Research Report

Water as an Economic Good:A Solution, or a Problem?

C. J. Perry, Michael Rock, and D. Seckler

INTERNATIONAL IRRIGATION MANAGEMENT INSTITUTEP O Box 2075, Colombo, Sri Lanka

Tel (94-1) 867404 • Fax (94-1) 866854 • E-mail IIMI@cgnet.comInternet Home Page http: //www.cgiar.org/iimi

ISSN 1026-0862ISBN 92-9090-351-1

Research Reports

IIMI’s mission is to foster and support sustainable increases in the productivity of irrigatedagriculture within the overall context of the water basin. In serving this mission, IIMI con-centrates on the integration of policies, technologies and management systems to achieveworkable solutions to real problems—practical, relevant results in the field of irrigation andwater resources.

The publications in this series cover a wide range of subjects—from computer mod-eling to experience with water users associations—and vary in content from directly ap-plicable research to more basic studies, on which applied work ultimately depends. Someresearch reports are narrowly focused, analytical, and detailed empirical studies; others arewide-ranging and synthetic overviews of generic problems.

Although most of the reports are published by IIMI staff and their collaborators, wewelcome contributions from others. Each report is reviewed internally by IIMI’s own staffand Fellows, and by external reviewers. The reports are published and distributed both inhard copy and electronically (http://.www.cgiar.org/iimi) and where possible all data andanalyses will be available as separate downloadable files. Reports may be copied freelyand cited with due acknowledgment.

3iii

Research Report 14

Water as an Economic Good:A Solution, or a Problem?

C. J. Perry, Michael Rock, and D. Seckler

International Irrigation Management InstituteP O Box 2075, Colombo, Sri Lanka

The authors: C. J. Perry is Deputy Director General at the International Irrigation Manage-ment Institute (IIMI). Michael Rock is a senior economist at Winrock International Institutefor Agricultural Development, Arlington, USA, and D. Seckler is Director General, IIMI.

This report has benefited—directly and indirectly—from comments of, and discussionswith, many IIMI staff, IIMI Fellows, and other professionals involved in water resourcesmanagement and research. In particular, the authors would like to thank Mark Rosegrantfor his balanced and perceptive comments, which have made the views expressed clearerand the authors hope, more persuasive.

C. J. Perry, M. Rock, and D. Seckler. 1997. Water as an economic good: A solution, or a problem?Research Report 14. Colombo, Sri Lanka: International Irrigation Management Institute.

/ water resource management / water as an economic good / economic aspects / economic analysis/ irrigated farming / water rights / pricing / privatization / marginal analysis / water market /water policy /

ISBN 92-9090-351-1ISSN 1026-0862

© IIMI, 1997. All rights reserved.

Responsibility for the contents of this publication rests with the authors.

Editor: Kingsley Kurukulasuriya; Consultant Editor: Steven Breth; Artist: D. C. Karunaratne;Typesetter: Kithsiri Jayakody; Publications Manager: Nimal A. Fernando.

5iiiiii

Contents

Summary v

Introduction 1

Values and Facts 2

The Economic Analysis of Different Values 3

Facts: Public Failure and Market Failure 6

Water Resource Management and Market Failure 10

Summing Up 13

Conclusion: Toward Improved Water Resources Management 14

Literature Cited 16

v

Summary

There is wide interest in, and support for, the idea oftreating water as an economic good. However, therole of water—as a basic need, a merit good, and asocial, economic, financial, and environmental re-source—makes the selection of an appropriate set ofprices exceptionally difficult. Further, the applicationof price-based instruments, once an appropriate valuesystem has been agreed, is particularly difficult in thecase of water. This is so because the flow of waterthrough a basin is complex, and provides wide scope

for externalities, market failure, and high transactioncosts. While judiciously applied market tools can beexpected to have benefits, in many cases the neces-sary and sufficient conditions, especially defined andenforced water rights, are not yet in place. Priority at-tention to these essentials is likely to have high re-turns; pursuit of economic approaches in the absenceof such preconditions may have unpredictable, andpossible negative effects.

1

Water as an Economic Good: A Solution, or a Problem?

C. J. Perry, Michael Rock, and D. Seckler

Introduction

The now famous proclamation that watershould be treated “as an economic good”originated in the Dublin Conference (ICWE1992). Like many such proclamations, it hasthe virtue of being sufficiently vague to al-low agreement, while leaving the impliedoperational content—over which there maybe strong disagreement—unstated.

The proclamation was a compromisebetween those, mainly economists, whowanted to treat water in the same way asother private goods, subject to allocationthrough competitive market pricing, andthose who wanted to treat water as a basichuman need that should be largely ex-empted from competitive market pricingand allocation.

To further complicate matters, there isan important distinction in economics be-tween the true “economic” value of a goodand its “financial” value. The two valuesrarely correspond, and as will be argued,for water the divergences are exceptionallycomplex and important. Thus it does notfollow from the declaration that water is aneconomic good that should be allocated bycompetitive market prices that reflect onlyfinancial, not necessarily economic, values.

The question is not whether water is aneconomic good or not—it certainly is an

economic good in most cases, like almosteverything else we have to worry about.Rather the question is whether it is a purelyprivate good that can reasonably be left tofree market forces, or a public good thatrequires some amount of extra-marketmanagement to effectively and efficientlyserve social objectives. The answer to thisquestion lies not so much in lofty principlesbut in value judgments, and theirapplication to different conditions of timeand place. Thus we find ourselves favoringthe private good side of the argument insome cases and the public good side inothers. The task is to define precisely whatthese cases, value judgments, and specificconditions of time and place, are. Thisdefinition is, we believe, important for tworeasons: First, dogmatic posturing by theproponents of each perspective is a waste ofintellectual talent. Second, and moreimportantly, water is far too important to itsusers to be the basis for socioeconomicexperiments. Much is already known aboutthe nature of successful policies andprocedures for allocating water; un-derstanding and incorporating theimplications of this knowledge will avoidsome potentially enormous financial,economic, environmental, and social costs.

2

That water usually is an economic goodfollows directly from the definition ofeconomics as “the science which studieshuman behavior as a relationship betweenends and scarce means which have al-ternative uses” (Robbins 1935). Water meetsthese requirements: it serves a multiplicity ofends (ranging from drinking and bathing,through irrigation, recreation, and environ-mental use, to waste disposal), and thussatisfies the condition of “alternative uses.”In many cases, water is scarce in the sensethat it cannot fully satisfy all its alter-nativeuses simultaneously.

But nearly everything we need toworry about is “an economic good” in thissense—that is why we have to worry. Withhis definition, Lord Robbins established thefoundations of what has since been de-scribed as the “Imperial Science” ofeconomics. But the important question isnot whether water is an economic good ornot, but what kind of economic good it is. Webelieve that it is both a public and a privategood. If this is correct then the true“economic” value of water will differ fromits “financial” value—i.e., the value basedonly on market prices—as any competenteconomist will recognize. How preciselywill it differ? The answer to this questionlies in two separate but interrelated do-mains or universes of discourse: values andfacts.

In the debate over water as a private orpublic good, the differences between valuesand facts are easily defined. First, in termsof values, the proponents of water as aprivate good contend that water is just likeany other good, that its production allo-cation should be determined by theoverriding value of consumer’s sovereignty—i.e., by the amount that people are “ready,willing, and able to pay” for it. The

criterion of consumer’s sovereignty totallyignores the distribution of income in society.If the poor cannot pay as much for a unit ofwater as the rich they should get less water,even if the marginal value to them in termsof other values (or utility) is greater.

This is what the proponents of water asa public good object to. They contend thatsafe water is a basic need that should beavailable at reasonable levels to everyone.They also contend that water used for irri-gation can be a powerful means of reducingfood costs to poor people and, under theproper conditions, should be subsidized(Chambers 1988). Others believe that waterserves important ecological, environmental,and aesthetic benefits in many cases, andshould not be allocated to other uses simplyon grounds of willingness to pay. In theterminology of economics, this schoolbelieves that, at least up to some minimallevel of availability, water is a social goodwhose availability to certain groups and forcertain purposes at well below marketprices will serve the greater benefit ofsociety as a whole. Similarly, narcotics,tobacco, and alcohol are “unsocial” goods,priced (or controlled) at levels well inexcess of the prices that would prevail in afree market.

But even in these cases of conflicts overvalues, rational people can reach agreementin terms of relative quantities. No onewould deny provision of safe drinkingwater to a poverty-stricken village thatcould not pay full costs, or advocate drain-ing wetlands so that water could beobtained for swimming pools and golfcourses. On the other hand, most advocatesof water as a public good would agree thatafter a basic level of service—of drinkingwater, for example—is attained, additionalsupplies could properly be allocated by

Values and Facts

3

market forces. Or, in the case of food, oncea nation achieves a certain level of pros-perity, market forces supplemented bydirected poverty alleviation programs maybe more efficient ways to help poor peoplethan general subsidies on food. Conflictsover values often depend on situation-specific facts. Unfortunately, dogmatistsoften control the discussion simply byvirtue of their very single-mindedness.

However, even if everyone agrees onvalues, there can be rational disagreementson facts. This occurs, for example, in thecase of “market failure,” discussed below.Even if everyone would agree that con-sumer’s sovereignty should rule as the solevalue, there are several well-known con-ditions under which the free market willfail to effectively and efficiently serve thisvalue. In these cases, various kinds anddegrees of government intervention or otherkinds of collective intervention, or both, arerequired to make the market performeffectively to serve the value of consumer’ssovereignty.

In sum, water serves many differentobjectives and has properties that make itboth a private and a public good. Theappropriate blend of values and facts inproper policy formulation for water re-quires a much more sophisticated form ofanalysis than that allowed by the sim-pleminded dogmatism of proponents, either

of basic needs or of free markets. Waterpolicy must be formulated in terms ofmulti-objective decision making, recog-nizing that the relevance and importance ofvarious values and facts will vary sub-stantially over different conditions of timeand place.

In the following pages we address thequestion: what does it mean to treat waterin irrigated agriculture as an economicgood? We focus on irrigated agriculture forseveral reasons. First, it is the largest con-sumer of water virtually everywhere thatirrigation is practiced. Second, policies gov-erning water use in irrigated agriculture arefraught with disagreements over both val-ues (should it be treated as a purely privategood, a public good, or a basic humanneed) and facts (what is the most cost-effec-tive allocation policy and method if water istreated as a purely private good, or a pub-lic good, and in what instances should it betreated in one way rather than in the other).

The argument proceeds in three steps.First, we examine the economic analysis ofdifferent values involved in treating wateras a private and public good. Then the factsinvolved in this debate in terms of thecauses of both market failure and publicfailure are examined. Finally, we examinealternative ways out of the dilemma posedby these two forms of institutional failureand suggest a practical plan of action.

The Economic Analysis of Different Values

The difference that different values canmake to the interpretation of the same factsmay be illustrated from an interesting, pro-market paper by Briscoe (1996). SinceBriscoe presents the basic economics neededfor both sides of the discussion in a clear

and simple way, this part of his paper isquoted below at considerable length.

The idea of “water as an economic good”is simple. Like any other good, water has avalue to users, who are willing to pay for it.

4

Like any other good, consumers will use waterso long as the benefits from the use of anadditional cubic meter exceed the costs soincurred. This is illustrated graphically infigure (a), which shows that the optimalconsumption is X*. Figure (b) shows that if aconsumer is charged a price P’, which isdifferent from the marginal cost of supply, thenthe consumer will not consume X* but X’. Theincrease in costs (the area under the cost curve)exceeds the increase in benefits (the area underthe benefit curve) and there is a correspondingloss of net benefits (called the “deadweightloss”).

But what about groups of users, and how iswelfare maximized for the group (or society) asa whole? The simple logic of figure 1 applies in

the aggregate—for society as a whole, andwelfare is maximized when:

• water is priced at its marginal cost, and

• water is used until the marginal cost is

equal to the marginal benefit.

So far so good, but what actually do wemean by “benefits” and “costs,” how are thesedealt with in different water-using sectors, andwhat are the implications?

Briscoe directly answers this question:“The value of water to a user is the maxi-mum amount the user would be willing topay for the use of the resource.” But thosewho believe that water is a social goodwould say that this is an incomplete andmisleading economic analysis. Willingnessto pay depends largely on the ability to pay.Thus even with the same basic need for orvalue of water, the rich will get more andthe poor less. Thus the people between X*and X1 are priced out of the market for wa-ter—if not completely, then in terms of mar-ginal reductions in the amount they can af-ford. Thus, the higher the price, the greaterthe incidence of poverty in X.

In an enlightened humane society, well-to-do taxpayers want to help the poor ob-tain basic needs. Thus, in terms of figure 1,tax-payers are willing to subsidize water,effectively shifting the supply curve downto the point where it intersects the demandcurve at P1, X1, and many more poor peopleget water. Who is Briscoe, they may say, tocall this humane gesture a “deadweightloss?” Indeed the real loss to society as awhole would be to counter social demandby reducing water availability from X1 to X*through competitive prices.

Many economists, and many non-economists, do not seem to understand thatconsumer ’s sovereignty is just anothervalue judgment, like any other value

FIGURE 1.Optimal consumption and “deadweight losses” if water is underpriced.

5

judgment. Thus Briscoe can say that thevalue of water is measured by willingnessto pay only if consumer’s sovereignty isassumed to be the appropriate value sys-tem. However, if this problem can beresolved, economic analysis can help insorting out the basic issues more preciselyby analyzing different values in terms oftheir marginal, rather than their total, value.Indeed, a controversy over the value ofwater at least indirectly led to the develop-ment of marginal analysis (Barkley andSeckler 1972), which is one of the two orthree most important concepts in econom-ics. Markets reflect marginal values, andfunction best where such values arerelatively stable, or change progressively.Water is not such a good.

In some cases, the marginal utility ofwater is essentially infinite (or very high)for all practical purposes—for example, in adrought, when people are dying of thirst, orwhen a reservoir runs dry at the end of theseason after all inputs have been providedexcept the last irrigation turn. Conversely,once a person (or a crop) receives enoughwater sufficient to alleviate physical stressand strain, the utility of additional unitsrapidly plummets and can even becomenegative.

A case can be made that in such ex-treme instances, water, as other basic hu-man needs, ceases to be an economic goodunder Robbins’ definition. The condition ofalternative uses implies alternative ends.When there is only one end to which all ofthe resource (and all other resources) wouldbe applied under conditions of extremescarcity, there is only one option and onlyone choice to be made—get the water ordie, which closes all options. In any case,under the welfare criterion of “expandedchoice,” a minimum condition for an eco-nomic good would be two choices (otherthan death, which eliminates the dynamic

choice set). Humans are choosing-animals,and if they have no choices, they are forcedto live at a subhuman level.

Recognizing this fact, humane societiesattempt to protect their citizens against in-human situations (Serageldin 1996). Civi-lized societies provide basic rights protect-ing citizens against murder, imprisonmentwithout trial, torture, and even indenturedservitude. And, for the same reason, hu-mane societies attempt to assure their citi-zens with a minimum supply of basicneeds—starting with water, food, shelter,and medical care. Humane societies alsoaccept responsibility for providing minimallevels of education and employment oppor-tunities to their citizens. This distinction isrecognized in economics as the differencebetween ordinary goods (including ser-vices), “merit goods,” which people shouldbe encouraged and helped to consume, and“demerit goods,” like drugs, which theyshould be discouraged or prevented fromconsuming. Underlying these distinctions isthe value of expanded choice: will con-sumption of this particular good now ex-pand, contract, or not affect the future set ofchoices? The extent to which humane soci-eties can actually achieve these goals, and atwhat levels depends on their judgments ofwhere the marginal utility curves becometoo steep and their resource costs too dear.

On the fringes of these marginal curves,there is room for debate over values amongproponents of the humane society. But fewwould argue that people below the povertyline in India, where the poor spend 80 per-cent of their income just to satisfy minimumnutritional calorie and protein requirements,are in a satisfactory state. Nor would any-one deny that substantially reducing thepercentage of people in this miserable situ-ation in India, China, Indonesia, and manyother developing countries is one of the sig-nal accomplishments of this century.

6

Water also fulfills the criteria for beingconsidered a merit good: access to cleanwater for washing and personal hygienehas health benefits (reduced incapacity forwork; reduced medical costs) that generallyexceed the cost of providing the water.

But most proponents of the humanesociety are also willing to accept the rule ofconsumer’s sovereignty, to switch govern-ing value judgments, once the margin ofbasic needs has been satisfied. In otherwords, it is an obligation of humanesocieties to assure reasonable levels ofwater, food, shelter, and medical care toassure that basic human needs are met. Butit is not reasonable to assist individuals orfamilies in the acquisition of goods beyondthis level. This is tantamount to arguingthat the same goods should be treateddifferently at different levels of con-sumption. There are clear implications ofthis for water, including irrigation water.Supplies of water at a level of basic needs isan obligation of humane societies to

provide irrespective of the ability to pay. Ata higher level of supply, lower on themarginal utility curve, society has little orno interest at all and consumer ’s sov-ereignty should rule. Needless to say, bothof these statements are, of course, based onvalue judgments.

It may also be noted that in the eco-nomic evaluation of projects, the poor arediscriminated against since they are notable to pay as much as the rich. One ap-proach to this problem is to evaluateprojects that help the poor on the basis ofwhat they would be willing to pay if theywere, say, middle-income consumers.

In sum, depending on the quantitiessupplied to individuals, water can be eithera basic human need, a merit good, or anordinary private good; it can best be allo-cated by the public sector or by the privatesector depending on these quantities. This isan important example of how a fundamen-tal conflict of values can be potentially re-solved by facts—indeed, quantitatively!

Facts: Public Failure and Market Failure

Given the generally dismal record of thepublic sector in this field, one must sympa-thize with anyone who is interested in alter-native institutional arrangements for watermanagement. Many have understandablyturned to economic instruments, marketsystems, and prices as an alternative on thepersuasive grounds that the market worksso spectacularly well in a host of other areasof economic activity. But even if one acceptsthe value of consumer’s sovereignty with-out qualification, water is unfortunately afield beset with the classic problems of mar-ket failure. In this section, we first examinethe rationale for privatization, the ap-

proaches that can be followed, and theirappropriateness. We then turn to the ques-tion of market failure—a critical problem inthe debate about the role of economic in-struments in the management of water re-sources.

Privatizing irrigation systems

In the public sector truly, “the road to hell ispaved with good intentions.” While thereare some notable exceptions to the rule, thepublic sector has generally performed mis-erably in all forms of water management—

7

whether in irrigation, or in domestic andindustrial water supplies, or in protectingresources and environmental quality. Thebook on public failure has not yet beenwritten in any sense comparable with theelegance and rigor of the works on marketfailure (Pigou 1932; Mishan 1976; Little1950). But some of the well-known and in-terrelated ills of the public sector can belisted readily:

• rent seeking, either economically, in theform of direct bribes and corruption, orsocio-politically in the form of empirebuilding, high costs, and excessive sup-plies (Wade 1982; Repetto 1986)

• the divorce of incentives from perfor-mance—indeed, sometimes almost aninverse relationship

• capture of public agencies and funds bypolitically powerful interests and theirclients

• administrative operations, “by thebook,” rather than management interms of objectives and results

The kinds of perverse behavior thatemanate from public sector entities arelegendary. In irrigation, they create thenotorious head-end-tail-end problem, wherefarmers at the heads of the system can takeso much water that they actually reduceyields through waterlogging, while farmersat the tails receive only what is left overand suffer drought damage to their crops.In domestic water supplies, as Briscoe(1996) points out, the poor usually subsidizethe rich—in the name of the humanesociety! Water projects are subject tocorruption, resulting in massive costoverruns and windfall gains to favoredclients—even in direct violation of the laws,even in the USA (Seckler and Young 1978).

The problems with public sector man-agement and allocation of water have cre-ated the movement toward privatizing irri-gation systems (Seckler 1993). Privatizationcan take several forms, from turnover ofoperation and maintenance to farmer asso-ciations, to volumetric or quasi-volumetricpricing at the farm level, to development ofwater markets and tradable water rights1

that would allow water to flow to the high-est-value uses. In each case, the aim ofprivatization is simple—to institutionalize amechanism for the management and alloca-tion of water that approximates a conven-tional market, including a direct relation-ship between service provided and chargesfor water that approaches marginal costs,and to a mechanism for (re)allocating waterfrom lower- to higher-value uses.

But which kind of privatization worksbest in what circumstance? In terms of val-ues, many people would agree that theprivatization route should be an approachor a combination of approaches that satis-fies basic needs criteria and then optimizeseconomic returns to water in terms of con-sumer’s sovereignty. But this answer justrequires us to ask another question: Whatdo we know about the marginal benefitsand marginal costs of privatization? Theanswer is some, but not nearly enough. An-swering this question is therefore an impor-tant part of a research program that concen-trates on the analysis and quantification ofbenefits and costs in the following circum-stances.

Turnover

The benefits of turnover are expected to in-clude impacts on productivity and equity,gains from freeing up of government re-sources for use elsewhere, and more effi-cient operation and maintenance (O&M).

1We differentiate delib-erately, though there areareas of overlap, be-tween water marketsand tradable waterrights. The latter re-quires both the formaldefinition of entitle-ments, and the specifica-tion of the conditionsunder which the entitle-ment may be traded.Water markets, on theother hand, have devel-oped widely, for ex-ample in areas whereprivate tube wells pro-vide competitive ser-vices to numerouspotential buyers, in theabsence of the formaldefinition of waterrights, tradable or other-wise.

8

The costs of turnover include organizationaland transaction costs, including labor earn-ings foregone by farmers who participate,and those costs associated with overcomingcollective action problems. Those in favor ofprivatization expect that the benefits out-weigh organizational and transaction costsand that turnover is more cost-effective thansimple user charges tied to services pro-vided by a public sector irrigation agency.

The evidence to date (Vermillion 1996)is mixed; there are undoubtedly caseswhere the projected impacts (especially, re-ductions in government spending) havebeen achieved. There are also cases of fail-ure and reduced performance. Success orfailure seems to depend more on the qual-ity and commitment of the implementingagencies and local leadership factors in theprocess of turnover than on the nature ofturnover itself.

Service-related user charges

User charges link use benefits to service de-livery costs. This contributes to cost recov-ery and may also provide incentives forimproved irrigation services with attendanteffects on productivity, equity, and effi-ciency of resource use. Many services (elec-tricity, telephone, buses, etc.) link the chargeto the level of service (kilowatt-hours, call-distance-minutes, distance traveled). Thesimplest measure of service corresponds tothe volume of water delivered to irrigationusers, but capturing the potential benefits ofservice-related charges will often requiredifferentiated charges. For example, wheredrinking water is supplied through irriga-tion canals, the level of service is usually farsuperior to that offered to agricultural us-ers—priority allocations in times ofdrought, and regular deliveries at times oflow or zero irrigation demand. This en-

hanced level of service has attendantcosts—losses are a far higher proportion ofdeliveries at low flow rates, and so on. Thuswell-defined service-related charges woulddifferentiate sharply among users.

Volumetric pricing

The potential benefits of volumetric pricingor quasi-volumetric pricing (achieved, forexample, by a crop tax related to consump-tive use) are obvious. This is the most forth-right way to link water use benefits withcosts and the value of services provided. Bysetting volumetric prices equal to opportu-nity costs, water is efficiently allocated, staticallocative efficiency gains are reaped, anddeadweight losses are avoided. Because wa-ter rents are captured through such pricing,losses associated with rent-seeking are alsoavoided. The costs of volumetric pricing in-clude the capital, administrative, and insti-tutional costs associated with volumetricmetering, billing, and collections of watercharges at the farm level (Perry 1995).

Water markets

Water markets most commonly operate lo-cally to allow agricultural water suppliersand consumers to include the opportunitycost of water in their management deci-sions. Usually, this involves trading wateramong similar uses (for example, the sale orexchange of irrigation ‘turns’ in a rotationalsystem), or sale of water by the owner of atube well to nearby farmers. The price gov-erning in such local markets encourages di-version to higher-value uses. It should benoted that water markets can coexist withextreme levels of subsidy—for examplewhere there is no charge for the irrigationwater that farmers trade, or where there is

9

subsidized power supply to the wells thatprovide water to local farmers. The costsassociated with water markets are thosetransaction and physical system costs neces-sary for water markets to efficiently operate.

Tradable water rights

Tradable water rights allow the formaltransfer of water entitlements among users,and as such are more likely to involve inter-sectoral transfers than the local water mar-kets described above. While water marketscan function in the absence of formal waterrights, tradable water rights require a muchmore specific definition of the entitlement.As in the case of water markets, while trad-able water rights are likely to result in (in-ter-sectoral) reallocation of water from low-to high-value uses, there is no guarantee, inthe absence of other charging mechanisms,that the costs of providing the service willbe recovered. (Indeed, the failure to recovercosts simply increases the potential priceand incentive to trade.)

Which of these privatization options isbest? Again there is no simple answer tothis question, but some aspects are clear. Animportant size-problem—the small size ofmost farms in most irrigation command ar-eas in developing countries—virtually rulesout volumetric metering at the farm level.In most instances, the incremental costs ofthe infrastructure, management, and admin-istration required for volumetric metering atthe farm level will exceed marginal ben-efits—especially where systems have notbeen designed and constructed to meetthese objectives. Quasi-pricing schemes(such as a crop charge linked to evapotrans-piration) can, to some degree, overcomethese problems. An additional problem inexisting irrigation systems is that the valueof water rights is capitalized in the value of

agricultural land. When water is volumetri-cally priced by metering or quasi-priced,full marginal pricing amounts to the expro-priation of those rights (with consequentcapital losses in irrigated land). Farmerswho have purchased land on the basis ofthe accepted, if not legally specified, waterentitlement will strongly oppose this.

In principle, turnover is a desirable op-tion. It can reduce the financial burden of fi-nancially strapped public irrigation agencies.It can provide a locally negotiated basis tolink services to user charges, contributing toan improvement in the quality of services.And it can simplify irrigation users’ involve-ment in investment decisions: Two draw-backs characterize turnover. First, there islittle evidence that turnover increases pro-ductivity (Vermillion 1996). Second, there areserious questions regarding the long-termsustainability of turned-over systems, be-cause farmers may not pay their dues and thedues do not include adequate provision forreplacement of major facilities (Svendsen andVermillion 1994). And often the withdrawalof the government agency leaves importantgaps in oversight areas—overexploitation ofgroundwater; pollution of canals, drains, andaquifers; dam safety; competition among ur-ban and agricultural demands; and droughtplanning. Once these issues are appropriatelyaddressed, it is not at all clear whether thebenefits of turnover exceed its organizationaland transaction costs.

This leaves water markets. At the locallevel of farmers trading small amounts ofwater on the watercourse and buying andselling water from tube wells, water marketsare already thriving in most irrigation sys-tems and should be encouraged. Waterprices and markets can also serve a valuablefunction in larger transfers, whether withinor among sectors under the appropriate con-ditions, noted before, and with suitable regu-lation, as discussed in the last section.

10

In the following discussion, we shall exam-ine three of the major causes of market fail-ure in irrigation and water resource man-agement, “externalities,” “transaction costs,”and “property rights.”

Irrigation water and externalities

External effects may be defined as uncom-pensated costs or benefits incurred by oneparty by virtue of the activities of anotherparty. There are few, if any, economicactivities that have as high an incidence ofexternal effects, both costs and benefits, aswater.

One of the most important, yet leastappreciated, facts about water is that in abasin, a substantial amount of it is recycled.When water is diverted from a stream (orpumped from groundwater) for use inagriculture or other activities, some of it isconsumed—for example, through evapo-ration—and some, returns via drains orpercolation to the stream or aquifer, thusbecoming available for a further cycle ofdiversion at another time, another place,and at another quality. Eventually, asdemand increases, the fresh water availablein a stream is fully utilized—all outflowsare of sufficiently poor quality, or to placeswhere the cost of recovery is too high, to bejustified for potential uses. Once thishappens a water basin is “closed” and nousable water supply is left.

The description set out above is highlysimplified but it indicates that if we wish toprivatize the allocation of water, the firstchange that needs to be made is to base ac-counting not for diverted water, but forconsumed water, with additional adjust-ments for time, location, and quality. Anexample should make this clear.

Let us take the case of rice. Rice farm-ers in Asia and elsewhere use water as asubstitute for labor to control weeds, in ad-dition to meeting crop water needs. The ef-fect of this is to dramatically increase thevolume of water applied, but the impact ofthis very high application of water on con-sumptive use of water is minimal. This isbecause consumptive use can never exceedpotential evapotranspiration, and as long asthe crop is well watered, crop evapotranspi-ration will be at or close to its potential.

If water applied is now priced with apositive marginal cost, farmers will substi-tute some labor for water and in doing soreduce the amount of water applied, yield-ing what looks like real water savings. Butthese savings are ultimately illusory as longas the excess water applied is recycled andreused. Marginal pricing of water appliedwill primarily have accomplished a shift inthe demand for labor. Further, if the “ex-cess” water applied in one season to therice was stored in the aquifer for pumpingand reuse in a drier season, then an externalbenefit will have been lost.

The importance of accounting for waterin terms of consumption is well understoodin the western US. There, there is substan-tial resistance to transfer of water out of ir-rigation districts because it is realized thatonce trading starts all of the secondary ef-fects of recharging aquifers and recyclingare disturbed.

Because of this problem, some states inthe US now assign property rights on thebasis of consumptive use rather than waterdiversions, for purposes of sale and transfer(Rosegrant and Binswanger 1994). In someinstances, state laws expand this by re-quiring irrigation districts to documentreductions in consumption at the local levelbefore water can be traded. While this is a

Water Resource Management and Market Failure

11

step in the right direction, it does notcompletely solve the problem of exter-nalities. If farmers sell the consumptiveuse right, they will no longer irrigate.In this situation, losses in canals will tend toform a higher proportion of deliveries; thefixed operating costs of the system must beborne by fewer users; aquifer recharge isreduced, and other farmers, who dependedon return flows will have to reduce theirirrigation accordingly. Keeping theseaccounts in order presents a majorchallenge.

Let us now add some real-world com-plexities to our simplified distinction be-tween field-level and basin-level measuresof consumption. The non-consumed fractionof water applied at the field returns to thesystem, but it returns in an altered state.First, to the extent that the originally di-verted water contained some salt, the returnflow will have a higher concentration be-cause plants consume only water. Second,as a result of the application of fertilizersand pesticides the return flows may be fur-ther degraded by pollutants. Third, if thesoil in the irrigated area contained naturalsalts, these may also be leached by the re-turn flows, reducing its quality still further.Fourth, the return flows may go to a loca-tion where not all can be recovered—a deepaquifer, or into a stream below the lowestdiversion structure.

Finally, the return flows may arrive at atime when water is in temporary excess (di-versions during the dry season returningduring the rainy season), or vice versa.Thus setting the appropriate prices may beseen to be complex, and of course also de-pendent upon the proposed further use:farmers may be pleased to receive waterthat already contains fertilizers. Barge op-erators are concerned only about the spe-cific gravity of the effluent, not its chemicalcomposition, and so on.

The point of these examples is to dem-onstrate the need to pay attention to exter-nal effects in a water basin—that is wherewater is going, where it is being consumed,where it is being reused, and what is hap-pening to salt and pollution loading, andthe timing and location of return flows aswater is being recycled and reused. Theseeffects must be regulated to bring privatebehavior into accord with social interestsand to achieve an economic optimum. Un-less this is done, pricing and trading waterwill have uncertain impacts on economic,social, and environmental efficiency.

Privatization in water andtransactions costs

The second efficiency problem associatedwith treating irrigation water as an eco-nomic good relates to transaction costs. Inmany irrigation projects, the irrigation infra-structure—both the physical and manage-ment infrastructure—required to allow de-livery of water to serve market purposes—that is, to price water at its marginal cost—is entirely absent. Perry (1995) recently cal-culated the cost of introducing the infra-structure required to give farm-level mea-surements of water deliveries in Egypt. Byhis estimate, they were such as to morethan offset the benefits that would flowfrom being able to set prices closer to mar-ginal cost. Egypt is not the only example ofthis. In the warabandi system of irrigationwidely practiced in India and Pakistan,while system structure permits trading ofwater among farmers within a watercourse,trading across water courses or betweenagricultural and nonagricultural uses out-side individual water courses would requiresubstantial physical infrastructure invest-ments and institutional change. It is doubt-ful whether the marginal benefits of trading

12

across watercourses would justify the in-vestment in irrigation infrastructure neces-sary to accommodate it.

For the most part, in the real world,water is allocated first to municipal anddomestic use, second to industrial andcommercial use, and third to agriculture.(Environmental allocations are also growingin volume and priority.) This sequence ofpriorities is generally consistent with socialand economic objectives that many wouldshare; the questions then are what wouldthe incremental benefit be of fully liberatingthe allocation process (with agreed valuesstructures!), and what would the infra-structural and transaction costs be.

In the terms of classical economics, oneof the features of the “dead-weight loss”triangles shown in figure 1 is that, withinsectors they are generally very small inrelation to total benefits or costs (the areasunder the demand and supply curves).Frustrating as this fact is to market ad-vocates, it means that one cannot spendvery much to eliminate them. And, in thecase of inter-sectoral transfers, e.g., fromagriculture to domestic or industrial use,the triangles may be large but the sideeffects—in terms of externalities, secondarybenefits, transaction costs, changes in thephysical systems, etc.—will often outweighthe benefits.

Property rights

The third point that needs to be consideredis that effective water markets and waterpricing are utterly dependent on secure andeffective property rights in water. Yet manywould agree that the single greatest prob-lem in water resource management in thedeveloping world is that property rights inwater are very insecure and ineffective. Orsaid another way, a central problem of

irrigation performance is the ability ofsome farmers to steal water from otherfarmers.

This fact has been extensively doc-umented in numerous studies of irrigationperformance. The most common mani-festation of it is that farmers at the heads ofirrigation systems usually receive far morewater than they have a right to, whilefarmers at the tails receive less. If propertyrights in water formally recognize rights todiversions while denying historical rights toreturn flows, as has been done in Mexicoand Chile (Gazmuri and Rosegrant 1994),they institutionalize theft. The same prob-lem can occur at the level of entire waterbasins, as in Pakistan and Egypt. Asdevelopment proceeds upstream, down-stream users receive progressively less, andmore polluted, water.

Since property rights in water are not,in this first instance, secure or effective, it isdifficult to see how privatization willcontribute to more efficient allocations ofwater use unless substantial efforts (andcosts) are made in advance to establish andprotect property rights.

In countries where water markets andtrading of water have worked beyond thepurely local level, there are in place: lawsassigning rights, laws describing how rightsmay be traded; legal systems that enforcethe rights and punish infringements onthose rights; farmers who are accustomed toworking with laws and rules that areenforced; and a physical irrigation infra-structure and irrigation management systemcapable of allocating water in accordancewith market-friendly principles and thechanging needs and priorities that flowfrom these principles. None of this ischaracteristic in developing countries—indeed most lack the first requirement,water rights—nor is it costless or easy toestablish such frameworks.

13

This set of problems, particularly theproblem of governance (and the associatedtransaction costs accompanying establishinggood governance), is so apparent to thosewho work with irrigation systems in devel-oping countries that it is difficult to under-stand why the proponents of privatizationignore it. Paradoxically, in those countrieswhere these features are in place, many spe-cialists are unaware of (or blind to) their

existence. Recently, a study in Germany re-vealed a vast array of complex rules andtraditions governing the allocation of water,responsibilities, and costs in a major riverbasin, whose management is undertakenentirely by very old user organizations. Spe-cialist observers were unaware of theirexistence, because they had worked sosmoothly and unobtrusively for so long!(Wolff and König 1997).

Summing Up

How might these considerations of water asan economic good affect water policies, par-ticularly those in irrigated agriculture? It isimportant to recognize that much of the dis-cussion of irrigated agriculture is takingplace under conditions of conflict—wherethere is intense disagreement over values,and facts are unknown. This is what makesworking in water so interesting and excit-ing.

If basic human needs for safe domesticwater and food produced by irrigation areto be met in poor countries, some degree ofsubsidy may be necessary so that masses ofpoor people are not priced out of the mar-ket. In the case of food, for example, manyAsian countries deliberately created an“oversupply” of food (in relation to market-price demand) by subsidizing agriculture,in particular irrigation investments. Thispolicy has resulted in dramatic decreases inthe number of abjectly poor farmers andhungry people in these countries and is oneof the signal accomplishments of human-kind in all of world history.

At more micro levels—on farmers’fields, within individual command areas,and in water basins—externalities can beinternalized by paying close attention towhat is happening to the physical flow of

water: Where is water going? Where is itbeing lost? Where is it being consumed?Where is it being reused? What is happen-ing to salt and pollution loading? Doingthis requires paying close attention to hy-drology and integrating the engineering ofwater with the economics of water.

We now know that the consequences ofwater movement through water basins dif-fer significantly and that these differenceshave important effects on the design of wa-ter policies. In Egypt, high basin-wide effi-ciency and low salt loading, mean that in-vestments to improve local, or classical, ir-rigation efficiencies, often have low or nega-tive returns. In Pakistan, high basin effi-ciency and high salt loading, combinedwith recycling and reuse contribute to sig-nificant salinization and waterlogging.Here, investments in local irrigation im-provement projects can yield high returns,particularly if reductions in the negativeexternalities associated with salinizationand waterlogging from over-irrigation arecounted in benefits. We doubt that market-determined prices offer guidance to an ap-propriate water allocation strategy in eitherof these two cases.

Similar technical and economic prob-lems confound the wisdom of policy advice

14

that fails to take account of the movementand use of water through a farmer’s field.For example, it has become fashionable toadvise farmers to shift to wet-seeding ofrice. Adoption of this technology is (fromthe farmer’s perspective) a means of savinglabor. It is also seen as a means of reducingwater consumption. Wet-seeding reduceswater diversions, but it may not reduceconsumptive use because the large wettedarea covered by the dry seeding (as com-pared to rice nurseries) evaporates morewater. In this instance, pricing diverted wa-ter at marginal cost may actually encouragea water-inefficient practice (one with higherevapotranspiration).

Transaction costs can be taken into ac-count by trying to measure them prior tomaking changes in the design, construction,or management of irrigation systems. Thisis likely to be particularly important wherereliance on water markets is being proposedto improve the economic efficiency of wateruse. As we stated earlier, where marketshave worked well, laws assigning rights,laws describing how rights may be traded,legal systems that enforce the rights andpunish infringements on those rights, and aphysical irrigation infrastructure and irriga-

tion management system capable of allocat-ing water in accordance with water marketsall exist. For most of the developing world,these conditions do not. If they did, stealingof water from tail enders by head enderswould not be as prevalent as it is—and ex-isting systems would be managed muchbetter than they are.

Finally, it is important to take ad-vantage of opportunities to enhanceefficiency and productivity by not lettingthe best be the enemy of the good. Thismight require recognizing, for example, thatfixed user fees for operation and main-tenance costs and public sector manage-ment may be more efficient than turnoverin many situations, even though turnover isattractive on other grounds. Or it mightrequire recognizing the physical limits ofdelivery systems, at least in the short run,that can meet only the simplest ofschedules. If all of these things are doneand done well, we stand a better chance ofmeeting our objectives. But if they are not,or if the current fad or ideology (of gettingprices right in water) replaces a search formore understanding, we may find ourselvesno better off, or worse off, a decade fromnow.

Conclusion: Toward Improved Water Resources Management

A forthcoming study of the “Support Sys-tems” required for sustainable waterresources management indicates the com-plexity of the institutional arrangements re-quired for sustainable, productive water use(Vissia 1997). The study, of a project inColorado in the western USA, also showsthat such arrangements do indeed allow therational allocation of water among compet-ing uses through market mechanisms.

First, Colorado has a strict system ofwater rights. These rights are based on thedoctrine of “prior appropriation,” the firstin use, the first in right. Maas and Andersen(1978) have argued that this system ofrights constitutes an inequitable and ineffi-cient system of water allocation. Neverthe-less (and of crucial interest here) the systemis legally enforced and transparent, andprovides all users, favored or otherwise,

15

with information for planning their opera-tions.

Second, while there is an active marketfor water in Colorado, transactions arefirmly embedded in a legal and admin-istrative structure that carefully reg-ulates external effects. The office of theState Engineer consists of professionalengineers, hydrologists, and others whoinvestigate all technical aspects of proposednew developments and reallocations ofwater.

Third, each of the seven water basins inColorado has its own specialist WaterCourt, which only deals with water issues,and adjudicates all water disputes. Thus aperson who feels that he or she is to be ad-versely affected by a water transaction canlodge a suit in these courts and the courtcan draw on the expert advice of the officeof the State Engineer to advise on the factsof the case.

These few extracts serve to confirm thecomplexity of the institutional frameworksrequired for sound water resourcemanagement; especially when waterallocation and management are passed froma centralized bureaucracy to local entities,with consequent “privatization” of waterresource management. We offer these obser-vations not to recommend an “American”approach to water manage-ment, but ratherto highlight the necessary components ofsustainable, productive management.Absent these basic pre-requisites—the normin most developing countries—the moreextreme variants of privatization, such asfull water pricing and unregulated marketallocations, are likely to do more harm thangood.

But our position can be stated morepositively, and to do this we define a neces-sary and sequential set of preconditions forthe beneficial introduction of market forcesinto the allocation of water, namely that:

• the entitlements of all users under alllevels of resource availability are de-fined and include specified assignmentsto social and environmental uses

• infrastructure is in place to deliver thedefined entitlements

• measurement standards are acceptableto the delivering agency and users

• effective recourse is available to thosewho do not receive their entitlements

• reallocations of water can be measuredand delivered, and third-party impacts(in quality, quantity, time, and place)can be identified

• effective recourse is available to thirdparties affected by changes in use

• users must be legally obligated to paydefined user fees through effective legaland policy procedures

• large-scale transfers of water with andbetween sectors must be subject to ap-proval and relevant charges by regula-tory agencies

With these sequentially interdependentpreconditions in place, we believe that theprivatization of water allocation wouldhave significant benefits; in their absenceimpacts are uncertain. We also believe theimplications of running experiments withpeoples’ livelihoods, especially where wateris involved, to be unacceptable.

Our contention is that developmentand the efficient use of water will be betterserved by the widespread, indeed universal,introduction of the necessary underpinningsand prerequisites to good water manage-ment (assigned water rights, delivery of adefined service). With these extraordinarilydifficult steps in place, further pursuit ofmarket forces in the allocation of water willbe useful.

16

Privatizing water, in the sense of givingfarmers—and markets—a greater role inboth the financing and management ofirrigation, is a promising development. Itsmajor benefits are likely to be more in the

long run than in the short run, by“inducing technological and institutionalinnovations” in irrigation management(Hayami and Ruttan 1985).

Literature Cited

Barkley, P., and D. Seckler. 1972. Economic growth and environmental decay: The solution becomes the problem. NewYork: Harcourt, Brace, Jovanovich, Inc.

Briscoe, J. 1996. Water as an economic good: The idea and what it means in practice. Paper presented to WorldCongress of ICID, Cairo, Egypt.

Chambers, R. 1988. Irrigation against rural poverty. In National seminar on water management—The key to devel-oping agriculture, ed. J. S. Kanwar. New Delhi: Agricole Publishing Academy.

Gazmuri, R. S., and M. Rosegrant. 1994. Markets in tradable water rights: The Chilean experience. Washing-ton, D. C.: International Food Policy Research Institute. Duplicated.

Hayami, Y., and V. Ruttan. 1985. Agricultural development: An international perspective. Baltimore: Johns HopkinsUniversity Press.

ICWE (International Conference on Water and the Environment). 1992. The Dublin statement and report of theconference. Geneva: World Meteorological Organization.

Little, I.M.D. 1950. A critique of welfare economics. Cambridge: Oxford.Maas A., and R. Anderson. 1978. ... and the desert shall rejoice: Conflict, growth and justice in arid environments.

Cambridge: MIT Press.Mishan, E. J. 1976. Cost-benefit analysis. New York: Praeger.Perry, C. J. 1995. Alternative approaches to cost sharing for water service to agriculture in Egypt. Research report 2.

Colombo, Sri Lanka: International Irrigation Management Institute.Pigou, A. C. 1932. Economics of welfare. London: MacMillan.Repetto, R. 1986. Skimming the water: Rent-seeking and the performance of public irrigation systems. Washington,

D. C., USA: World Resources Institute.Robbins, L. 1935. An essay on the nature and significance of economic science. London: MacMillan.Rosegrant, M., and H. Binswanger. 1994. Markets in tradable water rights: Potential for efficiency gains in

developing country water allocation. World Development 22(22):1613-1625.Seckler, D. 1993. Privatizing irrigation systems. Discussion paper no. 12. Arlington, Virginia.: Center for Eco-

nomic Policy Studies, Winrock International.Seckler, D., and R. A. Young. 1978. Economic and policy implications of the 160-acre limitation in Federal

Reclamation Law. American Journal of Agricultural Economics 575-588 (November).Serageldin, I. 1996. Interview: de Borchgrave with Serageldin. The Washington Quarterly, Summer 1996.Svendsen, M., and D. Vermillion. 1994. Irrigation management transfer in the Columbia Basin: Lessons and international

implications. Research paper no. 12. Colombo, Sri Lanka: International Irrigation Management Institute.Vermillion, D. 1996. Impacts of irrigation management turnover: A review of the evidence to date. Draft. Co-

lombo, Sri Lanka: International Irrigation Management Institute.Vissia, R. 1997. Country review of turnover programs. Draft. Colombo, Sri Lanka: International Irrigation

Management Institute.Wade, R. 1982. The system of administrative and political corruption: Canal irrigation in South India. Journal of

Development Studies 18(3):287-325.Wolff, P., and O. König. 1997. Germany’s water and land management associations–performance determi-

nants and problems. In German Association for Water Resources and Land Improvement (DVWK), 1997, 20:Deregulation, decentralization and privatization in irrigation–state functions move the free market, pp217-233.(English version, ed. DVWK). Federal Republic of Germany: International Committee on Irrigation andDrainage.

Research Reports

1. The New Era of Water Resources Management: From “Dry” to “Wet” Water Savings.David Seckler, 1996.

2. Alternative Approaches to Cost Sharing for Water Service to Agriculture in Egypt.C. J. Perry, 1996.

3. Integrated Water Resource Systems: Theory and Policy Implications. Andrew Keller,Jack Keller, and David Seckler, 1996.

4. Results of Management Turnover in Two Irrigation Districts in Colombia. Douglas L.Vermillion, and Carlos Gracés-Restrepo, 1996.

5. The IIMI Water Balance Framework: A Model for Project Level Analysis. C. J. Perry, 1996.

6. Water and Salinity Balances for Irrigated Agriculture in Pakistan. Jacob W. Kijne, 1996.

7. Free-Riders or Victims: Women’s Nonparticipation in Irrigation Management in Nepal’sChhattis Mauja Irrigation Scheme. Margreet Zwarteveen, and Nita Neupane, 1996.

8. Institutional Design Principles for Accountability in Large Irrigation Systems. Douglas J.Merrey, 1996.

9. Satellite Remote Sensing Techniques to Aid Assessment of Irrigation System Performance:A Case Study in India. S. Thiruvengadachari, and R. Sakthivadivel, 1997.

10. A Plot of One's Own: Gender Relations and Irrigated Land Allocation Policies in BurkinaFaso. Margreet Zwarteveen, 1997.

11. Impacts of Irrigation Management Transfer: A Review of the Evidence. Douglas L.Vermillion, 1997.

12. Water Distribution Rules and Water Distribution Performance: A Case Study in theTambraparani Irrigation System. Jeffrey D. Brewer, R. Sakthivadivel, and K.V. Raju,1997.

13. Rehabilitation Planning for Small Tanks in Cascades: A Methodology Based on RapidAssessment. R. Sakthivadivel, Nihal Fernando, and Jeffrey D. Brewer, 1997.

14. Water as an Economic Good: A Solution, or a Problem? C. J. Perry, D. Seckler, and MichaelRock, 1997.

Research Report

Water as an Economic Good:A Solution, or a Problem?

C. J. Perry, Michael Rock, and D. Seckler

INTERNATIONAL IRRIGATION MANAGEMENT INSTITUTEP O Box 2075, Colombo, Sri Lanka

Tel (94-1) 867404 • Fax (94-1) 866854 • E-mail IIMI@cgnet.comInternet Home Page http: //www.cgiar.org/iimi

ISSN 1026-0862ISBN 92-9090-351-1