rise dramatically. Better

delivery and more efficient home water use willlead to some increase in the proportion of house-holds connected to piped water. Many households,however, will remain unconnected. Small priceincreases for industrial water, improvements inpollution control regulation and enforcement, andnew industrial technologies will cut industrial wateruse intensity (water demand per $1,000 of grossdomestic product). Yet industrial water prices willremain relatively low and pollution regulations willoften be poorly enforced.Thus, significant potentialgains will be lost.

Environmental and other interest groups willpress to increase the amount of waterallocated to preserving wetlands,diluting pollutants, maintaining riparianflora and other aquatic species, andsupporting tourism and recreation.Yetbecause of competition for water forother uses, the share of water devotedto environmental uses will notincrease.

The Water Situation Almost all users will place heavydemands on the world’s water supplyunder the business as usual scenario.Total global water withdrawals in 2025are projected to increase by 22percent above 1995 withdrawals, to4,772 km3 (see Figure 1, page 2).4

Projected withdrawals in developingcountries will increase 27 percentover the 30-year period, while devel-oped-country withdrawals willincrease by 11 percent.5

Together, consumption of water fordomestic, industrial, and livestockuses—that is, all nonirrigation uses—will increase dramatically, rising by 62percent from 1995 to 2025 (Figure 3).Because of rapid population growth andrising per capita water use (Figure 4),total domestic consumption willincrease by 71 percent, of which more

60

50

40

30

20

10

0

Figure 4 Per capita domestic water consumption by region, 1995 and 2025

Asi

a

Wor

ld

Lati

n A

mer

ica

Sub-

Saha

ran

Afr

ica

Wes

t Asi

a/

Nor

th A

fric

a


Dev

elop

ed

coun

trie

s

19952025

Cub

ic m

eter

s/pe

rson

/yea

r

Dev

elop

ing

coun

trie

s

600

500

400

300

200

100

0

Figure 3 Total nonirrigation water consumption by region, 1995 and 2025

Asi

a

Wor

ld

Lati

n A

mer

ica

Sub-

Saha

ran

Afr

ica

Wes

t Asi

a/

Nor

th A

fric

a


Dev

elop

ed

coun

trie

s

19952025

Cub

ic k

ilom

eter

s

Dev

elop

ing

coun

trie

s

GLO

BA

L W

AT

ER O

UT

LOO

K T

O 2

025

than 90 percent will be in developing countries.Conservation and technological improvements willlower per capita domestic water use in developedcountries with the highest per capita waterconsumption.

Industrial water use will grow much faster indeveloping countries than in developed countries.In 1995 industries in developed countriesconsumed much more water than industries in thedeveloping world. By 2025, however, developing-world industrial water demand is projected toincrease to 121 km3, 7 km3 greater than in thedeveloped world (Figure 5). The intensity of indus-

trial water use will decrease worldwide, especiallyin developing countries (where initial intensitylevels are very high), thanks to improvements inwater-saving technology and demand policy.Nonetheless, the sheer size of the increase in theworld’s industrial production will still lead to anincrease in total industrial water demand.

Direct water consumption by livestock is verysmall compared with other sectors. But the rapidincrease of livestock production, particularly indeveloping countries, means that livestock waterdemand is projected to increase 71 percentbetween 1995 and 2025. Whereas livestock water

demand will increase only 19 percentin the developed world between1995 and 2025, it is projected tomore than double in the developingworld, from 22 to 45 km3.

Although irrigation is by far thelargest user of the world’s water, useof irrigation water is projected to risemuch more slowly than othersectors. For irrigation water, we havecomputed both potential demand andactual consumption. Potential demandis the demand for irrigation water inthe absence of any water supplyconstraints, whereas actual consump-tion of irrigation water is the realizedwater demand, given the limitations ofwater supply for irrigation (Figure 6).The proportion of potential demandthat is realized in actual consumptionis the irrigation water supply relia-bility index (IWSR).6 An IWSR of 1.0would mean that all potential demandis being met.

Potential irrigation demand willgrow by 12 percent in developingcountries, while it will actually declinein developed countries by 1.5 percent.The fastest growth in potentialdemand for irrigation water will occurin Sub-Saharan Africa, with an increaseof 27 percent, and in Latin America,

2,500

2,000

1,500

1,000

500

0

Figure 6 Potential and actual consumption of irrigation water, 1995 and 2025

World


Developed countries

Cub

ic k

ilom

eter

s

Developingcountries

1995 Potential irrigation consumption2025 Potential irrigation consumption1995 Actual irrigation consumption2025 Actual irrigation consumption

2,500

2,000

1,500

1,000

500

0

Figure 5 Water consumption by sector, 1995 and 2025

World


Developed countries

LivestockIrrigation

Cub

ic k

ilom

eter

s

Developingcountries

DomesticIndustrial

1995 1995 199520252025 2025

GLO

BA

LW

AT

ERO

UT

LOO

K T

O 2

025

6

with an increase of 21 percent. Each of theseregions has a high percentage increase in irrigatedarea from a relatively low 1995 level. India isprojected to have the highest absolute growth inpotential irrigation water demand, 66 km3 (17percent), owing to relatively rapid growth in irri-gated area from an already high level in 1995. WestAsia and North Africa will increase by 18 percent(28 km3, mainly in Turkey), while China will experi-ence a much smaller increase of 4 percent (12 km3).In Asia as a region, potential irrigation water demandwill increase by 8 percent (100 km3).

Water scarcity for irrigation will intensify, withactual consumption of irrigation water worldwideprojected to grow more slowly than potentialconsumption, increasing only 4 percent between1995 and 2025. In developing countries a decliningfraction of potential demand will be met over time.The IWSR for developing countries will declinefrom 0.81 in 1995 to 0.75 in 2025, and in dry riverbasins the decline will be steeper (Table 1). Forexample, in the Haihe River Basin in China, which isan important wheat and maize producer and servesmajor metropolitan areas, the IWSR is projected todecline from 0.78 to 0.62, and in the Ganges ofIndia, the IWSR will decline from 0.83 to 0.67.

In the developed world, the situation is thereverse: the supply of irrigation water is projectedto grow faster than potential demand (althoughcertain basins will face increasing water scarcity).

Increases in river basinefficiency will more thanoffset the very smallincrease in irrigated area.As a result, after initiallydeclining from 0.87 to0.85 in 2010, the IWSRwill improve to 0.90 in2025 thanks to slowinggrowth of domestic andindustrial demand (andactual declines in totaldomestic and industrialwater use in the UnitedStates and Europe) and

more efficient use of irrigation water.

The Food SituationWater scarcity under business as usual will lead toslower growth of food production and substantialshifts in where the world’s food is grown.

Farmers will find themselves unable to raisecrop yields as quickly as in the past in the face of adecline in relative water supply.The global yieldgrowth rate for all cereals is projected to declinefrom 1.5 percent per year from 1982 to 1995 to1.0 percent per year from 1995 to 2025. In devel-oping countries, average crop yield growth willdecline from 1.9 percent per year to 1.2 percent.

The projected relative crop yields for irrigatedcereals show that scarce water is a significant causeof the slowdown in cereal yield growth in devel-oping countries. Relative crop yield is the ratio ofthe actual projected crop yield to the economicallyattainable yields at given crop and input prices underconditions of zero water stress. The relative cropyield for cereals in irrigated areas in developingcountries is projected to decline from 0.86 in 1995to 0.75 in 2025 (Figure 7).This fall in the relativecrop yield represents an annual loss in crop yieldsforgone due to increased water stress of 0.68metric tons per hectare in 2025, or an annual loss ofcereal production of 130 million metric tons, equiva-lent to the annual rice crop in China in late 1990sand double the U.S. wheat crop in the same period.7

Table 1 Irrigation water supply reliability by region, 1995 and 2025

IRRIGATION WATER SUPPLY RELIABILITY INDEX

REGION 1995 2025

Asia 0.81 0.76Latin America 0.83 0.75Sub-Saharan Africa 0.73 0.72West Asia/North Africa 0.78 0.74Developed countries 0.87 0.90Developing countries 0.81 0.75World 0.82 0.78

SOURCE: Authors’ estimates and IMPACT-WATER projections, June 2002.

NOTE: Projections for 2025 are for the business as usual scenario.

GLO

BA

L W

AT

ER O

UT

LOO

K T

O 2

025

7

Crop-harvested area is expected to grow evenmore slowly than crop yield in the comingdecades. Cereal-harvested area will rise by only64 million hectares by 2025, from 687 millionhectares in 1995. All of this growth is projected tooccur in developing countries, with a slight declinein cereal-harvested area in developed countries.

Growth in food demand will be concentratedin developing countries, where rising incomes andrapid urbanization will also cause people tochange the kinds of food they demand.

Consumers will shift from maize and coarse grainsto wheat and rice, livestock products, and fruitsand vegetables. In much of Asia, an additional shiftwill occur from rice to wheat.The projectedstrong growth in meat consumption, in turn, willsubstantially increase cereal consumption in theform of animal feed, particularly maize. Totalworld cereal demand is projected to grow by 828million tons, or 47 percent.

With slowing production growth, the pricesof most food commodities are projected to

decline, but far more slowly than in thepast two decades. Maize prices willincrease slightly, while rice, wheat, andother cereals will all decline in price.Rice prices show the biggest decline, adrop of $64 per ton, or 22 percent,between 1995 and 2025, but this is stillfar below the rate of decline in thepast three decades. Real world pricesof wheat, rice, and maize fell by 47, 59,and 61 percent respectively, between1970 and 2000.

Under the business as usualscenario, irrigated and rainfed produc-tion will each account for about one-half the increase in production between1995 and 2025 (Figure 8).The largecontribution to production fromrainfed areas may surprise someobservers. But more than 80 percentof cereal area in developed countries israinfed, and much of this area is highlyproductive maize and wheat land. Theaverage rainfed cereal yield in devel-oped countries was 3.2 tons perhectare in 1995, virtually as high as irri-gated cereal yields in developing coun-tries, and is projected to grow to 4.0tons per hectare by 2025. Moreover,whereas rainfed cereal yields in devel-oping countries are projected toincrease only from 1.5 tons per hectareto 2.1 tons per hectare by 2025, rainfedarea in developing countries will

1.2

1.0

0.8

0.6

0.4

0.2

0

Figure 7 Relative cereal yield by region, 1995 and 2025

Asi

a

Wor

ld

Lati

n A

mer

ica

Sub-

Saha

ran

Afr

ica

Wes

t Asi

a/

Nor

th A

fric

a


Dev

elop

ed

coun

trie

s

19952025

Dev

elop

ing

coun

trie

s

Figure 8 Share of irrigated and rainfed production in cereal production increase, 1995 and 2025


Rainfed developingcountries

30%

Irrigated developed countries

11%

Irrigated developing countries

39%

Rainfed developed countries

20%

GLO

BA

LW

AT

ERO

UT

LOO

K T

O 2

025

8

account for 62 percent of total cereal area indeveloping countries (Figure 9).

By substituting cereal and other food importsfor irrigated agricultural production (so-calledimports of virtual water), countries can effectivelyreduce their agricultural water use.8 Under busi-ness as usual, developing countries will dramaticallyincrease their reliance on food imports from 107million tons in 1995 to 245 million tons in 2025.The increase in developing-country cereal importsby 138 million tons between 1995 and 2025 is theequivalent of saving 147 km3 of water at 2025water productivity levels, or 8 percent of totalwater consumption and 12 percent of irrigationwater consumption in developing countries in 2025.

The water (and land) savings from the projectedlarge increases of food imports by the developingcountries are particularly beneficial if they are theresult of strong economic growth that generatesthe necessary foreign exchange to pay for the foodimports. But even when rapidly growing foodimports are primarily a result of rapid incomegrowth, national policymakers concerned aboutheavy reliance on world markets often see them asa signal to set trade restrictions that can slowgrowth and food security in the longer term. Moreserious food security problems arise when highfood imports are the result of slow agricultural andeconomic development that fails to keep pace withbasic food demand driven by population and

income growth. Under these condi-tions, countries may find it impossibleto finance the required imports on acontinuing basis, causing a furtherdeterioration in the ability to bridgethe gap between food consumption andthe food required for basic livelihood.

“Hot spots” for food trade gapsare Sub-Saharan Africa, where cerealimports are projected to more thantriple by 2025 to 35 million tons, andWest Asia and North Africa, wherecereal imports are projected toincrease from 38 million tons in 1995to 83 million tons in 2025.The relianceon water-saving cereal imports in WestAsia and North Africa makeseconomic and environmental sense,but it must be supported by fasternonagricultural growth. It is highlyunlikely that Sub-Saharan Africa couldfinance the projected level of importsinternally; instead international finan-cial or food aid would be required.Failure to finance these imports wouldfurther increase food insecurity andpressure on water resources in thisregion.

7

6

5

4

3

2

1

0

Figure 9 Irrigated and rainfed cereal yield by region, 1995 and 2025

Asi

a

Wor

ld

Lati

n A

mer

ica

Sub-

Saha

ran

Afr

ica

Wes

t Asi

a/

Nor

th A

fric

a


Dev

elop

ed

coun

trie

s

19952025

Met

ric

tons

/hec

tare

Dev

elop

ing

coun

trie

s

Irrigated yield

4.5

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5

0

Asi

a

Wor

ld

Lati

n A

mer

ica

Sub-

Saha

ran

Afr

ica

Wes

t Asi

a/

Nor

th A

fric

a

Dev

elop

ed

coun

trie

s

19952025

Met

ric

tons

/hec

tare

Dev

elop

ing

coun

trie

s

Rainfed yield

GLO

BA

LW

AT

ERO

UT

LOO

K T

O 2

025

9

Water Crisis ScenarioA moderate worsening of many of the currenttrends in water and food policy and in investmentcould build to a genuine water crisis. In the watercrisis scenario, government budget problemsworsen. Governments further cut their spendingon irrigation systems and accelerate the turnoverof irrigation systems to farmers and farmer groupsbut without the necessary reforms in water rights.Attempts to fund operations and maintenance inthe main water system, still operated by publicagencies, cause water prices to irrigators to rise.Water users fight price increases, and conflictspills over to local management and cost-sharingarrangements. Spending on the operation andmaintenance of secondary and tertiary systemsfalls dramatically, and deteriorating infrastructureand poor management lead to falling water useefficiency. Likewise, attempts to organize riverbasin organizations to coordinate water manage-ment fail because of inadequate funding and highlevels of conflict among water stakeholders withinthe basin.

National governments and internationaldonors will reduce their investments in cropbreeding for rainfed agriculture in developingcountries, especially for staple crops such as rice,wheat, maize, other coarse grains, potatoes,cassava, yams, and sweet potatoes. Private agricul-tural research will fail to fill the investment gap forthese commodities. This loss of research fundingwill lead to further declines in productivity growthin rainfed crop areas, particularly in more marginalareas. In search of improved incomes, people willturn to slash-and-burn agriculture, thereby defor-esting the upper watersheds of many basins.Erosion and sediment loads in rivers will rise, inturn causing faster sedimentation of reservoirstorage. People will increasingly encroach onwetlands for both land and water, and the integrityand health of aquatic ecosystems will be compro-mised. The amount of water reserved for environ-mental purposes will decline as unregulated andillegal withdrawals increase.

The cost of building new dams will soar,discouraging new investment in many proposeddam sites. At other sites indigenous groups andnongovernmental organizations (NGOs) willmount opposition, often violent, over the environ-mental and human impacts of new dams. Theseprotests and high costs will virtually halt newinvestment in medium and large dams and storagereservoirs. Net reservoir storage will decline indeveloping countries and remain constant indeveloped countries.

In the attempt to get enough water to growtheir crops, farmers will extract increasingamounts of groundwater for several years, drivingdown water tables. But because of the acceler-ated pumping, after 2010 key aquifers in northernChina, northern and northwestern India, and WestAsia and North Africa will begin to fail.Withdeclining water tables, farmers will find the cost ofextracting water too high, and a big drop ingroundwater extraction from these regions willfurther reduce water availability for all uses.

As in the business as usual scenario, the rapidincrease in urban populations will quickly raisedemand for domestic water. But governments willlack the funds to extend piped water and sewagedisposal to newcomers. Governments will respondby privatizing urban water and sanitation servicesin a rushed and poorly planned fashion. The newprivate water and sanitation firms will be under-capitalized and able to do little to connect addi-tional populations to piped water. An increasingnumber and percentage of the urban populationmust rely on high-priced water from vendors orspend many hours fetching often-dirty water fromstandpipes and wells.

The Water SituationThe developing world will pay the highest pricefor the water crisis scenario. Total worldwidewater consumption in 2025 will be 261 km3 higherthan under the business as usual scenario—a 13percent increase—but much of this water will bewasted, of no benefit to anyone (Figure 10).G

LOB

AL

WA

TER

OU

TLO

OK

TO

202

5

10

GLO

BA

LW

AT

ERO

UT

LOO

K T

O 2

025

11

Virtually all of the increase will go toirrigation, mainly because farmers willuse water less efficiently and with-draw more water to compensate forwater losses.The supply of irrigationwater will be less reliable, except inregions where so much water isdiverted from environmental uses toirrigation that it compensates for thelower water use efficiency.

For most regions, per capitademand for domestic water will besignificantly lower than under thebusiness as usual scenario, in bothrural and urban areas.The result isthat people will not have access tothe water they need for drinking andsanitation. The total domestic demandunder the water crisis scenario will be162 km3 in developing countries, 28percent less than under business asusual; 64 km3 in developed countries,7 percent less than under business asusual; and 226 km3 in the world, 23percent less than under business asusual (FIgure 11).

Demand for industrial water, onthe other hand, will increase, owing tofailed technological improvements andeconomic measures. In 2025 the totalindustrial water demand worldwidewill be 80 km3 higher than under thebusiness as usual scenario—a 33percent rise—without generatingadditional industrial production.

With water diverted to make upfor less efficient water use, the watercrisis scenario will hit environmentaluses particularly hard. Comparedwith business as usual, environmentalflows will drop significantly by 2025,with 380 km3 less environmental flowin the developing world, 80 km3 lessin the developed world, and 460 km3

less globally.

350

300

250

200

150

100

50

0

Figure 11 Domestic consumption of water by region, business as usual and crisis scenarios, 2025

Asi

a

Wor

ld

Lati

n A

mer

ica

Sub-

Saha

ran

Afr

ica

Wes

t Asi

a/

Nor

th A

fric

a

SOURCE: Authors' estimates and IMPACT-WATER projections, June 2002.

Dev

elop

ed

coun

trie

s

Business as usualCrisis

Cub

ic k

ilom

eter

s

Dev

elop

ing

coun

trie

s

2,500

2,000

1,500

1,000

500

0

Figure 10 Total water consumption by region, business as usual and crisis scenarios, 2025

Asi

a

Wor

ld

Lati

n A

mer

ica

Sub-

Saha

ran

Afr

ica

Wes

t Asi

a/

Nor

th A

fric

a


Dev

elop

ed

coun

trie

s


Cub

ic k

ilom

eter

s

Dev

elop

ing

coun

trie

s

3,000

2,500

2,000

1,500

1,000

500

0

Figure 12 Total cereal production by region, business as usual and crisis scenarios, 2025

Asi

a

Wor

ld

Lati

n A

mer

ica

Sub-

Saha

ran

Afr

ica

Wes

t Asi

a/

Nor

th A

fric

a


Dev

elop

ed

coun

trie

s


Dev

elop

ing

coun

trie

s

Mill

ion

met

ric

tons

The Food SituationThe water crisis scenario will havesevere consequences for food produc-tion.Total cereal production underwater crisis, for example, will be 249million metric tons, or 10 percent, lessthan business as usual—the result ofdeclines in both cultivated area andyields (Figure 12). This reduction is theequivalent of annually losing the entirecereal crop of India, or the combinedannual harvest of Sub-Saharan Africaand West Asia and North Africa.

Compared with business asusual, the total cereal-harvested areaunder the water crisis scenario is17.7 million hectares, or 3 percent,lower than business as usual in thedeveloping world, 8.9 millionhectares, or 4 percent, lower in thedeveloped world, and 26.6 millionhectares, or 4 percent, lower globally

Yields will fall for both irrigatedand rainfed crops. The average totalcereal yield in 2025 is 216 kilogramsper hectare, or 6 percent, lowerunder the water crisis scenario thanunder business as usual. Because thesupply of irrigation water in mostregions is less reliable under thewater crisis scenario, the averageyield of irrigated cereals will be lowerfor developing and developed coun-tries and the world as a whole in2025. Globally, irrigated cereal yieldswill be 4 percent lower under watercrisis than under business as usual.

Since farmers fail to harvest morerainfall and crop research cutbacksslow yield growth, rainfed crops willyield 191 kilograms per hectare lessthan business as usual, a 7 percentdecrease from 1995 to 2025.

The decline in food productionwill help push up food prices sharply

Figure 13 World prices for rice, wheat, and maize, business as usual and crisis scenarios, 1995–2025


450

400

350

300

250

200

150

100

50

0


US$

/met

ric

ton

Rice

1995 2000 2005 2010 2015 2020 2025

300

250

200

150

100

50

0


US$

/met

ric

ton

Wheat

1995 2000 2005 2010 2015 2020 2025

250

200

150

100

50

0


US$

/met

ric

ton

Maize

1995 2000 2005 2010 2015 2020 2025

GLO

BA

LW

AT

ERO

UT

LOO

K T

O 2

025

12

under the water crisis scenario (Figure 13,previous page).The price of rice will rise by 40percent, wheat by 80 percent, maize by 120percent, other coarse grains by 85 percent,soybeans by 70, and potatoes, sweet potatoes, andother roots and tubers by 50 to 70 percent. Cropprices under the water crisis scenario in 2025 are1.8 times that of business as usual for rice, 1.7times for potatoes, 1.6 times for soybeans, andmore than double for all other crops.

These high prices will dampen food demand.Under the water crisis scenario, cereal demandin 2025 will decline by 55 million tons, or 7percent, compared with the business as usualscenario in the developed world, and 192 milliontons, or 11 percent, in the developing world.

Compared with business as usual, net tradewill decline in the water crisis scenario.Developing countries will import 58 million tons,or 23 percent, less cereal than under business asusual.This decline implies that high prices dampencrop demand and thus lead to trade reductions.

The ultimate result of this scenario is growingfood insecurity, especially in developing countries.Per capita cereal consumption in 2025 in the devel-oping world is 2 percent lower than 1995 levels.

This scenario makes it clear that increasingwater scarcity, combined with poor water policiesand inadequate investment in water, has the poten-tial to generate sharp increases in cereal foodprices over the coming decades. Price increases ofthis magnitude will take a significant bite out of thereal income of poor consumers. Malnutrition willincrease substantially, given that the poorestpeople in low-income developing countries spendmore than half their income on food. Sharp priceincreases can also fuel inflation, place severe pres-sure on foreign exchange reserves, and haveadverse impacts on macroeconomic stability andinvestment in developing countries.

Sustainable Water ScenarioA sustainable water scenario would dramaticallyincrease the amount of water allocated to envi-ronmental uses, connect all urban households to

piped water, and achieve higher per capitadomestic water consumption, while maintainingfood production at the levels described in thebusiness as usual scenario. It would achievegreater social equity and environmental protec-tion through both careful reform in the watersector and sound government action.

Governments and international donors willincrease their investments in crop research, tech-nological change, and reform of water manage-ment to boost water productivity and the growthof crop yields in rainfed agriculture. Accumulatingevidence shows that even drought-prone and high-temperature rainfed environments have the poten-tial for dramatic increases in yield. Breedingstrategies will directly target these rainfed areas.Improved policies and increased investment inrural infrastructure will help link remote farmersto markets and reduce the risks of rainfed farming.

To stimulate water conservation and free upagricultural water for environmental, domestic,and industrial uses, the effective price of water tothe agricultural sector will be gradually increased.Agricultural water price increases will be imple-mented through incentive programs that providefarmers income for the water that they save, suchas charge-subsidy schemes that pay farmers forreducing water use, and through the establish-ment, purchase, and trading of water use rights. By2025 agricultural water prices will be twice ashigh in developed countries and three times ashigh in developing countries as in the business asusual scenario.The government will simultane-ously transfer water rights and the responsibilityfor operation and management of irrigationsystems to communities and water user associa-tions in many countries and regions. The transferof rights and systems will be facilitated with animproved legal and institutional environment forpreventing and eliminating conflict and with tech-nical and organizational training and support.As aresult, farmers will increase their on-farm invest-ments in irrigation and water management tech-nology, and the efficiency of irrigation systems andbasin water use will improve significantly.

GLO

BA

LW

AT

ERO

UT

LOO

K T

O 2

025

13

River basin organizations will be established inmany water-scarce basins to allocate mainstreamwater among stakeholder interests. Higherfunding and reduced conflict over water, thanks tobetter water management, will facilitate effectivestakeholder participation in these organizations.

Farmers will be able to make more effectiveuse of rainfall in crop production, thanks tobreakthroughs in water harvesting systems andthe adoption of advanced farming techniques, likeprecision agriculture, contour plowing, precisionland leveling, and minimum-till and no-till tech-nologies. These technologies will increase theshare of rainfall that goes to infiltration and evap-otranspiration.

Spurred by the rapidly escalating costs ofbuilding new dams and the increasingly apparentenvironmental and human resettlement costs,developing and developed countries will reassesstheir reservoir construction plans, with compre-hensive analysis of the costs and benefits,including environmental and social effects, ofproposed projects. As a result, many plannedstorage projects will be canceled, but others willproceed with support from civil society groups.Yet new storage capacity will be less necessarybecause rapid growth in rainfed crop yields willhelp reduce rates of reservoir sedimentationfrom erosion due to slash-and-burn cultivation.

Policy toward groundwater extraction willchange significantly. Market-based approacheswill assign rights to groundwater based on bothannual withdrawals and the renewable stock ofgroundwater. This step will be combined withstricter regulations and better enforcement ofthese regulations. Groundwater overdrafts will bephased out in countries and regions that previ-ously pumped groundwater unsustainably.

Domestic and industrial water use will alsobe subject to reforms in pricing and regulation.Water prices for connected households willdouble, with targeted subsidies for low-incomehouseholds. Revenues from price increases willbe invested to reduce water losses in existingsystems and to extend piped water to previouslyunconnected households. By 2025 all householdswill be connected. Industries will respond tohigher prices, particularly in developing countries,by increasing in-plant recycling of water, whichreduces consumption of water.

With strong societal pressure for improvedenvironmental quality, allocations for environ-mental uses of water will increase. Moreover, thereforms in agricultural and nonagricultural watersectors will reduce pressure on wetlands andother environmental uses of water. Greaterinvestments and better water management willimprove the efficiency of water use, leaving more

water instream for environmentalpurposes. All reductions in domesticand urban water use, due to higherwater prices, will be allocated toinstream environmental uses.

The Water SituationIn the sustainable water scenario theworld consumes less water butreaps greater benefits than underbusiness as usual, especially in devel-oping countries. In 2025 total world-wide water consumption is 408 km3,or 20 percent, lower under thesustainable scenario than under busi-ness as usual (Figure 14). This reduc-

2,500

2,000

1,500

1,000

500

0

Figure 14 Total and irrigation water consumption, by region, business as usual and sustainable scenarios, 2025

World


Cub

ic k

ilom

eter

s

Developing countries

IrrigationOther uses

Business as usual

Developed countries

Business as usual

Business as usual

Sustainable Sustainable Sustainable

GLO

BA

LW

AT

ERO

UT

LOO

K T

O 2

025

14

tion in consumption frees up water for environ-mental uses. Higher water prices and higherwater use efficiency reduces consumption of irri-gation water by 296 km3 compared with businessas usual. The reliability of irrigation water supplyis reduced slightly in the sustainable scenariocompared with business as usual, because thisscenario places a high priority on environmentalflows. Over time, however, more efficient wateruse in this scenario counterbalances the transferof water to the environment and results in animprovement in the reliability of supply of irriga-tion water by 2025.

This scenario will improve the domesticwater supply through universal access to pipedwater for rural and urban households. Globally,potential domestic water demand under thesustainable water scenario will decrease 9percent compared with business as usual, owingto higher water prices. However, potential percapita domestic demand for connected house-holds in rural areas will be 12 percent higherthan that under business as usual in the devel-oping world, and 5 percent higher in the devel-oped world. This increase is accomplished byexpanding universal access to piped water in

rural areas even with higher pricesfor water. And in urban areas,potential per capita water consump-tion for poor households sharplyimproves through connection topiped water, while the initiallyconnected households reduceconsumption in response to higherprices and improved water-savingtechnology (Figure 15).

Through technological improve-ments and effective economic incen-tives, the sustainable water scenariowill reduce industrial water demand.In 2025 total industrial waterdemand worldwide under thesustainable scenario will be 85 km3,or 35 percent, lower than under

business as usual.The environment is a major beneficiary of the

sustainable water scenario, with large increases inthe amount of water reserved for wetlands,instream flows, and other environmentalpurposes. Compared with the business as usualscenario, the sustainable scenario will also resultin an increase in the environmental flow of 850km3 in the developing world, 180 km3 in thedeveloped world, and 1,030 km3 globally. This isthe equivalent of transferring 22 percent of globalwater withdrawals under business as usual toenvironmental purposes.

The Food SituationThe sustainable water scenario can raise foodproduction slightly over the business as usualscenario, while achieving much greater gains fordomestic water use and the environment.

The total harvested area under the sustain-able water scenario in 2025 will be slightly lowerthan under business as usual owing to less waterfor irrigation and barely lower crop prices.Because the supply of irrigation water will be lessreliable under the sustainable water scenariothan under business as usual, the yield of irrigated

120

100

80

60

40

20

0

Figure 15 Potential per capita domestic water consumption by connected rural and urban households, business as usual and sustainable scenarios, 2025

World


Cub

ic m

eter

s/pe

rson

/yea

r


RuralUrban

Business as usual

Developed countries

Business as usual

Business as usual

Sustainable Sustainable Sustainable

GLO

BA

LW

AT

ERO

UT

LOO

K T

O 2

025

15

cereals will be 2 percent lower for the world as awhole in 2025. On the other hand, global rainfedyields under the sustainable water scenario willbe 7 percent higher than business as usual, owingto higher agricultural research investment and alarger improvement in rainfall harvesting.Withthe faster growth in rainfed yields making up forslower growth in harvested area and irrigatedyields, total cereal production in 2025 will be 19million tons, or 1 percent, more under thesustainable scenario than under business as usual.Crop prices under this scenario will declineslowly from 1995 to 2025 except for slightincreases for maize and soybeans due to heavydemand for livestock feeds (Figure 16).

As in the water crisis scenario, net trade inthe sustainable scenario is lower than that underbusiness as usual, with cereal imports from thedeveloping world declining by 14 million tons,or 6 percent. This decline reflects the differentrates of adjustment of food production betweenfood-importing and -exporting countries. Cerealproduction under the sustainable water scenariois 10 million tons less in developed countries and29 million tons more in developing countries thanunder the business as usual scenario.

The sustainable scenario shows that withimproved water policies, investments, and rainfedcereal crop management and technology, growthin food production can be maintained whileuniversal access to piped water is achieved andenvironmental flows are increased dramatically.Compared with the water crisis scenario, theincrease in environmental flows under thesustainable water scenario is about 1,490 km3,equivalent to 5 times the annual flow of theMississippi River, 20 times the annual flow of theYellow River, and 4 times the annual flow of theGanges River.

300

250

200

150

100

50

0

Figure 16 World food prices, business as usual and sustainable scenarios, 2025

Ric

e

Swee

tpo

tato

es

Whe

at

Mai

zeO

ther

coa

rse

grai

ns


Soyb

eans

Business as usualSustainable

US$

/met

ric

ton

Pota

toes

Oth

er r

oots

and

tube

rs

GLO

BA

LW

AT

ERO

UT

LOO

K T

O 2

025

16

Raising Water Prices Raising water prices seems to make sense forseveral reasons. Higher water prices not onlyencourage all users to use water more efficiently,but also could generate funds to maintain existingwater infrastructure and to build new infrastruc-ture.Yet because of perceived political risks andconcern that higher prices would hurt poorfarmers and consumers, there have been fewattempts to implement higher water prices. Infact, in most instances the poor suffer fromcurrent subsidized water prices because watersubsidies in most countries go disproportionatelyto the better off: urban water users connected tothe public system and irrigated farmers.

Well-designed price hikes for water cancreate incentives for people to use water effi-ciently and recover at least operation and main-tenance costs, while protecting and evenincreasing farm incomes. But would such waterprice increases save significant amounts of waterthat could be left instream for environmentalpurposes without reducing food production?

To address this question, we examine twoscenarios with higher water pricing. One imple-ments higher water prices but with water useefficiency remaining the same as under the busi-ness as usual scenario. The other scenarioassumes higher water prices but also assumesthat higher water prices induce improvements inbasin efficiency through better crop watermanagement and investments in new irrigationtechnology compared with the business as usualscenario. Under both scenarios, water prices foragriculture, industry, and connected householdsare assumed to increase gradually over the

period from 2000 to 2025. By 2025 water pricesfor industrial water use are 1.75 times higherthan prices under the business as usual scenarioin developed countries and 2.25 times higher indeveloping countries. For domestic water uses,water prices are 1.5 times higher in developedcountries and double in developing countries.For agricultural water uses, prices double by2025 in developed countries and triple in devel-oping countries compared with the business asusual prices.

The higher-price scenarios result in a reduc-tion of water withdrawals of 839 km3 (18percent) and of total water consumption of 287km3 (14 percent) compared with business asusual, with more than half of the reduction occur-ring in developing countries. Withdrawals andconsumption are the same in the two scenarios;higher efficiency influences the proportion ofwithdrawals and consumption that are usedbeneficially.The impact on water withdrawal iseven greater in some regions, with reductions ofmore than 20 percent in China, Southeast Asia,Latin America, and West Asia and North Africa,and between 14 and 20 percent in other coun-tries and regions. The reduction in withdrawalsand consumption thus represents a major benefitfor the environment, dramatically increasing envi-ronmental flows.

Demand for water for all nonirrigation uses—that is, for industrial, domestic, and livestock use—falls dramatically under the higher-price scenarioscompared with business as usual. Total nonirriga-tion consumption of water falls from 599 km3

under the business as usual scenario to 449 km3

globally, from 395 km3 to 285 km3 in developing

Consequences of Key Policy Changes

If policymakers do not make a holistic change in water policies and investments but

simply change certain key factors, could this produce substantial benefits? We examine

what would happen if policymakers and water users raised water prices, shifted to sustain-

able groundwater use, or better exploited the potential of rainfed agriculture.

GLO

BA

LW

AT

ERO

UT

LOO

K T

O 2

025

17

countries, and from 204 km3 to 164 km3 in devel-oped countries.Water withdrawals for thesepurposes also fall significantly.

Total consumption of irrigation water underboth the higher-price and the higher-price/higher-efficiency scenarios is about 100 km3 less thanunder the business as usual scenario. Yet becauseof more efficient water use in the higher-price/higher-efficiency scenario, more of the irri-gation water is consumed beneficially by crops.Whereas cereal production declines by about 5percent in the higher-price scenario comparedwith business as usual, under the higher-price/higher-efficiency scenario the change in cerealproduction is slight.

With the decline in production under thehigher-price scenario, world food prices increase,with the greatest increase for rice (10 percent)and price increases of 4–8 percent for othercereals. When water prices induce higher basinefficiency, however, food prices are the same orlower (Figure 17).

These results show that higher water pricesfor industry, domestic, and agricultural sectorswould result in large water savings that can beused for environmental purposes. Making wateruse more efficient in conjunction with higher

prices is critical to maintaining or increasing thereliability of irrigation water supply and foodproduction compared with business as usual.

Shifting to Sustainable Groundwater UseA number of basins and countries are pumpinggroundwater in excess of natural recharge rates.These include the Rio Grande and ColoradoRiver Basins in the western United States, theYellow and Haihe River Basins in northern China,and several river basins in northern and westernIndia, Egypt, and West Asia and North Africa.What would happen to water and food if theseregions stopped overpumping and returned tosustainable water use?

The low-groundwater-pumping scenarioassumes that all countries and regions will phaseout unsustainable groundwater overdraft overthe next 25 years.Areas with more plentifulgroundwater will increase their pumping almostas much as in the business as usual scenario.Total global groundwater pumping will fall to 753km3 in 2021–25, a decline from the 1995 value of817 km3 and from the 2025 business as usualvalue of 922 km3. Compared with business asusual, global consumptive water use will decline

5.6 percent in the irrigation sector,0.5 percent in the livestock sector,0.1 percent in the domestic sector,and 0.1 percent in the industrialsector, with most of this changeoccurring in developing countries.Virtually no change in consumptionwill occur in other sectors.

In 2021–25 the total area plantedto cereals will be 730,000 hectaresless under the low-groundwaterscenario than under business asusual. Although rainfed area willincrease, it will not be enough tooffset the large decline in irrigatedarea. Most of the change in irrigatedarea will occur in developing coun-tries, especially in China. In most

Figure 17 World cereal prices under business as usual and alternative water price scenarios, 2021-25

300

250

200

150

100

50

0Rice Wheat Other coarse

grains


Business as usualHigher priceHigher price/higher efficiency

US$

/met

ric

ton

Maize

GLO

BA

LW

AT

ERO

UT

LOO

K T

O 2

025

18

regions yields for irrigated cereals will fall andyields for rainfed cereals will rise slightly.

Total cereal production will decline by anannual average of 18 million tons from businessas usual projections in 2021–25. The fall in irri-gated cereal production will lead to priceincreases that stimulate increased rainfed produc-tion, but, once again, the increase will not beenough to overcome the decline in irrigatedcereal production. Crop prices under the low-groundwater scenario are projected to be 5–10percent higher in 2021–25 than under business asusual projections. Although total developing-country cereal production will decline in the low-groundwater scenario compared with business asusual in 2021–25, these price rises will actuallycause farmers in developed countries to producemore cereals and thus lead to an overall increasein cereal production there compared with busi-ness as usual.

Not surprisingly, the low-groundwaterscenario projects the biggest drops in cerealproduction to be concentrated in the basins thatcurrently experience large overdrafts, especiallyChina and India. As a result, the developing worldas a whole will increase its net imports, withmajor increases concentrated in China and India,and developed countries will increase their netexports.

These country-level shortfalls in demand andincreases in imports could be serious, but theymay be a worthwhile trade-off for restoringsustainable groundwater supplies. More impor-tant, countries must combine a phase-out ofgroundwater overdrafting with policies to miti-gate the impacts on the overdrafting regions inorder to maintain income growth. Countriesshould increase their agricultural research invest-ments, and, particularly in the hardest-hit riverbasins, make investments and implement policyreforms to increase basin efficiency, andencourage diversification from irrigated cerealsto crops that give more value per unit of water.

Exploiting the Potential of Rainfed AgricultureCould more rapid growth in rainfed cerealproduction—through either research andtechnology-driven growth in cereal yield and areaor through increased rainfall harvesting—compensate for significant reductions in irrigationand water supply investment compared withbusiness as usual? We explore these questions intwo alternative scenarios.

The low-investment scenario assumes thatbasin efficiency will not increase above 1995levels, that the rate of increase in potential irri-gated area will be about one-third of the rateunder the business as usual scenario, that theincrease in reservoir storage will be 40 percentof business as usual, and that increases inmaximum allowable water withdrawals will be 30percent as high. This reduction in investment ininfrastructure and management seriouslyconstrains growth in food production, causing anannual drop in irrigated cereal production of 120million tons (11 percent) in 2021–25 and drivingcereal prices up by 25–35 percent.

We then estimate the required rainfed areaand yield increases to offset the reduction of irri-gated production and maintain essentially thesame international trade prices. A larger increaseis assigned to rainfed yield than area (because oflimited potential for area expansion), and a largerincrease is assigned to those basins, countries, orregions where irrigation effects are greater.

Under this scenario the international pricewill maintain approximately the same level as thebusiness as usual scenario for all cereal cropsexcept rice. It proved impossible to fullycompensate for the loss of rice production, whichhas a high proportion of irrigated area.

Compared with business as usual, this scenarioresults in a decline in global irrigation waterconsumption of 240 km3, or 16 percent, and adecline in irrigated cereal production of 153million tons. Rainfed area, however, increases by

GLO

BA

LW

AT

ERO

UT

LOO

K T

O 2

025

19

10 million hectares, mostly in developing countries.Rainfed yields increase by 11 percent, and rainfedproduction increases by 187 million tonscompared with business as usual. The share of theworld’s cereal produced on rainfed lands willincrease significantly, to 62 percent globally, 51percent in developing countries, and 78 percent indeveloped countries, compared with 56, 43, and 74percent, respectively, under business as usual.

The second scenario looks at the possibilityof increasing effective rainfall use through waterharvesting, conservation tillage, and precisionfarming to counteract the reduction of irrigatedproduction due to low investment in irrigationdevelopment and water supply. Effective rainfalluse increases by 10–15 percent above 1995 levelsfrom 1995 to 2025 in those basins and countrieswith rainwater shortages for crop production,including river basins in the western UnitedStates, northern and western China, northern andwestern India, and countries in West Asia andNorth Africa. An increase ranging from 5 to 10percent is projected for other regions. Theserates compare to 3–5 percent increases underbusiness as usual.

Under this scenario world cereal prices(especially for rice) are higher than those under

business as usual. The projected increase in effec-tive rainfall water use cannot fully compensatefor the irrigation decline. Although the globalproduction of rainfed cereals is 126 million tonsmore than under business as usual, irrigatedproduction is 131 million tons lower (Figure 18).Developing countries are harder hit, with a short-fall in total cereal production of 42 million tons(2.8 percent) that causes a reduction in demandof 26 million tons and an increase in imports of16 million tons. Nevertheless, the results showsignificant benefits from better management thatgenerates higher effective rainfall.

These scenarios show that there is significantpotential for increasing rainfed production tocompensate for lower investment in irrigation.Appropriate investments and policy reforms,however, will be required to enhance the contri-bution of rainfed agriculture. In some regionswater harvesting has the potential to improverainfed crop yields. But crop breeding for rainfedenvironments is crucial to future cereal yieldgrowth. Strong progress has been made inbreeding for enhanced crop yields in rainfedareas, even in the less favorable environments.The continued application of conventionalbreeding and the recent developments in

nonconventional breeding offerconsiderable potential for improvingcereal yield growth in rainfed envi-ronments. Progress could behastened by extending research tofarmers and by using tools derivedfrom biotechnology to assist conven-tional breeding (Table 2).

Governments must also combinecrop research targeted to rainfedareas with increased investment inrural infrastructure and policies toclose the gap between potential andactual yields in rainfed areas.Important policies include higherpriority for rainfed areas in agricul-tural extension services and accessto markets, credit, and input supplies.

3,000

2,500

2,000

1,500

1,000

500

0

Figure 18 Rainfed and irrigated cereal production, business as usual and lower irrigation investment and higher effective rainfall use scenarios, 2021–25

World

SOURCE: Authors' estimates and IMPACT-WATER projections, June 2002.NOTE: LINV-HIER stands for lower irrigation investment and higher effective rainfall use.

Mill

ion

met

ric

tons


Rainfed productionIrrigated production

Business as usual

Developed countries

Business as usual

Business as usual

LINV-HIER LINV-HIER LINV-HIER

GLO

BA

LW

AT

ERO

UT

LOO

K T

O 2

025

20

Table 2 Rainfed and irrigated cereal yield by region, for three scenarios: business as usual, lower irrigation investment and higher rainfed area and yield, and lower irrigation investment and higher effective rainfall use, 2021–25


NOTE: BAU stands for business as usual, LINV-HRF stands for lower irrigation investment and higher rainfall area and yield, andLINV-HIER stands for lower irrigation investment and higher effective rainfall use.

Rainfed yield (metric tons/hectare) Irrigated yield (metric tons/hectare)

REGION BAU LINV-HRF LINV-HIER BAU LINV-HRF LINV-HIER

Asia 2.46 2.96 2.49 4.50 4.05 4.13

Latin America 2.92 3.13 3.08 5.46 4.85 4.92

Sub-Saharan Africa 1.19 1.22 1.30 3.08 2.95 2.99

West Asia/North Africa 1.75 1.93 1.89 4.86 4.55 4.61

Developed countries 3.89 4.24 3.95 5.97 5.54 5.59

Developing countries 2.08 2.36 2.18 4.53 4.09 4.16

World 2.77 3.07 2.86 4.80 4.37 4.43

GLO

BA

LW

AT

ERO

UT

LOO

K T

O 2

025

21

Water scarcity will get much worse if policy and investment commitments from

national governments and international donors and development banks weaken

further. The water crisis scenario—predicated on the worsening of a number of already

evident trends—would lead to a breakdown in domestic water service for hundreds of

millions of people, devastating loss of wetlands, serious reductions in food production, and

skyrocketing food prices that would force declining per capita food consumption in much

of the world. Failure to adopt water-saving technology improvements and policy reforms

could make demand for nonirrigation water grow even faster than we projected, further

worsening water scarcity.

Implications for the Future

Water scarcity can lead to declining fooddemand and increasing food prices. As shown inthe water crisis scenario, major cereal crop pricesmay be more than double the projections underthe business as usual scenario, and at the same timefood demand may be significantly reduced, espe-cially in developing countries. Moreover, priceincreases can have an even larger impact on low-income consumers.

Excessive diversion of water flows and over-draft of groundwater have already caused environ-mental problems in many regions around theworld. Our analysis shows that the problems, froma local to a worldwide scale, will likely be evenmore serious in the future. If current investmentplans and recent trends in the water and foodsectors continue, expanding the environmental usesof water would require reducing the consumptionof irrigation water or domestic and municipalwater or both. Thus, in the absence of policy andinvestment reform, competition over waterbetween households and industries and betweenfarmers and environmental uses will increase inmany parts of the world.

With water becoming increasingly scarce,continued high flow diversions would become self-defeating. Excess extraction speeds the recessionof ecological systems and lowers water quality,finally reducing the qualified water supply for

human uses.This has already occurred in the AralSea Basin in Central Asia. Groundwater overdraftcan likewise lead to the loss of an important watersource for human uses, as is already happening inmany regions.

However, the analysis also reveals cause forhope. The scenarios explored in this report pointto three broad strategies that can address the chal-lenge posed by water scarcity for food production:

1. invest in infrastructure to increase the supply ofwater for irrigation, domestic, and industrialpurposes;

2. conserve water and improve the efficiency ofwater use in existing systems through reforms inwater management and policy; and

3. improve crop productivity per unit of water andland through integrated water management andagricultural research and policy efforts, includingcrop breeding and water management for rainfedagriculture.

Although the financial, environmental, and socialcosts of new water supply projects are high, insome regions, especially in developing countries, itis still crucial to selectively expand water supply,storage, and withdrawal capacities. Storage andwater distribution systems (such as water lift proj-ects and canals) are particularly needed for Sub-

GLO

BA

LW

AT

ERO

UT

LOO

K T

O 2

025

22

Saharan Africa, some countries in South andSoutheast Asia (such as Bangladesh, India, and VietNam), and some countries in Latin America. Thesecountries must consider not only the full social,economic, and environmental costs of develop-ment, but also the costs of failure to develop newwater sources. Projects must be designed toaccount for full costs and benefits, including notonly irrigation benefits, but also health, householdwater use, and catchment improvement benefits. Itis also essential to improve compensationprograms for those who are displaced or negativelyaffected by water projects.

Expanding water supplies can help alleviatewater scarcity, but the results show that the mostpromising avenue is likely to be water managementreforms, incentive policies, and investments in infra-structure and technology to enhance efficiency inexisting uses. Throughout this report, we haveshown that feasible improvements in the efficiencyof basin-scale irrigation water use can, on a globalscale, compensate for irrigation reduction resultingfrom (1) the phasing out of groundwater overdraftworldwide; (2) increased committed environmentalflows; (3) higher prices for agricultural water use(which themselves encourage investments inimproved efficiency); and (4) low irrigated areadevelopment. We have also shown that improvingirrigation water use efficiency is an effective way toincrease water productivity.

In severely water-scarce basins, however, rela-tively little room exists for improving water useefficiency, and food production and farm incomescould fall significantly if water for irrigation is trans-ferred to other uses. In these basins, governmentswill need to compensate for the negative impact ofgrowing water scarcity on agriculture by alternativemeans, such as investing in agriculture to obtainmore rapid growth in crop yields, promoting thediversification of farming into less water-intensivecrops, and diversifying the economy to reduce theeconomic role of agriculture over time.

Making big improvements in river basin effi-ciency in specific river basins will require site-specific analysis and implementation. Basin

efficiency depends on improvements both in water-saving technologies and in the institutionsgoverning water allocation, water rights, and waterquality. Industrial water recycling, such as recircula-tion of cooling water, can be a major source ofwater savings in many countries. Much potentialalso exists for improving the efficiency of domesticwater use. Steps may include anything fromdetecting and repairing leaks in municipal systemsto installing low-flow showerheads and low-wateror waterless toilets. Treated wastewater can beused for a variety of nonpotable purposes includinglandscape and recreational irrigation, maintenanceof urban stream flows and wetlands, wastewater-fed aquaculture, and toilet flushing. To encouragewater-saving innovation, domestic and industrialwater prices should be increased. Generalizedsubsidies should be replaced with subsidiestargeted to the poor.Water providers shouldcharge low prices for a basic entitlement of water,with increasing prices for greater amounts ofwater.

Improvements in the irrigation sector can bemade at the technical, managerial, and institutionallevels. Technical improvements include advancedirrigation systems such as drip irrigation, sprinklers,conjunctive use of surface and groundwater, andprecision agriculture, including computer moni-toring of crop water demand. Managerial improve-ments include the adoption of demand-basedirrigation scheduling systems and improved equip-ment maintenance. Institutional improvementsinvolve the establishment of effective water userassociations and water rights, the creation of abetter legal environment for water allocation, andthe introduction of higher water prices. Great caremust be taken in designing a water pricing systemfor agriculture. Direct water price increases arelikely to be punitive to farmers because water playssuch a large role in their cost of production. Betteralternatives would be pricing schemes that payfarmers for reducing water use, and water rightsand water trading arrangements that providefarmers or water user associations with incentivesto reduce wasteful water use.

GLO

BA

LW

AT

ERO

UT

LOO

K T

O 2

025

23

Rainfed agriculture emerges from the analysisas a potential key to sustainable development ofwater and food. Rainfed agriculture still producesabout 60 percent of total cereals, and its roleremains very important in both the business asusual and the sustainable water scenarios.Improved water management and crop produc-tivity in rainfed areas would relieve considerablepressure on irrigated agriculture and on waterresources. Exploiting the full potential of rainfedagriculture, however, will require investing in waterharvesting technologies, crop breeding targeted torainfed environments, agricultural extension serv-ices, and access to markets, credit, and inputsupplies in rainfed areas.

A large part of the world is facing severewater scarcity, but the impending water crisis canbe averted. The precise mix of water policy andmanagement reforms and investments, and thefeasible institutional arrangements and policyinstruments to be used, must be tailored tospecific countries and basins.They will vary basedon level of development, agroclimatic conditions,relative water scarcity, level of agricultural intensifi-cation, and degree of competition for water. Butthese solutions are not easy, and they take time,political commitment, and money. Fundamentalreform of the water sector must start now.

GLO

BA

LW

AT

ERO

UT

LOO

K T

O 2

025

24

GLO

BA

LW

AT

ERO

UT

LOO

K T

O 2

025

25

Notes1. W. J. Cosgrove and F. Rijsberman, World Water Vision: Making Water Everybody’s Business (London:World

Water Council and World Water Vision and Earthscan, 2000); I. A. Shiklomanov,“Electronic DataProvided to the Scenario Development Panel,World Commission on Water for the 21st Century”(State Hydrological Institute, St. Petersburg, Russia, 1999), mimeo.

2. E. Bos and G. Bergkamp,“Water and the Environment,” in Overcoming Water Scarcity and QualityContraints, 2020 Focus 9, ed. R. S. Meinzen-Dick and M.W. Rosegrant (Washington, D.C.: InternationalFood Policy Research Institute, 2001).

3. The business as usual, crisis, and sustainable scenarios are compared using average 2025 results gener-ated from 30 hydrologic scenarios. The other scenarios are compared with business as usual based ona single 30-year hydrologic sequence drawn from 1961–90, and results are shown as the average of theyears 2021–25.

4. Water demand can be defined and measured in terms of withdrawals and actual consumption. Whilewater withdrawal is the most commonly estimated figure, consumption best captures actual water use,and most of our analysis will utilize this concept.

5. The global projection is broadly consistent with other recent projections to 2025, including the 4,580km3 in the medium scenario of J.Alcamo, P. Döll, F. Kaspar, and S. Sieberg, Global Change and GlobalScenarios of Water Use and Availability:An Application of Water GAP 1.0 (Kassel, Germany: Center forEnvironmental System Research, University of Kassel, 1998), the 4,569 km3 in the “business-as-usual”scenario of D. Seckler, U.Amarasinghe, D. Molden, S. Rhadika, and R. Barker, World Water Demand andSupply, 1990 to 2025: Scenarios and Issues, Research Report Number 19 (Colombo, Sri Lanka:International Water Management Institute, 1998), and the forecast of 4,966 km3 (not including reser-voir evaporation) of Shiklomanov,“Electronic Data.”

6. Compared with other sectors, the growth of irrigation water potential demand is much lower, with 12percent growth in potential demand between 1995 and 2025 in developing countries and a slightdecline in potential demand in developed countries.

7. All tons mentioned in this report are metric tons.

8. J. A.Allan,“Water Security Policies and Global Systems for Water Scarce Regions,” in Sustainability ofIrrigated Agriculture—Transactions,Vol. 1E, special session:The Future of Irrigation under IncreasedDemand From Competitive Uses of Water and Greater Needs for Food Supply—R.7 in the sympo-sium on Management Information Systems in Irrigation and Drainage, 16th Congress on Irrigation andDrainage, Cairo (New Delhi: International Commission on Irrigation and Drainage, 1996).

To explore the relationships among water, environment, and food production, we developeda global modeling framework, IMPACT-WATER, that combines an extension of theInternational Model for Policy Analysis of Agricultural Commodities and Trade (IMPACT)with a newly developed Water Simulation Model (WSM).1 IMPACT is a partial equilibriummodel of the agricultural sector, representing a competitive agricultural market for cropsand livestock. Demand is a function of prices, income, and population growth. Growth incrop production in each country is determined by crop and input prices and the rate ofproductivity growth. World agricultural commodity prices are determined annually at levelsthat clear international markets. IMPACT generates projections for crop area; yield;production; demand for food, feed, and other uses; prices; and trade. For livestock, IMPACTprojects numbers, yield, production, demand, prices, and trade.

For this study we integrated the IMPACT model with WSM, a basin-scale model of waterresource use, to create a linked model, IMPACT-WATER. We made the linkage by (1)incorporating water in the crop area and yield functions; and (2) simultaneously determiningwater availability at the river basin scale, water demand by irrigation and other sectors, andcrop production. IMPACT-WATER divides the world into 69 spatial units, including macroriver basins in China, India, and the United States and aggregated basins over othercountries and regions. Domestic and industrial water demands are estimated as a functionof population, income, and water prices. Water demand in agriculture is projected based onirrigation and livestock production growth, water prices, climate, and water use efficiency forirrigation at the basin level. Then water demand is incorporated as a variable in the cropyield and area functions for each of eight major food crops: wheat, rice, maize, other coarsegrains, soybeans, potatoes, yams and sweet potatoes, and cassava and other roots andtubers. Water requirements for all other crops are estimated as a single aggregate demand.

We treat water availability as a stochastic variable with observable probabilitydistributions. WSM simulates water availability for crops at a river basin scale, taking intoaccount precipitation and runoff, water use efficiency, flow regulation through reservoir andgroundwater storage, nonagricultural water demand, water supply infrastructure andwithdrawal capacity, and environmental requirements at the river basin, country, andregional levels. Environmental impacts can be explored through scenario analysis ofcommitted instream and environmental flows, salt leaching requirements for soil salinitycontrol, and alternative rates of groundwater pumping.

1For a more detailed description of the integrated model, see M.W. Rosegrant, X. Cai, and S.A. Cline, World Water andFood to 2025: Dealing with Scarcity (Washington, D.C.: International Food Policy Research Institute, 2002).

The IMPACT-WATER Model

GLO

BA

LW

AT

ERO

UT

LOO

K T

O 2

025

26

2020 VISION INITIATIVE INTERNATIONAL ADVISORY COMMITTEE

H.E.Yoweri K. Museveni (Chairman), President, Republic of UgandaH.E. Olusegun Obasanjo, President, Federal Republic of NigeriaMr. Fawzi Hamad Al-Sultan, Former President, International Fund for Agricultural Development (IFAD), ItalyMr. Sartaj Aziz, Former Finance Minister and Foreign Minister, Government of PakistanMr. David Beckmann, Director, Bread for the World, USAMs. Catherine Bertini, Executive Director,World Food Programme, ItalyDr. Keith A. Bezanson, Director, Institute of Development Studies, United KingdomMr. Norman E. Borlaug, Distinguished Professor of International Agriculture,Texas A&M University, USADr. Lester Brown, Chairman of the Board of Directors,Worldwatch Institute, USADr. Margaret Catley-Carlson, Chairperson, Global Water PartnershipProf. Chen Chunming, Senior Advisor and President, Chinese Academy of Preventive Medicine, ChinaDr. Gordon R. Conway, President,The Rockefeller Foundation, USADr. Bernd Eisenblätter, Managing Director, Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ), GermanyDr. Leonard Good, President, Canadian International Development Agency (CIDA), CanadaDr. Bo Göransson, Director General, Swedish International Development Cooperation Agency (SIDA), SwedenMr. Michel Griffon, Director, Centre de Coopération Internationale en Recherche Agronomique pour le

Développement (CIRAD), FranceProf. Kenzo Hemmi, Professor Emeritus, University of Tokyo, JapanDr. Robert W. Herdt,Vice President,The Rockefeller Foundation, USAMr. Dean R. Hirsch, President,World Vision International, USAMr. Johan Holmberg,Ambassador of Sweden to EthiopiaMr. Ian Johnson, Chairman of the CGIAR and Vice President, Environmentally and Socially Sustainable

Development Network,The World Bank, USAH.E. Speciosa Wandira Kazibwe,Vice President, Republic of UgandaDr. Justin Lin, Professor and Director, China Center for Economic Research, Peking University, ChinaMr. Mark Malloch Brown,Administrator, United Nations Development Programme (UNDP), USASra. Margarita Marino de Botero, Corporacion El Colegio Verde, ColombiaProf.Alexander F. McCalla, Professor Emeritus, University of California at Davis, USAMr. Robert S. McNamara, Global Coalition for Africa, USADr. Moïse Mensah, Former Minister of Finance, Government of BeninH.E. John Evans Atta Mills,Vice President, Republic of GhanaSra. D. Cecilia Lopez Montaño, Head, National Planning Department, ColombiaMr. Harris Mutio Mule, Executive Director,Top Investment and Management Services (TIMS), Limited, KenyaDr. Maureen O’Neil, President, International Development Research Centre (IDRC), CanadaProf. David Pimentel, Department of Entomology, Cornell University, USAMrs. Mary Robinson, U.N. High Commissioner for Human Rights, SwitzerlandMrs.Victoria Sekitoleko, Regional Representative, Food and Agriculture Organization of the United Nations (FAO),

ZimbabweProf.Amartya Sen,Trinity College, United KingdomDr. Ismail Serageldin, Director, Library of Alexandria, EgyptDr.Ammar Siamwalla, President,Thailand Development Research Institute Foundation,ThailandDr. M. S. Swaminathan, Chairman, M. S. Swaminathan Research Foundation, IndiaThe Honorable Youssef Wally, Deputy Prime Minister and Minister of Agriculture, EgyptMr. Klaus Winkel, Head, Department of Evaluation, Research, and Documentation, Danish International

Development Agency (DANIDA), DenmarkMr.Timothy Wirth, President, Better World Foundation, USAProf. Muhammad Yunus, Managing Director, Grameen Bank, Bangladesh

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE2033 K Street, NW Washington, DC 20006-1002 USATelephone: 001-202-862-5600 Fax: 001-202-467-4439Email: [email protected]

Website: www.ifpri.org

Also available from the 2020 Vision Initiative:

Rosegrant, Mark W., Ximing Cai, and Sarah A. Cline. 2002. World water and food to 2025: Dealing with scarcity.IFPRI-2020 Vision/International Water Management Institute book.Washington, D.C. U.S.A.: InternationalFood Policy Research Institute.

Rosegrant, Mark W., Michael S. Paisner, Siet Meijer, and Julie Witcover. 2001. 2020 Global food outlook:Trends,alternatives and choices. 2020 Vision Food Policy Report.Washington, D.C., U.S.A.: International FoodPolicy Research Institute.

Meinzen-Dick, Ruth S., and Mark W. Rosegrant, eds. 2001. Overcoming water scarcity and quality constraints.2020 Vision Focus 9.Washington, D.C., U.S.A.: International Food Policy Research Institute.

ABOUT THE AUTHORS

Mark W. Rosegrant is a senior research fellow in the Environment and Production Technology

Division of the International Food Policy Research Institute (IFPRI) and principal researcher

at the International Water Management Institute (IWMI). Ximing Cai is a research fellow at

IFPRI and a researcher at IWMI. Sarah A. Cline is a research analyst at IFPRI.

Documents

Ifpri global water outlook to 2025 averting an impending crisis