Institutional Adaptation to Changing Risk of Water Scarcity in the Lower Guadalquivir Basin

C. GIANSANTE,* M. AGUILAR,* L. BABIANO,** A. GARRIDO,*** A. G6MEZ,*** E. IGLESIAS,*** W. LISE,*** L. MORAL,* B. PEDREGAL* Institutional Adaptation to Changing Risk of Water Scarcity in the Lower Guadalquivir Basin* ABSTRACT Historically, the Spanish water management model's predominant goal has been resource augmentation. This mindset has had important consequences for the system's capacity to cope with droughts. It has impacted the system's overall vulnerability, the discourse of scarcity, the conceptualisation of risk, and the stakeholders' interests and their approach to risk. The aim of this article is to present the traditional hydraulic paradigm, and its current crisis and implications for present and fiuture risk management, and to explore stakeholder and institutional reactions and adaptation to changing risk scenarios. The adaptation process will be framed within the wider context of macro-trends, such as marketisation and re-scaling of institutions and global warming. INTRODUCTION TO DROUGHT AND SPANISH SOCIETY This article will explore the management of water scarcity and drought in the Lower Guadalquivir Basin within the broader framework of the societal and institutional responses to climate change and to the increased possibility of extreme hydrological events in Europe. The first * Department of Human Geography, Facultad de Geograffa e Historia, University of Seville, Spain; e-mail: [email protected]. ** Department of Anthropology, Facultad de Geografia e Historia, University of Seville, Spain. *** Department of Agricultural Economics, E.T.S. Ingenieros Agr6nomos, Universidad Politcnica de Madrid, Spain. * Institute for Environmental Studies (IES), Vrije Universiteit, Amsterdam * This paper is part of the research project Societal and Institutional Responses to Climate Change and Climatic Hazards: Managing Flood and Drought Risks (SIRCH), which has been funded by the Environment and Cliamte Change Research Programme of the European Commission (Contract no. ENV4-CT97-0447). Due to the uniquely European location of some of the sources cited in this article, the Natural Resources Journal was not in a position to verify some of the sources cited. Nonetheless, it is thought that the information and perspectives will be of interest to the reader and of importance in the ongoing dialogue regarding transboundary resources. For questions regarding the articles or the source materials, please contact the authors. NATURAL RESOURCES JOURNAL important factor to note is the marked spatial and temporal climatic variability of the Spanish climate. In Spain, the traditional water management model, "hydraulic paradigm," has been resource-oriented. 1 As a result, the overall spatial development model of the country, including agricultural, urban, and industrial sectors, is based on the social reconstruction of the water environment.2 This model has had important consequences for the water system's capacity to cope with risk. It has impacted the system's overall vulnerability, the discourse of scarcity, the conceptualisation of risk, and the stakeholders' interests and approaches to risk. The system has become more vulnerable or, at best, maintained its level of vulnerability after each drought episode in the past few decades. 3 The current strategy, which favours resource augmentation, has caused the increased vulnerability. Encouraged by the misplaced perception of greater security, abundance, and access, water "demand" has risen as additional water supply has been made available. 4 The socio-economic impacts of the 1990's drought demonstrate the strain on the water system. Several million urban Andalusian consumers experienced service cuts of a maximum of 12 hours per day and thousands of irrigation farmers went without any irrigation water for three consecutive years. 5 The lack of irrigation water caused direct losses of 3 to 4.2 billion Euros and also caused losses to the 20,000 farm workers who lost their jobs. 6 Due, in part, to the water crisis brought on by the 1990's drought, current water management practices have been intensely debated over the last decade. Furthermore, the idea that the water system is in crisis is gaining acceptance among both water managers and stakeholders. 7 But, the traditional paradigm still has institutional inertia and social support behind it. The tension between the view that the water system is in crisis and the institutional inertia and steadfast social support has resulted in an "unstable stagnation" in Spain's water policy.' This article examines the current water management paradigm's ability, to cope with present and future risks and explores the process of 1. Leandro del Moral & David Saurf, Changing Course: Water Policy in Spain, THE ENVIRONMENT, 12-15, July-Aug. 1999. 2. For the definition of social construction of the water environment, see Erik Swyngedouw, Modernity and Hybridity: The Production of Nature: Water and Modernisation in Spain 97-112 (1997) (unpublished report, on file with authors). 3. Monica Aguilar et al., Drought Risk in the Lower Guadalquivir Basin: Social and Institutional Responses to Climate Change and Climatic Hazards, in SIRCH PROJECT FINAL REPORT (Feb. 2001). 4. Id. 5. EMASESA, CRONICA DE UNA SEQUJA 1992-1995 49-50 (1997). 6. Id. at 105. 7. Aguilar et al., supra note 3, at 5.1. 8. Id. at ch. 3. [Vol. 42 INSTITUTIONAL ADAPTATION institutional adaptation to changing risk scenarios. We tested the traditional paradigm's ability to adapt by modelling different risk scenarios. These scenarios incorporated the possible impacts of climate change on the overall availability of water resources and on the frequency of extreme events. 9 They also incorporate other macro trends that generate future uncertainty."° THE INSTITUTIONAL LANDSCAPE Institutions, previously defined as regularised patterns of behaviour that can be formal, informal, strategic, or operational," are difficult to capture because of the difficulty of defining clear boundaries and separating different operational levels at the global, national, regional, river basin, and local levels.' 2 Therefore, the description of water management institutions in this section will not be systematic. Instead, we will provide an institutional landscape upon which it will be possible to observe the process of institutional adaptation and change. At the formal, legal institutional level, the cornerstone of Spanish water policy is the 1985 Water Act, which established the current water management framework and conferred a key role to hydrological planning. 3 The two main aspects of this legislation are the water planning framework and the financial and economic public water framework. The Water Planning Framework The main instruments of water management and planning established by the 1985 Water Act are the national water plan, Plan Hidrol6gico Nacional (PHN), and the river basin water plans, Planes Hidrol6gicos de Cuenca (PHCs), for Spain's 17 river basins. 4 Through the technical services of the Ministry of Environment, Department for Water 9. The Spanish climate change debate is still in an embryonic stage, but, if it has merit, its implications for Spain's water system could be enormous. Francisco J. Ayala-Carcedo, De la politica hidrdulica a la politica de agues sostenible, 90 TEcNOAMBIENTE 5-9 (1999). 10. Leandro del Moral et al., Global trends and management of climatic water risks, in SIRCH FINAL REPORT, supra note 8, at ch. 14. 11. This means that institutions/rules of behaviour can influence the general strategy/perspective that should be adopted to tackle a problem (in this particular case, natural resources management) or, alternatively, can mandate the operational rules that should be applied. 12. See generally SIRCH, Societal and Institutional Responses to Climate Change and Climatic Hazards: Managing Changing Flood and Drought Risk (Karen Bakker ed., 1998). Institutions can also be described as regularised patters of behaviour, which are socially constructed, both formal and informal, and relatively stable. 13. Ley De Aguas, B.O.E., 1985, 189. 14. Id. art. 38.2, 71.1. Summer 2002] NATURAL RESOURCES JOURNAL and Coasts, the central government is in charge of drawing up the national hydrological plan. The river basin water plans (PHCs) are drafted by the river basin authorities. According to water legislation, the main goals of these documents are to forecast future water demands, to design plans to meet those demands (generally ten and twenty-year planning horizons are used), as well as to establish qualitative objectives for the basin's surface water. 15 The legal framework also asserts that Spain's water policy should coordinate with other planning policies that may affect hydrological plans." The other major planning policies that affect water management plans are agricultural policy 7 and energy policy." 8 The legal framework also requires that national water policy not only coordinate with but must be compatible with environmental conservation policies. 9 The process of drafting, discussing, and approving national and regional river basin water plans is an extremely long, complex, and continuous process. A first draft of the national water plan was presented in April 1993.20 This draft was premised on the notion that the national hydrological imbalance between water-rich and water-poor regions should be corrected and proposed re-distributing water resources between basins via a national water grid, Sistema Integrado de Equilibrio Hidraulico Nacional (SIEHNA). 1 The 1993 draft also projected considerable increases in water demand based on a projected increase in irrigation surface of 600,000 ha by 2012. 22 This draft was widely criticized by various administrative sectors and by the public.' In addition to the numerous and varied environmental objections, the Ministry of Finance criticised the plan for its lack of financial rigour and for its failure to coordinate with related sectors. 24 Agricultural 15. Real Decreto de 29 de Julio, art.79.1, R.C.L., 1988,927. 16. Id. art. 95; Ley De Aguas, B.O.E., 1985, 189, art. 38.4. 17. Through the regional competence within the framework of the national economic objectives and the EU Common Agricultural Policy (CAP). 18. The use of large volumes of water for energy production purposes (hydropower dams and cooling systems) can generate conflicts with other use sectors, such as irrigation and urban supply. 19. Ley De Aguas, B.O.E., 1985, 189, art. 40(d). 20. MINISTERIO DE OBRAS POBLICAS Y TRANSPORTES, PLAN HIDROLOGIcO NACIONAL MEMORIA (1993). 21. Id. at 100-10. 22. 46 percent in domestic supply, 14 percent in irrigation agriculture and 25 percent in industry for the year 2012. 23. Ministerio de Medio Ambiente, Libro Blanco del Agua 19-20, at http://www.mma.es/rechid/librojb/sintesis.pdf (Dec. 4, 1998);VICrOR P aR-DIAZ ET AL., POLITICA Y ECONOMIA DEL AGUA EN ESPA4A 42-51 (1996). 24. PtREz-DIAZ Er AL., supra note 23, at 57. [Vol. 42 INSTITUTIONAL ADAPTATION experts, who criticized the Plan, were also concerned because this draft had completely overlooked various macro-economic factors such as the General Agreement on Tariffs and Trade (GATT) treaty.' At the same time, regional governments in "water deficit" river basins were clamouring for higher transfer volumes. 26 Meanwhile, outcry in water-rich river basins against the transfer plan grew. In fact, one of the "water-rich" regions, Arag6n, where most of the reservoirs of the Pyrenean mountainous region for the nationwide water transfer system would be located, has campaigned vigorously against the PHN project. As a result of the extensive criticism, the Ministry of Public Works and Transportation prepared a scenarios analysis document that slightly modified the criteria for future water demand scenarios while maintaining the general priorities of the 1993 draft. 27 Most importantly, the Spanish Parliament decided that approval of the PHN should be conditioned on the approval of both a national irrigation plan and on all the river basins' hydrological plans.' The years following the publication of these two official documents were characterised by instability and stagnation. During this period, which coincided with the long and severe 1991-1995 drought, the water debate became a high profile political issue.' The planning process had reached an impasse. 3 The main opposition party at the time, the Partido Popular (PP), blamed the planning impasse on the PSOE's 3' inappropriate water management approach?2 In April 1996, a new government, led by the conservative opposition party, Partido Popular, was elected with high expectations.' Representatives of the newly created Ministry of the Environment, MIMAM, promised rapid approval of the PHN during their term. The mechanism for the approval was supposed to be enhanced 25. Id. 26. See, for example, discussion of the Andalucia regional government's demands for higher transfer volumes in Joan Corominar, El Plan Hidrol6gico Nacional en Andalucia y la Directive Marco sobre politica deaguas in NUNOGRANDE, PEDRO ARROJO, &JAViERMARTINEZGIL, IL CONGRESO IeugIco SOBRE PLANIFICAION Y GESTION DE AGUAS 313-18 (2000), available at www.us.es/ciberico. 27. SECRETARIA DE ESTADO DE POLfTcA TERRITORIAL Y OBRAS POBUCAS, MINsrKO DE OBRAS POBuCAS, TRANSPORTEYMEDIOAMBIENTE, PLAN HIDROLOGIcONACIONAL, ANALISISDE EscENARIos (1994). This document maintained the notion of national hydrological imbalance, SIEHNA, and the priority given to large-scale infrastructure versus demand management and control and water saving. 28. Ministerio de Medio Ambiente, supra note 23, at 410. 29. VICTOR Pftz-DfAz ET AL, supra note 23, at 59-67. 30. Id. at 69-71. 31. The PSOE is the Spanish socialist party. 32. Borrel modifica el Plan Hidrol6gico ante las criticas at Anteproyecto: El PP exige 1a retirada del proyecto, EL PAIS, Mar. 23, 1994. 33. El gobierno desliza en el BOE sin debate un miniplan hidrolfgico de 216.000 millones, EL PAIS, Sept. 8, 1998, at 22. Summer 20021 NATURAL RESOURCES JOURNAL dialogue.' The enhanced dialogue was to be supported by a white paper on water. The white paper would be designed to reach consensus on what criteria and priorities should guide the subsequent elaboration of water plans.' But, at the same time, the Partido Popular was arguing that the very nature of the water-planning framework needed to become more flexible to allow for more room for economic management of the resource. 6 The Partido Popular never implemented the former approach.' In August 1998, before the white paper on water was published' and before the national irrigation plan was adopted," the government approved each and every river basin water plan.' As approved, the plans were virtually the same as the drafts from the early 1990s. 1 ' The new government justified its surprising decision by characterizing the water plans as non-binding lists of possible projects.' 2 This characterization implied a significant change in the role of hydrological planning, a change unsupported by any legislative reform. A few months later, in February 2000, the Ministry of Agriculture, Food and Fisheries, MAPYA, tabled the national irrigation plan after a brief round of informal contacts and consultation with the regions' and users' associations.' Since the water basin plans were approved, there has been a striking decrease in planned irrigation surface: the 1998 river basin water plans envisioned 1,200,000 ha of new irrigation surface, but the national irrigation plan of 2000 reduced this to 250,000 ha of new irrigation surface 34. Borrel modifica el Plan Hidroldgico ante las crfticas al Anteproyecto: El PP exige la retirada del proyecto, supra note 32. 35. Id. 36. Id. 37. Id. 38. The White Paper on Water, that was only released by the Ministry of the Environment (MIMAM) at the end of the same year (December 1998) after an extensive conceptual and historical revision, still held as valid the high projections for future irrigation water demand and surfaces included in previous water planning documents. In addition, it failed to provoke the announced open and public debate, especially because it came too late to be really relevant in the discussion on the river basin water plans. See generally MINSTERIO DE MEDIO AMBIENTE, LIBRO BLANCO DEL AGUA EN ESPAIRA (1998). 39. Actually, a first National Irrigation Plan was tabled by the former government in February 1996 and did not become operational due to the change in government in spring of 1996. 40. Real Decreto 1664, July 24,1998. 41. They had been previously discussed and endorsed by the National Water Council (CNA) in April 1998, with the only negative votes from the representatives of the environmental organisations. 42. Personal communication with Josd Luis Blanco, Water State Secretary (Mar. 1999). 43. This document was meant to comply with the March 22,1994, decision of the Spanish Parliament, which urged the government to present a National Irrigation Plan before the National Hydrological Plan could be approved, although the discussion in Parliament occurred after the approval of the river basin water plans. [Vol. 42 INSTITUTIONAL ADAPTATION by 2008." Furthermore, the new 250,000 ha figure includes 138,365 ha of previously planned and partially implemented irrigation surfaces (regadios en ejecuci6n); 79,426 ha of irrigation surfaces justified on social grounds (regadfos socials); and 25,000 ha of private initiative irrigation surfaces." The Ministry of Environment tabled a new draft of the PHN in September 2000. This draft had substantially reduced the planned volume of inter-basin transfers and concentrated the water transfer burden on one "donor" basin, the Ebro river basin.' The .1993 draft of the national water plan had originally envisioned anation-wide inter-basin transfer network that would have transferred a total of 3768 Mm 3. The 2000 PHN draft eliminates transfers from the Tagus and Douro basins, which are shared with Portugal, to Mediterranean basins. 47 The only transfer the 2000 PHN draft contemplates would be from the Ebro river basin to the Mediterranean basins. 4 According to the terms of the 2000 PHN, the Ebro basin would have to transfer 200 Mm 3 to the Catalonia basin, 400 Mm 3 to the Segura basin, 300 Mm 3 to the Jitcar basin, and 100 Mm 3 to the Sur basin, most notably to the Almeria irrigation surfaces. 49 The total transferred volume in the 2000 PHN is just over half of the transfer volume set out in the 1993 draft.' Even so, the proposal has reopened old inter-regional wounds and has aroused considerable political opposition from the regional government of Arag6n, the Ebro basin's largest region." The Economic Framework of Spain's Water Policy The economic framework of Spain's water policy has traditionally been based on two main features: a generous subsidy system and a water rights regime. The subsidy system has allowed the proliferation of extensive state-funded water regulation schemes and the water rights regime requires 44. Ministry of Agriculture, Food and Fisheries, Plan Nacional de Regadfos 2000, at http://www.mapya.es/indexa/busca.asp?sitio=http%3a2f%2fwww2emapya%2ees%2f desarr%2fpags%2fpnr%2frega%2ehtm (last visited Oct. 28,2002). 45. Id. 46. Ministry of Environment, 2000, at http://www.mma.es/rec-hid/plan._hidro/ planhidro.htm. 47. Id. 48. Id. 49. Id. 50. The 1993 draft planned on transferring 1855 Mm3, 1.5 million acre-feet, of water to water-poor basins, see MINISrERUO DE OBRAS PO UCAS Y TRANSPORTE, supra note 20, the new draft only plans to transfer 1000 Mm 3, 0.81 million acre-feet, of water solely from the Ebro basin. See Ministry of Environment, supra note 46, at 102-08. 51. Arag6n lider6 el "no" al transvase que inundd la Puerta del Sol, EL HERALDO DE ARAGON, Mar. 12,2001, available at http://www.heraldo.es. Summer 2002] NATURAL RESOURCES JOURNAL [Vol. 42 all users of the public water domain 52 to apply for a license to use water that governs the terms of that use. The river basin authority' can grant licenses for a maximum of 75 years and can only grant licenses for the uses set out in the licensing document.5 Both the subsidizing and the licensing system have been debated and partially revised in recent years. 55 Traditionally, Spanish economic water regulation has been in line with the principles of flexible supply and the belief that generous state subsidies were necessary to overcome the serious economic and social deficits affecting Spain at the beginning of this century.' The 1985 Water Act established storage charges and rates, the canon de regulaci6n y tarifas. 57 The income from these fees is designed to help the government recover the cost of its investments on dams and other hydrological works, mainly canals,' as well as to cover their operation and maintenance costs s9 Despite attempts to revise the old system of subsidies, the 1985 Water Act only succeeded in maintaining it; the end result of the charges and rates is to subsidise bulk water by about 40 percent of total investment.' In general, these charges have been hard to apply and difficult to collect. 61 Furthermore, the funds they generate are manifestly insufficient 52. The 1985 Water Act established that all renewable freshwater--both surface and groundwater-is part of the public domain. 53. The river basin authority grants licences based on a general order of preference established by the Water Law: public supply (including industries connected to municipal supply networks); irrigation and agricultural uses; electric supply; other industrial uses; fish farming; recreational uses; navigation and river transport; and other uses. 54. Ley De Aguas, B.O.E., 1985,189, art. 57.4. 55. 1999 Water Act, B.O.E., 1999,298. 56. The cornerstone of this policy was the Large Hydraulic Works Act of 7 July 1911, which defined four possible avenues for carrying out hydraulic works: (1) execution by the private sector, with a state subsidy of up to 30 percent of the cost of large-scale hydraulic works in addition to a premium payment (by the state) "per litre per second" for irrigation water used; (2) execution by associations of irrigators, with a subsidy of up to 50 percent of the cost; (3) execution by the state, with the support of the irrigators affected, with a subsidy of up to 50 percent and low interest loans of up to 40 percent (most common choice); (4) executed entirely by the state. 57. In addition, the 1985 Water Act also established the water domain occupation charge (candn de occupacidn) for the physical occupation of parts of the water domain (river beds and banks) and the water pollution charge (candn de vertido) for the protection of those watercourses receiving polluting discharges. 58. Ley De Aguas, B.O.E., 1985, 189, art. 106.1 (canon de regulaci6n). 59. Id., art. 106.2 (tarifas de utilizaci6n). 60. All water facilities supplied by waterworks (dams, channels, etc.) are built and run with public subsidies. This subsidy rises to 90 percent when the conditions of the Act of 1911 are applied. "Bulk water" refers to all water that is transported in large-scale networks, before it enters the end-user distribution networks. In the case of drinking water, this is before it is treated for drinking water purposes. 61. Ministry of the Environment, 1998, supra note 38, at 556. INSTITUTIONAL ADAPTATION to cover the amortized cost of investment in and operation of hydraulic systems. 2 In addition to the charges applied to bulk water, urban and agricultural end-users have to pay operation management and distribution costs to water companies or to irrigation communities.' In most irrigated zones, water prices are not assessed on the basis of the volume of water used, but as a per hectare charge.6' Rates range from 6 to 198 $/ha.1 Due to this variability in prices in each geographical area and the fact that each farmer applies different volumes of water per hectare, the average price of water, including bulk water and irrigation association charges, is difficult to calculate. Based on the average consumption of about 7000 m3 /ha, a rough estimate of the price per cubic meter is 0.0158 $/m3 (0,0167 Euro/m); however, this figure varies with crop differences.' In the various drafts of the amendment to the 1985 Water Act (1997-1999), several proposals were made to modify different aspects of the water tariff system. 6 ' But the only change in the final text of the amendment approved in 1999 was one that introduced the obligation to meter water and a system of incentives and penalties for users that deviate from a baseline volume of water use.' The opposition of business organisations, agricultural associations, and particularly of powerful hydropower interest groups prevented the introduction of more radical changes, such as full-cost recovery or a non-renewable resource/environmental impact surcharge for water use.' While the tariff system has not significantly changed, the mechanisms for financing public hydraulic works are likely to be affected 62. Ministry of the Environment, 1998, supra note 38, at 557. The amount is fixed and updated annually, taking into account operation costs, 4 percent of the value of investments made by the State-the technical amortisation of the works and installations--and the depreciation of the currency. The total period of technical amortisation for investments is set at 50 years for reservoirs and at 25 years for canals. 63. These costs, and the way they are calculated, vary greatly. In domestic uses, for instance, prices vary greatly from one place to another, depending on several factors: water municipal policy, bodies in charge of administration, costs of the service, level of waste water treatment, and others. 64. Jos9 MARIA SUMPSI E" AL., ECONOMIA Y POLICA DE GESTION DEL AGUA EN LA AGRICULTURA, at 148, 148-51 (1998). 65. Id. 66. Id. 67. Consuelo Giansante et al., L'" olution des modalits d'allocation de la resource en eau en Espagne, 48 REVUE D'ECONOMIE MERIDIONALE (REM) 235, 235-48 (2000). 68. Id. at 240. 69. So far the prices paid for water use are only aimed at recovering-partially or completely-the costs of provision, transportation, or treatment, but never to pay for the resource itself; the access to which, though regulated by a water right system, is considered to be free of charge. Summer 2002] NATURAL RESOURCES JOURNAL responsible for assisting the river basin authorities with the construction and management of public water works."9 The creation of such companies at the river basin level is based on a contract (convenio) between the state administration and the mixed capital company 20 The contract governs construction and management conditions for waterworks as well as financial contributions made by the state and private sector. Several of these mixed capital companies have already been created; for example, the Aguas de la Cuenca del Guadalquivir S.A.' 2 was set up in the Guadalquivir basin in 1999. This legislation reflects the impact of global changes that will be explored further in the next section. THE WIDER CONTEXT OF CHANGE The ongoing and likely future changes in the drought management practices in the Guadalquivir basin are embedded in the general processes occurring at the global scale, the macro trends that affect the basin in different ways." Out of the general landscape of change, five macro trends that are particularly relevant for the field of water management can be identified. The first macro trend, which is observable both in Spain as well as in other countries, is a building atmosphere that favours solutions that incorporate "scarcity indicators," market instruments, and gives private business a larger role, what could generally be defined as marketisation.'" These solutions include economic analysis of water uses and full cost recovery, both of which would change the current concept of water deficit and would also change how drought coping mechanisms are prioritised." These solutions have important consequences for infrastructure, which is 119. Art.158, Law 13/96 (auxiliary regulation of the General State Budget of 1997) of 1996; Karen Bakker, From state to market?: Water mercantilizacid6n in Spain, 34 ENV'T & PLAN. A 767, 777-79 (2002). 120. BAKKER ET AL., supra note 82, at 8. 121. The statutes of the company establish as one of its objectives the "management of works and water resources, including environmental management of aquifers, lagoons, reservoirs, rivers and stretches of river, as well as the preparatory or complementary activities, derived from the former ones," thus opening the field of water management to private regulation. 122. Following the structural theory of A. Giddens, locale is a consubstantial factor in the constitution of social action. It is not just a mere spatial parameter, a physical environment where interactions happens, it is formed by all the elements mobilized as part of the interaction. See generally del Moral et al., supra note 10. 123. Bakker, supra note 119. 124. Arrojo, supra note 106. [Vol. 42 INSTITUTIONAL ADAPTATION presently stagnating due to financial cutbacks." 2 At the same time, market instruments are gaining legitimacy, although the Spanish people still have strong objections to treating water just like any other commodity. 26 As a second general trend, decision-making powers are being transferred in two different directions. They are, moving towards the regional scale and towards the international scale. This process has been defined as re-scalingY1' The increasing political role of the regions makes water policy a flashpoint for confrontation between regional and central governments." l Meanwhile, increasing international interaction is also impacting Spanish water policy. First, the global trend towards liberalising the economy affects agriculture and water policy decisions and has consequences for the local economy.' 29 At the same time, international political power centres like the European Union have become key local actors." 0 The EU has two primary watermanagement responsibilities. Its first task is to grant subsidies that partially cover the costs of large water works."' But it has also made a general commitment to sustainable use of natural resources and is responsible for protecting the environments of designated sites that may be adversely affected by such waterworks. 2 125. The reduced willingness and ability of the state to fund waterworks is part of a general trend to reduce public sector spending and to de-invest in strategic sectors. Bakker, supra note 119. 126. Rechazo undnime a la reforma de la Ley de Aguas propuesta por Isabel Tocino, LA TIERRA, July-Aug. 1997, at 22; Diario de Sessiones del Congreso de Diputados, 17 junio 1999 (no. 247: 13157-13176); 30 de septiembre 1999 (no. 263:13988-14003). 127. E. Swyngedouw, Neither global nor local: globalisation and the politics of scale, in THE GLOBAL AND THE LOCAL: MAKING THE CONNECTION (K. Cox ed., 1997). 128. This is exemplified by the contrast between the expansive approach of the river basin water plans-where regional governments and local stakeholders are most influential--and the conservative national irrigation plan, which is more conditioned by the logic of market liberalisation. 129. Sumpsi, supra note 117. 130. Asger Olsen, The New Water Framework Directive for the European Union-Main principles and obligations. Prospects for a sustainable water policy for the coming decades, in O'PORTO, UNA CITA EUROPEA CON LA NUEVA CULTURA DEL AGUA: LA DIRECTIVA MARCO, 25 (2001). 131. The economic and social cohesion between the regions of the European Community was one of the objectives included in the Treaty of Rome in 1986. As a complement to the Treaty's single market, opening up frontiers would expose the most disadvantaged countries to further problems. This is pursued, among other instruments, through European Regional Development Funds (ERDF)-75 percent EU and 25 percent member state contributions-which are limited to less-favoured regions (such as Andalusia) and are focused mainly on productive investments and infrastructure. 132. Article 6 of the Treaty of Amsterdam of the European Union states that environmental protection requirements must be integrated into the definition and implementation of community policies. Consequently, the European Commission must ensure that projects developed under regional policy are respectful of the environment. An assessment of their Summer 2002] NATURAL RESOURCES JOURNAL The third macro trend relates to ongoing, worldwide changes in how nature is represented, particularly representations of the cultural and symbolic value of water.' 33 These changes are crucial in Spain and in the Guadalquivir basin because these values are the cornerstones of the traditional hydraulic paradigm. TM The traditional paradigm is founded on a system of deeply rooted values about the relationship between nature and society, 35 which views the natural water environment as a hostile setting whose splendour and beauty is enhanced by human intervention. 1"M Another cornerstone of the traditional hydraulic paradigm is the notion that nature is an obstacle to social and economic progress and must be transformed to achieve modemisation.'1 Hence the fascination with interbasin water transfers; the traditional view sees these transfers as-a way to progress by overcoming hydrological imbalances, a natural obstacle to socio-economic modernisation." The fourth macro trend is increased participation in the water policy arena.' In particular, the transition to democracy (in 1977-1978) had major impacts on the political, social, and institutional characteristics of Spanish life." 4 While this trend has allowed a greater number and variety of actors to feel involved in the water debate, the inclusion/exclusion process is dynamic and some stakeholders continue to be excluded. 4' Finally, there is an observable attempt to better coordinate different sector and governmental department policies-at least at the discursive level, both in Spain and in Europe as a whole.Y For example, the need to environmental impact must be conducted by the member states concerned. Additionally, DGXI (Environment) of the European Union is responsible for the drafting and implementation of a number of Directives on environmental protection and water, such as the Water Framework Directive. 133. Del Moral, supra note 10, at 4, 12. 134. Id. 135. Babiano, supra note 98, at 11-13. 136. This is how human-built water landscapes associated with irrigation (orchards and domesticated water) come to be valued as ideal and positive images of development, overriding the possible negative impacts on the wider natural environment. 137. E. Swyngedouw, Modernityand Hybridity: Regeneracionismo, the Production ofNature and the Spanish Waterscape, 1890-1930, 89 ANNALS ASS'N AM. GEOG. 443, 443-65 (1999); Babiano, supra note 98. 138. Id. 139. Del Moral, supra note 10, at 5, 19. 140. RAYMOND CARR & JUAN PABLO FUSI, ESPARIA: DE LA DICTADURA A LA DEMOCRACIA (1979); VicToR P9REZ DlAz, LA PRIMAcIA DE LA SOCIEDAD CIVIL (1993) 141. BAKKER zr AL., supra note 82, at 27. 142. At the European level, a number of initiatives are underway aimed at the coordination of environmental protection with other development and sectoral policies. See EUROPEAN CONSULTATIVE FORUM ON ENVIRONMENT & SUSTAINABLE DEv., THE EUROPEAN SPATIAL PERsPECnIvE(1999),at http://www.europa.eu.int/conmm./regional-policy/sources/docoffic/ [Vol. 42 INSTITUTIONAL ADAPTATION horizontally integrate water policy and agricultural policy is an important issue in Spain." 4 Integration is important enough that in 1994 parliament asked the Spanish government to set out a national irrigation plan before parliament made any decision about a holistic national hydrological policy.'" Of course this is not the only way to achieve policy integration; political and academic discourses often cite spatial planning as another means of achieving this goal. 4s PHYSICAL AND SOCIO-ECONOMIC BACKGROUND OF THE BASIN Physical Characteristics of the Lower Guadalquivir Basin The Guadalquivir Basin has a total catchment area of 57,071 km 2 and the river itself is 640 km long. The source and estuary of the river are both located within the Spanish region-of Andalusia, but a small part of the catchment area (about 10 percent) is located in three other Spanish regions, Extremadura, Castilla-La Mancha, and Murcia (Figure 1). In its last stretches, the Lower Guadalquivir Basin (LGB) covers an area of 17,085 km 2 that includes the Andalusian capital, Seville (Figure 1). The mean annual precipitation in the Guadalquivir basin is 596 mm, which is relatively high compared to the central Spanish mesetas or to southeastern Spain. However, the main feature of Andalusia generally and of the LGB is spatial and temporal irregularity. The mean annual precipitation ranges from about 1500 mm in the mountainous areas to 400-600 mm in the Guadalquivir depression.'" The lowlands in the eastern part of the basin receive less than 400 mm of mean precipitation a year.47 official/reports/pdf/p32_en.pdf; Cohesion Funds and the environment, at http://www. europa.eu.int / comm / regional-policy /sources/ docoffic /official/ communic/ envir/home.en.htm). At the Spanish level, the White Paper on Water, see Ministerio de Medio Ambiente, supra note 22, can be interpreted as an effort to coordinate different expertises and policy fields related with water. Finally, in Andalusia the creation of the Consejo Andaluz del Agua in 1996, Decreto 202/1995, de 1 de agosto, and, more recently of a specific Secretariat for Water Affairs, directly dependent from the Pridency of the Andalusian regional government (Conserifa de Presidencia de la Junta de Andalucia) are seen as efforts to coordinate the waterrelated responsibilities of the departments of Public Works, Agriculture, and Environment. 143. Irrigation agriculture uses about 80 percent of the total water resources in Spain; therefore, the agricultural policy will necessarily greatly affect water use. Ministry of the Environment, supra note 38, at 384419. 144. Id. at 410. 145. Federico Aguilera Klink, Hada una nueva economia del agua: cuestionesfundamentales, at http://habitat.aq.upm.es/boletin/n8/afagul.html (May 16,2002). 146. Aguilar et al., supra note 3, at 13. 147. Id. Summer 2002] 540 NATURAL RESOURCES JOURNAL [Vol, 42 Figure 1. Map of the Guadalquivir basin within Andalusia and Spain NI. I ,k i d" ' i ' .l -0 1 v n H ,, o I ca / 0 INSTITUTIONAL ADAPTATION This variability is aggravated by the fact that the region's mean potential evapotranspiration is inversely proportional to mean annual precipitation with extremely high values in the dry lowlands (sometimes more than 1000 mm) and lower values in the wetter mountainous regions. 1" Temporal irregularity has two components, seasonal or intraannual variability and inter-annual variability, and is similarly distributed both in the dry and in the humid areas of the Guadalquivir basin. 14 The intra-annual variability of precipitation is typical of the Mediterranean area." 5° The seasonal pattern includes a dry season of five to six months when evapotranspiration exceeds precipitation and a rainy winter season, whichbrings most of the annual precipitation.' The range, persistence, and extreme values of inter-annual variability are also extremely high. According to long series of data from meteorological stations, the precipitation values can range from 800 mm to 1100 mm from one year to another. 152 While precipitation can vary widely from year to year, it is also true that years with extreme (either low or high) precipitation values tend to be concentrated over time. This concentration aggravates the effects of droughts. Lastly, torrential precipitations can be extremely intense: 5 in most of the Lower Guadalquivir basin it can rain up to 150-200 mm in twenty-four hours (between one-third and one-half of the total annual precipitation).15 4 The mean annual renewable or natural water resources 55 are estimated at about 7000 million cubic meters/year (5.6 million acrefeet/year).'" But this varies widely from year to year and can range from 1 to 17,000 million cubic meters/year (13.6 million acre-feet/year). 1 5' Furthermore, the spatial and temporal climatic variability mean that only a portion of these natural resources are actually available at the time and in 148. Id. 149. Id. 150. Id. 151. Id. 152. Id. at 19. 153. Torrentiality is defined as the total precipitation in a year (nam) divided by the number of days of rain. Id. 154. Id. 155. For surface water, annual renewable or natural water resources can be defined as the longterm average freshwater volume supplied naturally by the hydrological cycle, derived from the total run-off (surface and underground) resulting from rainfall over a river basin territory minus evapotranspiration. 156. CONSEJERIA DE OBRAS POBLUcAs Y TRANSPORTE DE LA JUNTA DE ANDALUCA, PLAN DIREcTOR DE INFRAESTRUCTURAS DE ANDALUCIA 43 (1998). 157. Id. Summer 2002] NATURAL RESOURCES JOURNAL the place required.'" These available resources (recursos disponibles) are obviously affected by technological constraints, as well as by socioeconomic and institutional considerations.'" Most of the available water resources in the Guadalquivir Basin" 6 are derived from surface water, 161 while only a small portion comes from groundwater. 162 In order to deal with its irregular precipitation, the Guadalquivir River is regulated through a network of reservoirs throughout its basin."6 The reservoirs help to partially control the naturally erratic Mediterranean precipitation regime by storing winter flows for later use. With more than 60 reservoirs, the basin's reservoir capacity totals 6833 million cubic meters/year (5.5 million acre-feet/year)."' About one third of this volume (2255 million cubic meters/year 65 or 1.8 million acre-feet/year) is available as average annual regulated flow.'" Most of the reservoirs are located in the Sierra Morena area, whose impermeable soils and topographic conditions make it ideal for water storage. 67 Important Water-Users in the Lower Guadalquivir Basin The most important factor in the water resources use and planning in the Guadalquivir basin is the prime importance of irrigation agriculture, both in terms of water use'" and in terms of contribution to the Final Agricultural Production (PFA) of Andalusia.'" Data on the basin's irrigation 158. Available water resources, or the average annual regulated flow, can be defined as the average water volume stored in the reservoir in a year. It is determined by the reservoir capacity, its goals and management pattern, including the level of guarantee of the use for which it is aimed and the hydrological regime of the river basin in which it is located, taking into account the pre-existing reservoirs and other human modifications. 159. That is, availability depends on the technical capacity to build dams and inter-basin transfer systems as well as on the socio-economic and institutional factors (legal provision, financial support, agreements between regions, etc.) that allow for the actual implementation of these technical measures. 160. A total of 2700 million cubic meters/year, equivalent to 2.1 million acre-feet/year is usually available. 161. 84 percent or 2255 million cubic meters/year, equivalent to 1.8 million acre-feet/ year is available from surface regulation. 162. 16 percent or 437 million cubic meter/year (0.35 million acre-feet/year) comes from groundwater abstraction. 163. CONSEJERIA DE OBRAS POBLICAS Y TRANSPORTE DE LA JUNTA DE ANDALUCIA, supra note 156, at 44. 164. Id. at 43. 165. Id. 166. Id. 167. Id. at 44. 168. 85 percent or 2847 million cubic meter/year, equivalent to 2.30 million acre-feet/year. 169. With an irrigation surface that represents 19 percent of the total agricultural land, this sector contributes about 53 percent of the Andalusian PFA (Producto Final Agricola). [Vol, 42 INSTITUTIONAL ADAPTATION surface is disparate. Official.estimates have been about 668,000 ha, but another 20 percent of the surface might be used for unlicensed irrigation. 70 Within the basin's irrigation system, the irrigation of rice paddies in the Guadalquivir river estuary (about 35,000 ha) is crucial: its importance arises from its high unitary -water use (9.6,acre-feet/ha) and its location in the lower stretches of the Guadalquivir River, which is affected by the tides.7' From April to September, the rice irrigation period, large volumes of water must be released from the upstream reservoirs of the basin to lower the salinity in the estuary." During droughts, this release of water creates conflict between the rice growers and the rest of the basin's irrigation farmers.1 73 Although the urban sector demands less water (12 percent of total demand) than the irrigation sector, its needs are also critical to planners because it requires a guaranteed supply of high quality water year-round. Three water companies, EMASESA, Aljarafe S.A., and potentially the Consorcio del Huesna, supply the 1.3 million occupants of the Seville metropolitan area with 130-150 million cubic meters of water/year or 0.100.12 million acre-feet of water/year. 74 The Seville water system has experienced serious water shortage problems during each of the 1972-1974, 1981-1983, and 1991-1995 droughts. During the last drought, in addition to severe water rationing measures (between 6 and 12 hours a day), Seville had to use low-quality water from the Guadalquivir River. 75 While the river water was mixed with varying proportions of better quality reservoir water or groundwater, its use still had a negative impact on the quality of potable water. 76 The importance of accurately anticipating changes in demand is underscored by the pace of increased demand over the past few decades.'" While industrial demand represents only a minor proportion of total water use, there has been a considerable increase in agricultural and domestic demand over the past few decades.' These demands have been met by the 170. Moral Ituarte et al., supra note 104, at 207, 210. The illegal irrigation surface can be derived from the difference between the irrigation surface cited in water management documents (668,000 ha) and the 1999 Andalusian Inventory of Irrigation Agriculture, which estimates the total irrigation surface at 815,000 ha, thus including illegal (i.e. unlicensed) irrigation surfaces. 171. CONFEDERACION HIDROGRAFICA DEL GUADALQUVIR, PLAN HIDROLOGICO DEL GUADALQUIR 50(1995). 172. Id. at50. 173. del Moral Ituarte et al., supra note 104, at 211. 174. Giansante et al., supra note 98, at 6. 175. BAKKER ET At, supra note 82, at 28. 176. See id. (discussing flexibility in water quality standards during times of drought). 177. Ministry of Environment, supra note 38, at 339-42, 385. 178. Id. Summer 2002] NATURAL RESOURCES JOURNAL resources in the Basin's extensive network of reservoirs." The total available water resources in the Guadalquivir basin represent as much as 40 percent of the total natural water resources. 8' (See Table 1.) But, this is still insufficient to meet present demand. 18' Present demand outpaces available resources in the Guadalquivir river basin by 241 million cubic meters (0.2 million acre-feet/year) according to the Andalusian Department of Public Works and Transport. " Furthermore, draft national and river basin hydrological plans, and other strategic planning documents, project significant increases in domestic, agricultural, and industrial water demand both in the Guadalquivir river basin and in the rest of Spain.' ' Unsurprisingly, demand projections are the most controversial aspect of the water resources planning debate in Spain. Table 1. Basic Data on Water Resources in the Guadalquivir Basin Natural water resources 5.6 million acre-feet/year Total available water resources 2.1 million acre-feet/year - Surface water 1.8 million acre-feet/year - Groundwater 0.35 million acre-feet/year Total number of reservoirs 60 Reservoir capacity 5.5 million acre-feet/year Modelling the Economic Impact of Drought and Current Water Management Strategies on the LGB The LGB region's remarkable variability has historically had important implications for the institutions that must try to manage the irregularity and deal with the risk of water scarcity.' But so far, few evaluations of the socio-economic impacts of droughts exist. Furthermore, 179. CONFEDERACION HiDROGRAFICA DEL GUADALQUIVIR, supra note 171, at 72. 180. CONSEJERIA DEOBRAS PBLICASYTRANSPORTE DE LA JUNTA DE ANDALUCIA, supra note 156, at 43. 181. Id. at 61. 182. Id. These estimates are part of the political debate on water and vary according to the circumstances. For example, in a recent document the Andalusian regional administration accepts the estimate of the Guadalquivir River Basin Plan, according to which the water deficit in 1992 was 602 million cubic meters or 0.5 million acre-feet. CONSEJO ANDALUZ DEL AGUA, ANDALUCIA Y EL PLAN HIDROLOGICO NACIONAL (2000). 183. CONFEDERACION HIDROGRAFICA DEL GUADALQUwIR, supra note 171, at 51-63. 184. Aguilar et al., supra note 3, at 51-52. [Vol. 42 INSTITUTIONAL ADAPTATION other than some partial analyses of the urban supply service failures and drought impacts on the irrigation sector, no comprehensive effort has been made to completely evaluate the economic, impact of the 1991-1995 drought." 5 For the lower Guadalquivir basin, however, a mathematical programming model was recently developed to evaluate the economic impacts of the drought." 8 The model is based on realistic assumptions about farmers' ability to anticipate water shortages and labour and capital market rigidities." 8 7 It also makes assumptions about how farmers revise their production plans as new information becomes available." In Table 218 we summarise the results of the simulations carried out for representative farms in two irrigation districts of the lower Guadalquivir basin. The relationship between social droughts, failures in the water system that impact society, and meteorological droughts, precipitation irregularity, is non-linear and depends on social, institutional, and physical processes.se Run-off derived reservoir inflows largely depend on soil moisture and rainfall regimes. 1 91 Table 3 shows the complex links between purely meteorological processes and hydrological outcomes. A simple comparative analysis of the data regarding the conversion of one unit of rainfall into units of reservoir inflows between the year 1989-1990 and the year 1992-1993 shows a sharp contrast.192 In 1989-1990,1.41 units of rainfall generated one unit of reservoir inflow, but during the first year of the 1991-1995 drought 32 units of rainfall were required to generate a unit of 185. EMASESA, supra note 5. 186. See generally Eva Iglesias et al., Evaluation of drought management in irrigated areas, (Oct. 19,2000) (unpublished paper, on file with authors). 187. Ministry of the Environment, 1998, supra note 38, at 345. 188. Id. at 340. 189. As used in Table 2, "stock" stands for the percentage of stored volumes over storage capacity of the reservoir(s) that service each irrigation district at the beginning of February; "water allowance" is the per hectare irrigation water volume supplied in each irrigation season. Whereas the previous two variables are derived from actual data on the decisions made (water allowance) by water managers in the Guadalguivir basin based on specific reservoir levels (stock)-the following variables are calculated by the model. "Shadow price" represents the dual value associated to the water availability allocated to farmers each season, "gross margin" is the difference between total revenues and variable costs, "net benefits" are defined as the difference between total revenues and variable costs, and "value of production" stands for the market value of the agricultural output. Finally, "hired labour" is the amount of person-days per hectare generated by external workers. 190. Ministry of the Environment, 1998, supra note 38, at 167,250. 191. Ministry of the Environment, PLANHIDROLOGICONACIONAL, ANALJISDE LOS SISTEMA HiDRAuucos 210 (2000). 192. EMASESA, supra note 5, at 53. Sumnmer 20021 NATURAL RESOURCES JOURNAL fragmentation. 1 The main divisions in the traditional water community run across the following lines: (1) differences between legal and illegal irrigation farmers,; 2 (2) differences related to the size of agricultural plots; (3) differences between upstream and downstream irrigation farmer interests; (4) differences related to specific water requirements, especially to irrigation technology; (5) differences related to the crop types; and (6) differencesbetween the interests of surface water and groundwater users.' While the differences in agricultural plot-size were an important fragmentary force during the drought of the 1990s,2m at present the most salient fragmentary force comes from differences in water demand. Demand is directly related to an irrigator's levels of technology and efficiency and may be a key impetus for future change.m The technological divide within the irrigation community has major implications for the management of water stocks. New districts that have high efficiency conveyance and distribution systems can function better with limited resources, while older irrigation districts need larger water volumes to function well.' As a result, districts with newer technology tend to favour smaller water allowances and endorse more conservative strategies for managing water stocks while districts with older equipment oppose water allowance reductions because they lack the technology to adequately adapt to that change.'3 The irrigation sector's dynamism was demonstrated during the 1999 irrigation season, which was a breaking point for traditional water allocation practices.' As a result of a very dry autumn and winter, the water regulation system of the Guadalquivir basin had only 12,000 m3 /ha of water left at the beginning of the 1999 irrigation season.' Initially, the Guadalquivir river basin authority proposed releasing one-third of this volume, 4000 m3 /ha, during the current season, in exchange for a 231. del Moral Ituarte et al., supra note 104, at 292-94. 232. This is an extremely relevant issue due to the importance of illegal irrigation (both in terms of total surface and water use) and lack of control in the basin. 233. Generally, when this was possible, farmers opted for the drilling of (mostly illegal) wells in order to cushion the consequences of the restrictions or the general irrigation ban. Given the almost complete lack of control of groundwater abstractions, the Guadalquivir river basin authority (CHG) and its participatory bodies almost renounced consideration of the groundwater cycle, except in the last and most dramatic phases of the last drought. 234. del Moral Ituarte et al., supra note 104, at 294. The plot-size split was exemplified best by the debate on the allocation of water according to social criteria. 235. del Moral Ituarte et al., supra note 104, at 293. 236. Id. 237. Id. at 295 238. Id. at 298-99. 239. Minutes of the Water Release Commission (Mar. 1999) (confidential document, on file with authors); del Moral Ituarte et al., supra note 104, at 298-99. [V/ol. 42 INSTITUTIONAL ADAPTATION guaranteed irrigation supply of 4000 m3 /ha for the following two years.' This strategy would have reduced the future, risk to the irrigation community in the event of a new multi-annual dry spell. The irrigation sector rejected this proposal outright and pressured the river basin authority into granting a 6000 m3 /ha water allowance in March of 1999.4 In July of 1999, this volume was increased to 8000 m3 /ha. 242 For the first time in the Guadalquivir basin's recent history, the decision to increase the irrigation water allowance garnered harsh social and political criticism.' Although this started to open a fissure in the hegemonic discourse, the situation is still very fluid. All of these characteristics paint a portrait of an extremely dynamic irrigation sector that exhibits divergent approaches to water allocation, to its collective bargaining strategies, and to its discourses. It is important to note that although water allocation is a basic function of water management, the discourses favouring water re-allocation as an instrument to cope with water scarcity have only recently emerged. 2' The water reallocation discourse often diametrically opposes the dominant discourse, which favours resource augmentation as the definitive solution to Spain's structural water deficit. 2 Game Theory Analysis of Farming Sector Incentives and Strategies Using the water allocation proposals and practices from the 1999 irrigation season, we used game theory to analyse the incentives and strategies of the farming sector regarding water allowances.'" A static twoagent game where two farmers compete for water resources yielded 240. Minutes of the Water Release Commission, supra note 239; del Moral Ituarte et al., supra note 104, at 298-99. 241. Minutes of the Water Release Commission, supra note 239; del Moral Ituarte et al., supra note 104, at 298-99. 242. Minutes of the Water Release Commission, (July 1999) (confidential document, on file with authors); del Moral Ituarte et al., supra note 104, at 298-99. 243. A local newspaper published on its front page the following headline: "The same volume as a six year water supply for the city of Seville was diverted for irrigation purposes." See Desviada para regadlo tanta agua como Sevilla consume en seis ailos, DIARIO DE SEVILLA, Sept. 1, 1999. The new element is the shift in focus from a claim for more reservoirs-as a solution to the drought-to the critique of how the remaining water had been managed and allocated, with a special concern for the possibility of a failure in the urban supply system. 244. del Moral Ituarte et al., supra note 104, at 299-300. 245. The concept of structural water deficit and the treatment of uncertainty within Spanish water planning is analysed in Leandro del Moral Ituarte & Consuelo Giansante, Constraints to Drought Contingency Planning in Spain: the Hydraulic Paradigm and the Case of Seville, 8 J. CONTINGENCIES & CRIsIS MGmT. 93(2000). 246. See generally Weitze Lise et al., A game model of farmers' demands for irrigation water from reservoirs in Southern Spain (Mar. 23, 2000) (unpublished paper, on file with authors). Summer 2002] NATURAL RESOURCES JOURNAL different equilibria across the three irrigation districts to which the game was applied, Bajo Guadalquivir, El Viar, and Genil-Cabra. Farmers in a selfmanaged irrigation district such as El Viar tend to be conservative and save water in the reservoirs for the next seasons. But; in centrally managed irrigation districts, such as Bajo Guadalquivir and Genil-Cabra, farmers tend to demand water instead of saving it. 247 Modeling farmers' strategies as a repeated game enabled us to analyze how the preferences for a high-demand strategy depend on the inherent risk level of the hydrological variables. Under the current conditions, a repeated game demonstrates that a more demanding strategy is optimal for farmers in both types of irrigation districts. 28 However, an extreme climate change scenario that substantially increases the risk of drought creates an incentive for farmers to become more conservativeassuming there are enforceable rules to protect farmers' water rights. 249 This indicates that farmers actually do react to present hydrological risks, but just not enough to achieve equilibrium with climatic patterns. These findings support the hypothesis that more transparent water rights that spell out the rights and duties of the water system beneficiaries may create a greater incentive for farmers to accept proposals that manage water stocks over time, such as the one made by the Guadalquivir river basin authority in 1999.' Specifically, if farmers' water rights were guaranteed across seasons they would be much more willing to cooperate with more conservative strategies than in the current system that pools unused water rights allocations from previous years and new inflows before beginning the allocation process for a new year." EVALUATING THE WATER ALLOCATION DECISIONS IN THE LOWER GUADALQUIVIR BASIN To evaluate the efficiency of the water allocation decisions made by water managers in the lower Guadalquivir basin in the 1991-1998 period, we created an Economic Drought Management Index (EDMI) and applied 247. Self-managed irrigation districts (for example El Viar) are here defined as those districts whose supplies depend exclusively on one reservoir and are only aimed at irrigation purposes. On the other hand, centrally managed irrigation districts (for example Bajo Guadalquivir) depend on a large number of reservoirs throughout the basin, which are shared with other irrigation districts and water use sectors (for example, urban supply). In both cases, the management of water stocks is supervised by the river basin authority. 248. Self-managed districts and centrally-managed districts. 249. The farmers are facing a so-called prisoner's dilemma. 250. Lise, et al., supra note 246, at 244. 251. Id. [Vol. 42 INSTITUTIONAL ADAPTATION it to data from the area.- 2 The index adds an economic dimension to climatic and hydrologic information and can assist water managers by providing information about the economic risk of their strategies along with the cost of reducing that risk. The main appeal of the EDMI is that it combines four key pieces of information in an easily interpretable index: (1) the structural constraints of a supply system based on reservoirs, (2) the stochastic nature of natural run-offs flowing into the storage facilities, (3) the institutional rules that have been followed by water managers during the periods of historical record, and (4) the benefits accruing to the consumptive users. An EDMI result uses the ratio of the present use-value of water and the expected use-value of the water left in stock to provide a measure of the opportunity cost of employing a more conservative, water-saving management strategy. An EDMI ratio of one indicates that the water management strategy is perfectly efficient because the opportunity-cost of saving the water equals its expected use-value for the following irrigation season. If the EDMI ratio is less than one, the cost of saving one water unit is less than the expected use-value for the following year and reveals that the water managers are having an exceedingly risky management behaviour. If the EDMI ratio is greater than one, the opportunity-cost of saving one water unit is greater than the benefits it would generate the following year and water should be used in the current season. Table 4 reports the EDMI ratios calculated for the El Viar (EV) and the Bajo Guadalquivir (BG) irrigation districts, using data available for each of the nine years from 1990 through 1998. Table 4. Actual EDMIs for the EV and BG Supply Systems (1990-1998) Year EV BG Stock on Water EDMI Stockon Water EDMI Feb. 1 (% of allowance Feb. 1 (% of allowance capacity), (acre-feet/ha) capacity)' (acre-feet/ha) 1990 91 7.54 0.170 37 5.01 0.030 1991 60 5.88 0.650 26 4.42 0.080 1992 42 4.60 0.640 21 2.54 1.040 1993 13 0.62 3.820 15 0.06 36.180 1994 35 3,86 0.806 18 0.73 6,350 1995 9 0 9.540 11 0 11.890 1996 95 5.80 0.760 36 4.76 0.66 1997 92 5.90 0.182 89 6.39 0.007 1998 93 5.90 0.750 87 7.05 0.940 'Source, Unpublished repot iuixvr river bas 252. See generally Eva Iglesias et al., An Economic Management Index to Evaluate Irrigators' Vulnerability to Water Shortages (Oct. 19, 2000) (unpublished paper, on file with authors). Summer 2002] NATURAL RESOURCES JOURNAL The EDMI values reported on Table 4 show several cases of inefficient management. To be fair, please note that the EDMI ratios were computed based on average inflow values that did not anticipate the early 1990s drought. Even so, EDMI ratios in 1991 are less than one in both districts, indicating a need for lower water allowances in the 1991 irrigation season. The reservoir inflows for the 12 months following February 1,1991, were below average, resulting in lower stock values on February 1, 1992, than in the previous year. However, the EDMI ratio for the Bajo Guadalquivir district hit its optimal value in 1992 with a water allowance of 2170 m3 /ha, which was historically low but efficient. El Viar's 1992 and 1994 decisions contrast sharply with that efficiency. The 1992 EDMI ratio was 0.64; the 1994 (a year between.two extremely dry ones) EDMI ratio was 0.80. Although Bajo Guadalquivir EDMI'ratios have been very high since the hydrological drought started in 1992 and until its end in 1995, when rainfall levels recovered, throughout the drought period no action could be taken to move them closer to one simply because there was no water left in the reservoirs. That said, the Bajo Guadalquivir district in 1990 and 1991 was managed with EDMI ratios close to zero, when stocks were at 37 and 26 percent of reservoir capacity respectively. If water managers had reacted at the time, the effects of the severe 1992 to 1995 droughts might have been partially mitigated or put off until the last two years of the drought period. All EDMI ratios after 1996 are less than one in both irrigation districts. This indicates that both systems are running inefficiently and at very high risk levels. These results also indicate that, so far, water managers have taken entirely too much risk. The high risk level suggests that lower maximum water allowances are acceptable and that a conservative, water saving strategy should be implemented as soon as reservoir network water levels start to drop. The EDMI ratios confirm our initial hypothesis that the current rules for irrigation water allowances and for managing reservoir water stocks are highly inefficient. It also suggests that the recent amendment to the Water Act, which timidly introduces water markets, may fall short of expectations because water right holders are not given sufficient guarantees of ownership over the resources left in the reservoir. This uncertainty makes it unclear whether the water left will be available either for sale in the market or for future consumption. This effectively limits the role of the market to the allocation of the resources available in a single irrigation season. The single season focus distorts scarcity signals, which may give incentives to defer consumption at the onset of droughts and soften the supply variability inherent to the Mediterranean climate. [Vol. 42 INSTITUTIONAL ADAPTATION CATALYSTS AND INHIBITORS OF INSTITUTIONAL CHANGE There are identifiable catalysts for change within the lower Guadalquivir basin that are likely to influence the water system's ability to cope with present and future risks. In this regard, even though climate change is not yet a major catalyst for change in Spain, it may significantly affect future water management scenarios.' Climate change has either been ignored or written off as an environmentalist concern. Although it has recently aroused greater interest,' its possible effects continue to be excluded from discussions of present water management options.'s This attitude is fuelled by the perception that the existing uncertainties the water planning process faces, such as estimating present water resources, projecting socio-economic trends, projecting water demands, predicting short term horizons, lack of control, and lack of credibility are greater than climate change uncertainties.' In point of fact, water mangers tend to view the current imbalance as a good proxy for an intermediate climate change scenario.2 However, in the short term, climate change is not likely to be a catalyst for change, as what is perceived as a chronic water deficit seems to be a more urgent issue to tackle. One of the factors likely to enhance adaptation in the drought management practices is growing public, expert, political, and media awareness of the continuing crisis the present water system faces.' This crisis is often blamed on the uncontrolled expansion of irrigated land over the past two decades.' The most recent drought has played a key role in 253. Ayala, supra note 215. 254. The issue of climate change is considered in the White Paper on Water and in the last draft of the National Hydrological Plan (Sept. 2000) and is receiving increasing attention in the media and expert circles. 255. The climate change scenarios have not been considered in the estimate of future available water resources in any planning document. See CONFEDERACION HIDROGRARCA DEL GUADALQUIviR, supra note 171. 256. These ideas are derived from a seminar on climate change organised by the Seville SIRCH team, to which several experts and water managers from the basin were invited. See SIRCH, WORKSHOP ON CLIMATE CHANGE AND WATER MANAGEMENT SCENARIOS IN THE GUADALQUIVIR BASIN (June 6,2000) (unpublished report, on file with authors). 257. From this perspective, the problems related to climate change are perceived as present and urgent. This paradoxical relationship with the subject shows the very imprecise information generally available and the lack of reflection regarding climate change. At present, the idea that the variability of rainfall and temperature and the persistence of dry spells have increased over the last three decades in the Guadalquivir river basin has been accepted in climate research and water management circles. 258. See generally, Juan L6pez Martos, Los problemas del agua y su gesti6n en el territorio andaluz, in ll CONGRESO IBtlCO SOBRE GEsTR)N Y PLANFICAcKiN DEL AGUA (2000), available at www.us.es/ciberico. 259. Id. at 256. Summer 2002] NATURAL RESOURCES JOURNAL increasing awareness of the weakness of the present management system.0 Although extreme events and crises can trigger change, they exist within a specific social-institutional setting. In Spain, the backdrop is a "structural deficit" one in which droughts are viewed as circumstantial expressions of a chronic water shortage.' This social understanding of droughts as a symptom or expression of chronic shortage makes it easy to neglect important planning areas like contingency planning and crisis management. This view has also made it easier to rationalize and to justify both a dangerous demand-hungry mindset 262 and a vigorous infrastructure building strategy by the state. 2" However, the pervasive confidence in the system's robustness that existed before the 1990's drought contrasts sharply with the present awareness of vulnerability.' As a result of that drought, two innovative approaches have gained support and may enhance future adaptation to uncertainty and risk. The two approaches are drought contingency planning and the reallocation of water between sectors and user groups.' The latter mechanism falls under the more general marketisation trend,' but they also reflect a stronger commitment to guaranteeing a basic urban water supply in drought situations even in the face of pressure from the irrigation sector. 267 Possibly the greatest potential for the water system in the lower Guadalquivir basin to adapt to scarcity comes from reduced irrigation that could result from agricultural policy changes or water rate increases. 26 Although full cost recovery, FCR, 26 is unlikely in either the short or medium term, some price increases and changes in the Common Agricultural Policy (CAP), which would liberate significant water volumes, 260. del Moral Ituarte et al., supra note 104, at 298-99. 261. del Moral Ituarte & Giansante, supra note 245, at 94-95. 262. Id. at 95-96. A similar analysis is performed on discursive mediation of droughts in California in L. Nevarez, Just wait until there is a drought: mediating environmental crisis for urban growth, 28 ANTIPODE 246-72 (1996). 263. del Moral Ituarte & Giansante, supra note 245, at 99-100. 264. Id. at 98. 265. BAKKER ET AL., supra note 82, at 27,63. 266. Discourses favouring the reallocation of water according to social criteria have given way to the introduction of market instruments as the main methods for enhancing flexibility and efficient resource allocation. 267. The experience of the 1981-1983 and 1991-1995 droughts-when millions of households suffered from water cut-offs and poor water quality-and of the 1999 irrigation season-when water managers' decisions on the allocation of water to farmers have been harshly criticised-played a key role. 268. Sumpsi, supra note 117. 269. About 80 percent of the value of production in half of the 600,000 ha of irrigation surfaces in the Guadalquivir basin comes directly from subsidies of the Common Agricultural Policy (CAP) and only about 25 percent of these irrigation surfaces would be able to subsist under a scenario of full cost recovery (FCR). [Vol. 42 INSTITUTIONAL ADAPTATION are very likely.' These changes, however, are highly dependent on institutional arrangements and are causing and will continue to cause great social resistance. 27' The greater awareness of vulnerability, the prospect of changes in water use patterns, and the promotion of inter-sectoral water allocation and contingency planning notwithstanding, significant portions of the traditional water paradigm persist unchanged and are poised to resist change, even when it appears necessary to cope with water scarcity. The main sources of resistance; in the Guadalquivir basin are aligned with various features of the traditional water policy community. The traditional engineering mindset and long-standing support of structural solutions to water scarcity continue to hinder serious consideration of other alternatives, such as temporary re-allocation, contingency planning, or managing demand. 2' In particular, structural solution supporters view demand management as possible and even necessary but ultimately insufficient to address the "structural deficit." ' Additionally, the still very low water prices and lack of awareness of environmental impacts provide little incentive to users to choose good scarcity responses or technology and discourage demand reductions. 274 The implementation of new policy preferences like water markets, metering, and price incentives are also seriously hampered by weak management and a general lack of monitoring and control over water uses.'n In particular, the present licensing system provides neither incentive nor legal guarantee of a minimum supply of water for the following year (inter-temporal allocation) for irrigation farmers who are willing to reduce water consumption during droughts. Also, lack of institutional transparency and conflict between distinct parts of the administration with 270. Sumpsi, supra note 117. 271. Los agentes sociales rechazan la tasa destinada a los gestores de agua, EL PAIS, July 16,1998. 272. In particular, the attitude of the Seville water company toward this option is still hesitant, despite the recent legal changes opening up new possibilities for inter-user transfers and the inclusion of this option in the 1998 Drought Manual. At the same time, water conservation measures are not developed within integrated programs of demand management but as isolated measures without defined objectives and calendars. On the other hand, the stable reduction of demand would have clear repercussions on a water company's financial balance. So far, the sustained low level of consumption that has persisted in Seville even after the drought emergencies has been absorbed by water price increases. However, this solution has high political costs for the municipal government that must approve them and results in the municipal water company's ambivalent attitude. See EMASESA, MANUAL DESEQUA (1998). 273. See Plataforma del Guadalquivir, at http://www.sevsigloxxi.org/plataforma (Mar. 15, 2002). 274. Id. 275. Professor Ramon Llamas asserts that the state of underground waters in Spain is in chaos, Arag6n Press, at http://www.circe.cps.unizar.es/spanish/waterweb/dossier.html (June 14, 2001). Summer 20021 NATURAL RESOURCES JOURNAL different political affiliations and between the administration and users or citizens also contribute to the maintenance of the status quo. 6 But, of all the features of the traditional water community, the social hegemony of the irrigation sector is the most important. It explains the emphasis placed on the negative effects of full cost recovery water tariffs for agriculture as well as the sustained increase in irrigation lands 2' that fly in the face of more general trends and the preferences set out in the National Irrigation Plan. After emerging from the last drought streamlined and strong, the agrarian sector can now rely on a well-consolidated structure, as well as new ad hoc pressure groups and alliances to push its agenda.m SOCIAL AND INSTITUTIONAL ADAPTIVE RESPONSES The most promising responses to water scarcity are likely to result from demand management, contingency planning, and inter-sector and inter-temporal allocation of water stocks. As a result of these catalysts for institutional change, in the lower Guadalquivir basin innovative water management practices (whose degrees of development vary) have emerged.7 9 For example, demand management measures have been promoted with greater energy in the Seville water supply system since the mid1990s. Similarly, a general downward trend in urban water use could be related to increased public awareness and could become the basis for a long-term water conservation strategy.' As regards contingency planning, the main development in this field is the elaboration of a drought manual by the Seville water company. 82 The manual proposes criteria for defining 276. PWREZ Diaz Er AL, 1996, supra note 23, at 61-64. 277. To a large extent, this increase has been uncontrolled-without the corresponding water licences. In some cases the Andalusian regional government has officially supported it, though. One of the main factors behind this growth is the current European agricultural subsidisation policy. 278. Although their infrastructure agenda has dominated the discourses of the main political parties and media, none of the water schemes that they were demanding have been carried out since 1995. Thus, a paradoxical situation exists in which an apparent social and political quasi-unanimity about the need for hydraulic infrastructure exists, but there is also a real stagnation in its implementation. 279. del Moral Ituarte et al., supra note 104, at 297-98. 280. del Moral Ituarte & Giansante, supra note 245, at 99. 281. However, the urban development model (suburbanisation) currently in force in Seville, as in other Spanish and European cities, also represents a clear obstacle to the reduction of urban water demand. 282. See EMASESA, supra note 272. [Vol. 42 INSTITUTIONAL ADAPTATION different phases of a drought according to increasing severity and also proposes correspondingly severe conservation measures.2" But the response to water scarcity that lies at the centre of the recent water debate concerns inter-sector and inter-temporal water re-allocation.2" This response came to the forefront because of the unequal distribution of water resources among sectors with different levels of economic and social status and because of the 1999 amendment to the Water Act, which prepares the ground for water transfers between users.' The idea that these transfers may increase the overall efficiency of water allocation is beginning to be accepted, albeit with some reticence.' But more turmoil lies ahead as the introduction of water markets is likely to elicit divergent reactions from differently situated irrigation farmers. Socio-demographic features, farm size, crop type, technology, source of supply, and geographical position in the water system will all impact different irrigation farmers in different ways and the group could experience more tension, especially if water were to become even more scarce.' CONCLUSIONS In the Lower Guadalquivir basin, as in the rest of Spain, a resourceoriented water management model with deep historical roots, the "hydraulic paradigm," has dominated so far.' This model has had a distinct impact on how the risk of water scarcity is perceived and on the vulnerability of the water system, which is increasing as a result of users' demand-hungry attitude. In this paradigm, droughts have traditionally been viewed as expressions of a "structural deficit," a permanent imbalance between water demand and water supply. This view has led water managers and water planners to neglect the crisis management and contingency planning areas. Nonetheless, the idea that the water management system faces a serious crisis is becoming more widespread among water managers and stakeholders alike. The last drought increased awareness of the water system's vulnerability and became a catalyst for social participation in the water debate. The uncontrolled expansion of irrigated lands over the past two decades also had important ramifications for current water allocation 283. Ministry of the Environment, supra note, 38, at 839. 284. 1999 Water Act, B.O.E., 1999, 298, art. 61 bis. 285. Id. 286. del Moral Ituarte et al., supra note 104, at 299-300. 287. Id. at 292-94. 288. This term describes a type of water development policy that advocates building new water infrastructure (dams that make more water available) instead of managing demand (by increasing water efficiency or campaigning in favour of water savings). Summer 2002]