Markets for managing natural resources have existed for many decades and have gradually made their way into the mix of discourses on water policy. However, there are not many established water markets functioning worldwide and little understanding about how and why water markets emerge as allocating institutions. In order to understand the dynamics of the evolution of water markets, the experiences of selected cases with relatively mature water market systems were analyzed, namely: the Murray–Darling Basin in Australia; the Colorado-Big Thompson Project and the transfers between the Palo Verde and Metropolitan Water Districts in the United States of America; and Spain. We found that formal markets emerged in water scarcity situations where water rights already existed and were sometimes exchanged informally. Water markets have not always moved to reduce transaction costs, as some of those costs were necessary to achieve societal goals beyond economic efficiency. There is a significant difference between the idea of water markets as proposed by economic theory and actual practice in the water sector. As institutions, markets are humanly devised rules embedded in a social and political context and do not always lead to efficient or effective solutions for the management of resources.
Water scarcity is a significant problem in many parts of the world, affecting human development and wealth creation (Rijsberman, 2006; Tropp, 2007). The United Nations 2030 Development Agenda set Sustainable Development Goal (SDG) No. 6 to ‘ensure the availability and sustainable management of water and sanitation for all’, but water also has implications for several other SDGs. However, scarcities are expected to further increase in severity and frequency both in current and new scarcity-prone areas, particularly due to increasing demand from population and economic growth, as well as due to climate change (Schewe et al., 2014; Richey et al., 2015; Distefano & Kelly, 2017); this scarcity alert was reinforced by the 8th World Water Forum in Brasilia in March 2018 (World Water Forum, 2018). Such a scenario presents a major challenge for water users and public authorities. It is no longer possible to solve scarcity problems solely through an engineering approach, with supply-side measures (Zetland, 2014); demand management strategies should also be entertained (Grimble, 1999). In this context, economic instruments have gradually made their way into the mix of water policies (Garrick et al., 2009). Some jurisdictions have already established water markets (WMs)1 to deal with droughts and increase water use efficiency, with different results (Johansson et al., 2002; Bitran et al., 2014; Saleth et al., 2016). Nevertheless, experience of WMs is still relatively rare. Markets are more common to manage other natural resources (such as fisheries) or environmental issues (such as greenhouse gases). In order to better design WM institutions, the objective of this paper is to understand the evolution of WMs as institutions, studying well-established WMs that can provide lessons for the establishment of new ones. Thus, we have studied the examples of: the Murray–Darling Basin (MDB) in Australia, the Colorado-Big Thompson Project (C-BT) and the transfers between the Palo Verde and Metropolitan Water Districts in the United States of America; and Spain (nationally).
Markets are not created solely for achieving economic goals; they are socio-political institutions and aim at broader societal objectives. Markets are the result of socio-political processes and should be embedded in societal needs, instead of being an imposition from bureaucrats, business people, investors, corporations or politicians. They also need effective regulatory institutions to work well (Bitran et al., 2014). Against this background, our paper attempts to offer a fresh perspective about the use of WMs, understanding how and why they emerge and what features make them effective institutions for water management.
We make several contributions to building knowledge about WMs. First, we provide an interdisciplinary study, combining economic theory with insights from other disciplines such as management, political science and public administration. Secondly, unlike traditional economic studies which focus on prices and economic efficiency, we are more interested in examining the conditions under which WMs have emerged and evolved as solutions to water scarcity, with all their limitations. Thirdly, we analyze empirical examples of ‘mature’ functioning WMs for surface water (which are relatively rare) where we can assess how they have developed and their results so far. Thus, we pinpoint the characteristics that make markets effective institutions to deal with water issues. Finally, we provide some recommendations for the establishment of WMs based on empirical studies and not only on economic theory. Instead of assuming markets as the solution to water issues based on economic principles, we deliver some insights to assess under what social and political circumstances markets can offer a viable alternative for water management.
Market-based instruments applied to water
Markets are institutions formed by formal and informal rules with different degrees of enforcement (North, 1990). For a long time, markets have been proposed as the most efficient solution to manage environmental issues and natural resources, including water (Anderson & Leal, 1992; TEEB, 2010). However, there are few established WM mechanisms functioning worldwide and thus much of what is propagated by economic theory cannot be verified by practice. Furthermore, institutions do not always evolve to be efficient and effective in the allocation of resources (Mellah, 2018) or to reduce transaction costs (North, 1990; Bitran et al., 2014; Loch, et al., 2018). Therefore, there is little empirical understanding about how and why markets rather than other institutions (such as command-and-control systems) emerge to manage natural resources.
Water has a unique set of characteristics: its high mobility; uncertain quality, quantity and location; high costs of transportation; and decisions concerning its extraction, use and reuse affect all the users in a hydrologic system (Backeberg, 1997). Demand and supply change over time due to seasonality, climate and socio-economic dynamics. Connected to this, water management requires some flexibility to be able to give incentives for the wise use of water. Command-and-control systems, however, generally offer little flexibility and often fail to ensure that the most efficient water use practices and technologies are adopted (Cantin et al., 2005). Conversely, water scarcity can be addressed by changing the incentives users face by managing demand with economic instruments (Zetland, 2011b; Jamshidi et al., 2016). Indeed, market mechanisms to allocate and reallocate water between competing uses can be seen, at least in theory, as more conducive towards efficiency gains than a rigid, centrally controlled process not reliant on a price system (Johansson et al., 2002). Occasionally, market-based approaches seem to draw more public support as tools for water management (Rissman et al., 2017), but can also have negative welfare effects (Juana et al., 2011).
Without prices that adequately signal how scarce water is, extraction and consumption tend to occur at rates above what is desired; over extended periods, the absence of this signal negatively influences decisions about how to use other resources (Olmstead & Stavins, 2007). However, increased water use efficiency may reduce return flows in the short term but, in some instances, aggravates scarcity in the long term (Scott et al., 2014). This so-called ‘rebound effect’ can be particularly detrimental to environmental water uses (Loch & Adamson, 2015). A simple way to prevent such an effect is the periodic revision of the total amount of water available for consumptive use (Young, 2014).
WMs, as discussed here, do not involve transferring water per se, but rather the right to use or access the resource; such a right usually implies a given quantity to be used in certain places and moments. Through voluntary trades, both buyer and seller (whether individual users, associations or even water districts) are made better off (Horbulyk & Adamowicz, 1997). When jurisdictions move from a centralized allocation of water resources towards market instruments, power is transferred from regulators and bureaucrats to users. Often, governments also relieve themselves of the need to make large investments in infrastructural projects to increase water supply (Rosegrant & Gazmuri, 1994). This does not mean that governments have no role to play whenever WMs are in place. Public authorities have to ensure that water rights are well defined, secure and reflect actual uses, that buyers and sellers comply with the trade, and that no third party will be negatively affected (at least without proper compensation) (Rosegrant & Gazmuri, 1994; Johansson, et al., 2002; Bitran et al., 2014). Provided that administrative processes are clear and consistent, governments can also supply water users with platforms to meet and trade, and offer relevant information about expected water availability and market transactions (e.g. prices, volumes, deadlines) (Livingston, 1995; Easter et al., 1997). These measures can limit any asymmetries between users and help reduce transaction costs. Still, even if high transaction costs do exist, and sometimes even increase over time (Loch et al., 2018), and that some market imperfections can be found, markets remain a useful way to allocate water (Backeberg, 1997).
Despite these strong theoretical arguments, only a few successful experiences of WMs worldwide exist (Garrick, et al., 2009; Grafton et al., 2011; Loch et al., 2018). Understanding how different arrangements alter the costs (actual or perceived) of engaging in WMs and influence users' incentives remains unclear, particularly given the scarce literature on this topic (see, for example, Garrick et al., 2013). There is also some resistance to WMs based on ideology or previous failed efforts. Different interest groups may have divergent views on how to regulate water access (Esteban et al., 2018). The literature is particularly limited with regard to the institutional evolution of markets to deal with important issues so that they can last and increase efficiency, such as understanding how transaction costs are reduced and non-consumption is treated (Debaere, et al., 2014). This paper addresses these gaps by empirically analyzing some of the best examples of lasting markets with some positive outcomes.
Examples of formal WMs functioning for a long time are relatively rare (Griffin, 2016). In most countries, water rights were conceived within previous contexts of relative abundance (Young, 2014) and there are costs associated with changing existing institutions and altering rights already granted. Still, more attention has been devoted to market mechanisms as water has become scarcer in some regions (Zetland, 2011a). This paper is directed towards areas with frequent and intense scarcity events with established markets for surface water. We conduct a comparative analysis of meaningful experiences in the following locations: the MDB in Australia; C-BT and Palo Verde Irrigation District and Metropolitan Water District Program (PVID/MWD) in Western USA; and the national experience in Spain. Initially, information was collected from a detailed literature review, from technical, legal and academic sources. Secondly, between May and July of 2016, we also conducted field visits to interview experts in WMs, with experience in these four regions, to verify that the information gathered, plus our interpretation of it, was pertinent and accurate. Some follow-up interviews were carried out by phone or email. In total, we interviewed thirteen experts familiar with the studied markets. The results were revised and discussed with experts in an ad-hoc manner during 2017 and 2018. References to interviewees are made anonymously, using the abbreviation ‘INT’ and numbered in order of appearance.
Water management is context-specific and the few occurrences of formal WMs restrict our attention to these four cases. Although we can extrapolate some general messages from a limited sample (Ragin & Becker, 1992), the absence of objective and comparable data between initiatives limits our analysis to the qualitative realm (Baglione, 2012). Indeed, knowledge that can be transferred between jurisdictions tends to be broader and more general in nature (Rose, 1991).
Australia: the Murray–Darling Basin
Australia has a few distinct markets for water rights. The country has also gone through a series of water reforms (Crase et al., 2008). By the mid-1990s, the Council of Australian Governments (COAG) was already incentivizing states to develop WMs as a way to deal with scarcity events, manage the growing demand and ensure water availability for environmental uses (Bennett, 2015); some WMs in Australia even precede this national orientation, with the states of New South Wales and South Australia already having formal WMs in 1987 and 1991, respectively (Garrick, et al., 2009). The market in the Murray–Darling Basin (MDB), particularly its Southern portion, is by far the most active and relevant in the country, accounting for more than 85% of permanent and 98% of temporary transfers (NWC, 2013).
The MDB occupies a large portion of Southern Australia (see Figure 1) and encompasses four states (New South Wales, Victoria, Queensland and South Australia) and one territory (Australian Capital Territory); it relies on an extensive system of reservoirs and interconnected canals (Shi, 2006; Grafton et al., 2015). Irrigated agriculture represents nearly 80% of total water consumption in the Basin, mostly from rice, dairy and horticulture production (MDBA, 2010). Hence, its WM is almost exclusively composed of commercial farmers (Grafton et al., 2015).
The most important characteristic of the Australian experience is how its water rights are designed. First, they are entirely separated from land titles. Second, they are divided into long-term access rights to a certain volume of water (entitlements) and annual allocations, stipulated in percentage terms from the nominal volume in entitlements. Both entitlements and allocations are tradable (NWC, 2013). As the amount of water in the reservoirs goes down, annual allocations are reduced and users have the options of decreasing their consumption or buying/selling allocations in order to match their water needs during scarcity events. This clear definition of rights was initially encouraged by economists and later by Australian governments (COAG, 1994) in order to improve water allocation (Freebairn & Quiggin, 2006), particularly concerning the MDB, since new entitlements were not to be permitted in the area, to prevent (further) over-allocation of the existing resources (Shi, 2006).
Indeed, the MDB market emerged from the need to join individual state initiatives and manage the basin as a whole, an approach intensified after the Murray–Darling Basin Authority (MDBA) was created in 2007. The MDB's most recent Basin Plan explicitly vows to facilitate transactions within and between states (MDBA, 2012). Since then, market activity has increased consistently (ABARES, 2016) and the MDB can be seen as ‘a leading example of the benefits of implementing widespread water reform’ (Grafton et al., 2011), one that has seen meaningful transformations in water consumption patterns after water rights were changed and allowed to be traded (Bennett, 2015). Even the federal and local governments also participate in the market, purchasing entitlements to guarantee environmental uses (Wheeler, 2014).
USA: Colorado and California
Western USA has a dry climate and rivers that are mostly fully or over-allocated, with a vastly developed water infrastructure (Culp et al., 2014). Water rights are use rights that follow the prior appropriation doctrine, reflecting the historical need to divert water from rivers to remote locations (Carey & Sunding, 2001; Howe, 2011). According to this doctrine, the first user to extract water for beneficial and reasonable purposes is granted the water right. Older (senior) rights are the first to be attended during droughts (Grafton et al., 2011). These rights are separated from land property and theoretically tradable (Howe, 2011; Lantz et al., 2014). In practice, however, regulatory doctrines such as ‘no injury rules’ (which usually mean that rights can only be traded if they do not negatively affect other downstream users, for instance by causing changes in return flows; see Squillace & McLeod, 2016) represent a significant barrier for trading (Libecap, 2010; Culp et al., 2014), particularly the expert and legal fees, and increased uncertainty and time required to close a deal (Squillace, 2013; Squillace & McLeod, 2016).
Generally, water rights can still be considered fragile given the requirement of beneficial use (commonly open to interpretation) and disputes over the proper level of seniority (Libecap, 2010; Gray, 2015). They are also not homogeneous between states, limiting WMs to smaller scale (Squillace, 2013). Nonetheless, there are instances in which transaction costs have been reduced and some local markets emerged.
The Colorado Big-Thompson project (C-BT)
The most active and mature WM in the US can be found in Northern Colorado (Libecap, 2010; see Figure 2), enabled by the Colorado Big-Thompson project (C-BT) (Griffin, 2006). This infrastructure project diverts water from the West to the East side of the Rocky Mountains through a series of tunnels and reservoirs (NCWCD, 2013). Once the C-BT was concluded in 1957, the ‘new’ water in the eastern slope was allocated between users in quotas (shares) nominally equivalent to 1 acre-foot (1,233.48 m³) of water from all the 313 thousand acre-feet the project was expected to deliver annually (Debaere et al., 2014). Quotas are adjusted according to actual water availability, are homogeneous and easily tradable between users, both permanently and temporarily, irrespective of location and type of use (Carey & Sunding, 2001). Such a market closely resembles the textbook definition of cap and trade programs.
The Pal Verde irrigation district and metropolitan water district program
The state of California recognizes two forms of water trade uncommon to other regions: (i) resting part of an agricultural field and trading the water that would have been used if that land had been cultivated (fallowing) and (ii) moving to a culture that requires less water and trading the volume saved from that change. Thus, conservation efforts are accepted in the state Water Code as beneficial uses (California State Legislature, 2016). This facilitates transfers and the emergence of ‘markets for conserved water’ (Squillace & McLeod, 2016). Within this context, two water districts in Southern California – Palo Verde Irrigation District (PVID) and Metropolitan Water District (MWD) – established a program that allows farmers who voluntarily rest some portion of their lands in PVID to sell water to MWD, expanding water supply for urban uses around Los Angeles and San Diego (MWD, 2016). Both districts receive water from the Colorado River. PVID has one of the most senior rights in the state (PVID, 2005) and is located downstream from the extraction point and infrastructure that supplies MWD, the Colorado River Aqueduct (CRA); see Figure 3 (MWD, 2013). Such a configuration allowed the former to temporarily sell some of its rights to the latter, which increases its extractions directly through the CRA. The program started in 2005 and will last until 2040, transferring between 30–120 thousand acre-feet per year of water annually (WestGov & WSWC, 2012). Each farmer can rest from 7–23% of their land (Smith, 2011) and receive US$ 3,170 per acre (4,047 m2) tagged for (possible) resting when adhering to the Program, plus US$ 710 (in 2012, value adjusted for inflation on an annual basis) per acre annually if ‘abandoned’ (MWD, 2013).
Formal WMs have not been common in Europe (Zetland, 2011a). Spain stands out as an exception and presents a legal framework that was partly constructed to make such markets viable. The country suffers more from scarcity than other European countries, has a vast and mature water infrastructure, and has had experience of water auctions and informal WMs (Berbel, et al., 2014; De Stefano & Hernández-Mora, 2016). Among all the WM experiences studied here, this is the one that least resembles a ‘pure’ market, relying on intense governmental intervention (Embid Irujo, 2013).
Water rights in Spain (concesiones) specify nominal volumes to be extracted (and returned) by each user in years with average or above average precipitation/flows (Palomo-Hierro et al., 2015). They can be revoked, reviewed or suspended if public authorities deem it necessary (Rey et al., 2014). Since 1999, with a new Water Law entering into action, two forms of markets can be observed in the country: temporary transfers of concessions directly between users; and water banks, in which transfers are managed by public authorities (Gómez et al., 2011; Berbel et al., 2014). The fact that these provisions are present in a national law suggests that our analysis should focus at the national level, rather than in individual states (comunidades autónomas). Moreover, given the level of interregional connectivity through infrastructure, scarcity issues are not restricted to a single basin (for accounts of conflicts between states, see Palomo-Hierro et al., 2015). Indeed, recent moves in Spanish legislation have tried to harmonize regional approaches to water management, although there are different views regarding the adequacy and success of these reforms (Esteban et al., 2018).
Some joint analyses of WMs can already be found in the existing literature (see, for instance, Grafton et al., 2011), chiefly focusing on their performance. Here, our contribution is mainly to further the understanding of how and why markets have emerged and evolved institutionally to tackle issues that hinder market functioning, such as reducing transaction costs and dealing with non-consumption – aspects that have received insufficient attention so far.
Markets have evolved from existing social, political and physical conditions
WMs emerged in the three regions due to different reasons. There were existing social, political and physical conditions that facilitated or catalyzed the markets to start to operate, such as the existence of tradable rights and the need to deal with scarcity. They were not simply top-down impositions from governments, but social-political processes that evolved over time to achieve certain local needs with support from governmental inputs.
In the MDB in Australia, WMs emerged as a way to equalize different water needs among the states, and between them and the federal government. Markets developed under an increasing demand for water in different states in order to: (i) improve allocative efficiency under growing water demand by allowing users to trade and use water to generate more economic value for the country (COAG, 1994, 2004) and (ii) reduce the over-allocation of resources, by governmental purchases of entitlements for environmental uses (Wheeler, 2014).
In the Western USA, WMs were not intentionally designed at the conception of the infrastructure projects (Carey & Sunding, 2001), but naturally emerged within a clear cap-and-trade structure. Water rights that follow the prior appropriation doctrine are, at least in theory, tradable (Howe, 2011; Culp et al., 2014). In the C-BT, it was expected that project quotas would attend different classes of use, which merited the creation of homogeneous and interchangeable titles (INT1). In California, State Law explicitly states that water conserved through land fallowing is beneficial and tradable. That, in conjunction with the existing infrastructure that connects water districts and the seniority of PVID's rights, enabled the establishment of an umbrella contract between MWD and PVID to increase the supply of water for the urban areas of Southern California.
Spain has extensive water infrastructure and the costs of expanding water supply through an engineering approach are increasingly high. Therefore, WMs appeared to be a feasible alternative in order to manage water demand in the country, which already had previous experience of informal WMs (Embid Irujo, 2013). However, Spanish water rights were not designed to be traded, except in crises, and remain attached to land property (Berbel et al., 2012), even after a set of legal reforms started in 1999, partly motivated by severe droughts (Garrido et al., 2012); still, divergent views on water policy reform prevails (Esteban et al., 2018) and the risks associated with market operations within a large geographic area are perceived to be high (INT2).
Market activity varies according to the established institutions
WMs do not exist in a vacuum. They are formalized within a set of different institutions, environments and cultures that can improve or hinder their acceptance by water users (Esteban et al., 2018) and have different effects on solving water problems and user welfare (Juana et al., 2011; Bitran et al., 2014); in all cases, some adjustments were necessary either to encourage engagement in the market or allow transactions to take place. It is hard to find appropriate evidence to understand how water would have been used in the absence of WMs. But it is possible, nonetheless, to see if users are in fact trading their rights in order to better manage their demands and reduce exposure to scarcity. Market activity serves as an indicator that the benefits of trading outweigh the transaction costs necessary to finalize a deal.
In the Southern MDB, farmers have increasingly resorted to allocation trading as a way to conduct their operations, improve cash flows and reduce scarcity-related losses (Bennett, 2015). Already in Fiscal Year (FY) 2010–11, more than 60% of farmers in New South Wales, Victoria and South Australia had engaged in at least one market transaction (Wheeler, et al., 2014); that number has been roughly consistent since then and in FY 2014–15, over 25,000 trades generated a turnover of more than AU$ 300 million (ABARES, 2016). Initial opposition to reforms was less about trading rights and more about the cap on total future extractions in the basin, even though increased efficiency and the development of groundwater resources allowed total water consumption to rise in the MDB (Young, 2010).
The C-BT's market is the most active in Western USA (Libecap, 2010; Squillace, 2013); indeed, ‘users have already incorporated the market into their culture’ (INT3). Although transfer records are not readily available, some sources suggest that up to 50% of all water annually available for users is leased, particularly with farmers temporarily using municipalities' quotas (Howe, 2011). Annual leases represent roughly 80% of all transactions (WestWater Research, 2016). Quota possession gradually shifted from rural to urban and industrial uses which now account for 65% of all quotas, up from 45% in 1987 and only 5% in 1957 (Debaere, et al., 2014); despite this shift, agricultural production oscillated around US$ 2,000 million per year between 1987 and 2008 (measured in 2009 US$) (Marchlik, 2014).
The MWD and PVID program managed to become a successful experience (WestGov & WSWC, 2012) with an ‘umbrella’ transfer that enabled voluntary participation from farmers. Close to 90% of all farmers in PVID chose to be part of the program (Perry, 2015) and ‘fallowing’ represents the second biggest ‘culture’ in the district, accounting for 12% of the area available for irrigation (PVID, 2015).
In Spain, data regarding water transfers is largely unavailable (Palomo-Hierro & Gómez-Limón, 2016). Formal transactions have been rare, even during dry years, ranging from 1–5% of total consumptive use in some basins (Berbel, et al., 2014; Palomo-Hierro & Gómez-Limón, 2016). Transfers have mainly been used during crises and have not constituted a key element for water management (Gómez, et al., 2011). Only four states adopted water banks, given that their operation is legally restricted to extreme situations (Embid Irujo, 2013; Palomo-Hierro et al., 2015).
Mechanisms have been created to avoid speculation or artificial scarcity
The market institutions we studied have mechanisms to avoid speculation. Stakeholders try to steer these WMs towards their initial social and economic needs, even when this increases market inefficiencies. In the MDB, initial opposition to reform was less about trading water rights and more about the limitations on total extractions that would be allowed in the future; that is, some ‘opposed the cap, but not the trade’ (INT4). In fact, agricultural users showed that they value the opportunities provided by the market in order to design their investment strategies and manage their needs (Bjornlund & McKay, 2002). Moreover, possible impacts of water trades on third parties or the environment are treated at the local level, when users apply for licenses to use water but outside of the titling and trading systems (Grafton & Horne, 2014). Not only do these effects tend to be small (Hanak, 2003) but this arrangement also prevents market transactions from becoming overly dependent on courts and the judicial system and deals more efficiently with local externalities (Young, 2014).
In Spain, an interesting feature with regard to water management is that irrigators can form water user associations (Asociaciones de Usuarios del Água – AUAs) and cede their individual rights to them, which then manage water use for their members, including participating in market transactions with decisions made in assemblies (Rey et al., 2014). These associations are indeed responsible for a large part of market activity in the country (Hernández-Mora & Del Moral, 2015). Once a transfer is submitted, government officials have up to two months to formally reject it, for instance if the transfer will likely harm third parties; otherwise, the negotiation is automatically approved (Embid Irujo, 2013). Given all the technical, legal and even cultural barriers currently in place with regards to WMs, it is possible to deduce that trades only take place in the presence of significant differences in marginal productivity between users that lead to efficiency gains high enough to overcome the transaction costs. Indeed, water transfers are mainly restricted within AUAs (Palomo-Hierro & Gómez-Limón, 2016) and during acute scarcity (Gómez, et al., 2011).
One aspect to avoid speculation that needs to be taken into consideration when evaluating the security of water rights is the risk a user faces of losing their right/quota if they fail to consume all of the water allocated to them in a given year. Provisions known as ‘use-it or lose-it’ are common in water laws; however, they may offer a perverse incentive to use water inefficiently rather than conserving it, if proper mechanisms are not in place. This is, for instance, the case in Spain where if water is not used in three consecutive years, a user may lose their rights (OECD, 2015). In fact, users might refrain from leasing their rights through the market (even partially), afraid that local authorities will deem that amount of water as surplus to their requirements and revoke or review the rights (Palomo-Hierro et al., 2015). In the MDB and the CB-T, on the other hand, any water not used in a given year (up to certain limits) is automatically carried over to the following year (NCWCD, 2004). Users can better manage their needs over an extended period because of the two types of rights (entitlements and annual allocations). Carried water (or carryover capacity) can also be traded. In California, non-use will only be respected and not lead to a review of the water right if it is as a result of an explicit conservation measures or the intentional resting of irrigated land in order to be traded. In any other circumstance, failure to use water can lead to the suspension or revision of the right. In all cases, the ‘risk of water speculators tends not to exist, [since] speculators make money by selling water [rights] and not by hoarding them’ (INT5).
Transaction costs evolve but do not necessarily decrease over time
One of the most important issues regarding WMs is a clear definition of water rights and transaction costs. Institutions, both public and private, have the key function of lowering transaction costs (Loch et al., 2018) to improve the efficiency of transactions and sustainability of policy choices (McCann et al., 2005). Rights for water use need to be flexible enough to evolve over time to adapt to new situations. Attaching water rights to land property may provide some clarity but can stifle change. The existence of a hierarchy of rights can help to prioritize use when scarcity becomes more significant. In this sense, Australia offers a valuable lesson in its Water Reform Framework by encouraging states to: (i) unbundle water and land titles; (ii) specify water entitlements and (annual) allocations with clear definitions of possession, volumes, priority, transferability and quality (if relevant) (COAG, 1994). This unbundling also raises the value of each component of what constitutes a water right (Young, 2015). Although nomenclature may vary between states in the MDB, all states differentiate their water rights according to security levels; thus, ‘high security’ entitlements have their allocations fully delivered before ‘low security’ rights receive their annual allocations (NWC, 2013). Clear priorities help users to know when they should engage in the market. Trades cannot be restricted based on the uses and volumes involved (Grafton & Horne, 2014). Since introduction of the framework, water trades have grown considerably (Grafton et al., 2012), though the transaction costs of entitlements and allocation trading have evolved in different directions (Loch et al., 2018). Additionally, water management has become more integrated in the MDB, both horizontally between states and vertically with the federal government, since the Commonwealth Water Act (2007) which appointed the MDBA with the mandate to develop an integrated Basin Plan, and which went through discussion and public consultation rounds before entering into law (MDBA, 2010). Even though there are claims that these consultations were flawed and were held merely to appease bureaucratic procedures, they ensured that all users, independent of location, have similar rights and can trade with an increasingly larger number of individuals. In 2014, the Federal Government introduced the National Water Market System with the goal of harmonizing and consolidating all market data into a single platform, though implementation has been slow and inconsistent (Grafton & Horne, 2014).
In the USA, the C-BT stands as a peculiar case, since it emerged already with a clear cap-and-trade structure, with interchangeable and well-defined quotas that can be easily traded with relatively low costs. The fact that any potential impacts of water trading accrue to the Northern Colorado Water Conservancy District, which owns all rights to return flows, since water diverted by the Project was considered ‘foreign’ and new to the eastern side of the Rocky Mountains (Carey & Sunding, 2001; Squillace, 2013), and not to other users, helps to reduce transaction costs (Howe, 2011), although some concerns have recently been raised over quotas being ‘transferred to users who fully consume the water, eliminating return flows’ (INT6). In California, the approval process for any given negotiation can be complex, especially when it comes to guaranteeing that a transfer will not negatively affect other users or the environment. This is the case even for short-term transfers, which increases transaction costs and discourages prospective buyers and sellers (Squillace & McLeod, 2016). Nevertheless, a long-term trade between water districts, such as the PVID/MWD Program, once established, becomes a platform for users to better manage their water needs with more flexibility; the umbrella contract between both districts already cleared the initial bureaucratic hurdles and legal obligations. The fact that PVID had one of the most senior rights in the state also helped to finalize the agreement.
Lastly, Spain has taken some steps to diminish transaction costs but these steps are still limited, as suggested by the small number of transfers recorded so far. Improving users' access to water information and establishing a time limit for public officials to reject a proposed transfer represented important actions; however, water rights are still rigid, somewhat tied to land, and use-it or lose-it provisions and other regulatory limits stifle market activity. When it comes to monitoring practices, all users with a private right to use water are required to install measurement systems to determine the actual volumes used and, eventually, returned to the rivers. Interestingly, such information was already used by some basin authorities to impede or limit water transfers (INT7). Additionally, irrigators can cede their individual rights to AUAs, which then manage water use between their members, including participating in market transactions; decisions are then made in assemblies (Rey et al., 2014). These associations are responsible for most of the market activity, reducing overall transaction costs (Hernández-Mora & Del Moral, 2015; Palomo-Hierro & Gómez-Limón, 2016).
Managing a scarce resource among multiple and competing uses presents a challenge to water users and policy makers. WMs, as an allocating mechanism, will probably become more attractive as scarcity events become more frequent and intense in certain regions. However, why WMs emerge in certain contexts and not in others is still under-researched. Also, there is a significant difference between the idea of markets as proposed by economic theory and practice in the water sector (Bitran et al., 2014). As institutions, markets are humanly devised rules for the use of resources and do not always lead to efficient or effective solutions for the management of resources, since high transaction costs and other market imperfections will always exist and may increase over time (Juana et al., 2011; Garrick, et al., 2013; Dale et al., 2016; Loch et al., 2018). There have been some failures in the establishment of WMs, or their operation can lead to unequal or ineffective solutions to improve water quality or availability in a sustainable manner (Bitran et al., 2014).
This paper sheds light on how and why WMs have been established and what the features are that make them effective, based on empirical analysis of four functioning markets. As suggested by some of the experiences studied, WMs are not a panacea. Still, there are solid theoretical and even empirical foundations to justify their consideration among the mix of strategies entertained by users and policy makers to better manage water resources, particularly in scarcity-prone areas (Jamshidi et al., 2016). As a first step, it is important to recognize that ‘governments alone do not set up WMs; rather they set up the institutional conditions that allow markets to emerge’ (INT8). For this to be possible, water rights need to be clear (Young, 2015) and separated from land titles, and users should know how and when they can either use or trade their rights. These predictions from economic theory seem to be supported by our sample of cases. Additionally, homogeneous water rights tend to reduce transaction costs and allow for more users to take part in a WM. The experiences of the MDB and CB-T provide good examples of how markets can flourish and help users satisfy their water needs, while limiting unsustainable levels of extraction.
As we move towards the concept of water governance (Tropp, 2007), which implies different stakeholders having interests (Esteban et al., 2018) and power over water management, markets understood as institutions can contribute to improving water governance together with other existing mechanisms. However, the idea of markets for natural resources like water has to evolve from one of being purely an economic mechanism to manage resources efficiently, to being an institution that is embedded in the socio-political context and which can help society to steer the management of the resources more effectively. The fundamentals of economics, such as low transaction costs or information symmetry, may be important to a certain degree but are not sufficient, or even necessary, for the functioning of existing WMs which can work quite reasonably, even with their imperfections. Institutional investments in the establishment of WMs can improve the benefits of trade (Loch et al., 2018) and solve scarcity problems. In any case, WMs are necessarily embedded within larger environments and cultures and, therefore, also need to reflect societal preferences. Those regions where users and managers are more open to the understanding that water is an economic good might make better use of market instruments to flexibly reduce overexploitation of existing resources, change consumption patterns, and increase water use efficiency. Wherever opposition exists, pilot programs and gradual implementation may help to dispel doubts and criticisms. As one expert (INT9) noted, with WMs ‘if you go too fast, you might go nowhere.’
Research for this paper was partially funded by Brazil's National Water Agency, as the fieldwork was part of a study on how to establish water markets in Brazil.
Throughout this paper, we apply the term “water markets” (WMs) in a broad sense as institutions (humanly devised rules to organize the exchange of water rights) concerning any situations in which water use rights are transferred between users, associations of users or water districts. Thus, terms such as “water markets” and “water trading” are used interchangeably, following Chong & Sunding (2006).