INSTITUTIONAL AND POLICY ANALYSIS OF RIVER BASIN MANAGEMENT DECENTRALIZATION

The Principle of Managing Water Resources at the Lowest Appropriate Level — When and Why Does It (Not) Work in Practice?

Edited by Karin Kemper, Ariel Dinar and William Blomquist

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

INSTITUTIONAL AND POLICY ANALYSIS

INSTITUTIONAL AND POLICY ANALYSIS OF RIVER BASIN MANAGEMENT DECENTRALIZATION

The Principle of Managing Water Resources at the Lowest Appropriate Level — When and Why Does It (Not) Work in Practice?

Edited by Karin Kemper, Ariel Dinar and William Blomquist

3

Acknowledgments

This document is a product of the study Integrated River Basin Management and the Principle of Managing Water Resources at the Lowest Appropriate Level – When and Why Does It (Not) Work in Practice?. The Research Support Budget of the World Bank provided major funding. The project was carried out by the Agriculture and Rural Development Department at the World Bank. The Water Resources Management Group and the South Asia Social and Environment Unit at the World Bank and the Bank-Netherlands Water Partnership Program have pro-vided additional support. The study was led by Karin Kemper and Ariel Dinar from the World Bank and William Blomquist (consultant, Indiana University). The team included Anjali Bhat (consultant, Indiana University), Michele Diez (World Bank), Rosa Maria Formiga Johnsson (consultant), William Fru (consultant), Gisèle Sine (International Network of Basin Organizations) and Corazon Solomon (World Bank).

The Technical Notes contained in this publication were prepared by Oliver Taft (consultant), with William Blomquist, Ariel Dinar, and Karin Kemper, based on the Working Papers with the same titles, available at www.worldbank.org/riverbasinmanagement. Melissa Williams managed design and publication arrangements of this document. Additional input from Rosa Maria Formiga Johnsson, Catherine Tovey, Jack Williams, Siet Meijer and Sachin Shahria is gratefully acknowledged.

Basin case study and background paper consultants include Maureen Ballestero (Tárcoles - Costa Rica), Ken Calbick and David Marshall (Fraser - Canada), Rosa Maria Formiga Johnsson (Alto Tietê and Jaguaribe - Brazil), Consuelo Giansante (Guadalquivir - Spain), Brian Haisman (Murray Darling - Australia), Kikkeri Ramu and Trie Mulat Sunaryo (Brantas - Indonesia), Andrzej Tonderski (Warta - Poland) and Jyothsna Mody (Accountability through Decentralization). We are grateful to the individuals whom we interviewed in the course of this research. None of those individuals is responsible for the findings and conclusions in this document. The views expressed in this document are those of the authors and should not be attributed to the World Bank.

5 Introduction

CASE STUDIES

8 Murray Darling (Australia)

12 Jaguaribe (Brazil)

18 Alto Tietê (Brazil)

23 Fraser (Canada)

26 Tárcoles (Costa Rica)

30 Brantas (Indonesia)

35 Warta (Poland)

39 Guadalquivir (Spain)

44 Comparative Analysis of River Basin Management

GLOBAL ANALYSIS

52 Decentralization of River Basin Management: A Global Institutional and Policy Analysis

57 Achieving Accountability through Decentralization: Lessons for Integrated River Basin Management

Table of Contents

4

5

Introduction

One of the internationally accepted principles of river basin management is to decentralize decision making to the lowest appropriate level. This principle has been promulgated extensively over the past 15 years, leading to the creation of many river basin management organizations structured around a broader participation of stakeholders from different user groups and sectors in order to achieve more integrated water resources management.

The study “River Basin Management at the Lowest Appropriate Level – When and Why Does It (Not) Work in Practice” originated from empirical observations, based on experience with World Bank-financed river basin management projects all over the world. Experience from these projects suggested that while the concept of management at the lowest appropriate level could be translated into laws and regulations relatively easily, its actual application often encounters obstacles due to the varying interests of different stakeholder groups, includ-ing those having to promote decentralization. In practice, this means that projects and policies based around integrated water resources management principles may not fully achieve their potential benefits. This observa-tion led this study to investigate the underlying reasons for (non-) achievement of decentralization in river basin management. The research was carried out over the period 2002 to 2005.

The study consists of three main parts: (1) a global survey of 83 river basin organizations around the world, (2) eight in-depth case studies and (3) an extensive literature review.

The case studies analyze the decentralization history and performance of basin organizations in the Murray Darling (Australia), Jaguaribe and Alto Tietê (Brazil), Fraser (Canada), Tárcoles (Costa Rica), Brantas (Indonesia), Warta (Poland) and Guadalquivir (Spain) river basins.

An econometric analysis of the large number of basin organizations included in the global survey in both developed and developing countries has also been carried out. It evaluates a range of factors that affect the outcomes of decentralization processes, including the way such processes are initiated (e.g. by governments or stakeholders), institutional and financing arrangements, climate, history, and basin hydrology.

This document presents the key results of the study in the form of 11 Notes which are based on the more exten-sive published Working Papers with the same titles, available at www.worldbank.org/riverbasinmanagement.

It is hoped that this publication will provide interested readers with a concise overview of the study results and contribute to the thinking and policy formulation regarding riverbasin management organizations, decentralization, stakeholder participation and what can be expected from such water resources management trends and reforms.

7

Case Studies

8

Murray-Darling River Basin, Australia

River basin governance and management in an industrialized country where water resources are scarce and extremely variable and where the evolution of basin management institutions has been as much a matter of integration as decentralization.

GEOGRAPHICAL OVERVIEW. The Murray-Darling Basin is an interior basin defined by the catchment areas of the Murray and Darling Rivers and their many tributaries. The Murray-Darling Basin lies to the west of the Great Dividing Range which runs the length of the east coast of Australia. The Basin extends across much of southeast-ern Australia, with the mouth of the Murray River on the southern coast of Australia near Adelaide. It includes over one million square kilometers, and about one-seventh of the land area of Australia. The Basin contains more than 20 major rivers as well as important groundwater systems. It is also an important source of freshwater for domestic consumption, agricultural production and industry.

The rivers of the Murray-Darling Basin are characterized by flat gradients, highly variable flows, and limited runoff. Average annual runoff is some 24 million cubic meters (m3) of which around half is lost to natural processes. Total runoff is the lowest of any of the world’s major basins and average annual flow to the sea is a mere 400 m3 per second. Much of the basin is semi-arid and some 86% of the area contributes no runoff. The basin covers 14% of Australia but receives only 6.1% of Australia’s mean annual runoff. There are about 30,000 wetlands in the basin.

The water resources of the basin are now highly developed. Annual diversions from the river system are 11.43 million m3, 96% of which is for irrigation. Total water storage capacity in the basin is 34.7 million m3, which supports some 1,470,000 hectares of irrigated crops and pastures. In 1996 the basin was home to nearly 2 million people and another one million people outside the basin were heavily dependent on its water resources. Around 40 percent of Australia’s

gross value of agricultural production originates from the basin.

History of the Basin Organization

The institutional arrangements in the Murray-Darling Basin have evolved through three major stages: (1) an intergovernmental agreement allocat-ing water flows of the Murray River and providing for the construction and operation of infrastructure on the River, benefiting three states (South Australia, New South Wales, and Victoria); (2) the extension of the scope and structure of the intergovernmental arrangements to the Darling River, as disputes among the States over river flows and water quality escalated, and unilateral state actions were implemented to restructure irrigation schemes and to limit water uses, and; (3) the emergence of integrated water resource management in the Basin with new organizational structures and relationships at the sub-basin and basin levels, and with leadership and financial support from both the national and State governments.

The states, prompted by substantial fiscal problems, reorganized water provision and water management opera-tions in the 1980s. The essence of the reorganizations was this: publicly provided services for which fees could be collected should be either corporatized (financially self-sufficient governmental bodies) or fully privatized. States and territories were offered financial incentives for the adoption of measures consistent with the National Competition Policy intended to improve public-sector efficiency in Australia. In 1988 the Murray-Darling Basin Commission (MDBC) was formed under a new Murray-Darling Basin Agreement (MDBA) and took over the

Darling

Murray

Murrumbidgee

Mur

ray

Darling

Lachla

n

Edward

Loddon

Macquarie

Namoi

Wei

r

Gwydir

CANBERRA

Broken Hill

Parkes

WaggaWagga

Bathurst Richmond

NowraGoulburn

Dubbo

Orange

Walgett

Leigh Creek

Port Augusta

Wilcannia

Thargomindah

Bourke

Adelaide

Melbourne

Wodonga

Albury

Wangaratta

Barnsdale

Sale

TraralgonMorwell

MoeMelton

Ballarat

Geelong

ColacWarrnamboolPortland

MountGambier

Murray Bridge

Mildura

Horsham Bendigo

Echuca

Sydney

Newcastle

Port Maquarie

Gold Coast

Grafton

Brisbane

Darling

Murray

Murrumbidgee

Mur

ray

Darling

Tasman

SeaLac

hlan

Edward

Loddon

Macquarie

Namoi

Wei

r

Gwydir

A U S T R A L I A

36

34

32

30

38

148 150 152 154

138 140 142

144 146AUSTRALIA

MURRAY-DARLING RIVER BASINSELECTED CITIES

NATIONAL CAPITAL

MAJOR ROADS

0 50

0 50 100 Miles

100 Kilometers This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other informationshown on this map do not imply, on the part of The World BankGroup, any judgment on the legal status of any territory, or anyendorsement or acceptance of such boundaries.

AUSTRALIAMurray-DarlingRiver Basin

Box 1

transboundary water management role plus took on a new responsibility for coordinating integrated catchment management across the whole Basin.

As of 2004, the institutional arrangements for governing and managing the Murray-Darling Basin have been modified in substantial ways. The state governments and the Murray-Darling Basin organizations have been supplemented at the sub-basin level with catchment management bodies that are still developing their own. At the national level, the Council of Australian Governments (COAG) and the Commonwealth government have become intensively involved in the development of national water policy reforms and initiatives that in some respects lead and in other respects follow the integrated water resource management direction taken during the 1990s by the Murray-Darling Basin Ministerial Council.

Salient Features of the Basin Organization

The states manage intrastate and interstate water uses and regulate water quality. The Commonwealth exercises some policy development and coordination functions through the Council of Australian Governments (COAG) including setting a national water policy reform agenda (1994) and a National Water Initiative (2004). The Commonwealth, which collects all income taxes, provides funds for natural resource management to the States as well well as to sub-state water management entities such as catchment management boards or authorities. State governments have full sovereign powers over land, water and natural resources. State law typically vests authority for the control and use of water in a ministry which is responsible for water rights systems etc. State governments built and still own and operate major dams on rivers. Irrigation schemes, with the exception of Victoria, are now all privatized. Local governments provide and operate water supply and sanitation infrastructure, generally with state financial assistance. Local governments, with state financial assistance, are responsible for flood protection.

The Murray-Darling Basin Ministerial Council, composed of ministers from the State and Commonwealth governments and a representative from the Australian Capital Territory (ACT), is the policy making body for the Murray-Darling Basin under the provisions of the 1992 Murray-Darling Basin Agreement. Council decisions require unanimity, giving participating governments an effective veto. The Murray-Darling Basin Commission is the executive body for implementing Council decisions on Basin policy and management. It also advises the Ministerial Council. The Commission consists of representatives from each basin government. Staff and opera-tions are funded under the cooperative agreement among the participating governments. The Murray-Darling Basin Community Advisory Committee (CAC) advises the Ministerial Council, representing the interests and concerns of local communities and stakeholder groups. On major policy issues, the Ministerial Council typically receives two reports—one from the Commission and one from the CAC. A “ring-fenced” business operation of the Murray-Darling Basin Commission, River Murray Water controls the flows of the transboundary Murray River as a bulk supplier, operating infrastructure facilities on the main stem of the river to assure the States of New South Wales, Victoria, and South Australia of their flows under provisions of the 1914 River Murray Water Agreement. Catchment Management Boards/ Committees/Authorities (depending on the state) are mostly coordinating and advisory sub-basin and sub-State bodies responsible for protecting water quality and riparian and floodplain conditions through efforts to improve land stewardship and through actions such as riverbank protection projects and tree-planting. Water Management Committees are community-based advisory commit-tees composed primarily of water users. They advise on water allocations, environmental flows, and in some cases flood protection, river facility operations, and/or water pricing.

Rural Water Authorities (Victoria only) are governmental, corporatized water suppliers managing all aspects of bulk water provision for municipalities and operating irrigation schemes (including major dams off the main stem of the Murray). They are virtually self-sufficient financially and set their own water prices. Irrigation companies are the fully privatized bodies operating in all states but Victoria, to which previously governmental irrigation infrastructure assets and operations have been transferred. Water user groups have been around for decades,

9

and take multiple forms. Nearly every river has at least one. Though none have formal powers, some are quite influential and play an important role in community participation.

Outcomes

CONTEXTUAL FACTORS AND INITIAL CONDITIONS. Initial conditions and contextual factors in the Murray-Darling were in most respects quite favorable for integrated water resource management. There are few class, religious, or other socio-cultural distinctions. Overall, the basin was quite favorable social and economic ter-rain for the development of Basin management institutions: its semi-arid climate makes water issues significant enough to stimulate action, and the relative wealth and homogeneity of its population present few barriers to such action. The initial distribution of resources among basin stakeholders clearly has favored irrigators in the basin, who account for more than 90% of water diversions. This has not stymied the emergence of basin-scale institutions or the development of integrated water resource management, but it has slowed the pace of reforms such as licensing restrictions and cost-recovery pricing, with the latter driven more by national eco-nomic policy reforms than by internal basin-scale reform efforts. Indeed, national reform efforts have provided leverage to policy actors within the basin trying to enact restrictions on water diversions and the reduction of agricultural water subsidies. Current reform efforts oriented toward implementation of the Living Murray initiative have entailed several concessions toward irrigation interests. THE DECENTRALIZATION PROCESS. Concerns about devolving decision making authority are of little consequence in Australia, where primary decision making authority predominantly rested at the sub-basin level with the state governments. Over time, and with the cooperation and consent of the national government, the states have constructed intergovernmental arrangements to control and operate Murray River flows and then to address other issues. The process has been as much a matter of integration as decentralization. Central-level recognition of basin governance and management has been complete and consistent. The Commonwealth government not only recognizes, but participates in and helps fund, basin-scale organizations such as the Ministerial Council and the Commission. Through financial incentives offered to the states and to sub-state catchment manage-ment authorities, and through establishment of and participation in bodies such as the Council of Australian Governments (CAG), the national government has actively encouraged the development of integrated water resource management in the basin. These commitments from the national government have remained consistent across elections and changes in party control.

CENTRAL-LOCAL RELATIONSHIPS AND CAPACITIES. One hindrance to basin-level integrated water management is the system of water rights. Entitlements to the use of water are issued, and vary, by state. Rights generally fall into three categories: licenses issued to organizations such as irrigation companies, trusts, or districts, licenses issued to individuals, and rights of riparian landowners. Each state has arguably over-allocated water licenses. Groundwater still has uncertain status in the states’ licensing arrangements and has not been fully integrated into the licensing system. Reforms will be needed if measures such as water trading and the protection of environmental flows are to be fully implemented. A second hindrance is that the organizations in the basin most directly associated with integrated resource management (eg., the sub-basin catchment management authorities) have virtually no financial resources of their own and are dependent on funding from state and Commonwealth governments.

Otherwise, central-local relationships and capacities are favorable to integrated water resource management. Indeed, the basin management participants have the ability to create and modify institutional arrangements. The States and the Commonwealth governments have amended and replaced the agreements for the Murray River and the Murray-Darling Basin during their existence resulting in their current structure of Ministerial Council, Commission, and Community Advisory Committee. In addition, there is considerable experience at the local and state levels with self-governance and service provision. By the time the Ministerial Council adopted its Natural Resource Management Strategy in 1990, there were already numerous sub-basin groups addressing issues of

10

improved land stewardship. With the existence of local government councils, irrigation trusts and districts, and rural water authorities (in Victoria), one finds that participatory catchment management has been introduced into a situation already rich with social and organizational capital.

BASIN-LEVEL INSTITUTIONAL ARRANGEMENTS. Basin-level institutions are conducive to integrated resource man-agement at the basin level. There are basin-level governance organizations and sub-basin organizations, each with firm recognition and considerable support from the state and commonwealth governments. The states themselves are recognized as communities of interest within the river basin, as are a number of stakeholder communities represented on the Community Advisory Committee. Basin users and policy makers have an array of means by which to negotiate and enter into agreement for committing and combining resources for projects and programs to improve basin conditions. Monitoring of basin conditions is performed regularly and then consolidated by the Commission staff.

Two basin-level institutional arrangements have led to limited uncertainty. First, the clarity of institutional bound-aries has been somewhat reduced by the introduction of the relatively new catchment management bodies. Local governments are not certain how the land and water management activities of these bodies will overlap with their own traditional land use regulatory authority. Some basin-level staff were unsure how coordination will occur between the new bodies’ water management activities and larger programs undertaken at the basin scale. Second, though mechanisms for conflict resolution exist to address conflicts between water users and states, it is less clear how conflicts between sub-state and sub-basin entities such as a local government and a catchment management body, or between catchment management bodies etc. would be addressed.

Murray-Darling Basin water resource management’s successes in gaining intergovernmental cooperation and commitment, instituting mechanisms for stakeholder participation, and generating trusted data are considerable. In terms of devolution of authority, stakeholder participation, and financial self-sufficiency, the arrangements have been generally successful. Water resource management is still driven by policy elites and audit groups in each state, but all actual management is carried out at regional levels in local offices with almost complete authority for policy implementation (including water sharing). Management and operation of dams and irrigation schemes has been transferred to entities designed for completely local-ized day-to-day management and financial sustainability. In all states but Victoria, this has included the privatization of irriga-tion schemes and their assets into the hands of the irrigators. Urban water and floodplain management have always been local responsibilities; this has continued and intensified in both technical and financial terms. All levels of water management are now supported by stakeholder advisory groups. This is complete in the case of privatized irrigation schemes where there is now no government involvement, but is also particularly well-developed for integrated catchment management. Public consultation is now the norm even for urban water and wastewater projects.

The national water reform agenda articulated in 1994 emphasized financial sustainability. Economic elements of water reform policy required removal of cross-subsidies, consumption-based water pricing, new investments only if they were economically viable and ecologically sustainable, better specification of water entitlements, and the encouragement of water trading. These reforms were accompanied by institutional reforms that separated regulatory roles from service provision, required greater local-level responsibilities for management, and encouraged public education and consultation. Generally, both urban and rural (irrigation) water supply infrastructure now gets no government funding for operations and mainte-nance and a very small and steadily decreasing amount of capital funding.

Today, the individuals and organizations in the Murray-Darling Basin management structure are incorporating sub-basin catchment organizations into the framework for integrated water resource management, while leaving the basin-level orga-nizations relatively unchanged. They are also attempting to achieve an ambitious portfolio of ecological restoration objectives in addition to their past focus on balancing water supply and demands for consumptive human uses. They undertake these efforts at a time when national-level bodies are becoming more actively involved in water policy, creating a national water policy framework into which the Murray-Darling will be expected to fit. These challenges will further test the robustness of the institutions for river basin governance and management in the basin which to this point have proved quite effective.

Conclusion

11

12

Jaguaribe River Basin, Ceará, Brazil

Overcoming poverty, politics, and inequality to successfully increase stakeholder participation in key water management issues.

GEOGRAPHICAL OVERVIEW. The Jaguaribe river basin is part of the Atlantic hydrographic region of Brazil’s Northeast, located entirely within the state of Ceará. It has a drainage area of 72,560 square kilometers, covering approximately 48% of the state’s territory. The principal river runs from south to north for about 610 km, flowing into the Atlantic Ocean. The basin has 80 municipalities and more than 2 million people, representing about a third of Ceará’s population. The majority of the basin’s population (over 55%) now lives in urban areas, still well below the state and national averages (of 72% and 81% respectively).

Most of the Jaguaribe basin falls within the semi-arid region known as the northeastern sertão (hinterland). Precipitation is highly variable, ranging from 400 mm in the hinterland to 1,200 mm along the coast. Although such rates of rainfall are higher than in many dry regions in the world, in Ceará the combination of impermeable crystalline rocks in the soil and high temperatures produce elevated rates of evapotranspiration — over 2,000 mm for the basin — and low levels of water retention and storage. Groundwater resources are considered of limited importance in most areas of the basin. Cyclical droughts occur at least every five years and can persist over a several year period.

Due to the climate, the Jaguaribe river and its tributaries are ephemeral rivers that only flow during the rainy season, typically but not always from January to July. The key water management challenge is to capture the water in reservoirs in rainy years and to manage it such that it will last for several years, in case the following years are drought years. The other important challenge is the increasing dependence of the state

capital Fortaleza, located in a different basin, on water from the Jaguaribe basin. For management purposes, the Jaguaribe basin has been divided into five hydrographic regions, corresponding to three parts of the basin to two sub-basins: Upper Jaguaribe, Middle Jaguaribe, Lower Jaguaribe, Salgado and Banabuiú.

In general, the basin is considered poor, even for Ceará, a state that contributes only 1.8% to the nation’s GNP. Indeed, 63% of the state’s income is generated by Greater Fortaleza, mostly from the service sector (56%) fol-lowed by the industrial and agricultural/ranching sectors (38% and 6% respectively). The Jaguaribe basin follows the state trend, with most of its income deriving from the service sector. Although agriculture accounts for only a small part of the basin’s income, it is of great social importance since subsistence agriculture (dependent on rainfall) still employs most of the rural and poor basin population.

History of the Basin Organization

Before the current reform began in the early 1990s, water resource policy and management in the Jaguaribe Basin, in Ceará state, and in the semi-arid region more generally, was traditionally the domain of federal initia-

Jaguarib

e

Jaguaribe

Palhan

o

Banabuiu

Juca

Conceica

o

Fortaleza

Santa Quitéria

Crateús

MoradaNova

Canindé

Quixadá

Encantado

Iguatu

Tauá

Sousa

Cajazeiras

PombalAntonina

Araripe

JuazeiroDo Norte

Solonópoles

Russas

Aracatí

ATLANTICOCEAN

Jaguarib

e

Jaguaribe

Palhan

o

Banabuiú

Juca

Conceica

o

B R A Z I L

383940

3839

4

5

6

7

5

4

6

7

8

IBRD 33728

BRAZIL

JAGUARIBE RIVER BASIN

SELECTED CITIES

MAJOR ROADS

0 10 20

0 10 20 30 Miles

30 Kilometers

This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other informationshown on this map do not imply, on the part of The World BankGroup, any judgment on the legal status of any territory, or anyendorsement or acceptance of such boundaries.

BRAZIL

JaguaribeRiver Basin

12

Box 2

13

tive. The main federal agency for drought prevention — the National Department of Drought Relief (DNOCS – Departamento Nacional de Combate das Secas)— was created in the 1910s. Since then, it has expended large amounts of federal money for massive construction of water storage infrastructure. There are close to 7,000 reservoirs in Ceará, about 130 of which are considered ‘strategic’ and serve multiple water uses. This supply-based approach was characterized by a centralized, rigid and nontransparent decision-making structure that favored the interests of large landowners and the irrigation projects financed by DNOCS.

The first steps toward institutional change in Ceará were the creation of the Water Resources Secretariat and the approval of the state Water Law in 1992. The law embraces the main ideas of modern water resource man-agement, following the principles of the – only subsequently developed - federal water law of 1997 and those of other state laws: integrated water management with the river basin as the planning unit; water as a finite, fragile and economic resource, and decentralized and participatory management. Likewise, Ceará included the same management instruments later instituted by the federal law: state and basin water resources plans; bulk water use permits; bulk water charges; and a water resources information system. While most states relied on existing environmental or water agencies funded through the general state budget, in Ceará a strong, indepen-dent and (eventually) self-financed state Water Resources Management Company (COGERH – Companhia de Gestão dos Recursos Hídricos) was created in 1994 to carry out management, monitoring and enforcement functions. In addition to affirming a new approach to the management of state water resources, the creation of COGERH also represented the beginning of a process of taking over control of federal infrastructure in the state, until then mostly governed by DNOCS.

Since COGERH was created, the predominant logic has been to centralize the technical and social aspects of water management as well as the collection of water charges in that agency, with the objective of financing both its administrative expenses and the operation and maintenance of the water infrastructure for which it is responsible. One of the reasons for the decision to centralize water management functions, rather than to operate on a basin-by-basin basis as the federal framework proposes, is the need for the redistribution of resources among basins in the state, since – except for the Greater Fortaleza Basin - none could be expected to maintain their own management agency and cover their own operating and maintenance expenses. Another reason is the fact that the metropolitan area is dependent and will become even more dependent on the Jaguaribe basin for its water supply. This need for ‘integration of supplies’ between basins resulted in an attitude within the State that the state government needed to maintain an overriding control on the manner in which rights were allocated and distributed within the state.

Salient Features of the Basin Organization

The creation of basin institutions has occurred gradually over more than ten years, under the initiative and coordination of COGERH and with the support of the State Water Resources Secretariat (SRH – Secretaria dos Recursos Hídricos). The result has been the emergence of various types of local organizations. The Jaguaribe-Banabuiú User Commission basically defines the annual operating rules of the three major reser-voirs of the basin according to the negotiated water allocation between the users of the “regulated valleys”, i.e. along the mainstem of the rivers. Thirty-six user commissions of “isolated strategic reservoirs” guarantee multiple water use in locally important reservoirs during drought periods. Together with the five sub-basin committees, these institutions cover the entire territory of the Jaguaribe basin, but they also overlap, leading to frictions in recent times.

In the search for solutions to the conflicts that arose during the major 1992-1994 drought, COGERH officials and local stakeholders decided to create a commission of water users to conceive and manage an emergency operation plan for the seasonal water allocation in the basin. Based on the way local organizations were already organized, thirty-six commissions of reservoir users and one commission of users in the regulated valleys of the Jaguaribe and Banabuiú Rivers were eventually created out of this pioneering experiment, which was a major

13

1414

departure from previous policies under the federal DNOCS. While their attribution and composition are simi-lar (water users, civil society, and representatives of the key institutions debate and deliberate, together, on the release of water stored in the reservoirs with strategic purposes for integrated water management) the reser-voir commission’s main purpose is to guarantee multiple water use in the immediate surroundings of the reser-voir during drought periods when rivers dry up. The five sub-basin committees that exist today in the Jaguaribe Basin were created only several years after stakeholder participation was established through the commissions. The committees have broader water management attributions than the commission, such as setting guidelines, approving river basin plans and conflict resolution. Their creation, which occurred between 1998 and 2001, was a much more formalized process which had to comply with both national and state regulations.

Outcomes

CONTEXTUAL FACTORS AND INITIAL CONDITIONS. At the time that reform began, local conditions in the Jaguaribe basin would have appeared for a number of reasons to be somewhat unfavorable to the development of decentralized and integrated water resources management. First, the basin is relatively poor. Second, par-ticipatory water management ran contrary to the prevailing political culture of the Ceará hinterlands (sertão), since until the reforms began, water there had historically been considered either a private good, the property of the owners of the lands through which it flowed or under the control of the government agencies of the reservoirs within which it lay (usually DNOCS). Third, Ceará had one of the most entrenched oligarchies in the Northeast.

OTHER FACTORS FAVORING REFORM. At the national level, the transition to democracy, that followed the fall of Brazil’s military regime in the mid-1980s, was accompanied by broadened acceptance throughout Brazilian society of values such as democratization, decentralization, and participation in policy making. An unprecedented movement within the technical water resource community began to promote integrated water resources man-agement. This movement, led in large part by the Brazilian Water Resources Association, had a strong impact on the dissemination throughout the country of certain management models, such as water use rights, pricing, and river basin level management.

In Ceará, specifically, the key enabling factor was the existence of strong political will in the form of advocates for the adoption of a new form of management of water resources. A succession of Governors and leading government officials over a time period of over 12 years furnished the stable political support required for the adoption of such radical changes in water policy. This coupled with the recruitment and maintenance of staff who were strong from a technical point of view and extremely dedicated to the new concepts of water resources management aided greatly in the process of developing the new policies and convincing the water user public to accept these principles. Thus, the change in the state political environment in the mid-Eighties helped overcome the other rather unfavorable factors and was crucial break the path dependence of water resources management in the state.

THE DECENTRALIZATION PROCESS. The decentralization in the Jaguaribe basin was marked by two distinct pro-cesses: i) decentralization from federal to state level, a result of the increased technical, institutional and financial capacity of Ceará’s water resource management agencies; and ii) decentralization from state to local level, which occurred through the creation of deliberative bodies at the river basin and lower territorial levels. The creation of COGERH was not part of the original design called for in the state water law, but resulted from the World Bank’s insistence that the state create a water agency with management, monitoring and enforcement functions, including pricing and the involvement and organization of local stakeholders in order to be able to implement the law. The fact that in 1997 COGERH took over some of DNOCS’s management responsibilities represented a major step towards decentralization from the federal to state level.

15

There are signs that federal-to-state devolu-tion is increasing, such as a recent agreement between COGERH and the National Water Agency (ANA) delegating authority for issuing use permits for waters in federal reservoirs. Even DNOCS never acquired such authority. The state-to-local decentralization process has been more complex. It is a top-down initia-tive, in which COGERH’s User Mobilization Department (DOU) played a central role. The Department employed a group of sociologists who, in the effort to implement participatory management practices, were able to manage a delicate balance between the sometimes cen-tralizing tendencies and interests of a state-level institution, and local interests and customs. In 2003, the DOU was dismantled, revealing that the priorities of the current state government subscribes rather to the old philosophy of cen-tralized management with the least participation possible from users. Accordingly, upper echelons at COGERH and the State Water Resources Secretariat seem intent to minimize the par-ticipatory effort and have, in fact disbanded the liaison efforts of the administration with the basin committees, commissions and user community. Although the long-term impact of these changes on the user commissions and river basin committees is still unclear, reform-oriented officials hope that the high level of mobilization achieved in the basin over the last ten years will make it difficult to undo the advances made thus far.

CENTRAL-LOCAL RELATIONSHIPS AND CAPACITIES. The devolution of some of DNOCS’s authority over the management and control of reservoirs to Ceará state has been highly effective, since COGERH has developed substantial technical, administrative and financial management capacities. Currently, COGERH operates and man-ages, through its agreement with DNOCS, all major reservoirs in the state, accounting for over 90% of the state’s water storage. But other aspects of water management remain underdeveloped. The state has proceeded only slowly with the implementation of groundwater management. The development of a new long-term water use permit system has also been slow.

For climatic and cultural reasons, the state has emphasized reservoir operation and water allocation over other important issues such as the management and control of groundwater use, implementation of water use per-mits and the extension of bulk water charges to the irrigation sector, etc. The process of devolving authority and responsibility from the state to local levels is less easy to characterize. The basin committees (which were supposed to be the most important basin institution) are formal institutions that still have not found a de facto place in the water management system. They lack effective technical, administrative and financial support. Contrary to the national framework, which places river basin committees and agencies at the heart of deci-

15

1616

sion-making about water pricing, Ceará has centralized bulkwater charging at the state level. This means that, in Ceará, basin committees will not have their own executive arm (basin agencies), nor will they have financial resources of their own. With most of their expenses provided by members themselves or by the occasional financial support of COGERH, the committees still have little influence over water resource management. Their activities have been limited to information dissemination, consciousness-raising and capacity-building among local actors and the resolution of water use conflicts. Major structural projects continue to be decided solely by the state government, especially with respect to supplying water to the state capital Fortaleza. A COGERH plan is being formulated which will define an executive support structure to replace the partial and low-level technical support that its regional offices have been providing the committees. Importantly, however, in addition to the political change in leadership in the state and its agencies, such as COGERH, the agency has also come under tremendous pressure with the creation of the National Water Agency ANA. At least 80% of the most competent staff of the State of Ceará in the area of water resources management have left the state and are now employed at the national level. This is in part due to the significant efforts of the Federal Government in the recruitment of the well-recognized staff of Ceará and partly due to the dissatisfaction of the staff with the new policy directions.

Conversely, decentralization in the Jaguaribe Basin has gone furthest with the user commissions, especially through the negotiated allocation of water. However, with resistance within COGERH and the State Water Resources Secretariat (to which COGERH is subordinate) to give decentralized bodies greater power over water management, the result is that only the sub-basin committees have been legally created, and even these have received little real support or authority. Meanwhile the Jaguaribe-Banabuiú Valley Commission — where the process of participatory decision-making began and has continued with great intensity — is still only an informal institution. This contradictory situation has created tensions between the sub-basin committees and the Jaguaribe-Banabuiú Valley Commission. Transformation of water management practices in Ceará appears to need more time.

BASIN-LEVEL INSTITUTIONAL ARRANGEMENTS. Water allocation is the most prestigious activity in the basin, with the greatest local impact and hence the greatest capacity to mobilize local actors. This importance has recently led some representatives of the Executive Board of most sub-basin committees to argue that the Jaguaribe-Banabuiú Valley Commission should be dismantled, with the transfer of its responsibilities to the committees, as occurred in Ceará’s Curu River Basin. This would involve a fundamental rearrangement in basin institutions, since four of the five sub-basin committees are involved with the regulated valleys. In the Curu basin, the transfer of power from commission to committee was simpler because the territorial jurisdiction of the user commission fell entirely within the geographical area of a single committee.

While the committees and the Jaguaribe-Banabuiú User Commission are sources of controversy, all actors in the basin and at the state level have shown support for the 36 reservoir user commissions. The allocation process they engage in is similar to that carried out by the Jaguaribe-Banabuiú Valley Commission, but the decisions have only very localized impact and consequently transaction costs are lower. Usually, the commis-sions include only users or groups of users directly affected by water allocations in the area of hydrological influence of a single reservoir, since members are mostly made up of organizations working in the perimeter of the reservoirs, and in the immediate downstream area. Despite the current uncertainties concerning insti-tutional boundaries, both user commissions and basin committees have been promoting the resolution of water use conflicts, with the support of COGERH. They have come to be perceived as the legitimate forum for negotiating conflicts in terms of water allocation and quantitative use, and for airing other controversial issues related to water quality and environmental degradation.

17

The devolution of federal authority over the management and operation of federal reservoirs to Ceará State has been highly effective. Federal institutions continue to develop, support and finance specific drought relief programs in the semi-arid region, together with state governments and sometimes with international organizations.

Decentralization from state to local level has been more partial. Although COGERH has decentralized the allocation of strategic reservoir waters to local institutions, many traditional water management attributions continue under COGERH’s and the state’s purview, such as water permits, bulk water pricing, planning, operation and maintenance of hydraulic infrastructure, groundwater management and control, etc. Furthermore, none of the changes introduced thus far have affected municipalities, which are still fully responsible for land use and urban drainage.

The creation of sub-basin committees and user commissions, under COGERH’s coordination with the support of the State Water Resources Secretariat, has increased stakeholder participation of all types, such as municipalities, public and large private irrigators, fishermen, and industry leaders. Although so far stakeholder involvement has been limited largely to the negotiated allocation of water and to conflict resolution, these experiences represent a radical transformation in management practices, transforming water users from uninformed takers of water management decisions to informed and aware participants in the management process. That said, local stakeholders still have no say in some decision-mak-ing processes that affect them directly, such as bulk water pricing or inter-basin transfers to Greater Fortaleza, which continue solely under the control of state government agencies. Also, the concerns of local stakeholders with respect to water quality problems and broader environmental problems related to water have yet to find a place on the agenda of the state water institutions.

At the basin level, the financial resources of decentralized institutions are both precarious and insecure. Neither the basin committees nor user commissions have their own financial resources, depending entirely on contributions from the state government and from their own members. At the state level, though, bulk water pricing has represented an important change in terms of financing water management. Until pricing was introduced, bulk water supply services were partially or fully subsidized by the general state budget. The pricing system has enabled COGERH to gradually achieve financial sustainability for its operation and maintenance costs and for investments in new water infrastructure. Although Ceará is one of Brazil's poorest states, collected revenues reached 83% of the total revenues in 2003, when 10 of the expected 12 million reais (USD4.6 million) were actually collected. The Metropolitan Basin —the state’s principal urban and industrial area—contributes over 90% of the total revenues from bulk water pricing. This means that, since most of the Jaguaribe basin’s users are irrigators who pay almost symbolic amounts, the operation and maintenance costs of the large water infrastructure in the Jaguaribe River Basin are currently subsidized by users in the Metropolitan Basin.

Much remains to be done, especially with respect to building a more holistic management system that incorporates efforts to promote better water quality and to coordinate water and environmental management. Nonetheless, the achievements made thus far are remarkable when compared to the problems and practices that seemed, until recently, impossible to overcome. Water rationing in the Jaguaribe Basin used to be an almost permanent state of affairs. Traditional institutions used to privilege the interests of entrenched oligarchies. Civil society and small users were excluded from water-related decision-making. Water was, in general, managed and protected in only the most precarious and unsustain-able of ways. All these unfavorable factors have been strongly challenged. The achievements made thus far, therefore, are quite impressive in spite of the current backpedaling.

The case of the Jaguaribe basin powerfully illustrates that (i) longstanding political support is of major importance in the development and implementation of water resources management reform, (ii) that institutional arrangements for water resources management can successfully be adapted to local conditions in order to achieve positive outcomes, (iii) that even with initial conditions that seem to not favor change, decentralization can be achieved.

One lesson, however, cannot be drawn as yet and only the future will tell if the strength of a 12-year long decentraliza-tion process in water resources management, initiated and strongly supported by a succession of political leaders, and implemented by highly professional and dedicated staff, is sufficient to overcome the current ‘political drought period’ that started with the change in state and institutional leadership in 2003.

Conclusion

17

Alto Tietê River Basin, São Paulo, Brazil

Illustrates the problems faced by water management in highly urbanized and industrialized humid regions in Brazil.

GEOGRAPHICAL OVERVIEW. The Tietê —São Paulo state’s largest river— runs 1,100 km from its eastern source in the São Paulo Metropolitan Region to the western border of the state where it joins the Paraná river, which then runs southward, toward the Rio de la Plata estuary between Argentina and Uruguay. The Alto-Tietê cor-responds to the upper part of the basin, from the headwaters of the Tietê River in Salesópolis city to the Rasgão Reservoir. Most of the Alto Tietê’s urban area lies on sedimentary soils. The climate in the basin is typical of tropi-cal high plain savannas, with a temperate summer. Precipitation varies little throughout the basin, averaging 1,400 mm per year.

The area covered by the Alto Tietê basin is almost coterminous with the Metropolitan Region of São Paulo. With a drainage area of 5,985 square kilometers (2.4% of the state’s territory), the basin encompasses 35 of the 39 municipalities and 99.5% of the population of Greater São Paulo. 37% of the basin’s territory is urbanized. Population growth and urban sprawl in Greater São Paulo have been rapid and uncontrolled in recent decades. In 2000, 17.8 million people lived in the basin and estimates are that in 2010, the population will reach 20 mil-lion. This massive human occupation was accompanied by the large-scale construction of water infrastructure, including dams, pumping stations, canals, tunnels and inter-basin transfers to and from neighboring basins. These projects were usually built to serve multiple purposes, especially hydropower, urban supply, and flood control. Today, the Alto-Tietê basin is served by a complex hydraulic and hydrological system. Despite this extensive water infrastructure, the water availability of the region is still very low (201 m3/hab/an) and even lower than the semi-arid regions of the Brazilian Northeast. Accordingly, the two key management issues to be addressed in the

Alto Tietê basin are water quantity to supply a burgeoning population and water quality, which is deteriorating to a point where water avail-ability for a range of uses is severely impacted. Urban flooding control and mitigation repre-sents another major challenge in the basin.

History of the Basin Organization

Contrary to the semi-arid region which has had a long history of federal intervention, water management practices in São Paulo State have historically been a local affair, even for the federal waters crossing through the state. This is especially the case in the Alto-Tietê Basin where all waters are under state dominion.

In technical, human, and financial terms, São Paulo’s water management and environmen-tal agencies are the best equipped in the country. The water resources management agency (Departamento de Água e Energia Elétrica, DAEE) was created in the 1950ies

Alto Tiete

São Paulo

Santos

Campinas

Sorocaba

S.João del Rei

BeloHorizonte

Marília

Assis

Londrina

Maringá

PresidentePrudente

Lins

Bauru

AraraquaraPoços deCaldas

PousoAlegre

S. Josédos Campos

Guaratinguetá

Araxá

São José

Curitiba

Ponta Grossa

RibeirãoPretoˆ

Uberlandiaˆ

Tres Lagoasˆ

Aracatuba

ATLANTIC OCEAN

Alto Tietê

Peixe

Paraná

Ivai

Paranapanema

Aquapei

Pa

ranaiba

Grande

B R A Z I L

52

20

22

24

22

24

48

52 50 48 46

46 44

BRAZIL

ALTO TIETÊ RIVER BASINSELECTED CITIES

MAJOR ROADS

0 25 50

0 25 50 75 Miles

75 Kilometers

This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other informationshown on this map do not imply, on the part of The World BankGroup, any judgment on the legal status of any territory, or anyendorsement or acceptance of such boundaries.

BRAZIL

Alto TietêRiver Basin

18

Box 3

19

and was the first to issue water use permits in Brazil. Nevertheless, it was unable to prevent or reduce the proliferation of policies affecting water use by other sectoral agencies, with little coordination. The São Paulo State Environment Agency (Companhia de Tecnologia de Saneamento Ambiental, CETESB) was created in the 1970ies. Although responsible for all kinds of pollution control, CETESB’s command-and-control regulations have been mostly limited to the state’s largest industries and worst polluters. As a result, some other players, for example, water and sanitation companies, have been subject to much weaker regulation and enforcement. There has been little coordination and integration between DAEE and CETESB to manage water quantity and quality, and between those agencies and many others directly involved with problems of water use and planning in Greater São Paulo such as the State Agency for Metropolitan Development and Planning (Empresa Paulista de Desenvolvimento Metropolitano, EMPLASA).

Prior to the reform, the system could be largely characterized as: (i) compartmentalized (quantity separate from quality, surface separate from groundwater resources); (ii) centralized finances, planning and decision-mak-ing at the state level, since the municipalities, private users, and civil society had no say in how to manage the water resources; and (iii) inadequate, considering the insufficient technical, administrative, and financial resources available for planning, control, and enforcement activities when compared to the complexity and magnitude of the problems.

Salient Features of Basin Organization

The proposal for creating a new state system for water resource management ultimately resulted in Law 7.663 of 1991 which along with complementary legislation became the embodiment of the State Water Resources Policy. The objective of this policy was to provide reliability of water availability for current and future genera-tions at the desired level of water quality. The policy was based on four principles, specifically: integrated water management, with the river basin as the planning unit; water as a finite and fragile resource; water as an eco-nomic good, and decentralized and participatory management. These principles were to be implemented by a number of organizations which are described below.

A State Water Resource Council, a deliberative body, with stakeholder participation, would be in charge of supervising and regulating the State Water Resources Management System. CORHI, the Council’s Executive secretariat, would act as a technical body in charge of elaborating the state water resources plan and promot-ing institutional integration among all the state water related institutions. State Water Management Agencies would be in charge of water use and pollution control and of implementing the Water Resources Management System. River Basin Committees would act as deliberative stakeholder bodies with decision-making and regula-tory powers. Finally, River Basin Agencies would act as their technical and administrative arms. The five instru-ments they would use would be: basin and state water resources plans; bulk water use permits; bulk water charges; the classification of water bodies according to predominant use and water quality standards, and a state water resource information system.

As in the French system, the center of gravity of the new management system would be the river basin com-mittees. These have a tripartite structure, with members representing the state government (including public water users), municipalities and civil society (including those representing private water users). The committees were expected to initiate a new approach in planning and management, with the technical and administrative support of basin agencies. The key to this new approach would be the implementation of bulk water use charges for waters under state dominion. The allocation of revenues would follow investment plans approved by the committees and be operationalized by the agencies, thus guaranteeing the financial sustainability of the new basin institutions. A state fund to finance water management was also established (Fundo Estadual

20

de Recursos Hídricos - FEHIDRO). This entire management structure is supervised and regulated by the State Water Resources Council . Both CETESB and DAEE continue to carry out their traditional functions as before the legislation. Legally, DAEE would also exercise the role of the basin agencies only until the latter are created.

Outcomes

CONTEXTUAL FACTORS AND INITIAL CONDITIONS. In the Alto-Tietê river basin the context seemed favorable to the development of a decentralized and integrated water resources management system. First, São Paulo state is by far the most important industrial and financial center of Brazil and the Alto-Tietê river basin is the richest basin. Second, in social terms, both the basin and the state were fertile terrain for reform: motivated by worsening water-related problems, a result of intense urbanization and industrialization, social movements demanding improved water policies emerged in the mid-1980s. At the national level an unprecedented movement within the technical water resource community began to promote integrated water resources management. Ideas and experiments related to integrating sectoral policies involved in water management at the river basin level began in the 1970s. During the period of democratic reform in the mid-1980s, ideas about integrated river basin management were associated with democratization, decentralization, and civic participation in policy making. It was on the basis of this technical and social capital that the water manage-ment agencies of São Paulo State began to pioneer a process of institutional transformation, introducing new legal concepts such as water use rights, pricing, river basin level management and planning, and stakeholder participation.

The initial distribution of resources among basin stakeholders also seemed to favor reform. The main users are the urban water supply companies, which face serious problems of supply in the face of growing demand. These problems should be an incentive for changes in water management practices. The irrigators, traditionally the most resistant to change in the water sector and especially to water charging, use an almost insignificant amount of water in the basin. The industries are the second largest user in the basin.

Despite this favorable context, the political will to advance the changes has proved insufficient. The political and environmental complexities of the Alto-Tietê basin seem to make it particularly difficult to implement practices involving integration and participation of decision making and implementation of administrative and demand-management instruments that could address São Paulo’s worsening water situation.

THE DECENTRALIZATION PROCESS. The decentralization process in the Alto-Tietê basin is marked by two dis-tinct processes: i) decentralization from the state to the basin level, which occurred with the creation of the Alto-Tietê Committee in 1994 and, more recently, its water agency, and; ii) further decentralization within the basin in 1997/1998, which resulted in five sub-committees at lower territorial levels.

While the devolution of authority and responsibility from the state level to the basin level was desired by both state government and local stakeholders, there is no agreement about the extent of this decentraliza-tion, i.e. the financial autonomy and capacities of the river basin bodies. The need to decentralize within the basin, however, had consensus from the earliest stages of mobilization for the creation of the Alto-Tietê basin committee. Despite the basin’s small geographical size, state and local stakeholders recognized that the com-plexity of such an intensely urbanized and industrialized region required smaller scale management. Some have disagreed with the methodology used to define the five sub-committees that were created, because they were not based on purely hydrological criteria. But no one questions either the need to create complementary deliberative bodies at lower levels or the fact that basin management participants should be allowed to create and modify institutional arrangements according to their needs and circumstances. The fact that there is no

21

conflict between the responsibilities of the central committee and the subcommittees demonstrates that the decentralization that occurred is satisfactory for both sides (the center and the local levels), despite evidence of some coordination difficulties.

CENTRAL-LOCAL RELATIONSHIPS AND CAPACITIES. The water reform in São Paulo State has been as much a matter of building new institutional capacities of central government and of integrating sectoral policies, as it has been about decentralizing water management. Prior to reform, São Paulo already had management institutions well-equipped with a highly-qualified technical corps. The main issue was not to initiate water management alto-gether, but to develop a culture of integrated management among technical state officials and to build capacity for shared decision-making. A decade after the process began, great improvements can be observed, ranging from the integration of the information systems and actions of the various agencies at the level of common territorial boundaries, to the elaboration of water resources plans at both the state and basin levels. Above all, considerable advances occurred in water use controls through the implementation of a new water permit system, even though monitoring and control is still not systematic. Water permits will take on even greater importance with the implementation of bulk water charges. Volumetric charges will not apply to actual uses but rather to the size of water use permits.

In brief, the advances in state water management capacity have been considerable and in some cases crucial for the survival of the basin committees in this transitory phase. However, tensions and problems exist between the central authorities, and the local bodies and basin committees are not always effective. Indeed, the São Paulo water resources management system as a whole is beginning to show signs of breakdown in face of the state government’s incapacity and unwillingness to make it fully operational, especially by implementing bulk water charges. Water reform in São Paulo seems to need much more time. Considering that the reform process is almost fifteen years old, it is becoming clear that transaction costs are very high in terms of time and money. So much so, that the pioneer state in water reform has begun to lag behind others.

BASIN-LEVEL INSTITUTIONAL ARRANGEMENTS. All of the new institutions defined in the water law have been formally implemented in the Alto-Tietê basin. However, their operation is still marked by imprecision and institu-tional overlap largely as a result of the varying performance of the public representatives holding the committee’s presidency and executive secretary of the committee. The committee has generally had difficulty in becoming a forum for the debate and design of solutions for water-related problems. State institutions (including SABESP and DAEE) make major decisions on water development, use protection and on important infrastructure without going through the committee. Gaining influence over state programs is the main challenge for all basin commit-tees in Brazil, especially for those with little or no capacity for implementing a water pricing system.

The main committee principally discusses the allocation of FEHIDRO funds while the basin agency, which does not have much technical capacity, is taking on the role of the committee for other issues, actively participating in debates about water management questions affecting the basin. DAEE provides technical and administrative support to the main committee (which should be the role of the water agency) and will probably continue to do so until the agency becomes financially independent —and thus, effectively operational— with the institution of water pricing.

The subcommittees are generally considered more dynamic, more effective and locally more important than the main committee although with strong variations among them. The most important role of the subcommittees is to deal with making water resources protection and urban expansion compatible through the implementa-tion of the State Water Source Protection Law of 1997. This includes conceptualizing a broad policy for water source protection and restoration through the elaboration of specific laws for each sub-basin. The Water Source Protection law recognizes that simple prohibition and policing measures for protecting strategic water supply

The analytical framework developed for the larger research project of which this study is part suggests that the political and institutional conditions in São Paulo and the Alto-Tietê basin should have been favorable to the development of integrated and participatory management. However, despite a general underestimation of the time needed to complete water resources management reforms that has occurred worldwide, almost 15 years of reform have not been enough to make the new water resources management system fully operational anywhere in the state. The inertia and lack of political will on the part of the São Paulo state government is the main impediment to further progress in the Alto-Tietê basin and in the state as a whole. Implementation has taken the “path of least resistance”, advancing only in areas that have been less costly in political terms, such as creating regulatory bodies (the state council and the basin and sub-basin committees), the elaboration of water resources plans, and the execution of a new water permit system. However, with more controversial issues such as water pricing, political will weakened in the face of resistance from various govern-ment actors and organized user groups.

It is not hard to identify the presence of committed institutions and, above all, committed water resources profession-als and civil society organizations. However, all this was not sufficient to go further because water management never reached the top of the state government’s political agenda. The lack of government commitment is not, however, enough to explain the lackluster performance of the Alto Tietê committee. Several peculiarities of the Alto-Tietê context made it even more difficult for river basin bodies to take advantage of their favorable conditions and to take the lead in coordinating water management. First, the extent and intensity of water related problems (and solutions), typical of highly dense and industrialized regions, represent an enormous technical, political and financial challenge. Under these conditions, it is harder for stakeholders to identify common interests. Second, the peculiar composition of the AT com-mittee -- which included among its members powerful state government agencies and the municipal government of São Paulo municipality – has so far proved to be more of a problem than an advantage. Thus far these powerful institutions have not taken the committee seriously and it is unlikely that they will throw their energies into its activities until the pricing system is implemented.

Both the intensity of problems and the lack of mobilization of crucial committee members seem, however, to lose importance at lower levels of management. While the main committee still tries to define its roles and powers, the sub-committees have found strong reasons for working. As fora for elaborating and implementing the water source protection policy at the local level (among other attributions), the sub-committees serve as strong building blocks for integrated management in basin. Indeed, the lowest appropriate level for many water management functions turned out to be even smaller than the original division of the Tietê river basin into five regions. The sub-regions that were cre-ated in the Alto Tietê basin can be defined as “social catchment” areas, combining socio-economic and environmental interests and identities with the region’s political and natural hydrological divisions.

Although important achievements have been made, the decentralization process has yet to reveal measurable physical results such as the improvement of water quality or the rationalization of water use. It is undeniable that the Alto-Tietê committee and its subcommittees have already played an important leadership role around several issues, a fact that in and of itself provides ample justification for replacing the old system. Above all, an extraordinary mobilization around water issues, problems and management has occurred, even though solving many water-related problems may be beyond the capacity of the committees or even of the water resources management system as a whole. Charging for water remains one of the key issues in making the Alto Tietê Committee more relevant and giving it more say in water investment and management decisions. As long as such decisions remain at the individual agency level (both state and municipal), decision making will remain fragmented and it is unlikely that key policy instruments to curb water demand increases and pollution will be implemented.

Conclusion

22

sources have had perverse effects. This new approach also represents a dramatic departure from São Paulo’s tradi-tional sectoral approach to water quantity and quality, which separated the management of water from its environ-mental aspects, especially water pollution and land use. However, the implementation of such policies will likely face significant difficulties, since reaching the goals proposed depends on the capacity and will of municipal authorities to improve their urban regulations so as to guarantee the control and monitoring of land use in the sub-basins.

The Fraser River Basin, Canada

A locally-initiated, non-governmental river basin organization that has pursued a broad set of objectives related to a concept of basin “sustainability,” which includes social and economic as well as environmental aspects.

GEOGRAPHICAL OVERVIEW. T he Fraser River drains a 238,000 km portion of British Columbia, an area about the size of California. The Fraser River basin supports a population of more than 2.7 million residents, includ-ing approximately 50,000 First Nations’ peoples, and an economy that includes forestry and ranching, pulp and paper production, fishing, mining and other industry, and recreation and tourism. Seventy-eight percent of the basin’s population lives in the lower Fraser River valley and estuary region where the Vancouver metropolitan area is located. The river itself is 1,399 km long, originating in the Rocky Mountains and emptying into the Strait of Georgia and the Pacific Ocean. The Fraser basin has 13 principal watersheds or sub-basins, but on a broader scale, one can identify three main hydrologic regions—the coastal mountains, the interior plateau, and the eastern (Columbia and Rocky) mountains. The interior plateau is the driest of these regions, the coastal mountains the wettest. Millions of dollars of dyke works have been constructed and maintained, with financial assistance from the federal and provincial governments, to keep streams in the lower river area within their banks and to preserve agricultural lands and building foundations from seepage. Although there are serious water resource management issues in the Fraser basin, there are also favorable situa-tions not found in some of the other cases studied for this project. The main stem of the Fraser River has never been dammed, and will not be due to its designation as a Heritage River. The river basin’s large size, its large volume of flow on the main stem in normal years, and the fact that urban development has been concentrated mainly near the river’s mouth have reduced the negative impacts on the basin as a whole.

History of the Fraser Basin Council and Society

The Fraser River Basin has a pair of non-governmental organizations as its principal governance and management institutions—the Fraser Basin Society and the Fraser Basin Council. The society is a Canadian non-profit organization governed by a board of directors and able to receive funding from public and private sources. The Fraser Basin Council is a multi-orga-nizational, multi-interest planning body composed of 36 representatives drawn from diverse geographical and sectoral communities within the basin as well as from all four levels of Canadian government—federal, provincial, local/regional, and First Nations (aboriginal peoples). The council is the body that truly addresses, discusses, and decides upon basin planning and management priorities and activities and, as such, is the focus of this analysis.

The society and council grew out of the Fraser Basin Management Board, which existed from 1992 through 1997. Efforts to focus on developing a plan for improving the conditions of the entire Fraser River began with the creation of the Fraser Basin Start-Up Committee, the Fraser Basin Action Plan, and the Fraser Basin Management Program in 1990-91. At the same time, the idea emerged of a basin management board respon-sible for planning and executing projects with input and funding from governmental agencies and with the participation of First Nations and nongovernmental bodies. In May 1992, the principal federal agencies and the provincial government agreed to initiate a five-year Fraser Basin Management Program (FBMP) to be adminis-tered by the Fraser Basin Management Board (FBMB). The FBMB was to be a multi-organizational, multi-interest committee designed to encourage, and itself employ, consensus-based decision making about basin activities. In

Burns Lake

Bella Coola

Victoria

Powell River

Hope

Kelowna

RevelstokeCalgary

Jasper

Lillooet

Little Fort

Cranbrook

WinterHarbour

Prince George

Kamloops

Vancouver

Williams Lake

Fraser

Fraser

Chilcotin

West Road

Nechako

Quesnel

Thompson

Columbia

PACIFICOCEAN

Fraser

Fraser

Chilcotin

West Road

Nechako

Quesnel

Thompson

Columbia

128 126 122 120 118 116

128 126 122124 116 114

54

54

52

50

52

50

CANADAFraserRiverBasin

C A N A D A

UNITED STATES

Vancouver Is land

CANADA

FRASER RIVER BASINSELECTED CITIES

MAJOR ROADS

INTERNATIONAL BOUNDARIES

0 25 50

0 25 50 75 Miles

75 Kilometers

This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other informationshown on this map do not imply, on the part of The World BankGroup, any judgment on the legal status of any territory, or anyendorsement or acceptance of such boundaries.

Box 4

23

24

1997, at the end of this five-year period, the FBMB drew up a quasi-constitutional document for the Fraser River basin, called the ‘Charter for Sustainability.’ The FBMB’s successor, the Fraser Basin Council, published the charter in 1997. The most striking aspect of the charter is the tremendous breadth that it gives to the concept of basin sustainability and the effect that this has had on the council’s organization and indeed its mandate.

Salient Features of the Basin Organization

The council is composed of 36 members. Council seats are deliberately distributed in such a way that no sector of basin interests or level of government has a majority of members and non-water as well as water-related basin interests are represented. Principal stakeholder groups include: B.C. Hydro, First Nations, port authori-

ties and harbor commissions, mining, timber/forestry, pulp and paper manufacturing, industry, agriculture, commercial fishing, recreation and tourism, environmental organizations, and federal and provincial agencies. The Fraser Basin Council has continued the FBMB practice of employing consensus-based decision making and of maintaining regional coordinators in addition to its staff in Vancouver. The council has divided the Fraser basin into five regions based primarily on sub-basin groupings but also reflecting some jurisdictional boundaries (particularly regional districts). These are the basis not only for the assignment of regional coordinators but for the designation of regional representatives on the council.

Unlike its predecessor organization, the Fraser Basin Council has moved beyond total reliance on funding support from government agencies and added “project funding,” i.e., funding that comes from public and private organizations that contract with the council to perform a study, organize an event or program, administer a project, etc. The council still relies on core funding support from federal, provincial, and local government annual contributions, but that funding has declined, in percentage terms, from 95% of the council’s revenue in 1998 to 51% of its revenue in 2003, while project funding has grown from 4% of council revenue in 1998 to 36% in 2003. The council’s revenue has roughly doubled since 1998 due primarily to increased project funding.

Outcomes

CONTEXTUAL FACTORS AND INITIAL CONDITIONS. Canada generally, and the Fraser Basin in particular, are prosperous enough that policy makers and stakeholders have resources to devote to research, institution building, meetings, environmental improvement projects, and monitoring and assessment. The initial distribution of resources among basin stakeholders is also conducive to successful river basin management. As a matter of economic policy, the vast majority of land and water resources in the Fraser Basin are held by the province of British Columbia, and used by private individuals under lease arrangements with the government. This situation has allowed institutional arrangements to develop in the basin under conditions where no one interest or sector of basin users had legal immunity for their claims or titles to resource use.

THE DECENTRALIZATION PROCESS. The construction of basin-scale institutional arrangements in the Fraser Basin is a matter of integrating already decentralized organizations and jurisdictions rather than decentralizing previ-ously centralized ones. The extent of central government recognition of local-level basin governance has been extraordinarily positive. Not only did the Canadian national and British Columbia provincial governments join in the predecessor organizations in the basin (the estuary steering committee, the basin management board) and fund the Fraser River Action Plan from 1992 through 1998, but they have been original and consistent members of the Fraser Basin Council and have supported it financially through changes of government and administration at both the provincial and federal levels

24

25

CENTRAL-LOCAL RELATIONSHIPS AND CAPACITIES. The financial resources and the financial autonomy of the council are relatively strong. The council members have, through the Fraser Basin Society and the council’s own by-laws, the demonstrated ability to create and modify the institutional arrangements with which they work. The water rights system, on the other hand, is something of a mixed bag. The arrangements governing rights to water and land use allow for considerable management flexibility, however, the control of groundwater resources is particularly weak and represents a current and future vulnerability in terms of water resource management within the overall basin sustainability effort.

BASIN-LEVEL INSTITUTIONAL ARRANGEMENTS. The strongest features in this regard are: the availability of a basin-level governance body (the council), the recognition of sub-basin communities of interest through the compo-sition of the council with regional representatives and through its employment of regional coordinators, and the institutionalization of regular monitoring of basin conditions by means that are trusted by resource users. The council was designed quite deliberately to share information and promote communication among basin stakeholders, to provide means for basin stakeholders to enter into agreements to take actions for improve-ment of basin conditions, and to resolve conflicts. While the council has emerged as the paramount deliberative body in the basin, in its capacity as a nongovernmental organization funded through a nonprofit society, the council has limits in terms of its ability to implement projects. It often must hand off projects to other (usually governmental) entities for implementation, and at times even the council members are not entirely clear what actions are within the council’s scope.

The NGO model reduces some of the bureaucratic “turf battles” that one would expect to be associated with placing basin management responsibility in an existing agency, or creating an agency that would have authority and responsibili-ties that were transferred from, or overlapped with, existing agencies. The NGO approach also fits well with a federal system such as Canada’s, since it provides a means of crossing jurisdictional boundaries among levels of government in a context where a constitution divides authority and one level of government is not entirely superior or subordinate to another. It is also suited to a common-law cultural context where private organizations are free to do anything that is not expressly forbidden by law, and to take actions (including the raising and distribution of funds) up to the limits of public authority.

Furthermore, the NGO approach in the Fraser basin has allowed for the integration of First Nations communities and private stakeholders in ways that more traditional inter-governmental programs have often found difficult if not impos-sible. It has provided a good forum for information generation and sharing, since there is less concern over who “owns” the information. An NGO has the boundary flexibility to cover the whole basin (which no local government can do) but not more than the basin (as would be the case for a provincial or federal agency). An NGO of the Fraser Basin Council type also provides good political cover for agencies, who can justify actions that might otherwise be unpopular with a particular constituency.

The NGO approach epitomized by the Fraser Basin Council also has its weaknesses. The council is generally unable to implement the plans and programs it agrees upon, and must hand them off to others—usually governmental agen-cies—for implementation. Other vulnerabilities include the fact that the council’s consensus decision making approach, though helpful in a number of respects, can at times be inefficient. Also, as an NGO reliant upon contributions and funded projects, the council is vulnerable to “mission creep,” the temptation to move beyond its primary concerns and interests to follow the money.

On balance, the approach represented by the Fraser Basin Council has worked well as a means of bridging fragmented public authorities and integrating indigenous and other private stakeholders. It has succeeded in preserving a reputation for objectivity and in building a more diverse financial base. One key to this success has been the ability of the council to promote the concept of interdependency among various stakeholders.

Conclusion

25

The Tárcoles River Basin, Costa Rica

Examining the early life cycle stage of a river basin organization and some of the factors linked to its origin,early growth, and recent stagnation.

GEOGRAPHICAL OVERVIEW. The Grande de Tárcoles river basin—the drainage area of the Río Grande de Tárcoles—is located in the west-central portion of Costa Rica and extends from the mountain ranges in the middle of the country to the Pacific coast. The basin is fairly large—at 2,155 km2 total area, it represents

4.2% of the total land area, and the upper and middle areas (the Virilla and Grande watersheds) are two of the largest watersheds in the country. With only 1/25 of the land area, the Tárcoles basin is home to half the nation’s population, contains 80% of the nation’s indus-try, 80% of all the vehicles, and more than half of all the registered wells. Precipitation ranges from 948 mm per year to 5,409 mm per year, with an annual average of 2,364 mm. Flooding is a recurring problem.

The cities in the upper basin have grown into a large metropolis at the center of Costa Rica known as the Greater Metropolitan Area or GAM (Grande Área Metropolitana). This metropolitan area includes San José, the nation’s capital. The GAM is also the transportation center of the country, with the national highways that reach the other regions of Costa Rica converging and intersecting in the Tárcoles basin. Despite the growth of urban and industrial centers within the basin, 37% of the land use remains in crops and pasture. Coffee farming, other crops with and without irrigation, dairy farming, and livestock ranching occur throughout the basin.

History of the Basin Organization

Costa Rica has a centralized form of government. There has as yet been no overall, nationwide effort to decen-tralize water resource management to the river basin level. The creation of the Coordinating Commission for the Río Grande de Tárcoles was the first effort to establish a structure for river basin management. In the early 1990s, with the Municipality of San José taking the lead, negotiations began to form the Coordinating Commission for the Río Grande de Tárcoles, the CRGT, which was to have a participatory structure that was both inter-organizational and inter-disciplinary. It was officially recognized, though it lacked formal governmen-tal powers, in April 1993 by executive decree. The commission had 19 members, of which five were nongov-ernmental organizations and six were municipalities. Throughout its first two years of the CRGT’s existence, the Municipality of San José continued to provide its principal leadership and support.

During 1994-1998, the CRGT implemented four major action programs. A volunteer plan program was devised for businesses to incorporate waste treatment, and an ecological banner program was created to recognize these and other efforts to protect and conserve water resources. Reforestation programs were initiated to promote watershed protection. Lastly, the CRGT developed an ambitious program for the inte-

Tarcoles SAN JOSÉ

Cañas

San Ramón

PuertoViejo

Guácimo

Siquirres

Turrialba

San Isidro

Parrita

San Ignacio

SanMarcos

Dominical

PuertoQuepos

Santiago

Jacó

Tárcoles

Orotina

Quesada

San Rafael

Carmona

PlayaNaranjo

Puntarenas

AlajuelaHeredia

CartagoTárcoles

Chirr

ipó

del A

tlánti

co

General

Reve

ntazón

Chirr

ipó

San

Lago deArenal

Gol fo deNicoya

ATLANTICOCEAN

PACIFICOCEAN

C O S TA R I C A

10 10

85

85 84

84COSTA RICA

TÁRCOLES RIVER BASINSELECTED CITIES

NATIONAL CAPITAL

MAJOR ROADS

0 5 10 15

0 5 10 15 20 Miles

20 Kilometers

This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other informationshown on this map do not imply, on the part of The World BankGroup, any judgment on the legal status of any territory, or anyendorsement or acceptance of such boundaries.

COSTA RICA

TárcolesRiver Basin

Box 5

26

grated management of natural resources at the basin scale. A change of government in 1998 resulted in less generous support of the CRGT and the “re-centralizing” of the Commission. The first president had used an extensive network of top decision-makers of other key institutions such as MINAE (Ministry of Environment and Energy). CRGT had come into existence and flourished under his leadership, but it diminished quickly after he left. Currently, a MINAE official is president of CRGT. While he has kept the CRGT alive, it is clearly functioning at a quite minimal level. It is not clear how much participation in the CRGT is still committed by the member organizations and agencies. For example, there appears to have been little or no interest on MINAE’s part in following up on the Integrated Natural Resources Management Program for the basin, despite the Inter-American Development Bank’s investment of USD$1 million in a 1997-98 study.

Salient Features of Basin Organization

The current situation in the administration of water resources in Costa Rica remains characterized by frag-mentation and dispersion of responsibilities in a large number of institutions, several of which operate on a national scale. At least 15 agencies are involved in local and national water administration. As a result, there are serious problems in the distribution of responsibilities, with overlaps in some areas and vacuums in oth-ers. There is no coordination between these institutions, and their systems of administration differ. They were created to fulfill specific functions (irrigation, drinking water supply, hydroelectric generation, sanitation, etc.) and lack an outlook that envisions an integral approach to water resource management. Contributing to this fragmentation are the absence of a supervisory agency and a national water resource policy. Even though the law designates the supervision of this resource to MINAE, that ministry has limited its responsibility to grant-ing concessions, giving permits to exploit superficial and underground water and to establishing and collecting fees for such use. There is no national water policy and no national water plan; there is not even a national water budget more recent than 1990. The existing national water law dates from 1942, and modified the first regulation of 1884; a proposed new water law has been under consideration in the Costa Rican legislature for about two years.

Despite the fact that MINAE is responsible for supervising water resources, in 2002 the National Water Council was formed by Executive Decree, and charged with the “harmonization of water legislation and the coordination of research, uses, development, utilization and conservation of water in the different depart-ments and institutions of the state.” The Ministry of Public Health, an agency with responsibilities for water quality, pollution and health, was appointed to coordinate this Council, which has created distortions and overlapping roles for both ministries. The formation of this Council is a product of the leadership vacuum created by MINAE as the supervisory agency, is a temporary measure, and is no substitute for an adequate institutional framework for integrated water resource management.

Outcomes

CONTEXTUAL FACTORS AND INITIAL CONDITIONS. Costa Rica is a developing country that has experienced a 500% increase in population over the past half-century, while seeing gains in household income, education levels and other development indicators. Literacy rates and life expectancy are at developed-nation levels, and Costa Rica has a sizeable middle class. Economic growth has been accompanied by a stable political system for the past half-century. The Tárcoles basin is by far the most economically developed in the country and there do not appear to be substantial cultural or religious differences across groups of basin stakeholders that would inhibit cooperation. Financial resources are, however, a limiting factor. The Tárcoles basin is politically important and reflects inconclusive national policy. There seems to be overall reluctance towards decentraliza-tion to lower levels of decision-making.

27

The system of water concessions in the country has significant gaps that contribute to uncertainty about water availability, since it is hard to tell what uses are occurring already in the basin and to what extent total water demand exceeds or is exceeded by available supply. The largest hydroelectric power producers (the national institute ICE and its subsidiary CNFL), which use the greatest quantities of surface water, are not included in the concession system. Also, the concession system provides little control over groundwater use. Tariffs on agricultural water use are based on land area rather than on the volume of water used, providing little incen-tive for farmers to replace or upgrade aging and inefficient gravity-fed systems. The entire system of water charges fails to provide enough revenue to maintain infrastructure within the basin, let alone support needed improvements such as wastewater treatment plants. This situation does not correspond to Costa Rica’s image of an ecologically–aware and active country.

THE DECENTRALIZATION PROCESS. The CRGT was essentially a municipal initiative and took on a bold lead-ership role. The central government partially devolved authority, and was supportive of the CRGT’s efforts, but there was never a full recognition of the CRGT’s authority to manage the basin. Since 1998 the central government has neither pushed the devolution forward nor terminated the commission. It has kept the com-mission alive while rethinking and shifting focus concerning environmental and natural resource policy aspects. Thus, neither a complete handover nor a complete abandonment has resulted. River basin management is now widely perceived to be an issue under MINAE’s authority and direction. This perception, in addition to the current government’s lack of commitment to the issue, has served to marginalize the past efforts of the CRGT. Discussion of how to strengthen local government and decentralize Costa Rica’s system has been under way in earnest for more than 20 years, with various legislative proposals introduced in the national legislature. Recently, these efforts have begun to bear fruit, with the establishment of a new Municipal Code in 1998 that among other things provides for the direct election of mayors (which occurred for the first time in December 2002), and the approval in 2001 of a constitutional amendment assigning 10% of the revenue of the regular budget to the municipalities.

CENTRAL-LOCAL RELATIONSHIPS AND CAPACITIES. Financial resources for the basin management effort have always been limited, and CRGT has never had its own revenue stream. This has severely limited the commis-sion’s ability to evolve into something more than a meeting place. Cantons and municipalities do perform a number of functions, so there appears to be local-level experience with self-governance and service provi-sion, rather than an excessive centralization of public services. According to Article 169 of the Constitution, municipalities are autonomous and have complete authority to administer their territory. However, they have neither the fiscal autonomy nor the funding from the central government to carry out their responsibilities very effectively. Up to and including 2002, the revenue of municipalities represented less than 1% of the GDP and less than 2% of the total public expenditure, despite the importance they are given by the legal ordinances. This lack of resources and capacity at the local level, with no authority in an intermediate level of governance, reflects the generally high level of centralization in Costa Rica.

The ability of any river basin commission in Costa Rica to develop and implement effective water supply man-agement policies is likely to be hampered by the weak framework of water rights allocation. There is a consen-sus now that it is necessary to have a legal framework to regulate water, and in 1998 a process was initiated, promoted by diverse sectors, to approve a new General Water Law. Different actors (MINAE, Ombudsman’s Office and others) presented various proposals to the Costa Rican legislature. All of these included, to varying degrees and with various approaches, the decentralization of water administration and the formation of local structures. At the time of this study, there was a single text in discussion by the national legislature.

28

BASIN-LEVEL INSTITUTIONAL ARRANGEMENTS. Once it lost its central government support and its dynamic initial leadership, the CRGT’s status and composition left it vulnerable to becoming more of a discussion forum than a governing body. The prevailing and traditional view that water has to be managed by its uses (drinking, irrigation, hydropower, etc) rather than in an integrated fashion has been reflected and reinforced by Costa Rican laws. There is considerable fragmentation and territorialism among agencies and institutes at the central government level. Likewise, at the local level, there is little inter-jurisdictional cooperation and coordination among municipali-ties, which have been gaining interest in entering water planning and water service business activities.

The basin commission’s greatest strength was to serve as a forum for information sharing and communication, but this appears to have waned substantially since its time of peak activity. Basin stakeholders seem to have sector-based fora for information sharing (e.g., union of municipalities, chamber of industries, chamber of agriculture, etc.), which may serve operational purposes, but are not conducive to development of an agenda of basin activities.

The start-up of the commission in a bottom-up format initiated by some of the large stakeholders was initially very suc-cessful and quickly showed a number of results, indicating the possibilities for basin management. Nevertheless, it was heavily reliant on high-level support. The central government’s commitment to river basin management has been uneven and inconsistent. The basin management approach had a strong champion. Once the champion left, the institutional set-up became relatively ineffectual. The severity of problems in the basin, and the difficulty of marshaling the financial resources to address them, stretch the management challenges beyond the capabilities of local action without sustained commitment of central government or external support. Flaws in the basin organization structure and authority kept it from exercising autonomous authority to govern basin management, and diminished the commitment of some important local actors to it. The past and current water rights laws have been notably unhelpful to integrated water resource management in the Tárcoles or other basins and it is not clear if the Tárcoles experience has helped reshape these laws. However, other basins have learned from the Tárcoles basin and adopted different approaches. The biggest water interests are national-scale and have their own agenda. They either must have incentives to participate (which they currently do not have) or Government must act more forcefully if it wants to promote better river basin management. Pollution may not be perceived as acutely as a problem as water scarcity in other contexts, so that the political pressure to deal with the issue is relatively low.

The Tarcoles basin commission was for a period in the 1990s able to initiate and lead important basin improvement activities. Agribusiness contamination of water, especially from coffee processing operations, was reduced through the Voluntary Plan Program. Although forest and farm land are still being lost to urbanization—aggravating erosion, flooding, runoff and contami-nation problems—reforestation efforts championed by the basin commission certainly helped slow the degradation.

Changes of leadership at the CRGT and its changed relationship with MINAE—a change that resulted in more central government control but less central government support—are associated with a decline of CRGT activity, visibility, and stakeholder participation. A number of basin management issues remain unaddressed and unresolved in the aftermath of that change.

The Tarcoles river basin still lacks sewage treatment, and river water quality conditions therefore continue to worsen as the basin population grows. Septic tanks used by many households and businesses in the basin are not being replaced with a sewage collection and treatment system, and groundwater quality is increasingly threatened as a result of septic systems as well as agricultural and industrial water and land uses. Industrial waste treatment occurs in some locations in the basin, but coverage remains incomplete. The water rights system in Costa Rica inhibits effective demand management. The current concession system does not cover groundwater use, or surface water use by public hydroelectric suppliers. The tariff system for agricultural water use continues to impose fees based on cultivated area rather than metered water use, providing no economic incentive in favor of efficient water use. Furthermore, groundwater use appears to be subject to no control whatsoever though there is evidence of overdraft in the San Jose area.

Conclusion

29

30

Brantas River Basin, East Java, Indonesia

Integrated water resources management in a developing country undergoing rapid economic, political, and institutional change.

GEOGRAPHICAL OVERVIEW. The Brantas river basin is located within the province of East Java in Indonesia. It has an area of approximately 11,800 km2 and makes up 25% of East Java’s land area. It has 23% of the province’s forest land and 55.8% of its arable land. Most arable land in the basin is used for productive farming (38%), while the remainder is used for forest, settlement, and non-agricultural activities. The basin’s population, which amounts to nearly fifteen million, has increased by 53.4% over the past 30 years and represents 42.4% of East Java’s population (2000) with a density of 1,249. The basin is bounded by Mt. Bromo and Mt. Semeru on the east, a series of low ridges to the south, Mt. Wilis and its ridges on the west, and the Kedung low ridges and the Madura Strait on the north. The Brantas River is 320 km long, with its headwaters located in the Arjuno volcanic massif, a substantial topographic feature in the southeast portion of the basin. The river flows around the massif clockwise through the Malang Plateau, then west through the major dam and reservoir complex, made up of Sengguruh, Sutami/Lahor, Wlingi, and Lodoyo. At the southwestern portion of the basin, where the Ngrowo River joins it, the Brantas turns north towards the agricultural plains region, and then east through the delta regulated by the New Lengkong Barrage. This structure divides the Brantas into the Porong and Surabaya Rivers and the Porong and Mangetan Canals. The canals provide irrigation water for the paddy-growing region before discharging into the Madura Strait. The Porong River serves as a flood diversion canal during the rainy season, and the Surabaya River is a primary source of water for Surabaya City. There are two active volcanoes within the basin, one of which, Mount Kelud, has had a great effect upon river morphology. Variation of rainfall from year to year is large. The average surface water potential in the Brantas basin is estimated to be approximately 12 billion m3, with the average flow at an estimated 3 billion m3, or 25% of available surface water. Intense

industrialization, agricultural development, and population growth within the basin over the past three decades, combined with its climate and physical features, have contributed to several critical water resources problems, including pol-lution, floods, and seasonal water scarcity.

History of the Organization

Given the devastating prominence of flooding in the basin, flood prevention was given first prior-ity in the initial stage of the Brantas River Basin’s development. The basin’s first master plan was prepared in 1961 using Japanese post-war repa-ration funds and consisted of large technical developments–dam structures, flood diversions, retarding basins, and riverbed channels. The basic concept for the Brantas River Basin Flood Control Plan (Master Plan I) was “one river, one plan, one coordinated management.” At that time, the Ministry of Public Works (MPW) established and oversaw the Brantas River Basin Development Project to carry out these infra-structure-related efforts.

Strai t of MaduraStrai t of Madura

Java SeaJava Sea

Brant

as

Bran

tas

Pasuruan

Lawang

WlingiBlitar

Malang

Surabaya

SidoarjoMojokerto

GresikLamongan

BojonegoroCepu

Blora

Rembang

NgawiSragen

Klaten

Wonosari

SurakartaMadiun

Wonogiri

Ponorogo

Pati

Demak

Kepara

Kudus

Purwodadi

Bangkalan

SampangPamekasan

Jombang

Kediri

Nganjuk

Ngunut

Trenggalek Tulungagung

Probolinggo

JemberLumajang

Tempen

J A W A T I M U R

Madura

Barung

Strai t of Madura

Java Sea

INDIAN OCEAN

Brant

as

Bran

tas

WlingiReservoir Kesamben

ReservoirKarangkates

Reservoir

11200'

11300' 11330'

730'

830'

800'

700'

730'

800'

700'

630'

11130'

11200' 11230'11130'

INDONESIA

BrantasRiver Basin

INDONESIA

BRANTAS RIVER BASINSELECTED CITIES

MAJOR ROADS

0 10

0 5 10 15 20 Miles

20 Kilometers

This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other informationshown on this map do not imply, on the part of The World BankGroup, any judgment on the legal status of any territory, or anyendorsement or acceptance of such boundaries.

Box 6

30

3131

There was a strong emphasis on infrastructure development during the first three decades of the Brantas Project’s existence. Only in 1990, when it was fully acknowledged that the sustainability of these investments had not been achieved by the Brantas Project itself, due to a lack of incentives for maintenance, was a different approach sought. This search led to the establishment of the Perum Jasa Tirta I (PJT 1, also referred to as the Brantas River Basin Management Corporation), a national state-owned company for river basin management, independent from the Brantas Project. The founding of PJT 1 in 1990 initiated an emphasis upon river basin management to operate and maintain existing infrastructure, plan and implement the allocation of water, and address problems that affect basin-level water resources.

At the national level, Indonesia’s second long-term twenty-five-year development plan (PJPII: 1994-2019) emphasizes integrated development and management of water resources, with a greater focus upon the opera-tion and maintenance of infrastructure. This new plan illustrates the shift in mindset of Indonesia’s administra-tors from a single-purpose focus to a multi-sector river basin approach to promote integrated water resource management. It was decided that irrigation management, which had been the primary focus in the previous long-term plan, was to be transferred gradually to the district level as part of Government policy to increase regional autonomy, while the allocation of water among irrigation projects would make up a core function of basin management. Indonesia began to set up national policies towards organizing institutions and integrating management functions on the basis of hydrological boundaries.

Integrated basin-level management was implemented relatively early on in Indonesia. The Brantas Project had been responsible for planning for Master Plans I, II and III. In 1991, the preparation of river basin master plans became a PJT I task. PJT I was involved in generating Master Plan IV with the consultation of local government and users. Once the plan is accepted by the central government, PJT I can set up its long-term action plan to implement it. Master Plan IV emphasized conservation and basin water resource management–institutional approaches for proper water governance. PJT I also completed a long-term (1999-2020) plan with assistance from the Japan International Cooperation Agency (JICA), and are in the process of finalizing it for government approval. There are a number of agencies that are involved fully or partially, directly or indirectly in water resource related functions in the basin. The 1998 study prepared by JICA lists 35 organizations that have a role in water resources development and management.

Salient Features of the Basin Organization

The outsourcing of water resources and infrastructure management functions to a freestanding company is rather unique. PJT I manages catchment conditions, water supply allocation, water quality, flood control, river environmental management, and water resource infrastructure for 40 rivers, constituting the brunt of water resources in the basin. The remaining secondary, tertiary, quaternary rivers are maintained by the province. Brantas PJT I distributes bulk water to the irrigation systems and also operates, maintains, and manages the flood control infrastructure and the flood warning system. As a state-owned company, PJT I is supposed to return a profit to the central government and accordingly seeks to fund itself through fees from industry and the hydropower sector, as well as municipal water suppliers. As in many other countries, farmers are exempt from payments although they constitute the largest water user group. PJT I is subsidized by the central govern-ment because national funds pay the salaries of its staff. Due to the basin’s relatively high level of economic development the PJT I can achieve a reasonably high level of O&M cost recovery from water users.

For the technical aspects of river basin management, PJT I solicits guidance of the Ministry of Public Works (MPW), which supervises PJT I’s management and functions. The provincial Water Resources Services Office serves as a regulator for PJT I. District-level government provides support for operational matters, providing

32

enabling conditions at the local level for PJT I. The Ministry of Finance sets tariffs for hydropower users. The Governor, who serves as the President’s representative in the region, sets tariffs for municipal and industrial users, and the Minister ultimately proposes the rate by regulation, further highlighting the Central Government’s continued influence on basin management.

Management of Brantas PJT I is through a Supervisory Board with a President Director assisted by three direc-tors. Its structure indicates ministerial authority over its affairs, which is typical of Indonesian state corporations. The President Director and Director positions are appointed by the MPW. The Supervisory Board, which is answerable to the Minister of Public Works, carries out the general supervision of the basin agency, including implementation of its work plan and annual budget. Every three years, two to five members are chosen to sit on the board from MPW, Ministry of Finance (MOF), and agencies whose activities are related to the corpora-tion. These affiliated agency positions are appointed by the President on a proposal from MPW which has been cleared by the Ministry of Finance (MOF). The Governor of East Java also sits on the Board. The power of the Supervisory Board as stipulated in the regulation ensures a degree of management autonomy to the basin agency. Nonetheless, it is apparent that given the Supervisory Board’s structure, MPW wields influence over its operation. Both from the financial and overall decision making perspective, there is a clear and continued relationship and dependence by the company on the national, and to a certain extent, provincial government. To wit, national, provincial, and basin-level coordinating committees are largely handled by governmental agen-cies serving on behalf of stakeholders. Moreover, there is no uniform national-level water resources policy. Overlapping functions and conflicting objectives among agencies remain an unresolved issue.

Outcomes

CONTEXTUAL FACTORS AND INITIAL CONDITIONS. The steady economic growth during the Suharto regime (1969-1998) came to an end with the 1997 financial crises and the collapse of the regime in 1998. Indonesia’s weak institutions, a poor legal and regulatory framework, ineffective bureaucracy, and endemic corruption made it difficult for Indonesia to withstand political uncertainty and an economic downturn. Subsequently, the government launched a four-pronged strategy of policy and institutional reform based upon macro-economic management, financial and corporate restructuring, protection for the poor and of human assets, and reform of economic instruments and institutions. A number of donor agencies jointly participated in an IMF-led restruc-turing program in FY1998-1999. Since 1999, the country has gradually recovered macroeconomic and political stability, but it is still vulnerable and limited in its capacity to obtain development funding.

THE DECENTRALIZATION PROCESS. Indonesia’s 1945 Constitution bestows strong powers upon the executive branch, giving the President the authority to determine the nature of regional autonomy. The pursuit of a strong unitary government has historically been an executive response to the sprawling archipelago’s extreme-ly high level of cultural, ethnic, language, and economic diversity. During his reign (1969-1998), General Suharto consolidated powers at the center. The central government set policies and regulations, provinces undertook coordination and supervision duties as agents of the center, and districts were responsible for implementa-tion. In time, over 90% of government revenues accrued to the center, and large conditional central transfers limited local autonomy.

The decentralization process was clearly top-down. Law No. 22/1999 on Local Government devolved Central Government powers and responsibilities to district-level governments in many administrative sectors. Key water users and stakeholders adhere to PJT I’s “rules”, i.e. payments of water use rights fees, which permit the company to function. Stakeholders–notably industries and municipalities–have an interest in a function-ing water allocation and flood management system and are generally willing to contribute to it (albeit not

32

33

sufficiently to cover full O&M costs). An important contributing factor to this relative success of the PJT I seems to have been its outreach capacity towards stakeholders. Passage of the new Water Law, in March 2004, signals central government commit-ment to continued reform of the water resources sector in accordance with the agreed action plan developed under the World Bank assisted Water Sector Adjustment Loan (WATSAL). Dialogue and debate are taking place openly among a broader set of actors in Indonesian society.

CENTRAL-LOCAL RELATIONSHIPS AND CAPACITIES. Much power still resides with Central Government Ministries for planning and policy-making. The author-ity to oversee the management and functioning of PJT I lies with the center through the MPW with the Ministry of Finance exercising a fiscal oversight role. The Supervisory Board of PJT I, with five members comprised of Central and Provincial Government, undertakes general supervision of PJT I’s program, work plan, and budget, and answers to the Ministry of Finance and MPW. The Supervisory Board does not have a stakeholder advisory group to work with or any other form of stakeholder institution. It structurally reflects the authority of the MPW and Ministry of Finance over basin-level interests. The Governor of East Java also sits on the Board, which helps to facilitate coordination and implemen-tation locally.

PJT I is responsible for operating, maintaining, and managing water resources on behalf of the central government, which is the owner of the infrastructure in the Brantas basin. PJT I is assigned to plan and operate day-to-day activities, maintain records, undertake minor maintenance, and assume responsibility for operational management. The company can undertake some rehabilitation activities within its financial limits that are necessary for operational purposes. Any subsequent changes or expansions to these activities through Ministry of Public Works decrees have been introduced to suit ministry needs or water resource conditions within the basin (such as for PJT I’s water quality monitoring capacity), without offering any changes in authority roles, responsibility, or financing mechanisms. This is consis-tent with the Decentralization Law 22/1999, which provides specific functions for local government without general authority or new means of raising revenue. This indicates a deconcentration rather than the decentral-ization of Central Government activities to a basin level institution. Despite these limitations to its autonomy per se, PJT I has developed strong relationships with various stakeholder groups, universities, and NGOs and is well trusted to manage water supply and flood control issues.

BASIN-LEVEL INSTITUTIONAL ARRANGEMENTS. PJT I is a state-owned company with clearly delineated manage-ment responsibilities and a profit motive. This construct has permitted the company to (i) focus on the river

33

34

basin as the management unit and (ii) focus on management rather than development and construction. Importantly, this structure has also endowed the company with credibility that the funds it receives from water users will be reapplied in the basin, an important condition to ensure stakeholders’ willingness to financially contribute to basin management expenses. The weak point in the overall institutional arrangements is related to the following. PJT I does not have authority over two more recent water resources management chal-lenges in the basin, notably the high degree of point and non-point pollution as well as deforestation in the upper reaches of the basin, leading to major erosion problems. Further, PJT I shares responsibility with, and is ultimately subservient to other political entities at both provincial and national levels with respect to the development of water resources management policy and allocation priority. The coordination activities it must undertake in support of its water resource management objectives are consequently quite complicated.

A shift has taken place in Indonesia since the mid-1990s from emphasizing infrastructure development to strengthening institutional aspects (hydrology, flood fighting, flood warning, flood management, etc.). The institutional arrangement for water resources management in the Brantas basin through a state-owned corporation is an interesting model. PJT I has achieved results in implementing a reasonably good system of water allocation and management and a reliable flood forecasting system, as well as maintaining major infrastructure in fairly good condition. Managing water quality, catchment conditions, and the river environment, however, are the responsibility of many entities, and there is need for greater coordination and authority to address these issues.

Decentralization has been gradual and still largely reflects top-down arrangements with the central government as policy maker, PJT I as the executing agency and local government in an intermediate position. Decentralization in Indonesia has focused on devolving authority directly to district-level actors and this has created some level of confu-sion concerning relationships among, and degrees of authority between, the many central government, provincial, and local actors. PJT I has autonomy in day-to-day operational affairs, but is dependent upon financing that is not completely within their control as it is set through a political process. Since it does not collect sufficient funds to cover O&M costs, it relies upon the central government to cover flood control costs as a social welfare activity. Additionally, though PJT I rehabilitates infrastructure where it can afford to, it relies heavily upon the central government to carry out larger rehabilitation efforts through the Brantas Project to rehabilitate gradually deteriorating infrastructure. Despite PJT I’s financial shortcomings to cover costs, the structure of user charges as reflecting multiple use interests provides a strong base for eventual full cost recovery.

In spite of the limitations in financial and policy making autonomy and overall authority, PJT I is considered to be suc-cessful by stakeholders because it is committed to upholding a professional and neutral profile, which gives it a great deal of legitimacy among water users who seek unbiased information, expertise, and mediation assistance. PJT I focuses upon being a reliable and accessible service provider for tasks it has most authority over to invoke legitimacy: water allocation and supply and flood control. The insulation from policy making activities works in its favor because it places PJT I squarely in the business of operating and maintaining. Water users evaluate it as such, understanding that PJT I is not to blame for issues that they have limited authority and resources to address.

Conclusion

Warta River Basin, Poland

Top-down devolution of authority, in a relatively water-poor country pursuing accession to the European Union, pro-vides a recent example of a government-led and supported attempt at integrated water resource management.

GEOGRAPHICAL OVERVIEW. The Warta River, in western Poland, is the largest tributary of the Oder, which forms part of the boundary between Poland and Germany. The Warta is Poland’s third largest river after the Oder and the Vistula. The river flows north from its headwaters in the mountains of southern Poland, then west to the Oder, and is 808.2 km in length with approximately 735 km navigable. The Warta River basin’s 55,193 km2 area covers approxi-mately one-sixth of Poland. The basin is divisible into three major sub-basins: the Upper Warta sub-basin (including the Prosna River watershed) which cov-ers about 20,825 km2; the Middle and Lower Warta sub-basin (to the river mouth at the confluence with the Oder) which covers about 17,033 km2; and the Upper and Lower sub-basin which covers 17,333 km2.

In addition to streams and rivers, the basin contains numerous lakes and reservoirs. Small reservoirs are used for agricultural purposes, while larger ones in the upper sub-basin and on the major tributaries provide flood protection and water storage for recreation and for industry. Land use in the basin is 70% agriculture and forestry, 30% urban and indus-trial. The basin’s population is about 6,770,000, over 34% of which live in cities. By far the largest city in the region is Poznan, the capital city of Wielkopolska Voivodeship, with a population of approximately 600,000. In terms of water availability per capita, Poland is one of the most water-poor European countries, and precipitation and runoff in the Warta River basin are below even the national average. On the other extreme, legendary floods have been experienced in the Warta basin from the beginning of the 20th century to the very recent past, most notably in 1997.

History of Basin Organization

During the postwar decades of Soviet dominance, Polish governance and policy emphasized central govern-ment planning and control. From 1960 to 1972, a central Institute of Water Management was responsible for water planning and use, and analysis of water resource information. A restructuring in 1972-73 yielded a central Ministry of Administration, Country Planning, and Environmental Protection, and an Institute of Meteorology and Water Management. Throughout this postwar period, water management generally focused on technical plan-ning and the construction of physical facilities to support industrial and agricultural development. Water resource expenditures were almost exclusively for water works and relied heavily on central government plans and fund-ing. Centrally-appointed District Directorates of Water Management (DDWMs) were established beginning in 1964 to construct and operate water works. Domestic water supply, sanitation, and wastewater disposal were

I I I I I I I I I I I

Radom

Konin

Plock

Torun

Kalisz

Leszno

Legnica

Sieradz

Koszalin

Wloclawek

JeleniaGóra

Zielona GóraSkierniewice

PiotrkówTrybunalski

Pila

Elblag

CzestochowaWalbrzych

Dresden

Leipzig

Dessau

Cottbus

Görlitz

Frankfurt

HradecKralove

Opole

Kraków

Kielce

Olsztyn

Katowice

Szczecin

Bydgoszcz

GorzówWielkopolski

Lódz

Poznan

Wroclaw

WARSAW

BERLIN

PRAGUE

Warta

War

ta

Warta

Prosna

Notec ´

Ner Obra

Dra

w

Widawka

Liswarta

Bal t ic Sea

Warta

War

ta

Warta

Odra

Odra

Wista

Wista

W

ista

Prosna

Notec ´

Ner Obra

Dra

w

Widawka

Liswarta

14 15 17 18 19

17161514 18 19

21

51

52

53

51

52

53

POLANDWarta River

Basin

P O L A N DG E R M A N Y

CZECHREPUBLIC

POLAND

WARTA RIVER BASINSELECTED CITIES

NATIONAL CAPITALS

MAJOR ROADS

INTERNATIONAL BOUNDARIES

0

0 10 20 30 Miles

30 Kilometers

This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other informationshown on this map do not imply, on the part of The World BankGroup, any judgment on the legal status of any territory, or anyendorsement or acceptance of such boundaries.

35

Box 7

decentralized, local functions with no meaningful planning and management at a regional or river basin scale. The extent and quality of these services was especially problematic in rural regions. Until 1991, the main governmental entities responsible for water management at the sub-national level were not fitted to river basin boundaries. The DDWMs worked along the main stems of some rivers, but not on river basins as a whole. The other regional body was the voivodeship (provinces). The voivodes governing these provinces were responsible for rivers and streams that were not being managed by the DDWMs, for irrigation and drainage, and issuing water permits. By the late 1980s, the entire governmental system faced a period of crisis and transformation.

The 1989-91 period of the democratic transformation in Poland was an “open policy window.” In February 1991, the Polish government announced the creation of a system of Regional Boards of Water Management (RBWMs), conforming essentially to river basin boundaries. Their principal purposes were to arrest the further pollution of

water supplies, protect drinking water sources, and aid water users and water user organizations in developing and implementing rational water manage-ment. The RBWMs were related directly to the national government’s Ministry of Environment. Each RBWM director was an individual charged by the minis-try with management of the basin. There was little provision for public partici-pation or water user involvement in RBWM decision making. In late 1999, the Minister of Environment decreed a merger of the DDWMs and RBWMs and their separate operations into seven Regional Water Management Authorities (RWMAs) covering the entire country and roughly corresponding to its prin-cipal river basins. The RWMA in Poznan covers the Warta River. A decade-long delay in revising the national water law, from 1991 through 2001, created some significant problems for the basin management agencies. Toward the end of the 1990s, efforts to rationalize management functions and shift policy toward integrated water resources management (IWRM) resumed in earnest. Poland’s movement toward EU accession (which occurred May 1, 2004) also made it necessary to focus on IWRM in order to begin aligning Polish policy and practice with EU standards and expectations.

Salient Features of Basin Organization The 2001 water law, with 220 articles, is quite detailed and specific concern-

ing water management generally, and river basin management in particular. On behalf of the central government, RWMAs perform IWRM planning and coordinating functions within river basins, overseeing the actions of voivode-ship and local governments and private users for compatibility with basin water management plans, and maintaining speci fied water works and state-owned reservoirs and other facilities. RWMAs have a legally recognized role in the water use and discharge permitting procedures that are carried out by voivodeship or local (poviat or starost) offices. RWMA Poznan receives an annual budget allocation from the central government, distributed through the Ministry of Environment. Some of the RWMA’s functions in managing state-owned facilities generate fee revenues, but most of that revenue goes directly to the Ministry of Finance. Overall, 99.5% of the RWMA’s budget comes from the central government.

RWMA functions focus mainly on: IWRM planning; promoting compatibility of other governments’ water manage-ment actions with basin and national plans and policies through monitoring, or through review and comment; and operational functions concerning reservoirs and other facilities. Voivodeship level functions relate primarily to: issuing permits (done by the Environmental Protection and Agriculture Departments in 8 voivodeships in the Warta basin); funding environmental improvement projects (done by the Funds for Environmental Protection and Water Management in the 8 voivodeships); monitoring and enforcement (done by Inspectorates of Environmental

36

Protection in the 8 voivodeships); and management of irrigation facilities and primary drainage systems (done by voivodeship Boards of Land Improvement and Water Devices in the 8 voivodeships). Poviat level functions focus on: issuing water use and water discharge permits; monitoring and enforcement (done by environmental protection departments of poviat’s offices); flood protection activities; fisheries protection; and funding environmental improve-ment projects (done by the poviat Funds for Environmental Protection and Water Management). Gmina level functions include: water quality enforcement for providers of public water supplies and wastewater service; funding environmental improvement projects (done by gmina Funds for Environmental Protection and Water Management); and monitoring and enforcement (done by gmina environmental protection departments).

Outcomes

CONTEXTUAL FACTORS AND INITIAL CONDITIONS. The Warta basin does not feature significant cultural, religious, ethnic or other divisions within the population that hinder the prospects for successful river basin management. Similarly, asymmetries in the distribution of resources among basin stakeholders do not appear to have impeded the move toward the adoption of IWRM at the basin scale. Economic development of the basin and the country have had effects, however. Poland’s agricultural and industrial sectors emerged from the era of Soviet domination lagging behind the West. Support from international financial institutions and from the EU aided Poland’s economic and political transition, and also provided incentives for reforms such as IWRM and the creation of river basin agencies. Still, Poland’s economic conditions have led to financial constraints on the government sector, limiting its ability to provide either central funding or revenue autonomy adequate to the tasks of IWRM at the basin scale.

THE DECENTRALIZATION PROCESS. The decentralization of government in Poland, and the reform of water policy and water organizations, were attempted over the same (relatively short) period of time resulting in some uneven progress. Significant responsibilities for water resource planning and management have been spread across basin and sub-basin agencies, and water law reform took several years longer than originally envisioned. There appears to be no question, however, of the central government’s commitment to decentralization and democratization reforms, or about its recognition of the local and basin-scale organizations that it created. Central government officials have maintained that commitment throughout the post-Soviet period. However, they have held the pur-sue strings rather tightly in light of the limited financial resources available to the public sector in Poland.

CENTRAL-LOCAL RELATIONSHIPS AND CAPACITIES. Overall, the water law changes in 1997 and 2001, and the merger with the DDWMs in 1999, have given the RWMAs more responsibilities but not additional sources of revenue. The RWMAs have not fared well in the national government’s budgetary process. RWMA Poznan had a 2002 budget of USD$1.8 million, quite small for an organization covering such a large basin and employing so many individuals. Of this allocation, 73.8% is used for investments end planning in the basin, 5.9% for other devel-opment activities, 2.2% for water quality activities, 0.1% for operations and maintenance, and 18% for administra-tion and other categories. The small amount of financing has left the RWMAs unable to address the wide array of management concerns within the basin, or even to adequately fund maintenance of water works within the basin. The RWMA Poznan estimates that fulfilling all of its responsibilities would cost about 100 million zlotys per year. It receives a budget of about 5 million zlotys per year.

The water rights system is conducive to IWRM, and the reforms since the 1990s have attempted to add a basin-wide perspective to that system. Permits for water use and water discharge are limited in time and quantity, and approved only after consultation with stakeholders about basin conditions. Fees associated with non-permitted actions or with permit violations provide incentives to users and also a revenue source for environmental improvement projects. Other reforms (such as transferability of permits) have yet to be undertaken, but most elements of the institutional infrastructure of a water rights system compatible with IWRM are in place.

37

BASIN-LEVEL INSTITUTIONAL ARRANGEMENTS. Management of the Warta River basin is substantially dispersed and polycentric. Its federated nature is illustrated by the following comparison, provided by the RWMA Poznan director, of the distribution of water management responsibilities before and after the 1989-1991 period of initial decentralization reform that shifted Polish water management toward a river basin focus. Before creation of the RBWMs, water administration was 100% national, now it is 65% municipal and county, 30% basin, 4% voivode, and 1% national. Before creation of the RBWMs, infrastructure financing was 100% national, now it is 50% municipal and county, 30% voivodeship, and 20% national. Before creation of the RBWMs, water quality enforcement was 100% national, now it is 60% municipal and county, 30% voivodeship, and 10% national. Before the reorganization, voivodes were 100% responsible for issuing water and wastewater permits; now this is done by poviat or voivodeship officials with the advice and consent of RWMAs.

This federal approach, with responsibilities spread across levels and units of government, allows for the recogni-tion of sub-basin communities of interest, and provides overlapping layers of monitoring and enforcement of water management regulations. It does not, however, lend itself to clarity of institutional boundaries or a close matching of jurisdictional boundaries to basin boundaries. The RWMA in Poznan has to interact with several voivodeships that lie partly within and partly outside the Warta River basin. These interactions do not yet oper-ate seamlessly. The RCWMs (Regional Council of Water Management) and NCWM (National Council of Water Management) seem willing to aid in information sharing and communication roles, but they are so new that there is no record from which to judge their operation. Nor can the effectiveness of conflict resolution methods (which rely strongly on negotiations between governmental units) be assessed conclusively at this point.

38

The Warta basin illustrates how much institutional creation and policy reform can be accomplished in a relatively short period when a central government commits to decentralization and to IWRM. Fifteen years ago, Poland did not have a rational system of water tariffs, wastewater discharge controls, or water resource planning, let alone a set of river basin-scale organizations for water management. Now all of these are in place, albeit still quite new, along with bodies at the national, provincial and local levels for funding water quality improvements and other environmental protection projects.

The Warta basin case also illustrates, however, the gaps that can emerge between river basin management organizations and a policy of IWRM. From 1989-2001, the central government attempted to revise and reform the entire structure of general-purpose governments at the provincial and local levels, to decentralize several state functions to those levels, to create and then reorganize its system of river basin management agencies, and reform its policy approach to water resource management. While much has been accomplished, institutional boundaries have not always been clear, and some things have proceeded quite out of phase—principally, the establishment of the river basin agencies without a revenue source of their own, without a structure for basin stakeholder representation and participation, and a decade before the passage of the water law that largely defines and authorizes their activities.

Polish water policy has indeed embraced and moved toward IWRM, but the decentralization has spread water manage-ment responsibilities and authority across a large number of sub-basin entities. Organizational responsibilities and relation-ships appear to be substantially less integrated than policy. There are requirements for consultation of the RWMAs and conformity to basin plans, but until 2002 (with the creation of the RCWMs) there was no formal structure to integrate the general-purpose governments at the voivodeship and local levels into the RWMAs or vice versa. Currently there is a sub-stantial gap between the basin-scale organizations that have been created in Poland and the activities that comprise IWRM, most of which have been assigned to sub-basin governments. The results are that water quality remains a great challenge in the Warta basin. Rivers remain polluted with bacteriological and chemical contaminants. There was a national policy to promote water storage facilities such as small reservoirs, but funding has not been adequate to carry it out. Severe flooding occurred in the Warta basin as throughout Poland in 1997, and the risk of flooding has not been eliminated.

Conclusion

39

Guadalquivir River Basin, Spain

How political transformation, the addition of EU policies and regulations, and the dispersion of responsibili-ties into overlapping agencies and levels of government have produced an uncertain transition and gener-ated a mixed record of institutional performance in the seventy-five-year history of river basin governance and management.

GEOGRAPHICAL OVERVIEW. The Guadalquivir river basin extends westerly across southern Spain, and nearly all (90.2%) of its 57,017 km2 area lies within the region (Communidad Autonoma, or CA) of Andalucia. The entire 640-km main stem of the Guadalquivir River itself is located within the CA of Andalucia. As is the case throughout southern Spain, the Guadalquivir basin has a relatively small share of the nation’s water resources, despite having a substantial share of Spain’s population. The southern river basins of Guadalquivir, Guadiana, Sur, Segura, and Jucar contain 37% of Spain’s population and represent 41% of the Spanish land surface, but receive 19% of the country’s total precipitation and runoff.

Water users in the Guadalquivir basin have relied primarily upon surface water resources to supply their needs. There are 52 identified groundwater areas in the river basin, and groundwater overdrafting is an isolated rather than widespread problem in the basin. Precipitation is greatest in the mountains along the edges of the basin and lowest in the valley floor, where most of the population and irrigation are concentrated. Precipitation and stream flows are highly variable, exposing residents to risks of flooding as well as drought. Mean annual precipitation is 596 mm, but this has ranged from as little as 300 mm during drought to as much as 1100 mm. Years of high or low precipitation have often clustered together, compounding their effects. Seasonal variability is also significant, with most precipitation concentrated in the winter months and peak rainfall occurring from November through March. Long, dry summers follow, during which precipitation is virtually nil and evapo-transpiration soars.

History of the Guadalquivir River Basin Organization

Created by the central government in 1927, the Guadalquivir River Basin Organization (or, Confederacion Hidrografica (CH) Guadalquivir) was, for much of its existence, a hydro-tech-nical agency devoted to the construction of dams, reservoirs, and water conveyance facilities, while water law administration and management of water uses were handled by a separate agency. At times, however, the cen-tral government merged these functions into one basin-wide authority. Since 1985 the CHs have had combined responsibility for physical infrastructure and water use management in the basins, and thus are currently supposed to function as integrated water resource management (IWRM) agencies at the river basin scale. CHs are not autonomous. They are under the direction of the central government

Guadalquivir

Guadiam

ar

Viar

Corbones

Genil

Guadajoz

Gua

dalén

Guadalima

GuadianaMenor

Fard

es

Almería

Baza

Motril

Málaga

Cádiz

Jerez dela Frontera

Sevilla

Zafra

Trujillo

Beja

Córdoba

Granada

Ubeda

Ciudad Real

Ecija

Loja

Lorca

Cieza

Albacete

Huelva

Jaén

Bailén

Faro

GILBRALTAR, U.K.

PORTUGAL

MOROCCO

S P A I N

Mediterranean

Sea

Guadalquivir

Guadiam

ar

Viar

Corbones

Genil

Guadajoz

Gua

dalén

Guadalima

GuadianaMenor

Fard

es

7 5 4 3

56 4 3

8

78

36

37

38

37

38

3939

SPAIN

GUADALQUIVIR RIVER BASINSELECTED CITIES

MAJOR ROADS

INTERNATIONAL BOUNDARIES

0 10 20 30 40

0 10 20 30 40 Miles

50 Kilometers

This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other informationshown on this map do not imply, on the part of The World BankGroup, any judgment on the legal status of any territory, or anyendorsement or acceptance of such boundaries.

SPAIN

GuadalquivirRiver Basin

Box 8

40

or regional governments, depending on whether they cover an inter-regional river basin or an intra-re gio nal one. National water law adopted in 1985 (and an associated 1987 decree) recognized 13 CHs. The principal responsibilities of CH Guadalquivir are established by the 1985 national water law and 1999 amendments to it, by the Guadalquivir Basin Plan and National Water Plan, and by EU regulations that are implemented at the river basin level.

The national water law prior to 1985 had been in place since 1879. The 1985 water law represented a major reform of water policy in Spain. It was not a product of grass-roots reformism, however, having been designed primarily by expert personnel at the national level. Four principles of the 1985 law were: 1) integrated man-agement of water resources (surface water and groundwater, water quantity and water quality); 2) the river basin as the appropriate unit or scale of management; 3) user participation; and 4) reliance on water planning to balance social and economic development with ecological sustainability. The 1985 law called for a new and comprehensive round of water plans. The central government was made responsible for drafting a National Water Plan, and for providing technical assistance to the CHs as they drafted Basin Water Plans. The 1985 law broadened the definition of waters in the public domain to include essentially all surface waters and ground-water. Uses that had previously been regarded as “private” therefore became public, and the government (pri-marily through the CHs) assumed responsibilities for regulating those uses through licensing and registration. Issuing water licenses and maintaining water registries represented a new role for the CHs, and reinforced their new mission as integrated water management agencies. The 1985 law added economic tools for the regulation of water. It provided for water regulation rates and water discharge taxes for the use of water and the water domain. The 1985 water law conferred authority upon the CHs for the control of discharges to waters, and for the monitoring and management of river water quality.

The 1999 amendments did not alter the organizational structure of water management in Spain, but they exemplified the continued movement of Spanish water law and policy toward an integrated water resource management approach, strengthening public control over water use and the protection of water quality and environmental values, while adding economic incentives and some flexibility with regard to water demands and water uses. The combination represents a continuation of water policy reform. The revised national water plan presented in 2001, which raised conservation targets, and the announcement of the newly elected government in 2004 that it would cancel a major inter-basin water conveyance project, also reflect the reform trend.

Salient Features of Basin Organization

Internally, CH Guadalquivir is organized into staff offices plus a set of boards, councils, and commissions com-posed of basin stakeholder representatives and CH staff . The CH president serves as head of the CH staff and chairs the advisory bodies. Thus, the president has a strong role in guiding the CH’s activities. The president is appointed by the Council of Ministers and is effectively an official of the central government. The Governing Board is in charge of financial matters, approves action plans, and defines aquifer depletion and groundwa ter protection areas.

There are several Operation Boards which co-ordinate the manage ment of hydraulic works and water resources in specific catchment areas and/or hydrogeological units. They are composed of representatives of the administration and of the water users. The 1985 Waters Act establis hes the representation of each sec-tor on the boards. The Water Users’ Assembly, also headed by the CH President, is composed of all users that are part of the Operation Boards. It makes recommendations concerning CH policies for the coordi-nated management of hydraulic works and water resources throughout the basin. The Reservoir Releases

40

41

Commission makes recommendations to the President concerning water release from reser voirs. The Water Users’ Assembly proposes which users should be mem bers of this Commission. A Permanent Committee of this commission responds to emergency situations such as floods or drought. Water Works Commissions pro-vide water users who will be served by a particular project information and provide a forum in which to make recommendations. The Basin Water Council, headed by the CH President, approves the Basin Hydrological Plan, which is forwarded to the central Government. It is composed of representati ves of different depart-ments of central and regional governments, technical services, and basin stakeholders (at least 33% of council membership) including professional associations and environmental groups. The Pla nning Office is a CH staff office headed by the Chief of Water Planning. It is responsible for drafting, monitoring, and reviewing the Basin Hydrological Plan and providing tech nical support to the Basin Water Council.

Outcomes

CONTEXTUAL FACTORS AND INITIAL CONDITIONS. Water management issues and their resolution in the Guadalquivir case do not appear to have been driven by ethnic, religious, or class divisions in Andalusian society. On the other hand, the economic development of the nation and of the region have had notable effects. The very establishment of the CHs, with an emphasis on the construction of water works to promote land develop-ment, emanated from national policies to bolster economic development by promoting first the expansion of agriculture and later the expansion of industry. The Guadalquivir basin was poorer and more rural than most of the rest of the country, and these conditions contributed to an emphasis on the expansion and protection of irrigated agriculture as the central element of the region’s economic and social life. These contextual factors have shaped the perceptions of many Guadalquivir basin stakeholders and the CH staff about the principal purposes and appropriate focus of river basin management.

THE DECENTRALIZATION PROCESS. The CHs were created by the central government for its own purposes—neither because of local-level demands for greater autonomy nor because of a central-government desire to shed water management responsibilities, but as an organizational device for executing central government policy one river basin at a time. The CHs nonetheless provided a means for stakeholder participation through representation on boards and commissions. The establishment of basin management institutions in Spain thus carried the potential for greater water user involvement, but that was not the principal reason for which they were created. The CHs are best thought of as central government agencies with representative components, with the balance between central control and user participation varying over time. Central government officials have established, diminished, and resurrected the user representation components over the life of the CHs.

Taking the 1978-1985 period as the principal decentralization reform landmark, several changes in that period contributed to the current configuration of river basin management, both in terms of regional devolution and of participation in policy making: 1) the transition to democracy and the approval of the Spanish Constitution in 1978; 2) the creation of regional governments (Communidades Autonomas, CAs) with differing degrees of responsibilities in water management; 3) the issuance of the 1985 decree re-uniting hydraulic and water resource management into the functional responsibilities of the CHs; 4) the approval of the 1985 Water Act, establishing a modern framework of water management, linking surface and groundwater, encouraging user participation, and conferring a central role to water planning; and 5) the 1987 decree restructuring the internal composition of the CHs and their boards and commissions. While basin-scale institutions enjoy the recognition of central government officials as legitimate water resource management entities, such recognition has not been accompanied by an extensive devolution of authority to basin-scale institutions. The organization, responsibili-ties, and policy direction of CH Guadalquivir takes its lead from Madrid.

41

42

CENTRAL-LOCAL RELATIONSHIPS AND CAPACITIES. Basin level entities such as CH Guadalquivir develop basin level plans, but these plans must be submitted for national approval and be consistent with the national water plan. CHs collect and maintain revenue of their own for some of the services they provide but they also rely on central government funding for functions established and determined by central government officials. CHs have several advisory bodies composed of stakeholder representatives, but several of those councils also have

central government representatives and the CH president is still a central government designee. In addition, central government policy reforms since the 1970s have broad-ened the CHs’ portfolio of IWRM respon-sibilities in ways that CH Guadalquivir staff have found difficult to sustain despite CH Guadalquivir having relatively good financial resources and autonomy. The central gov-ernment’s new water rights regime (begun in 1985 and modified somewhat in 1999) is certainly beneficial from an IWRM stand-point—bringing more users into the system (through the expanded definition of “public domain”), quantifying licenses, maintaining a registry of users, and creating opportunities for trading. Implementation and enforce-ment has been delegated to the CH. The central government has the authority to alter the governance structure or deci-sion making processes of the CHs (and to appoint its leadership) with as much or as little stakeholder consultation as it chooses, but the basin stakeholders do not possess a comparable ability to tailor the institutions to their perceptions of needed or appropri-ate arrangements.

BASIN-LEVEL INSTITUTIONAL ARRANGEMENTS. Basin-level governance institutions correspond with the geo-graphic boundaries of the river basin. From 1927 to the present the river basin authority was primarily a water works construction and operation agency. CH Guadalquivir did not function as a basin governance entity through much of its existence. Changes in the responsibilities and the structure of the CHs in 1985 and 1987 appear to have been intended to transform them into basin governance organizations. But while geographic boundaries fit well, institutional boundaries have become unclear. The allocation of authority to regional government in the 1970s leaves room for interpretation about who is responsible for what or who can do what, and the changes in the CHs’ roles/responsibilities from the 1930s through the mid-1980s broadened (but did not necessarily sharpen) the understanding of what they may, must, or must not do. The potential for conflict between regional government policies and CH policies is heightened by social and political factors. For example, the CH’s representation and governance structure gives disproportional weight to irrigation users. The basin-level institutional arrangements do recognize sub-watershed communities of interest within the basin. However, only irrigation user communities have formal recognition in both national law and the CH organizational structure.

42

43

Basin-level institutional arrangements are structured to provide fora for information sharing and communica-tion among basin stakeholders and between stakeholders and CH staff. The effectiveness of these structures varies. It appears that meetings of the Water Users’ Assembly, for example, are few and rather far between. Operations boards and the CH governing board meet more regularly but their representation is not as broad. At the regional level, the Andalusian Water Council appears to have met regularly and has a broad represen-tative structure, but is not coordinated with the river basin authority. At the sub-basin level, users’ assemblies in the irrigation user communities may draw greater rates of participation but this varies a great deal across the hundreds of communities in the basin. The irrigation communities have irrigation courts to resolve dis-agreements among water users, but no comparable forum exists on the river basin scale. Monitoring of water deliveries, water use, and water quality occurs in the basin, with water quality monitoring performed principally by CH Guadalquivir and by urban water suppliers.

The 1985 water law, 1999 amendments, and 2001 national water plan, combined with the EU Water Framework Directive, have moved Spanish water policy away from a sole emphasis on supply augmentation and toward the incorporation of additional goals of water quality improvement, water demand management, water use efficiency, and environmental protection. At a system-wide level, Spanish water policy in 2004 much more closely resembles an IWRM approach than it did 20 years ago. In the Guadalquivir basin, these changes have been accompanied by greater friction between urban and irrigation water constituencies (particularly during and after the 1992-95 drought), and the growing interest of the Andalusian regional government in providing an alternative forum for water policy making. Water policy reform has resulted mainly in the addition of new and different responsibilities to the CHs, and some involvement of regional and local governments in water policy and politics, but not a transformation of the organizational approach to river basin management. Even today, almost 20 years since the water works and management functions were re-com-bined in 1985, the water works function seems to be what CH Guadalquivir officials and staff are most interested in and comfortable with. The management functions—water licensing, demand management—have been performed with less vigor and with less positive results.

The Guadalquivir basin “water deficit” has not been erased, and exposure to droughts remains a principal problem. The effectiveness of the structural approach to addressing supply-demand imbalances is now being called into question. Flooding remains an occasional problem, despite the number of facilities that have been built. Indeed, urban and agricul-tural development has encroached into floodplain areas, making the population and economic activities more vulnerable. Thus far, no systematic efforts to reduce non-point pollution such as that coming from runoff have been implemented in the basin. Many problems have emerged in implementing the 1985 law’s integration of groundwater pumping into the water licensing system, and integration remains a work in progress. Agricultural water use remains comparatively inef-ficient: most irrigation users pay water tariffs based on their land surface rather than water use, so the tariff does not provide a financial incentive to conserve.

Although stakeholder representation on the CH boards and councils was expanded in 1987 and 1989 central govern-ment decrees, the management structure and internal culture of CH Guadalquivir has been slow to change. Formal decision-making authority remains concentrated in the hands of the CH president and board. The current conflicts concerning water management appear to be taking place outside the CHs rather than finding expression within them. CH Guadalquivir may still be perceived as a relatively closed agency serving irrigators’ interests, and is not yet the forum within which a broader range of basin stakeholders express their views and determine basin policy direction. CH Guadalquivir was established as an agency of the central government and remains so today. The transformation of the Spanish political system, the addition of EU policies and regulations, and the dispersion of responsibilities into overlap-ping agencies and levels of government have made for an uncertain transition in the Guadalquivir basin and generated a mixed record of institutional performance.

Conclusion

43

44

Comparative Analysis of River Basin Management

Analyzing each case in light of the specific geographical, historical and organizational contexts and the evolution of institutional arrangements, and comparing the levels of success in achieving improved stakeholder participation and integrated water resources management (IWRM).

Characteristics of the River Basin Organizations

The initial conditions and the principal water management problems that the basin stakeholders confronted in each case varied substantially. These challenges included: pollution, urban development in headwaters, scarcity, flooding, seasonal water scarcity, intersectoral conflict, water storage, drought exposure and erosion. The cases also differed on whether the development of institutional arrangements for water management at the basin level had originated

as a central government reform effort or as an effort initi-ated by stakeholders within the basin, as was the case in the Fraser, Murray-Darling, and Tárcoles basins. Where a central government initiated the development of basin manage-ment arrangements, it could have occurred as a singular act applying to just one basin or as part of a broader water policy reform—for example, creating basin-scale organiza-tions throughout the country as Poland has done.

It also became apparent that, for some cases, supranational organizations such as the World Bank and the European Union influenced the development or modification of basin management programs or institutions. Both Spain and Poland have moved substantially toward IWRM (par-ticularly in regard to water quality protection and water pricing) in response to the EU Water Framework Directive. The World Bank’s promotion of IWRM and stakeholder involvement influenced the creation of the basin manage-ment organizations in Ceará, Brazil (the Jaguaribe case) and the continuity of the basin management corpora-

tion approach in Indonesia (the Brantas case). The Inter-American Development Bank supported the Tárcoles Commission and Sao Pãulo’s 1991 water law. A World Bank-financed project in the state of São Paulo supported the development of legislation that would have influence on the instruments for river basin management in the Alto Tietê basin and on its institutions, such as the Headwaters Protection Law and the draft water pricing law.

Basin-scale organizations have been created in each of the eight cases, but they differ in structure and type. Two of the cases featured state companies, two involved central government agencies operating within nationally-defined basin boundaries, and the other four were unique variations (one inter-governmental commission, one quasi-govern-mental commission, one nongovernmental basin council, and one hybrid basin committee/basin agency structure).

Because IWRM at the river basin level could involve a range of responsibilities and activities, it is not surprising that the cases studied differed in the functions they perform. Some had authority to allocate water to users and oth-ers did not. Many but not all were responsible for water quality. A few were engaged in setting and/or collecting water tariffs. Some operated dams, reservoirs, and other physical facilities. The only function performed by all was planning and coordination—all developed basin management plans and/or coordinated activities among multiple governmental and nongovernmental entities present within the basins.

Fraser RiverBasin

See IBRD 33725

Tárcoles RiverBasin

See IBRD 33723

JaguaribeRiver BasinSee IBRD 33728

Alto Tietê River BasinSee IBRD 33727

GuadalquivirRiver BasinSee IBRD 33722

Warta RiverBasin

See IBRD 33721

Brantas RiverBasin

See IBRD 33724

Murray-DarlingRiver BasinSee IBRD 33726

CANADA

BRAZIL

COSTA RICA

SPAIN

POLAND

AUSTRALIA

INDONESIA

This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other information shown on this map do not imply, on the part of The World Bank Group, any judgment on the legal status of any territory, o r any endo r s emen t o r a c c e p t a n c e o f s u c h boundaries.

Box 9

45

Although management at the basin level is uniformly promoted as a way of increasing stakeholder involvement, there is no similarly uniform prescription for how this should be done. As expected, the eight cases demonstrated a variety of means of organizing stakeholder participation and soliciting stakeholder input into basin manage-ment decisions. Two of the cases, Brantas and Warta, had no established stakeholder organization (committee or other type) during the time of the study, although one (Brantas) had a program of outreach and communication between basin agency personnel and various individual stakeholders in the basin and the other (Warta) is now developing a regional water management council under the direction of a national law governing the structure and operation of basin management agencies. Other cases, such as the Alto Tietê and Fraser basins, had elabo-rate and multi-scale structures. Jaguaribe had numerous sub-basin user committees and commissions, but only the State Water Resources Management Company, which provides the technical support to these, operates at the basin scale, and Murray-Darling has a basin-wide Community Advisory Committee but not sub-basin ones. The Guadalquivir and Tárcoles cases have representative structures incorporating a variety of stakeholders, but there were doubts about which of those bodies met regularly and whether the broadly representative ones had substantial input into basin management decisions.

Similarly varied are the means and sources of funding for the basin organizations in the eight cases. Three (Alto Tietê, Tárcoles, and Warta) rely solely on central government budget allocations at present, although there have been other sources for the Tárcoles basin commission in the past, and the basin agency and committee in the Alto Tietê are supposed to have revenue from water charges in the future. Three others (Brantas, Guadalquivir, and Murray-Darling) enjoy a combination of central government support and water user charges. One (Jaguaribe) is funded entirely by water user charges, although these are collected by the state water resources management agency and then reallocated to the basin. The nongovernmental Fraser Basin Council obviously lacks the author-ity to levy taxes or charges on water use, and instead receives annual financial support from governments and project funds from a variety of sources.

Start-up Factors

STAKEHOLDER INVOLVEMENT. In most cases, stakeholder involvement was secured. The river basin organizations in the Warta River in Poland and in the Brantas River in Indonesia lacked organized structures for stakeholder involvement, so communication between basin management staff and stakeholders in those cases was more individualized and less regular. In the Guadalquivir basin in Spain, the number and diversity of stakeholder repre-sentatives has been expanded with reform and reorganization of the basin management agency, but some stake-holders (particularly irrigation communities) are more actively involved than others, and there were expressions of dissatisfaction from some other stakeholders related to contact with the basin management agency. In the Tárcoles basin in Costa Rica, stakeholder involvement was quite active in the 1990s as the basin commission was getting started. In the Murray-Darling case in Australia, there are multiple layers of stakeholder involvement, including rural irrigation companies and districts, regional catchment authorities, water service provider associa-tions, a basinwide Community Advisory Committee, and state and federal government representation on the commission and ministerial council for the basin as a whole. Local stakeholders participated actively with state government personnel in establishing basin management organizations in the Alto Tietê basin in Brazil, where sub-basin committees organized around “social basin’ boundaries more than hydrological ones continue to enjoy active stakeholder participation. In the Jaguaribe basin in Brazil, participation takes place along the river valley where the key infrastructure is located and around strategic multi-year reservoirs. However, for strate-gic reasons, the state government has little interest in having stakeholders from the Jaguaribe basin become actively involved in longer-term allocation decisions regarding the transfer of water into the Greater Fortaleza Metropolitan basin to meet the state capital’s growing water needs. The Fraser Basin Council in Canada has succeeded through regular meeting schedules and a combination of basin wide and sub-basin regional councils to secure and maintain a high level of stakeholder engagement since its establishment in 1997.

46

INCENTIVES. A highly common incentive was the presence of severe water resource problems. The presence, or the prospect, of valuable infrastructure investments became a point of stakeholder interest from the beginning of the Guadalquivir and Murray-Darling cases. The absence of significant cultural conflicts among basin stakeholders in most of the cases helped as well.

CHAMPIONS. The commitment of governmental support to the creation of stakeholder-based or stakeholder-involved organizations was a positive factor at the outset of the Tárcoles, Fraser, Jaguaribe and Alto Tietê experi-ences. In some cases, the process was strengthened by the presence of an influential individual champion who drew attention to basin problems and conditions, and who possessed the entrepreneurial skill and/or political power to bring about the combination of people, information, and finances or to use the mandate to get a basin-level orga-nization started within a wider context of water resources management reform. This was particularly important in the Brantas, Jaguaribe, Fraser, and Tárcoles cases, and to some extent with the widening of the scope of basin management issues in the Murray-Darling basin. A third factor was the influence of supra-national entities such as the World Bank or the European Union. Conditions or advice accompanying Bank assistance, and the desire for policies that are consistent with the EU Water Framework Directive, shaped the establishment of basin organiza-tions, the broadening of water management issues toward an IWRM approach, or both.

Sustainability Factors

KEEPING STAKEHOLDERS ENGAGED. Where sustained active involvement was observed, it appeared to be connected with stakeholder perceptions that the basin management organizations were engaged in important issues and were making (or had made) a positive difference in basin conditions, with consistency of governmental support, and with some of the same factors that stimulated stakeholder involvement in the first place such as regular and frequent interaction. These factors were present in the Fraser and Murray-Darling cases, at the sub-basin levels in the Alto Tietê and Jaguaribe cases, and with respect to irrigation communities in the Guadalquivir basin.

Stakeholder involvement has waned during the past five years in the Tárcoles case, as well as in the Jaguaribe and Alto Tietê cases, and is lower among non-irrigation stakeholders in the Guadalquivir case, for similar reasons. Irregular and less frequent commission meetings in recent years and a perceived inconsistency of government support were noted in the Tárcoles basin. In the Guadalquivir basin, some non-irrigation stakeholders expressed the view that they participated less because they believed their views and interests were not as welcome. Partly in reaction to this perception, the Andalusian regional government has even constructed an alternative stakeholder forum in that basin. In the state of Ceará, a change of government took place with the current government return-ing to more centralized policies. And in the Alto Tietê case, state government has never given full political support to some key reforms on the agenda. Over time, therefore, the activities of the Alto Tietê Committee seem to have lost enthusiasm and strength.

PARTICIPATORY DECISION-MAKING. Relatively new as well as long-lived river basin organizations had engaged stakeholders in substantive basin management decisions. Stakeholder involvement was more common with respect to basin planning, water supply allocation, and infrastructure operation, and less common with respect to setting water charges, collecting fees, flood control, monitoring basin conditions, altering land uses or infrastruc-ture construction decisions.

In the Guadalquivir basin, operations boards that include stakeholders and agency staff make reservoir manage-ment decisions, a basin-wide commission that includes stakeholder representatives decides upon the schedule and volume of water storage releases each year, and water supplies are allocated among individual farmers by management boards in the many irrigation communities of the basin. In the Jaguaribe basin, user committees at the local reservoir scale have a strong say in annual water storage and releases. In the Alto Tietê basin, the river

47

basin committee and sub-basin committees are involved in designating headwater protection areas. Catchment management authorities in the Murray-Darling basin are increasingly involved in review of land-use changes for possible effects on water and other natural resources, irrigation companies or districts set water rates and make allocation decisions, and representatives from the three states along the main stem of the Murray River participate in decisions about the operation of dams and the management of flows in the river. During the 1990s, stakehold-ers in the Tárcoles basin were actively involved in a reforestation program and in the recruitment and recognition of businesses for participation in a basin-wide voluntary program to reduce discharge of contaminants to the river and tributaries. Even the nongovernmental Fraser Basin Council—which does not execute policy or implement projects directly—holds “State of the Basin” conferences, issues “State of the Basin” reports, forms partnerships with other gov-ernmental and nongovernmental entities in the basin for specific projects, and lends its endorsement to an agenda of basin sustain-ability initiatives that appear to influence the decisions of officials at the local, provincial, and federal government levels.

Stakeholder involvement in basin management in the Brantas and Warta basin cases has been more extensive with respect to the review of basin plans prepared by agency staff than in the making of basin management decisions or the operation of projects. At the province (voivode) level in Poland, representative boards do participate in decisions about the allocation of funding for waste-water treatment and other environmental improvement projects, although this activity is not organized at the basin scale nor does the regional water management authority for the river basin have a direct role in it. Nevertheless, the financial investments made by these provincial bodies are a very important element of improving water quality conditions throughout the river basin.

BALANCING STAKEHOLDER INCENTIVES WITH DESIRED OUTCOMES. With respect to reducing exposure to flooding and better management of releases from water storage reservoirs, stake-holder involvement and performance improvements in the Brantas, Guadalquivir, Jaguaribe, and Murray-Darling cases have gone hand in hand. The same can be said of reduction in the rate of deforestation in the Tárcoles basin, of improved treatment of industrial wastewater and reduced use of the river for waste discharge in the Fraser basin, and of headwater area protection in the Alto Tietê basin. Measurable improvements to wastewater treatment in the Warta basin have resulted from the financial investments of the provincial funds for environmental protection and water management.

On the other hand, stakeholder involvement can perpetuate impediments to improved water resource manage-ment. Irrigation communities and basin agency personnel in the Guadalquivir basin, for example, have devoted decades to the lobbying for construction of additional dams and reservoirs, yet the basin’s annual “water deficit” remains in place. Furthermore, improvements to the efficiency of agricultural water use there have been piece-meal, unlicensed uses of water for irrigation continue to grow, and the issue of agricultural contributions to water pollution has been addressed only recently and partially. In the Brantas basin, agricultural users consume the largest share of water supply but have managed to remain exempt from water tariffs that are paid by municipal, industrial, and hydropower users. In the Tárcoles basin, public hydropower producers make the most, albeit non-consumptive, use of surface water yet are exempt from water licensing and tariff requirements that apply to other users (including private hydropower producers).

48

RESPONSIVENESS TO ENVIRONMENTAL CHANGE. Despite improvements, significant water resource management problems remain in all of the cases. In several cases, this is due to changing conditions to which management entities have not adapted fully. For instance, the long-lived management regime of the Murray-Darling basin is wrestling with more recently recognized problems of dryland salinity and deteriorated river ecology. The equally long-lived Guadalquivir basin agency has yet to fully cope with its enlarged responsibilities for water licensing and demand management, and with water supply reliability and flood protection for a growing and rapidly urbanizing population. In the Warta case, flood risks have diminished and the quality of large-scale municipal and industrial wastewater discharges has improved, but lack of integration of water supply management with land use devel-opment and jurisdictional gaps between the issuance of permits and the enforcement of compliance with them have contributed to the rise of seasonal water scarcity and accumulating groundwater overdraft in portions of the basin. In the Tárcoles case, early successes in raising awareness of the negative impacts of deforestation and industrial water discharges have not been sustained, and in the meantime the basin’s population and economy have grown rapidly and the problems of land use, erosion, and water pollution have continued to increase. In the Jaguaribe case, water security in drought conditions for all stakeholders has significantly increased by more transparent and effective planning, and operation and maintenance of infrastructure. Environmental problems, which used to not even be contemplated due to the overriding issue of water scarcity and allocation are now a second-generation issue being raised and starting to be addressed. In the Alto Tietê case, the main problems of São Paulo’s water supply have not been sufficiently tackled and keep growing. Only at the subbasin levels are some activities carried out, but these are not enough for such a metropolitan area.

CONSISTENCY OF GOVERNMENT SUPPORT. The consistency of central government support for basin management, stakeholder involvement, and water policy reform has emerged as one of the most important factors distinguish-ing cases with greater levels of success and stakeholder participation from those with less. Consistency of support may be as important as magnitude of support over the long run. Magnitude matters, as can be seen in the negative effects of insufficient central government funding on basin management in the Warta and Tárcoles (post-1998) cases. By contrast, organizations such as the Murray-Darling Basin Commission, the Fraser Basin Council, and the Confederacion Hidrografica del Guadalquivir have been able to sustain multi-year basin planning and projects thanks to relatively consistent levels of central (and in the Fraser and Murray-Darling cases, state/provincial and local) support for their work.

Related variables concerning financial resources and financial autonomy mattered to longevity as well. Most rev-enue generated from water users in the Guadalquivir, Murray-Darling, and Brantas basins, for example, remains within or is returned to the basin for improvements and operations there. Basins that are wholly dependent on central government allocations (Warta and Alto Tietê, for example) have had a more difficult time establishing their own priorities and undertaking substantial projects.

The matter of top-down versus bottom-up initiation of reform seems to make a difference, as predicted, but this needs to be considered along with other variables—prior experience at the local level with self-governance and service provision, and consistency of central government support through periods of transition. Thus, while stakeholders initiated the creation of basin organizations in the Fraser, Murray-Darling, and Tárcoles basins, the first two are in countries with more experience of local autonomy than the third, and central government support has been more variable in the latter as well.

MANAGING CONFLICT. The relatively low levels of conflict among stakeholders in most of these cases was a positive factor. In the Jaguaribe case, differences between the metropolitan area of Fortaleza in an adjacent basin and the agricultural communities in the Jaguaribe basin from which water is being withdrawn to supply Fortaleza have contributed to the persistence of state-level decision making and to the delay in full devolution of manage-ment functions to the basin level. The Fraser Basin is another exception, where decades of tension between First

Conclusion

Nations and the Canadian federal and provincial governments was one of the factors that contributed to the decision to form a non-governmental basin organization that could incorporate representatives from each.

Finally, we noted earlier the importance of “champions” in getting basin management institutions started. Here, we add that the departure of such a “champion” is also a delicate moment in the life of those institutions. This has been evident in the Fraser, Tárcoles, Jaguaribe and to a certain extent Alto Tietê cases. Nevertheless, the institu-tions in most of those cases have survived and basin management activities continue, so it is equally important not to presume that the departure of a champion must spell doom for a basin management effort.

While decentralization holds out the promise of reduced financial and transaction costs, increased flexibility and effi-ciency, local control and accountability, the preconditions for realizing it are significant especially given the historical, political and socio-economic conditions facing many developing countries. Successful decentralization depends on the regions or districts concerned having a large enough revenue base to undertake activities; it requires the existence of clearly defined property and water rights among potential stakeholders (for effective negotiation and commitment to operations and maintenance); it requires the socio-political pre-conditions for participation and the resistance of elite capture (e.g., unstratified social structures); and demands transparency, clear roles and responsibilities with legal authority, and quality information. For those anticipating that decentralization might solve problems associated with weak centers and fiscal constraints, it is perhaps worth noting that decentralization works bests in the context of a sufficiently strong central government to that can support decentralization over time, including financially. What follows are a few of the study’s most relevant findings.

While the level of economic development of the nation and the basin can clearly make the creation and sustainability of basin-level institutions more or less difficult, there is no reason to believe that improvements in water resource management are limited to wealthy basins. Notable improvements can be and have been realized in a variety of settings, and sometimes very early in the life of basin organizations and stakeholder participation initiatives.

Although river basins are important hydrologically, ecologically, and economically, not all aspects of stakeholder par-ticipation and not all decisions and activities that contribute to IWRM need to be organized at the basin scale. The “lowest appropriate level” for some water resource management functions may be a sub-basin, a local or regional unit of government, or a hybrid unit sometimes referred to as the “social basin,” (e.g., the basin subcommittees in the Alto Tiete case).

The establishment of participatory and decision making structures involves shifts of power, something which can be a controversial and complicating factor. Even in settings where there is a desire for decentralized basin management, the political dimensions of public policy play a key role. Efforts by the Andalusian regional government to exercise more leadership over basin management in the Guadalquivir case, the efforts of users in the Jaguaribe basin to gain more influence in decision making, especially with regard to infrastructure, the desire of the state company in Indonesia to also take on pollution control which is currently the responsibility of provincial government, are only a few examples of the ways in which jurisdictional and other power-related considerations are likely to arise.

Finally, decentralization reforms and the establishment of river basin management with active stakeholder involvement are processes that take time, sometimes even decades. In order to sustain the reform process, consistent support is vital, as is the ability to adapt and modify basin management arrangements in response to changed conditions. Central governments and external organizations promoting integrated water resources management on a river basin scale must be prepared to sustain their commitment to reform, both through changes of administration and over the long haul.

49

50

51

Global Analyses

52

Decentralization of River Basin Management:

A Global Institutional and Policy Analysis

Global experience can help policymakers assess the most appropriate level to which they should decentralize river basin management. Many case study analyses, including the case studies highlighted here, shed light on the factors affecting basin decentralization. However, because they are case-study specific, they do not yet allow one to identify the generic reasons and forces behind decentralization, to characterize the process and the level of success of the decentralization process. Based on an analytical framework for relating decentralization and stakeholder involvement to improved river basin management, which was developed in the study underly-ing this text, this section suggests several hypotheses about factors associated with greater or lesser likelihood of decentralization success using statistical procedures that were applied to data collected from 83 river basins worldwide.

Methodology

A questionnaire consisting of 47 questions producing 226 primary variables was sent to nearly 200 river basin management organizations. Of the 103 returned, 83 questionnaires were used in the analysis. The variables were divided into four groups: (a) general data variables that provide contact information of the River Basin Organization’s (RBO); (b) institutional setup variables that describe various aspects of the institutional arrange-ments in the basin before and after the decentralization process; (c) finance variables addressing aspects of the RBO’s budget, and; (d) performance indicators as measures for performance of the RBO.

In terms of the general framework for the statistical analysis, the note reports two types of distinct relationships. The first is a relationship that explains the characterization of the decentralization process such as its length, participation, political hardship etc. It is done using a set of explanatory variables related to: (a) contextual factors and initial conditions in the basin, such as the level of economic development, the initial distribution of water resources among basin stakeholders and water availability, extent of already existing decentralization and level of reliance on surface water ; (b) characteristics of government/basin-level relationships and capacities, such as whether the decentralization initiative was top-down, bottom-up or mutually desired devolution; (c) the internal configuration of basin-level institutional arrangements, such as user groups and the existence of conflict resolution mechanisms, and; (d) ‘other’ variables identified as necessary.

The second relationship explains the level of success/progress of the decentralization process, measured by vari-ables that express incremental improvements in local responsibility between before and after decentralization, success level of major and minor reform objectives, and whether or not the decentralization yielded a discern-able institutional change. In addition to the variables mentioned above in the context of the first relationship, in this second relationship, the vector of characteristics of the decentralization process, such as the extent of initial decentralization and the share of surface water in the available water resources in the basin, is included as an independent variable.

These two relationships were estimated separately using statistical procedures, however, they are linked by the fact that the characteristics of the decentralization are used both to explain the nature of the decentralization process and to explain the performance of the decentralization.

53

Results

The outcomes of the survey can be divided into four sections corresponding to the variables explained above, that is: contextual factors and initial conditions; the decentralization process; central-local relationships and capacities, and basin-level institutional arrangements. The results are to be interpreted in the context of a partial equilibrium, that is, the impact of one explanatory variable on the dependent variable while holding everything else constant. In a later phase of the work, interactions with the various variables will be analyzed to allow recommendations regarding sequencing and packaging of policy interventions.

CONTEXTUAL FACTORS AND INITIAL CONDITIONS. Variables in this group significantly affect the nature of the decentralization process and its performance. Several results suggest that:

• The greater the extent of initial decentralization in the basin, the less time the decentralization process took

• The greater the reliance on surface water in the basin; the higher the degree of water user involvement and participation, the larger the number of institutions that were created during the decentralization process

• The greater the political transaction costs associated with the decentralization process, the greater the extent of reported improvement between “before” and “after” decentralization

• The greater the number of major problems in the basin prior to decentralization, the greater the extent of reported improvement between “before” and “after” decentralization

The basin water scarcity variable was the most robust variable in the various analyses that were conducted. Several of these findings suggest that:

• The greater the water scarcity level in the basin, the less time the decentralization process took

• The greater the water scarcity in the basin, the greater the extent of reported improvement between “before” and “after” decentralization

• The greater the water scarcity in the basin, the greater the extent of reported success with respect to the major objectives of basin management.

The starting point and level of the natural resource endowments in the basin do matter and thus, where one stand dictates how one should implement the reform and thus, how one may end up. One’s take from this set of findings is that rich and well endowed basins do not necessarily have an advantage over less endowed basins. Stressed resource conditions and the presence of multiple major problems can be stimulants to effective action and not only obstacles. Approaches for decentralization would address such differences and may lead to similar performances.

54

THE DECENTRALIZATION PROCESS. Variables in this group, such as the political economy of the process, participation, compli-ance and governance level suggest an array of supporting results some of which are as follows:

• The greater the extent of tariff compliance the lower the political transaction costs of the decentralization process; and the smaller the number of institutions that were dis-mantled during the decentralization process

• The greater the availability of fora for dispute resolution, the greater the extent of water user involvement and participation

• The greater the political transaction costs associated with the decentralization process, the smaller the reported improvement between “before” and “after” decentralization; and the less likely that some form of institutional change was associated with the decentralization process

• The longer the decentralization process took, the greater the extent of reported improvement between “before” and “after” decentralization.

• The more comprehensive the basin management objec-tives were, the greater the extent of reported success with respect to the major objectives of basin management; the greater the reported improvement between “before” and “after” decentralization.

The results associated with this set of variables suggest that indeed political economy plays a role in the decentralization

process and affects its performance level, implying a need for compromise and increased transaction costs. But a relevant and coherent decentralization agenda that addresses all stakeholders’ concerns proved to be effec-tive and successful. This finding suggests that diverse and ‘crowded’ basins do not necessarily have to face higher political cost and lower levels of performance of the reform, if an appropriate set of mechanisms and objec-tives, such as forums for dispute resolution, and a coherent reform agenda, are put in place at the appropriate time. Another important result is that longer is not uglier, and depending on the length of the process, some dividends can become evident.

CENTRAL-LOCAL RELATIONSHIPS AND CAPACITIES. Variables included in this set, such as budget and funding by the government agencies, and the initiation of the reform process also suggest plausible explanations. The common findings include:

• The larger the share of the RBO budget received from external governmental agencies, the smaller the num-ber of institutions that were created during the decentralization process

55

• The greater the share of the RBO budget coming from external governmental agencies, the greater the reported improvement between “before” and “after” decentralization

• The more “top-down” the decentralization process was, the smaller the extent of reported success with respect to the major objectives of basin management

The results of this group of variables suggest that government support is an important factor that has to be included at the right dose. On the other hand, the experience in the basins analyzed suggest that initiation of the decentralization process by governments is counter productive and ends in lower levels of reform performance. It therefore appears that supportive governmental involvement is good as long as it allows the stakeholders to initiate and lead the reform process.

BASIN-LEVEL INSTITUTIONAL ARRANGEMENTS. The set of variables includes local configurations such as: user group presence, budget sources and usage. Main findings suggest that:

• The greater the share of the RBO budget contributed by other sources, the smaller the extent of water user involvement and participation

• The larger the share of the RBO budget collected from basin stakeholders, the greater the number of institutions that were dismantled during the decentralization process; and the longer the decentralization process took

• The greater the presence of existing user groups in the basin, the greater the reported improvement between “before” and “after” decentralization

• The greater the share of the RBO budget spent within/returned to the basin, the greater the extent of reported improvement between “before” and “after” decentralization; the more likely it was that some form of institutional change was associated with the decentralization process; and the greater the extent of reported success with respect to the major objectives of basin management

• The greater the RBO budget per capita, the lower the extent of reported success with respect to the major objectives of basin management.

The overall results of this group of variables indicate the importance of the presence of water user organiza-tions in making the difference between pre and post decentralization reforms. Involvement of such groups, as other participatory processes, may make the process longer, but longer does not necessarily mean worse. The results also provide additional support to the old ‘mantra’ that budget that goes back to the source has a greater impact on stakeholder involvement and system performance—in this case the decentralization reform. One interesting result is that richer basins—this time measured in budget per capita—are not neces-sarily more successful. As with our earlier findings about natural resource endowments, success in river basin management is not necessarily confined to well-endowed basins. Our take from these results is that the RBO budget is an important tool for management, enhancement of participation and, if well designed and managed, could promote the decentralization process.

Conclusion

The results of the statistical estimates suggest that the process of decentralization and the performance level of the decentralization process, expressed by a variety of variables, are well explained by a set of explanatory variables. Several independent variables provide a robust explanation regardless of the relationship selected and the estimation procedure used. Our conclusions focus on several of the variables that worked well throughout the various sets of estimates.

Water scarcity is an important variable that affects the process as well as the performance of decentralization. As water in the basin is less abundant, incentives for a simpler decentralization process and a more successful outcome is more likely. Scarcity was positively associated with several aspects of the decentralization process and with the decentralization performance. The presence of scarcity may therefore be a stimulus to reform, uniting the stakeholders in the basin.

The number and severity of water resource problems (quality, allocation) present in a basin prior to decentralization was, perhaps surprisingly, a positive factor with respect to both the initiation of decentralization reforms and the level of performance.The more ambitious and comprehensive the decentralization effort was, and the greater the problems users faced, the more likely they were to see the effort as worthwhile and effective. This corroborates the widely held belief by practitioners that reforms can best be carried out in crisis situations.

The existence and number of organized user groups was positively associated with the initiation of decentralization reforms, but also with the costs and difficulty of achieving decentralization. Existence of dispute resolution mechanisms was positively associated with water user involvement and with decentralization performance. Length of the decentral-ization process was positively associated with perceptions of decentralization success and with tariff compliance and share of the RBO budget contributed by stakeholders. A decentralization process that was characterized by protracted political struggle leaves a negative impact on decentralization performance.

Dismantling institutions during the decentralization process contributes to the performance of the decentralization pro-cess. Combined with the two preceding findings, it appears that complexity and conflict are two distinct characteristics and work in opposite ways. The mere presence of a larger number of organizations within a river basin, and the length of time a decentralization reform takes, do not appear to be substantial negative factors. On the other hand, highly con-flictual decentralization processes are associated with poorer performance, and some elimination of previously existing institutional arrangements may be a positive factor. Thus what matters is not so much how complicated or lengthy the process is, but the degree of conflict and the ability to make organizational changes along the way.

River basins with higher percentages of their budgets from external governmental sources (such as the local and federal governments) benefit from better stability and support and it shows in the performance of the decentralization process, although the same relationship does not hold for the budget share contributed by other outside sources.

In basins where stakeholders accepted greater financial responsibility, complying with tariffs and contributing to the budget for basin management, decentralization process and performance measures increased. Combined with the pre-ceding finding, it appears that the financial dimensions of decentralized river basin management are both important and complex: success is associated with central government support as well as water user financial responsibility and with revenues generated within the basin remaining in the basin. Thus it is the combination of financial responsibility (on the part of water users), financial autonomy (basin revenues remaining in the basin), and central government support that is associated with success, and not necessarily one element alone.

56

57

Achieving Accountability through Decentralization:

Lessons for Integrated River Basin Management

Identifying the key factors that influence successful decentralization in river basin management from experiences in other sectors which include initial conditions, political leadership, subsidiarity, and allocation of property rights.

The central goal of decentralization is to improve service provision by making it more responsive to those being served. This note briefly outlines the goals of decentralization, the primary organizational models that have been adopted, a set of common challenges, and a number of issues that help frame the debate between centralization and decentralization. Then, to inform the process of productive decentralization in integrated river basin management, this note briefly examines some lessons learned from the decentralization experience in several other sectors (education, healthcare, roads, irrigation and water supply, electric power and anti-pov-erty programs).

EXPECTATIONS OF DECENTRALIZATION. Decentralization is a process of transitioning from a governance struc-ture in which power is concentrated at the central or national level to one in which the authority to make decisions and implement them is shifted to lower level governments or agencies (including parastatal organiza-tions). The resulting governing structure is anticipated to deliver public services more efficiently and equitably. Because of proximity to the locus of action, decentralization offers the prospect of lower transactions costs and the generation of information most relevant for serving the consumer of public services. As such, it is expected that decision-makers at decentralized levels may be held more directly accountable for the outcomes of their actions than an anonymous bureaucrat in the central government. In addition to accountability, success-ful decentralization depends on a number of other factors including negotiated voluntary arrangements, conflict resolution mechanisms, and the institutions necessary to support them.

FORMS OF DECENTRALIZATION. Organizational models adopted by a number of river basin organizations may be classified roughly as those that are large organizations which perform planning, regulatory, and operational functions involving infrastructure, and those that are small agencies concentrating on planning, policy, and coordination only. Administrative decentraliza-tion can take place within an organizational or in a socio-political context. Functions are said to be deconcentrated when some of them are shifted within an organization, as from the center to field offices at lower levels. In this case, either all authority is centralized or some flexibility is allowed so that field offices can adjust central plans to suit local conditions. Alternatively, the local government is responsible for delivery of services under central guidelines and supervi-sion. This is essentially an administrative restruc-turing wherein central control is maintained. A different kind of structure that aims to facilitate local coordination is an integrated model in which local bodies may plan and implement

programs but may not raise the necessary revenue to run them. In this instance, these bodies do not constitute an independent local government. They continue to receive policy and technical advice from the ministries.

Sometimes authority to make decisions is delegated to indepen-dent corporations or agencies who also manage specific func-tions. These may include revenue collection and allocation, and construction and management of infrastructure undertakings, such as dams, school buildings, and roads. When local communi-ties, in geographically delineated areas, manage their own neces-sities by electing or appointing local representatives who function without direct central control then authority is considered to be devolved to that level. Devolution is more complete when local governments may also raise revenue and decide on expenditures based on local needs.

COMMON CHALLENGES. Common challenges to many forms of decentralization include: (1) inadequate financing; (2) paucity in skills, particularly with respect to management and supervi-sion; (3) resistance from those who benefit from the central-ized structure; (4) how to sustain interest in the participatory process for the long term. Leadership is also critical to ensuring that administrative, political, and fiscal decentralization operate in tandem. Given these common challenges, this note examines some lessons from the decentralization experience in several other sectors (education, healthcare, roads, irrigation and water supply, electric power and anti-poverty programs) in order to draw lessons for productive decentralization in integrated river basin management,.

There are also a number of common themes associated with successful decentralization. First, the historical context and initial

conditions, including geographical and climatic features, existing patterns of service provision or traditional methods of addressing the user’s need, and economic conditions significantly influence the pattern of decentral-ization. Second, political leadership is key to overcoming the implementation hurdles that arise in the transition to decentralization.

Central Control versus Decentralization

The essential trade-off is between efficiency achieved through economies of scale that favors centralization and efficiency and equity benefits through local or decentralized decision-making. The following considerations are useful in framing the debate.

TECHNOLOGICAL ECONOMIES OF SCALE. These are a key justification for centralization. Such scale economies vary considerably by sector and within sectors by type of activity. There is also some evidence that scale econo-mies have declined recently in infrastructure sectors, like power and telephony, due to technological innovations thus weakening the case for central control.

58

59

TRANSACTION COSTS. Proponents argue that a key efficiency gain from decentralization is lower transaction costs. This may occur for a variety of reasons. Availability of better information at the local level with regard to feasible technologies and the specific needs of customers, where properly used, can improve efficiency. A related gain is greater accountability when responsibility for decision-making and implementation of programs is placed in the hands of local leaders. Participation at the local level can also improve skill levels and through this and other mechanisms enhance equity. In practice, the net effect on transaction costs may vary.

EQUITY. While decentralization of infrastructure and social sectors is popular in many areas of the world, there are some inherent obstacles in terms of implementation. If the institutional structure for local accountability is not already in place in developing countries and unless the local power structures are changed by some means, infrastructure delivery could be diverted to local elites who capture the government or agency.

ACCOUNTABILITY. Realizing accountability depends on the quality of supporting institutions both to enforce fulfillment of commitments to local beneficiaries, for conflict resolution, for financial accounting, and for tech-nological assistance and expertise.

CONFLICT RESOLUTION. In the context of river basin management conflicts arise among competing users of the water in rivers and in a decentralized management structure these have to be resolved through negotiation. One argument is that negotiation is effective when there are clearly defined property or priority rights. If this is not the case, then a higher authority would have to intervene to defuse the situation.

TECHNICAL EXPERTISE. Under centralization, large investments are made in central bureaucracies in research and development. Qualified people, interested in public service are drawn to these bureaucracies in the hope of superior career prospects, varied job opportunities, and better resources. With decentralization, the accu-mulated pool of expertise is dissipated because the roles of central bureaucracies are shrunk and there is job loss. It is unclear whether the associated loss in efficiency can be made up by local governments or even by the private sector given the smaller capacity and shortage of skills at the local level and the uncertainty of the private sector being able to mobilize the kind of investment that central governments can.

SERVICE PROVISION. Services, such as, power production and roads, requiring advanced technical expertise, man-agement and administrative skills are less suitable for efficient local production than, for example, solid waste management. Technological advancements allow vertical disintegration (unbundling) in each sector permitting some aspects of service provision to be local while others are central. Where there is greater potential for competition, there is less need for regulation. Therefore, factors affecting competition determine whether regu-lation in the sector is more efficient locally or not. The greater the economies of scale and dependence on expertise, the lower the potential for competition and the less likely that regulation can be handled locally. On the other hand, if information about local conditions and more direct monitoring are important in the sector then local regulation may be preferable.

FINANCING OF SERVICES. Generally and especially in developing countries, the tax base on which local govern-ments depend is insufficient to cover expenditures. This implies that local governments depend on transfers from the center. Central governments can use transfers to induce a level of service or to work jointly with local governments to take advantage of economies of scale. Transfers, however, do not enhance allocation efficiencies because they distort decision-making. The idea is to balance the power between the central and lower levels in such a manner as to retain a central government that is strong enough to perform its duties and at the same time to transfer authority to local governments which have the ability to use that power efficiently. In general, capital infrastructure investments and design, and auditing are more efficient under central control. Other functions that are more dependent on local information and subject to local accountability, such as, design selection and who should benefit from it, collection of fees, operation, maintenance and sometimes

monitoring and regulation, setting of standards and addressing complaints from users can all be decentralized with adequate training or through contracts with private providers wherever feasible.

Lessons Learned From Other Sectors Relevant To River Basins

Decentralization in its various forms—technological, regulatory, administrative, political, and fiscal—is being experienced in sectors such as electric power supply, roads, targeted assistance programs, health, education, and irrigation management. The twin objectives of improving efficiency in governance and increasing equity at local levels remain the unifying themes. To implement them the underlying principle common to all these sectors and to river basin management is that of subsidiarity, which argues that governance is more effective if the deci-sion-making authority is located where the pertinent knowledge exists and where decision-makers are directly responsible for the outcomes of actions taken to the community they serve. In general, capital infrastructure investments and design, and auditing are more efficient under central control. Other functions that are more dependent on local information and subject to local accountability, such as, design selection and who should benefit from it, collection of fees, operation, maintenance and sometimes monitoring and regulation, setting of standards and addressing complaints from users can all be decentralized with adequate training or through con-tracts with private providers wherever feasible.

The set of initial conditions shapes the seriousness with which decentralization objectives can be pursued; the political leadership displayed in implementing decentralization is key to building the necessary trust and ensur-ing that difficult decisions are legitimized; and the technical arrangements that generate accountability include effective subsidiarity, transparency of objectives and responsibilities, and the allocation of property rights. First, it appears that among initial conditions the physical characteristics of the area, such as the extent of aridity, are not necessarily consequential. Second, implementing the principle of subsidiarity is essential if decentralization is to achieve successful management outcomes. Third, integrated river basin management faces significant hurdles when there is no transparency.

EFFICIENCY AND EQUITY. The following aspects of initial conditions and history influence the efficiency and equity outcomes of decentralization. First, gains from lower transactions costs when decisions are based on better local information are more likely in a large country setting rather than in a small one. In large federal systems such as in the United States and Brazil, considerable authority is devolved to the individual states, however, this seems counterproductive in small countries where regions or districts are not likely to have the necessary tax base (insufficient population and economic activity especially in agrarian economies) to support public services without dependence on central transfers. Decentralization is in fact, desirable in large countries because under central control districts are too remote from the center for adequate financial accountability in terms of how transferred funds have been spent. Further, the incentive to utilize local information is greater for decision mak-ers at the lower levels than it is at the center. The issue of sufficient size (in an economic sense) is relevant for river basin management because the hierarchy of agencies that are set up may also have to depend on revenues and/or compete with other sectors for financial resources. Second, history plays an important role in shaping the decentralization reform in the water resources sector. Third, conditions that encourage cooperation among small groups are crucial. Fourth, initial conditions also imply existing legislation, especially relating to property and water rights and institutions such as effective voting mechanisms at the local level. Fifth, the issue of capture is directly linked to the extent of economic inequality existing before the decentralization reform.

SUCCESSFUL DECENTRALIZATION. Successful decentralization depends on the regions or districts concerned having a large enough revenue base to undertake activities; it requires the existence of clearly defined property and water rights among potential stakeholders (for effective negotiation and commitment to operations and maintenance); it requires the socio-political pre-conditions for participation and the resistance of elite capture (e.g., unstratified social structures); and demands transparency, clear roles and responsibilities with legal authority,

60

61

and quality information. For those anticipating that decentralization might solve problems associated with weak centers and fiscal constraints, it is salutary to learn that decentralization works bests in the context of strong central government and economically prosperous regions with a high education and skills base. The case studies examined reveal the serious problems that occur when regulatory, administrative, political or fiscal mechanisms have not evolved sufficiently to support decentralization; or where skills accumulated in central bureaucracies are lost through retrenchment. More intractable still are factors of historical context and political leadership. Many of the factors affecting degrees of elite capture in decentralization, such as land rights, power inequalities, and political resistance, are not easily amenable to policy interventions.

HIGH PRIORITY ISSUES. The literature survey reveals that some debates, such as whether professional expertise is neces-sary to maintain minimum or desired standards in education, or of geographic boundaries in health system organiza-tion, are sector specific. However, others, such as the need for political and financial accountability, the related data require-ments, and educating stakeholders and potential beneficiaries of the new system and how they can participate in it, are true of decentralization wherever it is to unfold. In particular, this note identifies four issues to be tackled that demand high priority. These are (1) devising ways to overcome financial inadequacy at the lower level; (2) making a commitment to incorporating opportunities to upgrade skills, particularly management skills, when designing programs while also ensuring that the expertise accumulated in central bureaucracies is not dissipated; simultane-ously encouraging those facing retrench-ment to contribute to the new systems wherever feasible; (3) assuring beneficiaries of the pre-reform structures that their rights would be protected; and (4) planning to sustain a long-term commitment to the decentralization process as it is likely to be slow and drawn out, perhaps by demonstrating positive outcomes in a key element of the sector in question.

FINANCE. If revenue decentralization is not devolved to local governments along with authority over expen-diture choices then the volume of service delivery is adjusted downward whether the local government is accountable (that is, captured) or not. Financing through central grants does not devolve fiscal authority to local governments and is therefore less efficient compared to charging user fees. With central grants, if the funds are insufficient to provide service to everyone then either provision favors elites depending on extent of capture by them or with a severe financing constraint both elites and others suffer from a lack of infrastructure services. Central grants may be qualified as unrestricted, tied to expenditures, or matched to particular services. Again, their success with respect to service delivery will depend on how much flexibility they allow to respond to local needs and if at the same time they cannot be easily allocated elsewhere.

ConclusionThe main conclusions of the literature survey come as a caution to anyone who believes that decentralization is, in itself, a solution to problems of inefficiency and inequity in developing countries. The report shows that the case for decentraliza-tion as against central control is not itself unambiguous. Central control offers distinct advantages in certain circumstances; there are trade-offs and tensions to be reconciled (e.g., economies of scale versus local monitoring; integrated manage-ment or inter-regional equity versus local control). Nonetheless, decentralization (in its various manifestations) holds out the promise of reduced financial and transaction costs, increased flexibility and efficiency, local control and accountability.

The successful implementation of decentralization policies for all sectors generally requires four key issues to be tackled with high priority. These are:

1. devising ways to overcome financial inadequacy at the lower level;

2. making a commitment to incorporating opportunities to upgrade skills, particularly management skills, when designing programs while also ensuring that the expertise accumulated in central bureaucracies is not dissipated; simultaneously encouraging those facing retrenchment to contribute to the new systems wherever feasible;

3. assuring beneficiaries of the pre-reform structures that their rights would be protected;

4. planning to sustain a long-term commitment to the decentralization process as it is likely to be slow and drawn out, perhaps by demonstrating positive outcomes in a key element of the sector in question.

The experience of decentralization from a range of sectors therefore provides highly relevant lessons for integrated river basin management through decentralization. These lessons constitute the nitty-gritty details of the broader criteria for success identified in the context of river basins, namely, initial conditions, political leadership, subsidiarity, transparency, and property rights.

62

63

64

Photo Credits:

Cover: Main photo: Tran Thi Hoa

Small photos: clockwise from bottom left: Curt Carnemark, Karin Kemper and José Manuel Alonso Puchardes

Page 4: Curt Carnemark

Page 7, 15, 33, 47, 53, 57, 58: Karin Kemper

Page 24, 61: Fraser Basin Council

Page 36: Andrzej Tonderski

Page 42, 54: José Manuel Alonso Puchardes

Page 51, 58: Murray Darling Basin Commission

The World Bank

1818 H Street, NWWashington, DC 20433USAwww.worldbank.org/riverbasinmanagement

Prin

ted

on r

ecyc

led

pape

r w

ith s

oy b

ased

inks

May 2005