RESEARCH REPORT

72

IWMI International Water Management Institute

Development Trajectories of River Basins A Conceptual Framework

Fran~ois Molle

Resource

closure . (Degree of water scarcity)

COMPREHENSIVE t ASsessmen

of water management in agriculture

Potential 3

Time

F U T U R E­HARV EST

IWMI is a Future Harvest Center supported by the CGIAR

Research Reports

IWMI’s mission is to improve water and land resources management for food,livelihoods and nature. In serving this mission, IWMI concentrates on the integrationof policies, technologies and management systems to achieve workable solutions toreal problems—practical, relevant results in the field of irrigation and water and landresources.

The publications in this series cover a wide range of subjects—from computermodeling to experience with water user associations—and vary in content fromdirectly applicable research to more basic studies, on which applied work ultimatelydepends. Some research reports are narrowly focused, analytical and detailedempirical studies; others are wide-ranging and synthetic overviews of genericproblems.

Although most of the reports are published by IWMI staff and their collaborators,we welcome contributions from others. Each report is reviewed internally by IWMI’sown staff and Fellows, and by external reviewers. The reports are published anddistributed both in hard copy and electronically (www.iwmi.org) and where possibleall data and analyses will be available as separate downloadable files. Reports maybe copied freely and cited with due acknowledgment.

ii

Research Report 72

Development Trajectories of River Basins:A Conceptual Framework

François Molle

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

ii

The author: François Molle is a senior researcher at the Institut de Recherche pour leDéveloppement (IRD), France, currently seconded to the International Water ManagementInstitute, Colombo, Sri Lanka.

Acknowledgments: The author would like to thank David Molden, Madar Samad, DougVermillion and an anonymous reviewer for their suggestions and comments on earlier draftsof this report.

Molle, F. 2003. Development trajectories of river basins: A conceptual framework. ResearchReport 72. Colombo, Sri Lanka: International Water Management Institute.

/ natural resources / water resources development / water scarcity / river basin development/ crops / water management / aquifers / water use efficiency / agriculture / water availability/ farming / rain / groundwater / technology / environmental degradation / economicdevelopment / ecology / irrigation systems / land productivity / food production /

ISBN: 92- 9090- 524- 7

ISSN 1026-0862

Copyright © 2003 by IWMI. All rights reserved.

Please send inquiries and comments to iwmi@cgiar.org

IWMI receives its principal funding from 58 governments, private foundations, andinternational and regional organizations known as the Consultative Group on InternationalAgricultural Research (CGIAR). Support is also given by the Governments of Ghana,Pakistan, South Africa, Sri Lanka, and Thailand.

iiiiii

Contents

Summary v

Introduction 1

Common Views on River-Basin Development 2

Disaggregating Basin Trajectories 6

The Demand-Supply Equation and Adaptation to Scarcity 10

Basin Trajectories 26

Conclusions 28

Literature Cited 29

iv

v

Summary

The development of societies is, to a large extentdependent upon their resource-base, notably waterresources. Access to and control of water, dependsprimarily on the available technology and engineeringfeats, such as river diversion structures, canals anddams. As growing human pressure on waterresources brings actual water use closer to potentialceilings, societies usually respond by adoptingconservation measures and by reallocating watertowards more beneficial uses.

Several frameworks and diagrammaticrepresentations of great heuristic value have beenproposed to conceptualize the development of riverbasins along these lines. They have brought intosharp focus the crucial phenomenon of basinclosure, making it much more clearer andunderstandable. At the same time, the simplicity ofthese representations may not allow one to capturethe deeper heterogeneity of the processes thatunderlie the historical relationship between aparticular society and its water resources.

This report first critically reviews variousconceptualizations of river basin development thatcan be found in the literature on water resources. Itthen shows that distinguishing between severalcategories of water sources, instead of consideringthem as a whole, provides additional insight into howwater resources are put into use and are controlled.The report proposes a disaggregated view ofdifferent categories of water (rainwater, stream water,regulated surface water and underground water),allowing for a better understanding of how the actualand potential use of the different sources of waterrelate to each other, and of the scope forimprovement.

A typology of societal responses to waterscarcity is then presented. It emphasizes the needto distinguish between responses devised by thestate at the national level and those of individualfarmers and small groups or communities, although

both responses are partly interdependent. Whileemphasis is often placed on state policies,adjustments made by local actors also oftenappear to be very significant. Responses from boththe state and the local actors can be further brokendown into three types (supply augmentation,conservation and reallocation). It is shown,however, that these categories are not asstraightforward as they appear to be. Because ofthe interconnectedness of users throughout thehydrological cycle (particularly upstream/downstream and surface water/groundwaterlinkages) they are not purely additive. Whetherthese responses occur sequentially, is examined byreferring to several empirical situations whichillustrate that specific historical evolutions andpatterns that do not accord with the reviewedframeworks can frequently be encountered. A fewelements, which appear to be crucial in shapingresponses are then singled out, including the natureof the state and state/citizenry relationships, theimpact of “shock events,” the nature of the politicaleconomy, and the conditions of agrarian change.

Existing linear visions of basin developmenttend to be based on economic rationality or onconcepts of social adaptiveness that are toorestrictive or too difficult to evaluate. Societalresponses to the scarcity of resources, at both thelocal and the state level, are not driven solely byeconomic considerations or locally perceivedneeds. It is argued that they must be understoodnot only on the basis of hydrological, physical oreconomic constraints, but within a wider politicaleconomy framework that considers the distributionof human agency and power among actors, as wellas their respective interests and strategies. The lastsection attempts to devise a new framework that iscomprehensive enough to trace the evolution of awide variety of river basins and to avoid reducingthem to a single, oversimplified form.

vi

1

Development Trajectories of River Basins:A Conceptual Framework

François Molle

Introduction

The evolution of societies and, in particular, thedevelopment of their productive activities, arepartly determined by the available naturalresources. Water resources, defined within theirriver-basin environments, are mobilized fordomestic use and food production, and are,therefore, of paramount importance. In earlytimes, when population density was still low,people adapted agriculture to abundant land andwater resources. Crops, cropping calendars andelaborate subsistence techniques were attunedto natural conditions of soil, topography, climateand hydrology. Gradually, entire landscapes andwaterscapes were crafted, especially underconditions in which it was easy for a small groupof individuals to work together to construct atemporary diversion of river water, or underconditions in which early states exercised tightcontrol over large populations to develop large-scale schemes for flood control or irrigation.

In the face of growing human pressure andin the course of time, however, the supply ofnatural resources reaches its capacity. Riverbasins, and often land frontiers, “close,” in thesense that most water resources are committedor depleted with very few remaining untapped.Water needs come to exceed water availability inriver basins and people find that their productiveactivities are constrained by water shortages.This prompts crises that, in turn, lead to

technological innovations and to adjustments andinterventions, both in institutions and in theeconomy. Societal adaptations to changingrelationships between people and water-basinresources seem to move through a unilinearsequence of stages, similar to those that Rostow(1962) has theorized for economic growth.

This sequential model of the evolution of theriver-basin society has the merit of providing ageneric framework to address an issue ofworldwide relevance. It also has a great heuristicvalue, in that it brings into sharp focus thecrucial phenomenon of basin closure, making itunderstandable and straightforward. At the sametime, however, its simplicity may not allow us tocapture the contingency and deeperheterogeneity of the processes that underlie thehistorical development of societies. This reportfirst reviews several approaches to conceptualizeriver-basin development. It then shows thatdistinguishing between several categories ofwater sources provides additional insight intohow water resources are put into use and arecontrolled. A typology of responses to waterscarcity and a review of some crucial aspects ofriver basin development, are then used to devisea new framework, that is both comprehensiveenough to describe the evolution of a widevariety of river basins, and avoids reducing themto a single, oversimplified model.

2

The literature provides a few (and recent)attempts to theorize the phases of waterresources development. Molden et al. (2001b),for example, distinguishes between threephases: development, utilization and allocation(figure 1).

In the first phase of basin development,water use is limited to rain-fed agriculture andrun-off-river- water utilization. Dams areconstructed in the most convenient locations,either to produce energy or to irrigate, while thesupply for domestic purposes remainsquantitatively negligible. Large-scale irrigationsystems, therefore, may be constructed. Theamount of water effectively used for agricultureand other beneficial uses is less than what isavailable, much of which simply flows to the seaor to the next downstream basin. Watermanagement tends to be based on demand, andconflicts, therefore, rarely arise. Water qualityremains good and releasing more water fromreservoirs easily mitigates pollution. Ecologicalsystems and environmental functions are notsignificantly altered.

In the next phase, the utilization phase,shortages of water begin to appear in the driestyears and during unusually dry seasonal spells.Storage dams are added to the river as asafeguard against shortages, but adequate sitestend to be rare. Improving management,rehabilitating infrastructures and conservingwater become critical issues, while pollutionproblems and competition within irrigated areasbecome apparent.

As the basin nears closure, sectoralallocation becomes a point of tension (allocationphase). Efforts are directed at allocating watertowards the most economically valuable uses,and new institutions evolve to address inter-sectoral competition and manage river-basinresources in an integrated manner.

The account by Molden et al. (2001b)identifies consecutive responses to waterscarcity, starting with water-resourcesdevelopment, followed by improvements inefficiency and sectoral management, andculminating with modernization and inter-sectoralreallocation. The linear logic of this model is

Common Views on River-Basin Development

FIGURE 1.Schematic representation of river basin development.

Source: Molden et al. 2001b.

3

depicted in a table that compares the problemsfaced by, and the prominent issues associatedwith, each of the three different phases towardsa final situation, where water use is reserved foractivities with a high economic value,groundwater and pollution are regulated, andbasin-level coordination and integratedmanagement are used to mitigate conflicts.

The description of Molden et al. draws onearlier works of Keller et al. (1998) and Keller(2000), which also break down river-basindevelopment sequences into three phases, eachone being subdivided into two further subphases(figure 2). In the first phase (exploitation),demand is initially satisfied by the simplediversion of river water and by pumping waterfrom shallow aquifers. As pressure for waterincreases and these tactics become inadequate,people evolve a second subphase of exploitation,marked by the construction of large-scalestorage capacity and by a capacity to pump

water from deep aquifers (“a straightforwardengineering solution”). When the available waterresources in a basin approach full development,the conservation phase begins. In the earlystage, demand reduction and an increase inefficiency are the usual means to bring usageinto conformity with the availability of resources.Return flows are reused, but savings in quantitygenerally result in problems of water quality(pollution and salinity) that need to be addressedin a later phase (water treatment, reclamationand salt disposal). Despite these efforts, it is notunusual to see imbalances between theabstracted water and the annual renewablesupply, especially with the overexploitation ofaquifers. The third phase, or the augmentationphase, starts when the basin is fully closed andadditional supply must be brought from outsidethe basin, typically from neighboring basins withan excess supply (if any) or from the sea(desalinization).

FIGURE 2.River-basin development phases.

Source: Keller et al. 1998.

4

This framework is also based implicitly on alinear vision of the development of water usetowards a “mature” situation, where uses areattuned to the renewable supply, while waterquality is maintained and environmental-flowrequirements are ensured. The evolution isdepicted as a “natural progression of waterdevelopment in a river basin.” The twosequences are similar, but they differ in somedetails.1 In particular, the latter sequencespecifies a phase of augmentation after demandmanagement has been effectively implementedand it sees groundwater use as a characteristicof the earliest stage, unlike the former.

These stages are broadly consistent with avision of development, described in economicterms, as shown in figure 3 (Hayami et al. 1976;Kikuchi et al. 2002). Development starts with theexpansion of rain-fed agricultural land. A point is

reached (point t1) where the marginal cost ofopening up new lands exceeds the marginal costof developing irrigation. In the constructionphase, the cost of surface irrigation systemsrises as irrigation expands into increasinglymarginal areas. However, the development ofnew technologies, specifically adapted toirrigated agriculture (green-revolutiontechnologies), shifts the marginal cost2 curvedownward from I to I1. Eventually, a point t2 isreached where the cost of new irrigation exceedsthe cost of investment in effective utilization.Depending on the situation, more effectiveutilization and better performance can beachieved in a number of ways, such as throughbetter control and management of surface-waterflows, recycling water by pumping from drainageditches, and by developing groundwaterresources (curve W), (Barker and Molle 2003).

FIGURE 3.Hypothetical development paths of agriculture by means of land and water development.

1The former is presumably derived from a situation observed in developing countries, while the latter is clearly influenced by the experienceof the USA (more particularly that of California).

2This, of course, reflects the situation of the producer and does not account for the costs of developing such technologies.

5

This economic view focuses on technologyand on the costs of development and watermanagement. It neglects the transaction andpolitical costs of institutional reforms. Anotherlimitation of this economic view is that it doesnot account for the fact that upland expansionmay restart at a certain point, because possibleirrigated areas are already saturated. New landsare then put into use and a new phase ofagricultural expansion in marginal upland areasbegins, as seen, for example, in Vietnam (deKoninck 1997).

These generic descriptions of basindevelopment are based on the evidence thatwater resources gradually come under growingpressure, and that society devises strategies tocope with the new situation, thus relieving thetension, albeit temporarily. Another conceptualframework for water-resources development hasbeen developed by Turton and Ohlsson3 (1999).They also distinguish between three phases, inwhich water scarcity is answered by differentstrategies. During the first phase (supply), “theindividuals surrender their responsibility forproviding water for themselves to a centralauthority,” which embraces a hydraulic missiondevoted to large-scale hydraulic infrastructure.Later, water deficits arise and demandmanagement (demand phase) first consists ofraising the efficiency of use, and then ofoperating intra- and inter-sectoral reallocation ofwater resources. In a third phase (adaptive), thesociety has to cope with absolute water scarcityand must bring water demand in line with asustainable level, defined by annual renewableresources. This description is similar to the

preceding ones,4 but Turton and Ohlsson’s (ibid.)analysis goes beyond the quantitative analysis ofimbalances in supply and demand, anddiscusses the societal responses induced. Theyare interested in the links between the natureand legitimacy of the “coping strategies,” devisedby decision-making elites and by social stability.In particular, they see the emergence of waterdeficit conditions (demand phase) as a crucialperiod, where environmentalism is likely to gainmomentum, and where the adaptive capacity ofthe social entities concerned is put on trial.

Turton and Ohlsson (ibid.) distinguish betweena “first-order scarcity” of natural resources (water),and a “second-order scarcity” of the socialresources, required to adapt to the former. Theyhave reformulated their framework as the “turningof the water screw,” where engineeringdevelopment, improvements in end-use efficiency,and allocative reforms, are seen as the threesuccessive responses that are needed to addressa persistent water scarcity. Although they claim togo beyond the kind of linear vision describedabove, this reformulation does not really provide aconceptual alternative, although it adds a moreintuitive visualization of the fact that every turn ofthe water screw is “tighter.” The social resourcesneeded, therefore, to achieve this reformulation areincreasingly high.

In addition, Turton and Ohlsson (ibid.) arealso concerned with distinguishing betweensituations in which social resources enablesociety to adapt, and others, where conflicts andsocial unrest are likely to be the outcome. Inparticular, they do not necessarily posit thatcrises generated by water scarcity will be

3Their framework is not specifically focused on river basins but, rather, on water resources at the national level, because of the importancegiven, in their discussion, to state policies. However, water scarcity is discerned at the basin level and it is, therefore, consistent to apply theirconcepts at that level.

4Although the demand management phase includes both conservation and allocation strategies, the adaptive phase is construed as an addi-tional phase, where social stability is at risk and major societal adjustments are needed.

6

The first limitation to the development modelsreviewed above, is that they consider aggregate“annual supply” or “available resources” withoutdistinguishing between the different sources ofwater. This section proposes a disaggregationand categorization of different “sources of water”that may help better understand how a societymanages and uses its water resources. The fourcategories described here can also be viewed aswater sources that can be ranked based on thedegree of control that the society has on them.

Water-Source Categories

At the basin level, water availability isdetermined primarily by rainfall.6 Rainwater isintercepted and evaporates, or is captureddirectly by cultivated fields, either because itdirectly infiltrates the soil or because it does soafter being retained in the fields by furrows,bunds or small dikes. Part of the water that fillsthe soil reservoir is eventually consumed bycrops (effective rainfall), while the remainingwater infiltrates lower layers of the soil, partlyrecharging the streams. The total effectiverainfall depends, of course, not only on thecropping area but also on the terrain (soil type,

slope, etc.), and on how crops are cultivated(cropping techniques, mulching, etc.).

Stream water is unregulated water from the riversystem, generated by direct surface orsubsurface runoff. It may be diverted (by gravity)or abstracted (by pumping) for production orother needs. Human intervention is necessary todivert, convey and apply stream water. Streamwater occurs independently of reservoir operationalthough, in many cases, its flow adds to thatreleased from dams.

Controlled water is water that has been stored insurface reservoirs (dams, lakes, etc.),discounting the average loss by spill, evaporationand infiltration, or underground reservoirs(aquifers), and that can be used at will andpurposely, once released or extracted. Even for agiven set of reservoirs, the amount of controlledwater is not necessarily constant and maychange over time (e.g., change in dam inflow,siltation of reservoirs, etc.). Likewise, the safeyield of the aquifer may change slightly if therecharge is altered in consequence of changesin the hydrological regime (for example, if water-harvesting infrastructures are constructed).Average values are considered here, which

5While the other frameworks do not explicitly state that the evolution of river basins must be seen as a positive move towards a more desir-able state, it seems that the ideology of progress is so pervasive that river basin development is often implicitly interpreted as such. This wasillustrated, anecdotally, during a seminar in which different river basins were tentatively placed on the time axis of a chart. A Nepalese col-league, uncomfortable with seeing a Nepalese basin occupying the leftmost position, expressed his disagreement with the classification,before being explained that many people wish their basin were still at that stage…

6Aquifers may extend beyond the limits of the watershed and trans-basin diversions may also occur, which may alter the typical situationdescribed herein. Desalinization of seawater can also be considered as an inflow of freshwater, not pertaining to the basin itself.

eventually solved,5 especially with regard toecological impacts, thus rightly shifting the

attention to the adaptive capacity or socialcapital of societies.

Disaggregating Basin Trajectories

7

somehow obscure the interannual variability thatis paramount to the management of regulatedresources.

Potential controlled water is the average totalamount of water generated in a basin that canbe stored in reservoirs that are technically andeconomically feasible, societally acceptable, andenvironmentally suitable, plus the safe yield ofthe aquifer. This limit is partly relative andsusceptible to change. For example, uneconomicoptions at one point in time (e.g., trans-basindiversion), may become acceptable when thevalue of water rises or, on the other hand, whendam construction may run into too muchopposition from the civil society and may beruled out. Actual controlled water may exceedthe potential (renewable) controlled water ifgroundwater abstraction exceeds the safe yieldof the aquifer, and/or if dam management leadsto tapping carryover stocks (i.e., if the averageyearly release from the dam temporarily exceedsthe average net inflow).

Distinguishing between such categories isfurther complicated by the fact that both theabsolute value of the different categories ofwater and their relative share of the total amountof water used vary over time. If there is a rainyseason, rainwater will generally predominate,while in the dry season (or in arid regions),controlled water will be the primary source forthe crops. Though a period of one year allowsone to get an aggregated overview of thesituation, it does not provide any informationabout the constraints faced in particular periodsof that year.

General Evolution of Water Use

Figure 4 provides a view of (hypothetical)evolutionary patterns. It allows us to compare theamount of water, available between the differentcategories of water sources, used by differentactivities. It also indicates how the abstraction ofcontrolled water compares with the potentialvalues.

The rainfall quadrant (lower-left quadrant)shows the amount of rainfall that has beenutilized in both irrigated and nonirrigated areas(which may also include forests), while thestream-water quadrant (upper-left quadrant)indicates how much water has been divertedfrom streams for use.

The evolution of “surface water” (upper-rightquadrant) exemplifies a (particular) case in whichthe potential available water is slightly declining(e.g., dam siltation or upper catchmenthydrological changes), and in which the basin isreopened by a trans-basin diversion (indicatedby a last step, t), and also indicates theincidence of desalinization (slight increase afterstep d). The amount of water depleted orcommitted to in-stream and downstream needs(including environmental services) becomescloser to the potential value. The differencebetween these two curves accounts for the shareof dam releases, which is non-beneficial,7 losteither to the atmosphere (evaporation) or tosinks (including the sea).8 The groundwaterquadrant (lower-right) shows an increase inwithdrawals, which exceeds the aquifer safeyield before returning to a level under this limit(an optimistic scenario).

7The case of hydropower makes this distinction less straightforward, in that all the water released does provide the benefit of energy genera-tion. In a closed basin, however, providing that hydroelectricity is not the main source of energy generation in the country, the downstreamusers will generally be granted priority. Their needs will dictate water releases, notably their timing, and hydroelectricity will only be a byproductof these releases.

8For some ecologists, such losses do not exist, since the whole natural water regime is constitutive of ecological systems. We have assumedhere that minimal environmental flows can be defined and incorporated as one term of demand.

8

Figure 4.

Changes in water use at the basin level.

It must be made clear, at this stage, that theabove figure does not depict a water balance butthat it only compares the contribution of thedifferent sources of water utilized by humanactivities (or needed for environmental services).In that order, the four quadrants disaggregatewater sources, from the least to the mostcontrolled, where “control” is taken as a measureof the reliability of the source in terms of quantityand timing at the seasonal level. Undergroundwater is generally safe in the short term andusers can extract a predictable volume.Regulated surface water is also reliable, butincludes some uncertainty of the coming inflow,and its distribution over long distances mayresult in an uncertain supply. Stream water is

dependent on natural flow (i.e., rainfall andrunoff) and is, therefore, subject to little control.Finally, rainfall is the most unpredictable source,in terms of quantity and timing.

The return flow from the use of these fourdifferent categories of water may contribute tothe potential of another category. For example,the return flow from rainwater may accrue tothe stream, surface water or underground-water categories. Figure 5 is derived fromfigure 4, and shows both water that isabstracted and that which is effectivelydepleted. The ratio of the two values providesa measure of the efficiency of the use of eachcategory. For rainwater, the depleted fraction isthe effective rainfall.9

9There are some difficulties in establishing the amount of rainfall utilized by plants because they may also tap groundwater that entered theaquifer in other parts of the basin (or even outside of it). In basins with a certain degree of closure, the water table will often be low and theamount of water taken up by plants will generally originate from the in-situ infiltration of rainfall (or irrigation water).

9

FIGURE 5.Changes in water used and depleted at the basin level.

Breaking down the different types of wateruse highlights, not only their respective sharesand the degree of overall control on waterresources, but also the main sources, thepotential of which is not fully realized. Inpractice, sometimes it will be difficult to separate,not only stream water and rainwater (e.g., whenwater use is very scattered: numerous smalltanks and run-of-river systems in Sri Lanka), butalso stream water and controlled water (e.g.,when there are significant surface or sub-surfaceside-flows between the dams and the irrigatedareas, which add to dam releases). Yet, thisdistinction is essential for the betterunderstanding of water use: for example, it isoften the case that irrigation mainly uses stream

water that would otherwise be lost. The meaningand implications of stating that agriculture makesup, almost, 80 percent of total water use, thusdiffer from a situation in which mostly regulatedwater is used.

A water-accounting procedure10 thatconsiders these different categories of water, aswell as their spatiotemporal interrelationships, isa more complex exercise than figure 4 suggests.What is clear, however, is that the whole pictureof water accounting is changed significantly if weconsider controlled water in isolation, in order toestimate potential gains from improvedmanagement, or if we include stream water orrainwater in the picture (see Molle and Aloysius,forthcoming, for the case of the Chao Phraya

10See fundamentals of water accounting in Keller et al. 1996; Perry 1999; Molden and Sakthivadivel 1999.

10

and Walawe river basins). Therefore, it is usefulto disaggregate the temporal evolutions of thedifferent segments of water resources, to bettergrasp what the evolutions have been and wherewe are at present, thus avoiding the limitationsinherent in the representations described in thefirst section.11 Even with this disaggregation,

however, the figure referring to surface water doesnot indicate the water shortages that may occur inspecific periods of the year. Likewise, since thecurves are based on moving averages of yearlyvalues, they also do not account for shortagesinduced by the improper inter-seasonal andinterannual regulation of controlled water.

11For example, because surface water and groundwater are considered together, the difference between the actual and the utilized regulatedwater may not indicate where the remaining potential lies, or where excess abstraction occurs.

The Demand-Supply Equation and Adaptation to Scarcity

The conceptual frameworks, reviewed above,hypothesize the nature and the chronology of thechallenges and responses that society faces,with respect to water resources. The underlyingassumption is that population growth putspressure on water resources and that this, inturn, creates challenges for the society and leadsto a gradual closure of the basin. The focus here ison basins with significant and growinganthropogenic pressure, and it is reasonable toretain both the gradual closure and consequentadjustments/conflicts as general phenomena. Whatis debatable, however (and questioned in thissection), is whether the sequences described inthese frameworks are found in all basins, andwhether the responses are similar, or if a moreinclusive framework can be designed.

It is hypothesized here that societal responsesto water scarcity comprise a set of strategies,defined both at the individual/community level andat the state level, and is elaborated or induced,based on several location-specific factors, withoutany other assumption about a possible “natural”order or sequencing.

Range of Responses

While Turton and Ohlsson (1999) emphasize theway “technocratic elites” devise “copingstrategies” to respond to water scarcity, theytend to overlook the multileveled adjustment ofsociety to basin closure. In particular, they do notaccount for local adjustments made by individualusers, or groups of individuals, and by localmanagers/officials, in addition to the state. Thisconstrains the analysis of processes and leadsto envisioning future changes as being governedby the decisions of the state and its elites.Broadening the scope for analyzing actors’responses provides a richer understanding of theprocesses at work. The distinction betweenmicro/local and macro/global adjustments will beherein, identified.

These two types of adjustments can befurther broken down into three categories.(These are similar to the three consecutive stepsmentioned by Turton and Ohlsson (ibid.) but, atthis stage, no hypotheses are made on the orderin which they materialize).

11

Supply responses: These consist of solvingwater scarcity by augmenting the supply fromexisting sources (foremost, increasing thequantity of controlled water), as well as fromtapping additional sources. Typically, this is donenot only by constructing new reservoirs ordigging more tubewells, but also by divertingwater from neighboring basins, by desalinizingseawater (which is tantamount to importing andtreating water from a sink), by artificialgroundwater recharge or by cloud-seeding.

At the local level, farmers may tap shallow ordeep aquifers, and may also invest in localstorage facilities (most commonly farm ponds,which are used to store excess irrigation flows orrainfall). They also develop the conjunctive useof water, by using water from drains, rivers,ponds and even by pumping from irrigationcanals when the water level does not allow forgravity inflow to their plots. Thus, they broadentheir access to a variety of sources and augmenttheir individual potential supply of water.

It is already apparent here that therelationship between local dynamics and thebasin (macro level) must be considered. Forexample, pumping water from drains increasessupply locally, but not necessarily at the macroscale, if the pumped water was to be reuseddownstream.

Techniques and interventions aimed atcapturing more rainwater, for example, water-harvesting techniques aimed at increasinggroundwater recharge, can also be consideredas means to augmenting one’s effective supplyof water.

At a more global level, the import of foodstuffis also an indirect way to increase water supply,or at least the water supply as a necessaryfactor of food production and security. Thistransaction is often referred to as virtual water,because of the water used by the crops thatcomes embedded in foodstuff.

Conservation responses: The key phrase here is“efficiency in use.” Conservation refers to makinga better use of existing resources, withoutincreasing the supply or the sources of water.Line agencies may not only implement structuralmeasures, such as lining canals, controllingleakage in pipe systems, or treating and allowingthe reuse of wastewater, but may also resort tononstructural measures, such as improving damor canal management (so that unproductive ornon-beneficial releases are not lost12 to the sea)and establishing rotations or other arrangementsfor better scheduling. The state is alsoinstrumental in devising and enforcing policiesthat may elicit water savings, such as waterpricing, rationing and quotas. In these two lattercases, conservation aims at “doing as well asbefore with less supply,” rather than “doing morewith the same amount.”13 The state may alsosupply innovations derived from research (plot-level water management, improved varieties, andcultivation techniques).

Improved coordination between users andinnovative organizational patterns may alsocontribute to water savings and can bementioned here, particularly in the setting ofwater-user groups or river-basin organizations.

12Again, this does not mean that all water reaching the sea is lost, since minimum flows are essential to maintain ecosystems and controlsalinity intrusion in lower reaches of the rivers. These losses must be understood as the extra water that reaches the sea after abstraction byall other users, including environmental services.

13These two situations are rarely distinguished, although they differ fundamentally. In the latter case, the incentive to conserve water is strong,and users are prompted to make effective adjustments. In the former case, users may or may not be willing to take the necessary steps toconserve water.

12

Subsumed in the “conservation strategy” are,more generally, better management practicesthat not only tend to conserve water, but mayalso increase equity or reliability in the supply ofwater.

At the local level, farmers and groups offarmers are not passive.14 Water may be saved,for example, by shifting calendars or raisingbunds around rice fields (to make a better use ofdirect rainfall), adopting adequate cultivationtechniques (such as mulching, alternating thewet/dry water regime in rice farming, shorteningfurrows, etc.) or by choosing crop varieties witha shorter cycle. These farmers or groups offarmers may also invest in water-savingtechnologies, such as micro-irrigation. However,such technologies may sometimes result inbetter control of irrigation doses and, in anincrease in the amount of water depleted byevapotranspiration (see an example for Chile inCai et al. 2001). The water saved thereby mayalso be used to increase the farmers’ irrigatedarea (see Feuillette 2001, for an example inTunisia), further reducing the return flowavailable to downstream users. Therefore,expected water savings at the basin level may ormay not occur and may also have an adverseimpact on other users. Here, too, a case-by-casecautious analysis must be carried out to fullyunderstand the relationships between the locallevel and the macro level.

Allocation responses: A third type of strategyconsists of reallocating water from one user toanother, either within the same sector (e.g.,within or between irrigation schemes) or acrosssectors. This reallocation may be justified by the

need to raise water productivity, but may also beaimed at easing tension on the resources byfavoring uses which enhance land productivity,food security or equity, or which reduce conflictsand protests. This broadens the principle thatwater should be moved to uses that areeconomically more beneficial, to whichreallocation is usually attached, towards anapproach whereby reallocation is a means toreduce the overall pressure on water resources.

Reallocation can occur within the farm, whena farmer chooses to direct his/her limited waterresources to the crops that give him/her a higherreturn per m3 (assuming that risks and marketingconditions are similar for all crops).15 At theirrigation-system level, managers allocate water,based on a set of factors. This allocation can bemore or less even in terms of water duty perhectare, but it can also be driven by otherconsiderations: for example, areas that are toodistant or that are endowed with sandy soils maybe excluded, because they incur too high losses.On the other hand, reducing canal head-end/tail-end differences is conducive to higher equityand, often, to increased economic efficiency(Hussain et al. 2003).

At the basin level, managers also have theoption of reallocating water according to a givenpriority system. Within the agricultural sector,water can be shifted from one area to anotherwith comparative advantages regarding waterproductivity (typically, areas with orchards oraquaculture). The rationale for inter-sectoraltransfers is generally economic: water ischanneled to cities for domestic and industrialuses before to agriculture, where the economicreturn of one cubic meter of water is much

14Contrary to the central argument of demand management through water pricing, farmers facing water scarcity do not need to be taxed tomake sense of it. Experience with reduced supply quickly translates into responses that may, however, be encouraged or constrained bypublic policies (e.g., impact of free rural electricity supply in some states of India upon groundwater depletion).

15Again, clear-cut definitions are not always possible. A crop shift can be a conservation measure (less water used), or a reallocation to amore productive (crop) activity (more $/m3), or both.

13

lower.16 Such allocation decisions are, of course,not only potential sources of conflicts andtension, but are also sometimes mediated bymarket mechanisms or negotiations. Groups ofusers within an area (e.g., a catchment or anirrigation scheme) may also agree to renegotiaterights in order to ease tension. To some extent,giving up agriculture (or any other water-consuming activity) can also be interpreted as astrategy that allows one’s water allotment or rightto be reallocated to other uses.

At a more global level, water can bereallocated to nonagricultural uses at the cost oflower food production, if imports are increasedaccordingly (see virtual water mentioned earlier).

The conservation and allocation responsesare often pooled together under the concept ofwater demand management, which can betypified by “doing better with what we have,” asopposed to supply augmentation strategies.

Figure 6 synthesizes these differentresponses and shows how the three categoriescan be broken down into two sublevels. It mustbe emphasized that the definition of thesecategories cannot be totally freed from ambiguitybecause of the complex relationships betweenlocal and basin-level processes: a farm pond dugout by a farmer can be seen as conservation if itcaptures some canal water that would be furtherlost (to sinks, to non-beneficial evaporation or byflowing out of the system), but it may also be areappropriation if this water was ultimately to beused by downstream users. The pond can alsobe considered as an augmentation of the supplyto that particular farmer, if it captures runoff thatwas lost earlier. In practice, because it is notalways possible to establish what fraction of aspecific return flow is eventually lost or reused, itmay not be possible to define a preciseterminology.

16Contrary to the widespread idea that economic incentives are necessary to ensure the reallocation of water to economically more produc-tive uses, inter-sectoral reallocation is often well administered by the state, and it is rather the consequences of this reallocation, in particularupon agriculture, that need to be addressed.

FIGURE 6.Types of responses to water scarcity.

14

The same difficulty arises when new tanksare constructed in the catchment area of anexisting reservoir. Normally, isolated tankscapture runoff and transform it into controlledwater but, in that case, the flow captured bythese new tanks was already potentiallycontrolled by the reservoir into which it wouldhave flowed, if the new tanks had not beenadded. Overall, there is no or little increase17 ofcontrolled water; instead there is a mere spatialredistribution of the resource. Likewise, trans-basin diversions, and even cloud-seeding, canbe considered as supply augmentation from anarrow/local point of view, but they can also betreated as spatial redistribution (or reallocation)when seen on a wider scale.

These examples show that the measurestaken at the two levels (individual and collective)schematized in figure 6, are not purely additive.This fact, however, should not be treated as“noise.” The relationship between micro- andmacro-processes is, in fact, at the core of theanalysis. Because the closure of river basinsresults in a growing interdependence of theusers within the basin, one must carefullyanalyze how the paths of the different surfaceand underground flows are interrelated, and howany local intervention that modifies the quantity,the quality or the timing of one of these flows,impacts on the whole system.18 What is stored,conserved or depleted at point A, dictates whatis available at point B, further downstream.

Likewise, it is not always possible todissociate the decisions taken by farmers fromthe economic environment that is shaped bystate policies. The adoption of micro-irrigation,for example, might be an individual decision,influenced by the availability of subsidized creditsupplied by the state. (This is symbolized infigure 6 by the arrows that link the two layers).

The types of responses shown in figure 6can also be distinguished according to the timeframe of their implementation. Some of them,particularly those implemented by the state, needmany years to be in place (e.g., dams, trans-basin diversion, and water laws), while someresponses unfold within a more limited period(one season, one or two years). This is the caseof short-term responses to water crises (typically,implementing rotations, changing crops, drillingwells, etc.). In other words, one may distinguishbetween responses to long-term imbalancesbetween supply and demand, and responses toshort-term ones (punctual water shortages orcrises).

Actions taken by people may not bemotivated by the type of strategy they comeunder: conservation responses, for example, mayarise from an interest in increasing farm incomerather than in conserving water.19 Such actionsmay not be “responses” or “strategies” directlyrelated to water scarcity, but may have acoincidental effect on the basin hydrology. Moregenerally, responses also have an effect on

17An increase occurs if the amount of spill in the reservoir—now reduced—was formerly lost out of the system. This is more or less significantdepending on the dimensions of the reservoir in relation to the hydrologic regime (and, of course, depending also on the type of downstreamusers and ecosystems present).

18See Keller (2000) with regard to water management in California, and Molden et al. (2001a) who developed the concept of hydronomiczones, with a call for reason interventions, based on a zoning of river basins. “Hydronomic zones are defined to characterize the combinationof hydrologic and water-use settings within a basin. The zones are based, primarily, on the considerations of the outflow of water from theparticular areas” (Molden et al. 2001a).

19Micro-irrigation provides a good example of such a situation. In many cases (probably a majority of them), it can be shown that the adoptionof such technology is motivated not by water scarcity itself but, rather, by labor shortages, the need to better control the quality of production,the ease in applying (liquid) fertilizers, or the use of plastic mulch that precludes more rudimentary irrigation (e.g., Jordan).

15

factor uses and often change the distribution ofbenefits. For example, investments in capital andlabor-intensive water-saving technology have animpact on local suppliers of equipment, and onthe labor market.

Figure 6 also highlights that actors within thesystem are not passive and inactive. On thecontrary, they respond individually andcollectively to the growing scarcity of water, justas agrarian systems respond to changes in therelative availability of other production factors.This has been shown by case studies, such asthose by Zilberman et al. (1992) for Californiaand by Molle (forthcoming) for Thailand. State-driven responses are only a part of thetransformation, although officials tend to seerural areas as globally static, and malleablethrough public interventions (infrastructures orotherwise), overlooking the constant endogenousadjustment of rural households and communities,as well as of line managers, to changingconditions.

Context-Specific Responses

In addition to categorizing the diversity ofsocietal responses to water scarcity, aconceptual framework must also provide adegree of prediction on the range of adjustmentsa given society is likely to pick up or, rather, onthe measures that are made possible,constrained, or that are ruled out, in the contextunder consideration.

Keller et al. (1998) place emphasis on theeconomic logic of the sequence of development.At any point in time, the cheapest solutions areselected, from simple flow diversion throughdesalinization. Their sequence follows the cost ofthe main options available in large basins of thewestern USA. In Ohlsson and Turton’s (1999)approach, the logic of the succession is basedon a scale of complexity, with the solution ofwater-scarcity problems demanding ever-

increasing levels of social resources. It is thusassumed that hydraulic development is theeasiest response, and that its exhaustion leadsto conservation efforts, later followed byallocative decisions and adjustments. The lattertwo are regarded as much more sensitive andare prone to generating social conflicts andwidespread disruption. These two analytical gridsprobably apply to many cases, but may notcapture important nuances found in variedsituations. The following examples provideinstances in which the “natural” sequencing, orpart of it, as proposed by one of theabovementioned frameworks, does notsatisfactorily represent the historicaltransformations observed.

• Sakthivadivel and Molden (2001) havecompared five basins said to be atdifferent stages of evolution, and havefound that the problems faced by thesebasins were different. However, some ofthe problems encountered were notthose that would be typical of the phasein which each basin was assumed to be.For example, in the Singkarak-Ombilinbasin in Sumatra, which is considered tobe at the beginning of the utilization/conservation phase, it seems that waterallocated to nonagricultural activities andtrans-basin diversions threaten to throw thebasin directly into the third phase, wherewater rights and reallocation rules need tobe defined. The East Rapti basin in Nepalis an open basin, with only 5 percent ofthe water resources being used byagriculture. In spite of this, water pollutionfrom industries, competition for river waterduring the dry season among wildlifesanctuaries, tourist requirements,ecological requirements and human useare apparent problems that are normallyassociated with the later phases ofdevelopment.

16

• More generally, most of Africa, with theexception of northern and South Africa,is still characterized by limiteddevelopment of infrastructures. Thenumber of large dams in a continent thatmakes up 20 percent of the world areaamounts to only 1,192 out of a total of24,864 dams worldwide (IWMI 2002).Likewise, the irrigated area makes uponly 6 percent of the total cropping areaof the world. The reasons that seem tohave prevented the emergence of thedevelopment phase are not fullyelucidated.20 The decline in crop prices inthe recent years compounded thissituation, making both subsidized/publicand private investments in agriculture/infrastructure, economically unviable.This means that a conjunction ofendogenous and exogenous factors hasconstrained the large-scale developmentof water resources, while problemsrelated to other phases have emerged.

• In some basins, the tapping of shallowaquifers occurred in the inception phaseof development (Keller et al. 1998), butin many other cases, the spread of wellswas a late reaction to poor access tosurface water (see Thailand, Bangladesh,India, etc.). The availability of technologyand the investment capacity of usersmust also be considered.

• Very often too, large dams were notconstructed to increase supply, but tocontrol floods and generate electricity.Because their potential has been then—

often later—taken advantage of todevelop irrigation areas, demandconsequently built up as a result of thenew supply, rather than demand forwater being a driver of supplydevelopment. This has often been thecase in the river basins of the UnitedStates of America (Wengert 1985).

• Problems of pollution are generallyassociated with late phases in which thescarcity of the resource does not allowadequate mitigation by dilution, but itmay also happen very early if there aresignificant point sources of pollution, withlittle regulated water to ensure dilution(such as with mines in South Africa).

• The need to design more complex andintegrated forms of organization at thebasin level, are associated with anultimate phase of “allocation” of veryscarce water resources but, in somecases, as with the case of France in the1960s, it was the problem of waterquality, and not quantity, that was thedriving force behind the establishment ofthe “Basin Agencies” (despite bothaspects being interlinked).

• Trans-basin diversion is generallyconsidered as a possible way in which to“reopen” the basin after it has closed, butthis option is often taken at much earlierstages of development (at the firststage), especially in small and mediumbasins. This was commonly achieved inSri Lanka, at least as early as in the fifth

20Koning (2002) argues that this is partly because Africa’s elite is caught up in personal patronage relations and lack strong farmer move-ments that could ensure adequate support to agriculture, including trade protections.

17

century (Mendis 1993), and hasremained a basic principle of water-resources development, ever since. Suchtransfers are also typical of irrigation inmountain interfluves, where irrigatedareas straddle the boundary of twoadjacent basins (e.g., in the Andes).

• It seems that, in many cases, the laterphases of basin closure employ, not onlyallocation strategies but, morepragmatically, all the options that mayhelp relieve pressure. The case ofCalifornia, as described by Turral (1998),clearly shows not only that efficiency andallocative measures are bothsimultaneously sought, but also that thegains they provide are more limited thanare commonly believed and need to beaccompanied with a substantial amountof supply augmentation. Closure doesnot end up with allocative strategies but,rather, elicits continuous improvementson the three “fronts” (conservation,allocation, and supply).21

• In contrast to the impression that thecurrent focus on the economic value ofwater is characteristic of a late phase ofwater-resources development, the Britishperiod of the Indian history clearly showsthat all the questions currently debatedon the economics of irrigation werealready prominent ones. The questionsof who was to finance infrastructures(local revenue, the Crown, or privateinterests), whether and how a water feeshould be charged and what its impacton different categories of people is,whether it should be increased (opinions

diverged among the British Government,the Government of India, localgovernment, canal engineers, etc.),whether it could influence crop choice orwater-use behaviors, to cite a fewexamples, were fiercely debated from thebeginning of the nineteenth centuryonwards. Privatization, bulk volumetricpricing and crop-based differential rateswere experimented with.

• Not all trajectories are upward. Historicalexamples of civilizations that have notsuccessfully maintained their resourcesbase and that have collapsed can alsobe easily found. For example, aerialphotographs of some basins in Sri Lankareveal a very high density of small tanksthat have been abandoned, silted ordestroyed. After Independence (in 1948),new water-resources developments havesometimes been superimposed upon theolder systems, and larger dams havebeen built. Other examples includeancient Mesopotamia in the ninth century(cf. Pointing 1991) and, more generally,all the impacts of climatic change,salinization or tectonic change that areoften overlooked in historical studies(Brown 2001).

• The “averaged” vision, inherent in thetrajectory concept, may also obscure theheterogeneity of on-the-ground reality.Subareas in the basin are often at differentstages of evolution and the problems theyface, as well as their solutions, varysignificantly. Molden et al. (2001a) haveattempted to address this issue by defining“hydronomic zones” within a basin.

21If we reexamine figure 3 on such grounds, we might draw a different pattern, where the different curves tend to converge, in the long run, tothe same order of magnitude.

18

• Spatial heterogeneity is paralleled byhuman heterogeneity. When consideringthe diversity of farming systems, onefinds at the same time, differentindividual responses: water conservation,intensification, diversification, giving upagriculture, etc. Factor endowments,farmers’ agency and marketopportunities, among other things, shapeindividual strategies and, therefore, it isnot always possible to describe theresulting aggregated trend at the basinlevel in a simple way.

Critical Elements in BasinDevelopment

The above examples suggest that specifichistorical evolutions and patterns can be easilyencountered. Investigating in detail the causalcorrelations between particular physical/societalcontexts and historical transformations of riverbasins is, beyond the scope of this report, butthis section attempts to single out a fewelements that appear to be crucial in shapingresponses.

The Nature of the State and State/CitizenryRelationships: The state is often described asthe main actor that shapes all river-basinevolution by virtue of its investments andpolicies. Turton and Ohlsson (1999) identified thebuilding up the state’s power with the launch ofthe “hydraulic mission.”22 As mentioned earlier,this emphasis tends to obscure the magnitudeand significance of endogenous effortsundertaken by local actors, especially in view ofthe development of the conjunctive use of water.It also fails to account for situations in which

resources are principally managed by the usersthemselves, such as in the case of the tanksystems of south India or of the mountainousregions of the Andes, Nepal or upland southeastAsia.

The nature of the center/periphery and state/citizenry relationships defines the scope and theroom for maneuver and adjustment allowed tothe different actors in the system. Authoritarianor despotic states may be more oftenassociated with large-scale development andcentralized management (regardless of how thecausality is theorized), while weak states mayleave more scope for local initiatives. Thedegree of decentralization and democratizationalso obviously influences how negative impacts(particularly on health or the environment) areboth perceived and addressed.

The role of the state often changes overtime. Ruf’s (2001) description of the PradesValley in the south of France shows that, duringits five centuries of existence, the irrigationsystem has been managed, in turn, by the state,communal associations and privateentrepreneurs, and that these changes could betraced to the prevailing nature of the center/periphery relationships, defined within the widerpolitical and historical context. Theories ofinduced institutional change tend to seechanges as occurring by necessity, in responseto a mismatch between demand and supply, andthese theories do not account for the (numerous)cases where the power and political structureeventually dictates and supplies new institutionsand forms of organizations (see below).

Potkanski and Adams (1998) describe theresponse of the Sonjo, in Tanzania, to waterscarcity and show the interrelationship betweenagricultural commercialization, property rightsand water scarcity, on the one hand, and the

22It is not always very clear whether Turton and Ohlsson (1999) refer to river basins or countries.

19

complex and tightly interactive relations betweenlocal institutions and the state, on the other.23

Economic power and political will are oftenassociated together, and small communalschemes may be captured and enlarged whennew actors, such as local lords or colonialestates, come into play. Larger-scaleinterventions that are beyond the scope of localactors may be later undertaken by the state orby some capitalistic private entity.24 When basinsare developed in stages, defined by successivehistorical, economic and political contexts, thenseveral phases of supply augmentation, followedby growing water scarcity and resultingconservation/allocation measures, are likely totake place. This scenario differs from that of theabove frameworks, which tend to hypothesizeone, single, compact phase of state interventionand “heroic engineering” that exhausts (or nearlyexhausts) the physical opportunities for suchinterventions.

Turton (1999) posits that a popularlysupported legitimate government will be able tointroduce measures that an unpopulargovernment will not be able to introduce. This isconsistent with the hypothesis that reallocativestrategies are likely to be unpopular, but the linkwith the dichotomy of democratic versusauthoritarian regimes, is not developed. In fact, itis evident that the latter may sometimesreallocate water with more facility, irrespective ofwhether this is done in a sound manner or not.

Last, things get even more complex forinternational rivers, where political relations

between states come into play. Scenarios maydiffer significantly from those described above.

The Political Economy of Water ResourceDevelopment: Clearly, the economic rationalitydescribed by Keller et al. (1998) has a strongimpact on the choices that are made. However, itis necessary to go beyond a formal model ofrationality and towards a political and economicapproach, where decisions are understood notonly on the basis of their actual financial costs,but also on the basis of the benefits andincreased power that accrue to the differentcategories of actors within the society, andsometimes beyond.25 Costly (and even absurdlycostly) solutions, such as groundwater rechargeby injection or trans-basin diversion, aresometimes justified and implemented in lieu ofdemand-management options, because they fitthe logic of pork-barrel politics and serve assubstitutes of other more politically risky options.The conflicts between politicians, constructionfirms, consultants, local population andenvironmentalists, provide the best example ofhow the decision of building a dam is eventuallydebated in a political arena, where financial orpolitical interests coexist with environmentalism,communitarianism and concerns for locallivelihoods.

Another way to look at these societaldebates, is to weigh the monetary costs ofinfrastructural development (and their financialbenefits) against the transaction and politicalcosts of more demand-management-oriented

23This study shows, in particular, that not all endogenous institutional innovations provide benefits for the whole community, and that if thestate may be intrusive, it may also be instrumental in preventing the outright and inequitable appropriation of resources by any group ofindividuals.

24Ruf’s (2001) paper also provides such an example in the Ecuadorian Andes.

25Barker and Molle (2003) stress how investments in large-scale infrastructure during the cold war can be partly ascribed to geopolitical con-siderations. The “lending culture” of development banks has also been criticized as reflecting either the interest of western construction firmsor that of the banks in increasing the scale of their interventions.

20

options. The difficulty of reforming management,including efficiency and equity aspects, variesdepending on many cultural, social and politicalfactors. But it is recognized that “regionalpoliticians have a powerful intuition thateconomic principles and the allocative measures,which follow logically from them, must beavoided at all costs” (Allan 1999). This largelyexplains the persistent gap between therationality of consultants and experts and theactual adoption of policy measures in the realworld. It also provides hints on why stricteconomic logic is not always the best criteria tounderstand the succession of state investmentsand responses. Resource capture can occur atany time, depending on the power balance withinthe society, and is perhaps more frequent thanrational allocation. More generally, the responsesgiven to water shortages depend on whichconstituency succeeds in making its discourselegitimized and accepted (Ingram 1971). Theexample of water conservation in the ImperialValley Irrigation Scheme, southern California,given by Waller (1994), neatly shows how local,influential farming elites associated with irrigationmanagers could legitimize policies which shiftedthe cost of conservation onto other parties.Rather than accepting on-farm conservationsystems, which would have raised the amount oftime and effort needed in farming, as well as thereliance on hired labor, these elites successfullypushed for temporary measures of landfallowing, for which they would be compensated.

This leads Wester and Warner (2003) to callfor a vision of water as a politically contestedresource, and the closure of river basins as apolitical process, where the overexploitation ofresources is accompanied by the changing

patterns of access to water, and by thelegitimization of certain forms of basinmanagement through the production of dominantwater discourses.

It is interesting to note that the financialcosts of the three main responses (supplyaugmentation, conservation, and allocationoptions), are generally in a decreasing order.26

The lesson that can be drawn from this is thatdecision makers in most societies have aninverted perception of what the costs to societyare, or more to the point, of how the political andfinancial “benefits” accruing to them and to theirsupporters, compare with the political costs andthe risks of societal discontent incurred.

Shock Events: The responses and behavior ofwater users and of the society at large, regardingwater use and water-related problems, dependon their perception of the magnitude and severityof these problems. This perception, in turn, isoften sharply influenced by extreme naturalevents, such as typhoons, droughts and floods,which are generally accompanied by foodshortage, disasters, and the disruption oflivelihoods. Waves of dam construction are oftenlaunched after severe droughts or famines, suchas in Burkina Faso or northeast Brazil (see Molle1991). The drought periods of 1987-1991 inCalifornia (Zilberman et al. 1992; Keller 2000), of1992 in Turkey (GDRS and IWMI 2000), of 1982-83 in Australia (Turral 1998), of 1986 in Israel(Allan 1999), and of 1991-94 in Thailand (Molleet al. 2001), to mention a few examples, havecatalyzed a series of significant local and globalresponses. Dams are also designed in responseto floods, as the case of the Tennessee ValleyAuthority well exemplifies.

26Except for the case of water treatment. It can be posited that investments in urban water supply and sanitation create the potential for a“new construction bonanza” (Turral 1998), similar to that of the earlier “hydraulic mission.”

21

The most significant example of a crisis,however, remains the El-Niño-related climaticperturbation of 1972, which severely affectedgrain production and sent prices rocketing up.The psychological impact of this event, on bothnational decision makers and western countries(engaged in the Cold War and bent on investingin countries potentially threatened by the spreadof communism), was so high that much of thehuge investments in dams and irrigationinfrastructures that were to follow, can beascribed to the threat of food shortage in theparticular geopolitical context of the time (Barkerand Molle 2003).

Shock events often allow politicians toimpose policies that would have been otherwiseunpopular and opposed. Allan (1999) hasremarked that politicians are, therefore, likely towait for the exhaustion of resources and thesurge of crises, before embarking on draconianreforms.

Other shock events, which have a dramaticimpact on agricultural development in generaland water resource use, and development inparticular, are political events, such as wars,revolutions, agrarian reforms, etc. Theperiodization of water-infrastructure developmentin Vietnam (Tessier and Fontenelle 2000) or inChina (Lohmar et al. 2002), for example, neatlydovetails with that of political upheavals andreforms. Likewise, the collapse of theMesopotamian irrigation in the early Middle-Agesowes much to the impacts of wars andepidemics (plague), (Christensen 1998).

“Spatial Equity” and Regional Politics: Anotherimportant point, which drives the choice ofsupply augmentation as a response, is that theeconomic rationality at the basin or country levelhas to be combined with a notion of “equity” thatpervades regional politics and, more often thannot, conflicts with economic criteria. The regionswith lower comparative advantages stand to lagbehind other regions, and often display higher

levels of poverty. These “problem” regions,therefore, turn out to be the target of specialinvestments, which generally have a low return,and which are dictated by political andsocioeconomic concerns.

An example in point is that of “overbuiltbasins,” where regulated water resources areinsufficient to serve the existing irrigated areas,but where more irrigation schemes are built,based on the claim that the regions not benefitedhitherto, also have a right to receive investments.This, Ingram (1971) noted, is supported by thestrong local sentiment that water is alwayslocally thought of as “our” water. Localsociopolitical dynamics and strategies conflictwith macro considerations and logic, thusdepicting a typical example of a basin-levelissue. When decentralization means that localregions or provinces have the mandate to designlocal strategies without considering their impactat a wider level, these situations are often verysalient. The Mekong delta offers a good exampleof how provincial plans are based on the samefreshwater resource, and how their combinedimpact on the whole delta hydrology might havecatastrophic consequences on salinity intrusionin dry years (Can Tho University and IRD 2001).In other words, the micro/macro dialectic ofwater use within the basin is paralleled by asimilar spatial interconnectedness with regard toeconomic planning. The result is often “artificial”scarcity created by the overcommitment ofresources, fostered by flimsy knowledge orconsideration of hydrology, and promoted bydevelopmentalism. In Algeria, for example, theWorld Bank supported both irrigation projectsand urban water supply networks incompetition for the same scarce resource(Winpenny 1994).

A particular form of the impact of regionalpolitics on water resource development is wellexemplified by the case of the U.S.A. The states,wishing to see their local projects funded byfederal agencies, need to muster support from

22

other states in order to obtain congressionalacceptance. This “pork barrel” politics leads tothe spread of projects with very low returns and,often, with environmental impacts. The case ofthe Central Arizona Project (CAP), discussed byWelsh (1985), shows how Colorado garneredprojects for its vote authorizing the CAP, andhow other states like New Mexico, Utah, andCalifornia could take advantage of the CAP toclaim for their own projects. The CAP itself, mostparticularly the projected Orme dam that waspart of it, was an extravagant example of aproject with low returns, critical impact onwildlife, spoliation of Indian land, which waspushed by the Bureau of Reclamation, despitestrong opposition and numerous alternatives(Welsh 1985).

Agrarian Pressure and Agrarian Transition:Individual and societal responses are alsostrongly governed by the alternatives oflivelihoods offered to people, and to, what can betermed as “agrarian pressure.” If strongpopulation growth occurs in a context where,nonagricultural sectors are unable to absorb theexcess rural labor, then the pressure upon landand water resources is likely to increase. Thistranslates, not only into greater agriculturalintensification and water conservation, but alsointo water-quality degradation and conflicts. If thewater supply to agriculture is squeezed, then thesocial consequences on the rural world (and alsothe possible political clout of the farming sector)are likely to trigger state responses (asexemplified by the subsidizing of westernagriculture). On the other hand, if favorablealternatives are supplied within the widereconomy, water users with deficient supplies willbe encouraged to diversify their activities or tosimply give up farming, thus easing the tensionon resources.

Agrarian pressure is also, directly linked tohousehold incomes and, therefore, at least forthose products with a degree of import/export,

linked to the price of commodities in the worldmarkets. This price, in its turn, depends on amore complex and global equation that includesfood production, population growth, productivity,as well as all other variables that impact on, ordistort markets. The Middle East provides a goodexample of a region, where water policy is“subordinate to the political economy of globaltrade in staple food” (Allan 1996).

The intensity of societal demand for change,the individual responses to the deterioration ofaccess to water, the relative growth of thedifferent economic sectors and the impact onwater use, are all interrelated with macro-economic settings.

Demand/Supply-Driven Innovations

Molden et al. (2001b) argue that as basinsdevelop, new forms of organization and rules arerequired, but they take the view that new“institutions (inevitably) evolve to fulfillmanagement requirements,” and do not addressthe issue of whether, and how these institutionsemerge. The few points mentioned in the abovesection already suggest some agents of change.The more general question of the causation ofinstitutional and technical changes cannot beaddressed in detail in this report, but it can beuseful to recall here, a few theoretical postures,and to sketch out the driving forces at work.

Social Adaptive Capacity: The evidence thatsome societies seem to avert crises (as well asto “reconstruct” resources after their degradation)through the mobilization of appropriateinnovations and new institutions, while others donot, has led some social scientists to investigatewhich factors could be singled out to explain thissituation. Ohlsson (1999) has recently developedthe idea that some societies were endowed with“adaptive capacity,” which should be consideredas a resource, and that the lack of this resource

23

could be termed “second-order scarcity,” inaddition to the physical first-order (water)scarcity.27 Turton (1999) built on such conceptsand reasoned that this capacity has both supplyand demand aspects. He distinguishes first asocial component, “existing in the hearts andminds of the governed,” which basically definesthe willingness and ability of the people toaccept the measures to be taken, as well as theability and legitimacy of the regime to generatethe needed strategies. The second componentcan be engineered to some extent and consiststhe capacity to mobilize and analyze data andinformation, and to derive adequate strategies,all of which—according to Turton—can possiblybe provided, either by local or foreign sources.This division is useful in emphasizing somesupply and demand sides of the problem but,while stressing the importance of quantitativeinformation and human resources, does notfathom the societal term, which “is the difficultone to come to terms with.”28

Allan (1996) posits that the development of a“diverse and strong” economy is a prerequisite tothe implementation of the measures needed tomanage the closure of water resources (with theimplication that in the opposite case, there is afailure to manage it). Factors that are highlycorrelated with economic development include,for example, a certain degree of intellectualcapital to acquire and process data, and capitalavailability and institutional capacity to devisemeasures and to enforce them. Redressingenvironmental degradation and improving waterquality also require both, very high outlays of

capital and the political capacity to enforce alegal framework, where these funds areeffectively mobilized (from both public andprivate sources).29 In addition, there is a degreeof formalization of rights, of recourses availableto citizens to voice their opinions, and altogethera more democratic participation of civil society,all of which are also paramount in shaping thedecisions of politicians and policymakers.

However, it may be argued that developedcountries are also those in which pressure overresources, water-quality problems andenvironmental degradation has first occurred,precisely because of their earlier economicdevelopment. Therefore, their capacity tomanage basin closure could be seen as inducedby pressing needs and enabled by their financialcapacity, rather than as an intrinsic preexistingquality. It might be hypothesized too, thatchanges become easier to implement whenstructural changes in the economy have reducedthe importance of agriculture and, therefore,have reduced its political weight, and sometimesthe very amount of water it needs (in whichcase, policy changes are facilitated by structuralchanges). Finally, the very perception ofenvironmental damages, by people (andconsequently their motivation to act againstthem), is highly linked to their standard of livingand is, therefore, weaker in developing countries.In other words, it might well be that it is theconsequences of economic development that(partly) facilitates basin-closure management,rather than the factors that are believed to beconducive to development.

27Ohlsson (1999), however, does not provide many clues on what makes societies endowed or deficient in “adaptive capacity.” Therefore,stating that some societies adapt because they have enough “adaptive capacity” might be seen as tautological.

28This can be compared with Ruttan’s (1989) recognition that, if it is intuitively admitted that cultural factors exert major impact on behaviorsand responses to opportunities, the lack of analytical means to take this factor into consideration, is a “deficiency in professional capacity.”

29In the two decades following 1972, the USA has spent more than US$500 billion on preventing and cleaning up water pollution, so thatenvironmental costs have already exceeded the cost of subsidies in infrastructure (Frederick 1998). It is difficult to envision such a responsein developing countries.

24

Supply and Demand Models: Much of thetheories of institutional changes, derived from theworks by Hayami and Ruttan (see for example1985), as well as later New InstitutionalEconomics (NIE) approaches,30 tend to favor thedemand side of change, with emphasis onchanges in relative resource endowments andtechnical aspects. Individuals, seen as rationalentrepreneurs, respond to both “deadlocksituations” and opportunities. In the first case,the design of new institutions is needed to adaptto changing conditions, which have raisedtransaction costs. In the second case, individualsrecognize the existence of uncaptured potentialgains and devise arrangements that allow themto tap these gains. Regarding water resources,the first case is well exemplified by rotationalarrangements in irrigated schemes, which needto be implemented so as to deal with waterscarcity, or by basin-level management of waterquantities and quality. The second case can beexemplified by negotiations for trans-basintransfers or, more self-interest-centered, by thecapture of benefits by a given interest group(such as sometimes would occur in processes ofprivatization).

This last example shows that it would benaive to assume that the constraints faced by asociety give way, through a perfect, market-likemechanism, to a solution that minimizestransaction costs and improves efficiency(Wegerich 2001). It must be acknowledged thatthe distribution of power, embodied in the accessto (water) resources, is a societal, historicalconstruction, where equity and efficiency mightnot be the chief aspects. This takes us beyondthe economics of transaction costs, to includepolitical economy and rent-seeking behaviors(Roumasset 1995).

30For North (1995), “the fundamental source of change is learning by entrepreneurs of organisations (sic)… and the rate of learning reflectsthe intensity of competition amongst organisations (sic).”

This leads us to also give consideration totop-down approaches of institutional change thatemphasize the supply of innovations by the eliteor by outsiders. This is not the place to expandon theories of elite behavior and decisionmaking, but one may note, as emphasized byNorth (1995), that “institutions are notnecessarily, or even usually created to besocially efficient; rather they, or at least theformal rules, are created to serve the interests ofthose with the bargaining power to create newrules.” This is also emphasized by Feeny (1988),who sees the benefits and costs to the elite as amain determinant of change. A policy implicationof this is that ignoring power structures is likelyto doom bottom-up movements or policies, tofailure (Wegerich 2001; Schlager and Blomquist2000).

What triggers elite decision making is complex,but it seems that immediate political considerationgenerally overruns what some would like to see asthe result of rational planning (Schlager andBlomquist 2000). In practice, changes are verymuch demand-driven “by the need to find solutionsto relatively immediate, specific problems, notgrand visions of river basin or regionaldevelopment” (Whittington 1991; see alsoWinpenny 1994). Historical changes in waterinstitutions, for example, have often been spurredby flood or water-shortage crises (Livingston 1993;Wengert 1985).

Disaggregating “Costs and Benefits”: Here wemay build on Wegerich’s (2001) conjecture thatboth institutional change and adaptive capacityare linked to the power of stakeholders, whileredefining power as the possession of resources,information or legitimacy that allows actors toselect an option that best suits their interest.

25

From this viewpoint, both decision-making andthe implementation of reforms, are seen as theoutcome of a process in which the financial,economic, social (status) and political interests ofall actors are confronted. Actors typically includestates, other administrative units likemunicipalities, politicians, line agencies,construction companies or water utilities, banksand development agencies, experts, lawyers,environmentalists, lobbyists, NGO leaders andactivists, farmers31 and other water users.

Part of the relative “weight” of thesecategories of actors in the process of change (ornonchange) is embodied in the institutions. Theypartly define actors’ rights and, more generally,how they are allowed or able to, voice theirconcerns and have their opinions taken intoconsideration, as well as to exercise theiragency. This also means that part of the changeis due to the responses to specific water scarcityissues that individuals or small groups of peopleare able to design locally. As seen earlier, suchresponses include innovations in croppingtechniques, adjustments of farming activities,improved individual and collective water

management, tapping local water resources andalso attempts to participate or intervene in moremacro-level issues, where the fate of theiraccess to water is mainly determined.

We may also retain the idea of “costs” asdetermining societal responses, but “costs” mustencompass several dimensions: economic costs (ininfrastructure, of foregone benefits owing to waterdeficits, etc.), transaction costs (in decision making,conflict-solving, accessing information, etc.) orpolitical costs (unpopular measures). These costsare paralleled by “benefits,” which accrue toparticular actors (increased water supply, morecompanies, lawyers, politicians, etc.).

Last, this political economy of societalresponse to water scarcity, is obviously shapedby the physical/climatic context (aridity, humidity,variability in precipitation, etc.), the availabletechnology (defined as “tools plus knowledge”[Livingston 1993]), and the preexistingsociocultural and institutional context. Theunderstanding of institutional change in themanagement of water resources is a field ofinvestigation that clearly needs to be furtherresearched on (Ingram 1971; Livingston 1993).

31Listing categories does not mean that these are homogeneous: different farmers, for example, are in distinctly different positions, both inphysical terms (with regard to the water supply network) and in socioeconomic terms (see Schlager and Blomquist 2000).

Basin Trajectories

Acknowledging the variety of societal responsesto water scarcity and the complexity ofdetermining which particular response appears ata certain point in time, a river-basin developmenttrajectory may be represented by the graphshown in figure 7. The schematization retains thegeneral evolution towards basin closure,although it is also recognized that a basin can

be “reopened” (see below). The succession ofcircles (corresponding to that of figure 6)represents successive adjustments to waterscarcity, which are made by the implementationor the inducement of a range of individual,collective and state responses that come underthree categories (supply expansion,conservation, and allocation). (The varying

26

respective shares of the inner circle (local) andthe outer circle (global), as well as the varyingsizes of the arrows, symbolize that the relativeimportance of each response varies dependingon the point in time).32

Which strategies have been implemented ata given point in time, and similarly what optionsare more suitable for the current situation, canonly be determined by a sound analysis of allthe relevant physical, economic and societalfactors. The preceding section providedexamples of issues situated at the convergenceor interface of several of these important factors.

Figure 8 proposes a variant of figure 7 andintroduces two visual modifications. The first onerepresents a possible (albeit transient)breakdown of the system, while the second,indicates successive potential ceilings toward

“resource closure.” This accounts for the fact thatif total renewable basin resources can behydrologically defined, the share that is availableat one point in time, often depends on theexisting technological level. In other words,societies are faced with different ceilings thatmay shift with time, but that are sensed asabsolute ceilings for a certain period. Basins canbe reopened, and often are, by achieving a newincrement in supply through a costly investmentthat was formerly thought not to be technically orfinancially feasible. This, in particular, includestrans-basin diversions, which are much morecommon than is usually believed. Only 20percent of the water used by south California, forexample, is generated locally, as the region ismostly supplied by water from the ColoradoRiver (57%) and from the north of the state.

FIGURE 7.Basin trajectories (1).

32The sizes and shares indicated here are arbitrary. This graph should be taken as a conceptual representation, rather than as a concretevisualization of a particular change.

27

The figure also reminds us thatunsustainable management may lead to usingresources beyond the potential threshold. Thismeans that aquifers may be “mined” (or depletedfaster than they are replenished), and reservoirsecurity stocks can be tapped during a few

FIGURE 8.Basin trajectories (2)

consecutive seasons. It is not certain whetherunsustainable management will always betackled in an appropriate manner, since manyaquifers are probably doomed to be mined untilexhaustion (or, in practical terms, untilabstraction becomes economically unviable).

28

Conclusions

Population growth and economic developmenttranslate into growing pressure on waterresources. In the course of time, theinterdependence of users in the basin increasesand conflicts arise. Scarcity elicits adjustments inwater-supply augmentation, water-use efficiencyand in water allocation. Several existingconceptual frameworks that provide a descriptionof basin development have been examined. Thegreat merit of these frameworks is that they offersimplicity of reading that conveys the notion ofbasin closure with great strength. They arebased on a sequence of phases that is widelyrelevant to depict the evolutions observed duringthe twentieth century in many basins orcountries. The downside of these approaches,however, is that the simplification of reality doesnot always allow one to fully capture orunderstand the geographical and historicaldiversity of river-basin development.

A first drawback resides in the representationof the exploitation of water resources by a singlecurve. A “disaggregated” diagram has beenproposed, that distinguishes between differenttypes of resources (rainwater, stream water, andcontrolled [actual and potential] water), and thatallows a clearer view of the status of water usein relation to supply, and specifies the part of thesupply that is controlled by man.

A second limitation in these frameworksresides in their simplification of the succession ofdifferent types of responses in sequence. Theseresponses often occur concomitantly or

sequentially, but not in the order proposed.Adjustments in water-supply management,allocation and institutions are made, not only bythe state, but also by individuals and groups, andare characterized by their interrelatedness. Eachtime a decision, taken either locally or at theglobal level, impacts on water flow paths, interms of quantity, quality or timing, other usersare likely to be affected somewhere else in thebasin. The characteristics of each basincommand a particular pattern of evolution, whichmay include gridlocks, collapses and thediversion of water flows across basin boundaries.

Existing linear visions of basin developmenttend to be based on economic rationality or onconcepts of social adaptiveness that are difficultto evaluate. Societal responses to scarcity ofresources are driven not only by economicconsiderations or locally perceived needs.Rather, societal responses must also beunderstood in a wider political economyframework, in which costs and benefits areattributed to different categories of actors who,often, have antagonistic interests that are noteven internally homogeneous. The particularblend of responses, selected by a society at aparticular point in time of its history, to addresswater-resources problems must, therefore, beunderstood within a framework that spans notonly hydrological, physical or economicconstraints, but also the distribution of agencyand power among actors, and their respectiveinterests and strategies.

29

Literature Cited

Allan, J. A. 1996. Policy responses to the closure of water resources: Regional and global issues. In Water policy:Allocation and management in practice, 228-234, ed. P. Howsam and R. C. Carter. UK: Silsoe College, CranfieldUniversity.

Allan, J. A. 1999. Productive efficiency and allocative efficiency: Why better water management may not solve theproblem. Agricultural Water Management 40: 71-75.

Barker, R.; Molle, F. Forthcoming. Evolution of irrigation in Asia. Research Report, Comprehensive Assessment ofWater Management in Agriculture. Colombo: International Water Management Institute.

Brown, N. 2001. Wittfogel and hydraulic despotism. Paper presented at “The Role of Water in History and Development,”IWHA 2nd Conference, Bergen, Norway. 17 p.

Cai, X.; Ringler, C.; Rosegrant, M. 2001. Does efficient water management matter? Physical and economic efficiencyof water use in the river basin. EPTD Discussion Paper No. 72. Washington, D.C.: International Food Policy ResearchInstitute.

Can Tho University and IRD (Institut de Recherche pour le Développement) 2001. Agricultural diversification in theMekong river delta: Raised beds management, marketing and dynamics. DELTA Project - Mekong Delta: ResearchReport No. 3. Can Tho, Vietnam: Cantho University.

Christensen, P. 1998. Middle eastern irrigation: Legacies and lessons. Yale F&ES Bulletin, No. 103, pp. 15-31.

De Koninck, R. 1997. Agricultural expansion as a tool of population redistribution in Southeast Asia. Journal of SoutheastAsian Studies 28:26-44.

Feeny, D. 1988. The demand and supply of institutional arrangements. In Rethinking institutional analysis anddevelopment, 159-209, ed. Vincent Ostrom, D. Feeny and Hartmut Picht. San Francisco: International Center forEconomic Growth.

Feuillette, S. 2001. Vers une gestion de la demande sur une nappe en accés libre: exploration des interactionsresources usages par les systèmes multi-agents. Application à la nappe de Kairouan, Tunisie Centrale. Ph.D. diss.,Montpellier: Université de Montpellier II.

Frederick, K. D. 1998. Marketing water: The obstacles and the impetus. Resources for the Future, Issue 132. http://www.rff.org/resources_archive/pdf_files/132_water.pdf

General Directorate of Rural Services; IWMI (International Water Management Institute). 2000. Irrigation in the basincontext: The Gediz river study, Turkey. Colombo: International Water Management Institute. 105 p.

Hayami, Y.; David, C.; Flores, P.; Kikuchi, M. 1976. Agricultural growth against a land resource constraint: the Philippineexperience. Australian Journal of Agricultural Economics, 20(3): 144–159

Hayami,Y.; Ruttan, V. W. 1985. Agricultural development. Revised and expanded version. Baltimore and London: TheJohn Hopkins University Press, 506 p.

Hussain, I.; Sakthivadivel, R.; Amarasinghe, U.; Mudasser, M.; Molden, D. 2003. Land and water productivity of wheatin the western Indo-Gangetic plains of India and Pakistan: A comparative analysis. IWMI Research Report 65.Colombo, Sri Lanka: International Water Management Institute.

Ingram, H. 1971. Patterns of politics in water resources development. Natural Resources Journal 11:102-118.

International Water Management Institute (IWMI). 2002. World irrigation and water statistics 2002. Colombo, Sri Lanka:International Water Management Institute.

Keller, A.; Keller, J.; Seckler, D. 1996. Integrated water resources systems: Theory and policy implications. ResearchReport 3. Colombo, Sri Lanka: International Irrigation Management Institute.

30

Keller, J.; Keller, A.; Davids, G. 1998. River basin development phases and implications of closure. Journal of AppliedIrrigation Science 33(2):145-164.

Keller, J. 2000. Reengineering irrigation to meet growing freshwater demands. Proceedings of the 4th decennialsymposium of the American Society of Agricultural Engineers, pp. 21-39. St Joseph, Michigan: American Societyof Agricultural Engineers.

Kikuchi, M.; Barker, R.; Weligamage, P.; Samad, M. 2002. The irrigation sector in Sri Lanka: Recent investment trendsand the development path ahead. Research Report 62. Colombo, Sri Lanka: International Water ManagementInstitute.

Koning, N. 2002. Beyond bad governance: An agrarian reinterpretation of Africa’s problems. Department of SocialSciences, Wageningen University. Mimeo Duplicated.

Livingston, M. L. 1993. Normative and positive aspects of institutional economics: the implications for water policy.Water Resources Research 29(4): 815-821.

Lohmar, B.; Wang, J.; Rozelle, S.; Huang, J.; Dawe, D. 2002. Investment, conflicts and incentives: the role of institutionsand policies in China’s agricultural water management on the north China plain. Draft.

Mendis, D. L. O. 1993. Irrigation systems in the Walawe Ganga basin: An historical overview. In The south-east dryzone of Sri Lanka, Proceedings of a Symposium, 29-30 April 1992. Colombo, Sri Lanka: Agrarian Research andTraining Institute (ARTI), pp. 37-54.

Molden, D.; Sakthivadivel, R. 1999. Water accounting to assess use and productivity of water. Water ResourcesDevelopment 15(1/2): 55–71.

Molden, D.; Sakthivadivel, R.; Samad, M. 2001b. Accounting for changes in water use and the need for institutionaladaptation. In Intersectoral management of river basins: Proceedings of an international workshop on “IntegratedWater Management in Water-Stressed River Basins in Developing Countries: Strategies for Poverty Alleviation andAgricultural Growth,” Loskop Dam, South Africa, 16-21 October 2000, 73-87, ed. C. L. Abernethy. Colombo, SriLanka: International Water Management Institute; Deutsche Stiftung für Internationale Entwicklung.

Molden, D.; Sakthivadivel, R.; Keller, J. 2001a. Hydronomic zones for developing basin water conservation strategies.Research Report No. 56. Colombo, Sri Lanka: International Water Management Institute, 30 p.

Molle, F.; Aloysius, N. Forthcoming. Seasonal water accounting for river basins, with case studies from Thailand andSri Lanka. Colombo, Sri Lanka: International Water Management Institute.

Molle, F. 1991. Historical benchmarks and reflections on small tanks and their utilization. Mossoro, Brazil: CollectionMossoroense, 190 p. (in Portuguese)

Molle, F. Forthcoming. Technical and institutional responses to basin closure in the Chao Phraya river basin, Thailand.Accepted for publication in Water International.

Molle, F.; Chompadist, C.; Srijantr, T.; Keawkulaya, J. 2001. Dry-season water allocation and management in the ChaoPhraya Basin. Research report submitted to the European Union, Research Report No. 8, Bangkok, 278 p.

North, D. C. 1995. The new institutional economics and third world development. In The new institutional economicsand third world development, ed. J. Harriss, Janet Hunter and Colin M. Lewis, 17-26. London and New York:Routledge.

Ohlsson, L.; Turton, A. R. 1999. The turning of a screw: Social resource scarcity as a bottleneck in adaptation towater scarcity. Working Paper. London: School of Oriental and African Studies, University of London.

Ohlsson, L. 1999. Environment, scarcity and conflict: A study of Malthusian concerns. Göteborg: Department of Peaceand Development Research, Göteborg University.

31

Perry, C. J. 1999. The IWMI water resources paradigm: Definitions and implications. Agricultural Water Management40(1): 45–50.

Pointing, C. 1991. A green history of the world. USA: Penguin Books. 432 p.

Potkanski, T.; Adams, W. M. 1998. Water scarcity, property regimes and irrigation management in Sonjo. Tanzania.Journal of Development Studies 34(4): 86-116.

Rostow, W. W. 1962. The stages of economic growth: A non-Communist manifesto. New York: Cambridge UniversityPress.

Roumasset, J. A. 1995. Induced institutional change: A neoclassical synthesis. In Induced innovation theory andinternational agricultural development: A reassessment, 136-150, ed. B. M. Koppel. Baltimore and London: JohnsHopkins University Press.

Ruf, T. 2001. The institutional cycles of peasants’ irrigation: Historical debates on ownership and control of water inSouth of France and in the Ecuadorian Andes. Paper presented at “The Role of Water in History and Development”,IWHA 2nd conference, Bergen, Norway. 17 p.

Ruttan, V. W. 1989. Institutional innovation and agricultural development. World Development 17 (9): 1375-1387.

Sakthivadivel, R.; Molden, D. 2001. Linking water accounting analysis to institutions: Synthesis of five country studies.Colombo, Sri Lanka: International Water Management Institute.

Schlager, E.; Blomquist, W. 2000. Local communities, policy prescriptions, and watershed management in Arizona,California, and Colorado. Paper presented at “Constituting the Commons: Crafting Sustainable Commons in theNew Millenium,” the Eighth Conference of the International Association for the Study of Common Property,Bloomington, Indiana, USA, May 31-June 4.

Tessier, O.; Fontenelle, Jean-Philippe. 2000. ‘Pression démographique et contraintes politiques: la paysannerie dudelta du Fleuve Rouge dans la tourmente du 20ème siècle.’ In Population et développement au Viêt-nam, ed.Patrick Gubry. Paris: Karthala, 613 p.

Turral, H. 1998. Hydro-logic? Reform in water resources management in developed countries with major agriculturalwater use - Lessons for developing nations. London: Overseas Development Institute, 174 p.

Turton, A. R. 1999. Water scarcity and social adaptive capacity: Towards an understanding of the social dynamics ofwater demand management in developing countries. MEWREW Occasional Paper No. 9. London: School of Orientaland African Studies, University of London.

Turton, A. R.; Ohlsson, L. 1999. Water scarcity and social stability: towards a deeper understanding of the key conceptsneeded to manage water scarcity in developing countries. Working Paper. London: School of Oriental and AfricanStudies, University of London.

Waller, T. 1994. Expertise, elites, and resource management reform: Resisting agricultural water conservation inCalifornia’s Imperial Valley. Journal of Political Ecology 1: 13-42. (1994)

Wegerich, K. 2001. Institutional change: A theoretical approach. Occasional Paper No. 30. London: School of Orientaland African Studies, University of London.

Welsh, F. 1985. How to create a water crisis. Boulder: Johnson publishing Company. 238 p.

Wengert, N. 1985. The river basin concept as seen from a management perspective in USA. In Strategies for riverbasin development, 299-305, ed. J. Lohm, U. Lundqvist and M. Falkenmark. Dordrecht, Holland: Reidel.

Wester, P.; Warner, J. 2003. River basin management reconsidered. In Hydropolitics in the developing world: A SouthernAfrican perspective, 61-71, ed. A. Turton and R. Henwood. Pretoria, South Africa: African Water Issues ResearchUnit.

32

Whittington, D. 1991. A note on the development of USAID water resources policy for developing countries. Cited inWinpenny, 1994. Duplicated.

Winpenny, J. 1994. Managing water as an economic resource. Development Policies Studies. London: Routledgeand Overseas Development Institute, 133 p.

Zilberman, D.; Dinar, A.; MacDougall, N.; Khanna, M.; Brown, C.; Castillo, F. 1992. Individual and institutional responsesto the drought: The case of California agriculture. ERS Staff Paper. Washington, D.C.: US Department of Agriculture,70 p.

60. Institutional Alternatives in African Smallholder Irrigation: Lessons from InternationalExperience with Irrigation Management Transfer. Tushaar Shah, Barbara van Koppen,Douglas Merrey, Marna de Lange and Madar Samad. 2002.

61. Poverty Dimensions of Irrigation Management Transfer in Large-Scale Canal Irrigationin Andra Pradesh and Gujarat, India. Barbara van Koppen, R. Parthasarathy andConstantina Safiliou. 2002.

62. Irrigation Sector in Sri Lanka: Recent Investment Trends and the Development PathAhead. M. Kikuchi, R. Barker, P. Weligamage and M. Samad. 2002

63. Urban Wastewater: A Valuable Resource for Agriculture. Wim van der Hoek, MehmoodUl Hassan, Jeroen H. J. Ensink, Sabiena Feenstra, Liqa Raschid-Sally, Sarfraz Munir,Rizwan Aslam, Nazim Ali, Raheela Hussain and Yutaka Matsuno. 2002.

64. Use of Untreated Wastewater in Peri-Urban Agriculture in Pakistan: Risks andOpportunities. Jeroen H. J. Ensink, Wim van der Hoek, Yutaka Matsuno, SafrazMunir and M. Rizwan Aslam. 2002.

65. Land and Water Productivity of Wheat in the Western Indo-Gangetic Plains of Indiaand Pakistan A Comparative Analysis. Intizar Hussain,R.Sakthivadivel, UpaliAmarasinghe, Muhammad Mudasser and David Molden. 2003.

66. Agro-Well and Pump Diffusion in the Dry Zone of Sri Lanka: Past Trends, PresentStatus and Future Prospects. M. Kikuchi, P. Weligamage, R. Barker, M. Samad, H.Kono and H.M. Somaratne. 2003.

67. Water Productivity in the Syr-Darya River Basin. Hammond Murray-Rust, IskandarAbdullaev, Mehmood ul Hassan and Vilma Horinkova. 2003.

68. Malaria and Land Use: A Spatial and Temporal Risk Analysis in Southern Sri Lanka.Eveline Klinkenberg, Wim van der Hoek, Felix P. Amerasinghe, Gayathri Jayasinghe,Lal Mutuwatte and Dissanayake M. Gunawardena. 2003.

69. Tubewell Transfer in Gujarat: A Study of the GWRDC Approach. Aditi Mukherji andAvinash Kishore. 2003.

70. Energy-Irrigation Nexus in South Asia: Improving Groundwater Conservation andPower Sector Viability. Tushaar Shah, Christopher Scott, Avinash Kishore andAbhishek Sharma. 2003.

71. Policies Drain the North China Plain: Agricultural Policy and Groundwater Depletionin Luancheng County, 1949-2000. Eloise Kendy, David J. Molden, Tammo S. Steenhuisand Changming Liu. 2003.

72. Development Trajectories of River Basins: A Conceptual Framework. François Molle.2003.

Research Reports