This paper analyzes irrigator behavior in formal and informal water markets within an irrigation region of southeastern Australia to identify the extent to which these markets have facilitated the development of sustainable rural communities. The analyses show clear evidence that both the formal and informal markets have assisted irrigators in managing the significant adjustment pressure within the irrigation industry and the increased risk management burden placed on them by changing allocation policies. There is also evidence that many irrigators are active as both buyers and sellers and within both the formal and informal market, shifting their risk position within and between seasons and adjusting the allocation of their capital assets. All of these functions are essential to develop sustainable rural communities and retain community cohesion.
 Many rural communities are under stress. Their natural resource base is threatened by agricultural practices driven by short-term economic objectives rather than long-term sustainability. This is the case not least in areas dependent on irrigation. The construction of large dams and water delivery systems, extensive tree clearing, and modern intensive farming methods have all had a detrimental impact on the natural environment and its ability to produce and, consequently, a detrimental socioeconomic impact within affected rural communities [Postel, 1999; Sharma, 2000]. The pressure to expand production in order to pay for the cost of these developments has exacerbated this impact because new crop varieties have been introduced to increase yields, but they require a high level of input of water, fertilizer, and chemicals.
 Rural communities and policy makers are therefore at a crossroads, where a new path forward has to be defined that can reverse this degradation process and restore the health and vitality of the natural resource base to allow for future prosperity [Wolff and Gleick, 2002]. More water needs to be dedicated to the environment to restore ecosystems and improve water quality. Water needs to be applied more efficiently, so that less water augments the underlying water table and less water drains back into natural waterways, bringing with it residual fertilizers and chemicals as well as salt. As less water is available for consumptive uses, irrigators need to use what is left more efficiently, on higher-valued crops and more productive soils. To facilitate this process, the 1990s saw the emergence of a new water policy paradigm, treating water as an economic, rather than a social, good. This new paradigm was promoted by the United Nations at the Rio Convention on Environment and Development in 1992 [Sitarz, 1993] and by the World Bank in its Water Policy Document [World Bank, 1993]. Within this paradigm, great importance is placed on water markets and pricing policies as tools to facilitate the necessary reallocation of water in an era where no new supply is forthcoming.
 Two different water markets have emerged. In the international literature they are often referred to as formal and informal markets [Easter et al., 1999]. In the formal market the long-term entitlement to the water is traded, while in the informal market, only the right to use a given volume of water, for a given period of time, is traded. Informal markets have been widely implemented in many countries such as India, Pakistan, and Mexico [Saleth, 1998; Meinzen-Dick, 1998; Rosegrant and Binswanger, 1994]. They often take place as simple agreements between irrigators and require little administrative input. Formal markets, however, have been far less implemented because the underlying entitlement to the water is transferred, which requires that rights are registered and defined and their use monitored and measured [Bjornlund and McKay, 2002]. This requires far more complex institutional and administrative arrangements, which are not available in many developing countries [Appelgren and Klohn, 1999].
 Unregulated markets based on neoclassical economics, in which private interests, represented by willing buyers and sellers, are deciding the allocation of water, are not going to ensure water for vital social, cultural, and ecological purposes [Gleick et al., 2002]. In order to ensure this outcome, transparent markets guided by strong regulatory instruments are needed [Bjornlund, 2000; Bjornlund and McKay, 2003]. It has also been argued that pure economic and market-based approaches fall short of social and noneconomic aspirations of many developing countries [Appelgren and Klohn, 1999]. Likewise, research in the United States and Australia has shown that irrigation communities are not comfortable with water markets as the sole instrument for allocating and reallocating water [Syme et al., 1999; Keenan et al., 1999; Tisdell and Ward, 2003; Bjornlund, 2002a].
 Australia has in many ways been at the forefront of implementing the new policy paradigm and probably has the most active water markets in the world, almost entirely facilitating transfers between competing agricultural users. Markets are being relied on as a major instrument in reversing the trend of environmental degradation within the Murray-Darling Basin (MDB) [Murray-Darling Basin Ministerial Council (MDBMC), 2002]. This paper analyzes the early outcomes of an Australian water market and discusses the potential benefits and pitfalls of formal and informal water markets as drivers of sustainable rural communities. Given the significant and increasing international interests in water markets as instruments for allocating scarce water resources among competing users, the Australian experiences should be of interest to water managers and academics as well as irrigators throughout the world who are considering introducing water markets. Section 2 describes the study area and the data sources and methodologies used. Section 3 briefly defines sustainability in the context of rural communities, while sections 4 and 5 discuss the results for the formal and informal markets, respectively. Section 6 describes how the two markets interact, and section 7 outlines the pitfalls associated with the introduction of water markets.
2. Study Area, Data Sources, and Methodologies
2.1. Murray-Darling Basin
 The MDB is Australia's largest and most important river system (Figure 1). It covers most of the inland part of southeastern Australia, makes up 14% of the country's total area, supports 75% of irrigation, and provides just over 41% of total gross value of agricultural production, with a significant flow-on effect, supporting more than 1.5 million jobs, most of them in the major cities outside the basin. It also supports significant tourism, with 15 million visitors a year in its national parks and forests, and it also has important cultural, social, and environmental values. The basin thus has significant importance for all facets of Australian life [MDBMC, 2001].
 Jurisdictions within the basin have followed very different policy paths with respect to water allocation and management, which reflects the differences in agricultural productions. For the three southern states along the Murray River this is illustrated by the following facts.
 1. In South Australia (SA), irrigation almost entirely consists of horticulture and viticulture, which are dependent on a reliable water supply. The emphasis of water policies has therefore always been on security of supply rather than on maximizing the volume of water use every year. Entitlements in SA are therefore considered 100% secure.
 2. In New South Wales (NSW) most water is applied to annual crops with the ability to expand and contract production depending on water availability. Allocation policies have therefore aimed at maximizing annual access to water. This policy leaves little water in the storages at the end of each season; as a result, the security of seasonal supply is low with highly variable seasonal allocations that have a long-term average of 70%. To accommodate the horticultural industry, particularly in the downstream reaches, high-security licenses are available.
 3. In Victoria the largest water user is the dairy industry, which is heavily dependent on the security of supply because of investments in permanent pastures, herds, and dairy equipment. At the same time, they have the ability to benefit from a variable supply of extra water to produce additional feed. The allocation policy is therefore twofold. Irrigators have a water right, which is 96% secure, plus “sales” water that is announced each season as a percentage of water right (see section 2.2).
 The MDB came under stress during the 1970s, 1980s, and 1990s, with large blue-green algae blooms in the early 1990s bringing the issue to the fore. An audit of water use was therefore initiated in 1995 [MDBMC, 1995], and it concluded that the level of extraction for consumptive use was far in excess of what was ecologically sustainable and that use continued to increase. The basin spans four states, and all these jurisdictions have in the past issued large volumes of water entitlements, which have not been used or only partly used. As water markets take hold, this water is likely to be activated, further escalating the increase in use. The audit predicted significant environmental and economic impacts, if the estimated development continued. It was therefore decided to cap water extraction for consumptive use to the volume that would have been used at the 1993/1994 level of development [MDBMC, 1996]. It was recommended to accept existing entitlements, whether they had been developed or not, but it was left to the states to decide how they were going to implement the cap.
 It is generally accepted that the present cap will have to be revised and that the volume of water for consumptive uses will have to be further reduced [Department or Natural Resources and Environment (DNRE), 2001a], the question is by how much. The Murray−Darling Basin Commission (MDBC) has just started a community process “The Living Murray” [MDBMC, 2002] to determine how much more water should be set aside for environmental purposes, to secure continued prosperity within the basin, and to determine how such reduction should be paid for and implemented. These processes have generated policy uncertainty within the irrigation industry and have been a major impediment to the adoption of formal markets [Bjornlund, 2002a].
 Water markets have been an important instrument enabling irrigators to manage the impact of the cap and are anticipated to play an even more important role in facilitating the process of further reductions [MDBMC, 2002]. The MDBC has therefore established an “Inter-state Permanent Water Trading Pilot Project” within a part of the basin (see Figure 1) [Young et al., 2000] and is working on finding ways of expanding the program.
2.2. Goulburn−Murray Irrigation District (GMID) in Victoria
 The GMID is Australia's largest irrigation district with a total permanent entitlement of 1940 GL. More than 80% of the water is used within a number of gravity-fed irrigation areas, while private diverters, pumping their own water directly from the river, use the remaining water. The GMID consists of two main systems: the Goulburn System and the Murray System (Figure 2). The study area is located within the western part of the GMID and includes two sections: (1) the Torrumbarry irrigation district in the north, supplied from the Murray system and (2) the Pyramid Hill-Boort area (PHB) in the south, supplied by the Goulburn system (Figure 2). Trade is possible throughout the GMID, but a number of spatial restrictions exist so that trade can only take place if the water can be moved from the seller to the buyer. The major limitation is the restriction on trade between the Goulburn and the Murray Systems. Trade can take place from the Goulburn to the Murray System but not in the opposite direction. Initially, little trade took place in that direction because most of the high-value dairy producers are located in the Goulburn system, and allocations in the Murray system generally are higher (Table 1), which has eased demand and reduced prices [Bjornlund, 2002b]. In December 1994, the trading regulations were amended to allow trade from upstream to downstream of Nyah (Figure 1). From 1997 to 2000, 25,000 ML were traded from the Goulburn system to downstream of Nyah mainly to develop new vineyards [Bjornlund, 2003a]. Following this development, Goulburn-Murray Water (GMW) was given the authority to allow upstream trade into the Goulburn system in substitution for downstream trade. This rule has been invoked twice in 2001/2002 and 2002/2003.
Table 1. Relationship Between Seasonal Allocations and Extent of Tradea
This table is based on the work of DNRE [2001a] and updated on the basis of Goulburn-Murray Water's records.
Allocations are announced as a percent of water right at the beginning of the season and then adjusted monthly during the season. These are the maximum allocation level reached for each season.
This is the total temporary trade as percentage of total water use for each season.
 Both sections have large areas with very high soil salinity levels caused by the impact of inefficient irrigation and high saline water tables. This has had a significant impact on the productivity of the area and the ability of farmers to stay in business and prosper. Land uses in the area have traditionally been low-value mixed farming, with broad acre cropping and annual pastures for cattle and sheep production, with pockets of high-value dairy production, especially around Cohuna in the Torrumbarry section [Bjornlund, 1999]. Irrigators within the study region have therefore been under constant adjustment pressure.
 To understand the outcome of trade, it is important to discuss the structure of water entitlements. All irrigators have a water right with a high level of reliability of supply with full delivery in 96 out of 100 years; irrigators pay for this water every year, whether they use it or not. In addition, irrigators have access to “sales” water. The level of sales varies from year to year, depending on availability, and is announced as a percentage of water right; irrigators only pay for this water if they use it. The high level of reliability of water rights is maintained by only announcing sales once the next season's water rights are secure in the reservoirs. Historically, sales have been very high, exceeding 100% of water right in most years. Dairy farmers have therefore developed their properties with infrastructure depending on sales of 60%, giving them access to a volume of water equivalent to 160% of water right. However, in recent years, sales have been declining (Table 1) because of (1) general resource constraints, caused by a long period of drought; (2) water trading, which has activated previously unused water and thereby increased use; and (3) the impact of the MDB cap. To stay within the cap while maintaining a high level of supply certainty of water rights, sales have been reduced [MDBMC, 1996], and access to sales has been limited to 30% if the irrigator sells any water during a season [DNRE, 2001a].
 The water allocation is announced at the beginning of the season and is then revised on the 1st and 15th of each month as additional water enters the reservoirs. This represents a significant change in policy. Traditionally, the authorities made one allocation at the beginning of the season on the basis of what was available in the reservoirs and expected inflows. This has significantly shifted the risk management burden from the authorities to the individual irrigators.
2.3. Trade Within the GMID
 Informal markets were first introduced by a legislative amendment in 1987. Formal markets were introduced with the Water Act 1989 and made effective by regulations late in 1991. Trade on the informal market started at about 25,000 ML per year during the first 7 years and then increased markedly in 1994/1995 and has, since 1997/1998, stayed above 200,000 ML or around 10% of total entitlement (Figure 3). Trade on the formal market was also subdued during the first years, then surged in 1997/1998, and has since been at a level between 17,000 and 24,000 ML or just around 1% of total entitlements. The low volume of trade in the formal market should be taken into account when reading the discussions in section 4. The high volume of trade in the informal market indicates that this market has a far more profound impact within the irrigation industry, which places increased importance on the discussions in section 5. A water exchange was introduced in 1998/1999 to facilitate fast, easy, secure, and cheap trading in the informal market. This exchange has been very successful, and it has extended its market share from 15% during the first year to a level around one third of all informal trade [Bjornlund, 2003b].
 The surge in trade has been caused by a number of factors: (1) a 5 year drought with no sales within the Goulburn system, (2) the impact of the cap, (3) the relaxation of trading rules in 1994/1995 [Bjornlund, 2002b], (4) the fact that irrigators have increasingly become familiar with water trading and aware of the potential benefits [Bjornlund, 2003a, 2003b], and (5) the initial success of water markets activating large volumes of unused water, augmenting resource constraints.
2.4. Data Sources and Methodologies
 The discussions in sections 4, 5, and 6 will be supported by empirical and anecdotal evidence based on both quantitative and qualitative data and gathered over a 10 year period. As is apparent from the text a number of papers have been published based on the quantitative part of this material to illustrate various aspects of water markets and their operations. The strength of this paper is that it, for the first time, tries to consolidate all of these findings to explore the extent to which the operations of existing markets have facilitated the development of more sustainable rural communities. It does this by supporting the previous quantitative data with new qualitative data gathered during 2002 and 2003. As a consequence of this history a number of methodologies have been used to support the discussion in each section of the paper. The use of these methods has been justified and explained in the paper in which they were first published as referenced in the text.
 The material on which this paper is based is as follows: first is full day workshop with 14 key stakeholders in the irrigation industry including irrigators, water brokers, senior staff from the relevant government department, Goulburn-Murray Water (GMW, the authority managing the GMID), Victorian Farmers Federation, and one of the largest milk-processing factories in the region as well as a representative from the environmental movement held in Tatura, Victoria in March 2002. Second is 12 focus groups held during April 2003 with irrigators within the study region with different experiences in the water market. The objectives of the focus groups were to get a more detailed understanding of how irrigators used the formal and informal markets, what they expected to achieve, what triggered their decisions to buy and sell water in both markets, and how they perceived the two most contemporary and contentious issues: the separation of land and water rights and environmental flows. The focus groups followed a set structure to ensure that all the issues were covered in each group. The discussions in the focus groups were conducted by a professional facilitator with extensive experience in facilitating community meetings and focus groups with respect to natural resource issues. A summarized transcript of the discussions was entered directly into a computer and displayed by a data projector so that the participants could follow the discussions and the record that was being kept. Focus group transcripts were analyzed using Nudist N6 software. The third source is telephone interviews with (1) 100 buyers (49%) and 100 sellers (63%) using the informal market during 1998/1999, (2) 100 irrigators who have never participated in water trading as of July 1999 (26%), and (3) 100 sellers (41%) and 100 buyers (35%) using the formal market during 1994–1996. The samples interviewed represent a significant proportion of the respective populations (as indicated above 26–63%); the findings of the analyses should therefore reflect the outcome of trade within the study region. These quantitative data were entered into SPSS and analyzed using descriptive statistics and significance tests. These analyses are used to support the discussions in sections 4.2, 4.3, 4.4, 5.1, 5.2, and 6. Factor and cluster analyses were used to support the discussions in section 5.3; these methodologies were described by Bjornlund [2002c]. Next is the trading and water rights registers of GMW (used to produce Table 1 and the statistics in section 6). Last is the records of the Northern Victorian Water Exchange during the first 5 years of operation 1998/1999 to 2002/2003 (used to create Figures 4 and 5).
 Among many other questions the irrigators were asked to rate how important a number of reasons for buying and selling were in their decision making process to use the informal or formal market. A 1 to 5 scale was used, with 1 being not important and 5 being very important. In the discussions in sections 5 and 6, when reference is made to irrigators finding a certain reason important, they gave that reason a rating of 4 or 5.
3. Sustainable Rural Communities
 The issue of sustainable rural communities has been widely debated in recent times [e.g., Lockie and Bourke, 2001; Gray and Lawrence, 2001; Beer et al., 2003; Cheers, 1998; Martinez-Brawley, 2000]. Rural sociologists do not normally define sustainable development as some well-defined end-point; rather, they consider it as a multidimensional process toward socially, economically, and environmentally strong communities, with a flexible economy not totally dependent on the seasonal price for specific commodities [Gray and Lawrence, 2001]. Lockie [2001, p.229] provided a good definition of sustainability when he stated that “…in very general terms, sustainability is defined as the ability of current generations to meet their needs and aspirations without compromising the ability of future generations to do the same thing.” Following these definitions, policies that promote development of farming practices that reduce the negative environmental impact of farming and the abandonment of old damaging practices must be said to drive the process toward more environmentally sustainable communities; those that promote more efficient use of water as well as production of higher-valued commodities must be said to drive the process toward environmentally, socially, and economically stronger communities. These outcomes will increase the value of output per unit of water used and will reduce saline return flows to the river and reduce leaching of water into the aquifers. High-value commodities are often more labor intensive both on farm and off farm since they are often associated with processing, packaging, and transport industries. Such commodities also depend on more sophisticated service industries such as consultants, technicians, accountants, lawyers, etc. These professions offer alternative opportunities for the young generation and thereby make it more attractive for them to stay in the community. Policies that promote more flexible use of the farmer's resources must also be said to drive the process toward socially and economically stronger communities. More flexible resource management will enable the farmers to better manage their farm businesses on an ongoing basis and therefore make them less dependent on fluctuating annual commodity prices and climatic conditions. Such policies might also allow unviable farmers better avenues to fully or partly exit farming, while remaining within the local community. This will facilitate the process of maintaining community cohesion and thereby drive the process toward socially stronger communities. The discussions in section 4, 5, and 6 will analyze how water markets in Australia has assisted farmers and thereby their communities in the above processes and thereby have driven the process toward more sustainable rural communities.
4. Formal Markets
 It has been argued that irrigators are unlikely to make the significant financial commitments associated with a change to more efficient and higher-value use unless they have the long-term control of adequate water resources to protect the investment [Crase et al., 2000; Bjornlund and McKay, 2001]. It has therefore been of some concern that formal markets have not been more widely adopted, and it has been argued that economic benefits have been foregone or postponed as a result [Marsden Jacob and Associates, 1999]. The major benefits of the formal market are associated with a reallocation of water to (1) more productive soils in more suitable locations, (2) more efficient water users, (3) higher-valued uses, and (4) new developments and the consolidation of water into larger more viable units. The net effect of such reallocations is structural change and an increase in the volume of output as well as the gross margin per unit of water used. Such developments will increase employment and economic activity and improve the social and economic prospects of the affected communities. The reallocation to more suitable soils and more efficient irrigators will also produce environmental benefits [Cullen et al., 2000]. In other words, it will reduce the negative environmental impacts of irrigation. These developments, if achieved, will drive the process toward more sustainable rural communities.
4.1. A Reallocation of Water Into More Productive Soils in More Suitable Locations
 This outcome of trade is partly secured by regulation. Trade will only be allowed to move additional water onto a property if it has soils with a low salinity level. Further, some areas along the Murray River have been zoned into low- and high-impact zones. Trade will not be allowed into a high-impact zone, while trading out of these zones is encouraged by a government subsidy of A$50/ML. Trade into low-impact zones is allowed, but developers are required to pay a levy of A$130/ML to cover the capital cost of salt interception schemes to offset the impacts that might be caused by the increase in water use in that location [Government of Victoria, 1993]. Early markets showed a shift of water away from highly saline to low saline soils within the PHB and Torrumbarry areas and from saline areas in PHB into better soils in the eastern part of the Goulburn System [Bjornlund, 1999; DNRE, 2001a].
 To encourage the reallocation of water into more suitable soils, the Victorian government has promoted the development of salinity plans for some of the worst affected areas. One of the purposes of these plans is to identify the saline areas and the areas most suitable for development. Detailed farm analysis by Barr  showed that half the saline soils irrigated in 1989 were no longer irrigated 5 years later. Water had been reallocated to better areas on the same farm or had been sold to other farms with better soils. These results indicate that the combination of appropriate regulation and trade has driven the process toward environmentally sustainable rural communities.
4.2. Reallocation to More Efficient Water Users
 Within the study area most water is supplied by gravity and applied by flood irrigation. Water use efficiency is therefore determined by the extent and quality of laser grading, reuse systems, surface drains, and access to off-farm drainage. Again, the water-trading regulations, to some extent, ensure that the buyers are efficient users. The regulations set maximum water use limits for properties depending on water use efficiency. To be allowed to apply the maximum volume of water of 10,000 ML/ha, the irrigator must have access to off-farm drainage and have a reuse system [DNRE, 2001b]. For new developments the developer must produce a whole farm plan demonstrating best practice irrigation and drainage methods.
 The interviews of buyers and sellers in the formal market showed that (1) 82.6% of the water bought moved onto farms with a reuse system, while 63.5% of the water sold moved out of farms without it; (2) 90.7% of the water bought moved onto properties with laser grading, while 45.6% of the water sold moved out of farms without it; and (3) 89.4% of the water bought moved onto farms with surface drainage, while 29.8% of the water sold moved out of farms without it [Bjornlund and McKay, 2000]. Since 1996, there has been a trend toward water going to new greenfield developments such as vineyards, olives, and vegetable production [DNRE, 2001a]. These operators all have relatively high tech efficient irrigation. These results indicate that the combination of appropriate regulation and trade has driven the process toward environmentally, socially, and economically sustainable rural communities.
4.3. Reallocation to Higher-Value Uses
 The potential social gains from this reallocation are linked to the heterogeneity of water values across commodities. In Australia, if water moves out of grazing for sheep or cattle, which has a gross margin at about A$50/ML, or grain growing at about A$100/ML and into wine grapes at about A$2800/ML, various vegetable productions at A$400–1800/ML, or dairying at around A$500/ML, the economic benefit to the farm economy is evident. In addition, it has been estimated that for each 1000 ML traded into horticulture, 30 new jobs will be created on farm and within processing and support industries, while for dairying the job creation is about 15 per 1000 ML [DNRE, 2001a].
 Interviews with buyers and sellers in the formal market showed that most of the water traded had no impact on the production of the buyers and sellers: 58.3% of the water sold did not cause a reduction in the seller's irrigated area, while 56.8% of the water bought did not cause any expansion of the buyer's irrigated area. Sellers disposed of unused water or expected to increase their reliance on the informal market to maintain production, while buyers purchased water to maintain existing production because sales had declined. However, 75.4% of the water was sold out of cattle, sheep, and grain-producing properties, and 8.6% was sold out of properties with no irrigated production at all. Of the 41.7% of the water sold causing a reduction in the irrigated area, 63.3% of the reduction took place in cattle and cropping production. Likewise, 69.0% of the water was purchased by dairy farmers, which is the group that has traditionally developed their properties to depend on 60% sales and, as a consequence, is suffering most from reduced sales. Of the 43.2% of water purchased for expansion most of it, 67.4%, was used to expand dairy production. As mentioned above, since 1996, large volumes of water have gone into horticulture and vegetable production, which are higher-value uses than dairying.
 It also has to be acknowledged that not only water used for expanding production has this effect. If the dairy industry had not been able to purchase this water, it could not have maintained production, which would have resulted in loss of farm income and associated loss of employment and economic flow-on effects within the community. These results indicate that water markets have driven the process toward socially and economically sustainable rural communities.
4.4. New Developments and a Consolidation of Properties Into Larger More Viable Businesses
 Analyzing the property and entitlement sizes of the buying and selling properties as well as the farm expansion or reduction caused by trade, it is apparent that the buying properties have become larger and the selling properties have become smaller [Bjornlund and McKay, 1999]. There is, however, no evidence to suggest that as of 1996, trade has resulted in a movement of water from small family farms to large corporate farms. There is, however, evidence to suggest that trade has generated a polarization of the farming community into a group of larger more viable family farms, dependent on a nonfamily workforce, and a group of smaller unviable family farms, dependent on off-farm work. There is also evidence to suggest that the polarization is between larger water-rich farms, which can better sustain periods of resource scarcity, and smaller water-poor farms, which are dependent on annual purchases of water and therefore more at risk during periods of resource constraints and associated high water prices (during most of the season 2002/2003 prices on the exchange were between A$350/ML and A$500/ML, Figure 5).
 Among water buyers there is, however, also evidence of growth in the number of properties with an unviable size of <20 ha [Bjornlund and McKay, 1999]. This growth expresses an increase in the demand for life style farms coming from people working within the growing support industries for the irrigation sector. This in its own right has an important function in the process toward more sustainable rural communities. This demand for smaller properties has allowed exiting farmers to sell properties for which there would have been no demand from expanding commercial farmers, for whom additional improvements have little or no value. It has also allowed such farmers more flexibility in their exit strategy by selling their dwelling to life style farmers and the best agricultural land and the water to neighboring farmers, thereby maximizing the return on their asset and putting them in a better position when relocating. These results indicate that water markets have driven the process toward socially and economically sustainable rural communities but also indicate that there might be some less favorable outcomes with long-term socioeconomic impacts on some farmers selling water.
5. Informal Markets
 Seasonal transfers should be able to assist farmers in managing fluctuations in commodity prices and water supply from year to year and managing periods of personal hardship such as illness. It should also assist many farm families to remain on the land and within the community by generating an income from water sales, off-farm work, and possibly some farming, while assisting other expanding or new irrigators in the development phase by giving them access to water without having to invest their available capital in expensive long-term entitlements. Such outcomes would drive the process toward socially and economically sustainable communities.
5.1. Allow Irrigators to Respond to Annual Fluctuations in Commodity Prices
 When buyers and sellers were asked why they bought and sold water during 1998/1999, the impact of fluctuating commodity prices was apparent: 26% of the buyers quoted good commodity prices and 16% cited a one-off opportunity to sell more of a certain crop as important reasons for buying, while among the sellers, 36% referred to low commodity prices as an important reason for selling. During the workshop in March 2002 and the focus groups in 2003, the ability to respond to fluctuating commodity prices was also mentioned as an important function of informal markets. Commodity prices have two potential impacts: (1) irrigators react to the prices of the commodities they are growing and decide to sell when the price of water on the informal market exceeds their gross margin per megaliter, while they decide to buy when the price drops below that amount; and (2) dairy farmers react to the prices of substitute feed, such as feeding grain, silage, or hay. Extension officers had advised that at A$90/ML, dairy farmers would be better off buying feed; hence prices during 1998/1999 and 1999/2000 flattened out at A$90/ML (see Figure 4).
 Traditionally, grain, cattle and lamb producers are the major sellers of water. However, during 2000/2001, prices for cattle, lamb, and grain were high, causing these producers to be reluctant to sell. As a consequence, supply was low, forcing prices up. Not until prices reached A$130/ML in early January did sellers in significant numbers appear in the market. For several reasons, prices reached A$200/ML later that season, one of which was that prices of substitute feed were very high since supply was low because of the drought. This increased the comparable price of water to produce grass; at the same time, prices for dairy products were high, which increased dairy farmers' willingness to pay. In conclusion, informal markets have significantly improved irrigators' ability to respond to changing commodity prices, which has driven the process toward socially and economically sustainable rural communities.
5.2. Allow Irrigators to Respond to Fluctuations in Supply and Demand
 The workshop and focus groups emphasized that one of the most important functions of informal markets during the last 5 years has been to assist farmers managing the drought. When allocations are relatively high, trade accounts for around 5% of total water use, but during the last 6 years, when there have been no sales, trading has slowly increased its contribution to total water use to 24% (Table 1). Reflecting this, 82% of the buyers in the informal market during 1998/1999 said that an important reason for doing so was that sales were very low that year. At the beginning of 1998/1999 the allocation was at record low levels. As a result, horticultural growers bought in panic to secure their permanent plantings, which forced up prices. Once allocation levels increased, prices dropped back to A$90/ML (Figure 4). Horticultural growers again reacted to low allocations during 2002/2003 when opening allocations were very low and GMW gave a very low probability of allocations getting much higher. Most horticultural growers anticipated an allocation of 60–70% and entered the market to buy water based on that assumption. This demand caused prices to increase to record levels of A$500/ML, forcing most other irrigators out of the market (Figure 5). As the horticultural growers satisfied their demand, prices decreased. However, it soon became apparent that allocations would not reach 60%, and horticultural growers again entered the market to protect their permanent plantings, forcing prices back up to A$500/ML. When that demand was satisfied, prices again dropped but not down to the levels of the first 4 years.
 Resource constraint is not only caused by the allocation level but also by natural precipitation during the season. If rainfall is low or evaporation is high, demand for water will increase. Trade on the exchange over the first 5 years shows evidence of this (Figure 4). At the opening of 2000/2001, spring rain was good, reducing the need to irrigate, and prices therefore remained low, whereas at the opening of 2001/2002, the spring season was very dry. Water use was up by around 15%, compared with “normal” years, and this resulted in increased demand for water on the exchange. This was one of several coinciding factors causing prices to be exceptionally high during that year [Bjornlund, 2003b]. Also in that year, opening winter rains were late, contributing to prices remaining high toward the end of the season. When good rains finally arrived in May, prices dropped. In conclusion, informal markets have significantly improved irrigators' ability to respond to changing resource availability and have driven the process toward socially and economically sustainable rural communities.
5.3. Maintain Community Cohesion
 The use of multivariate techniques to analyze the responses from buyers and sellers using the informal market during 1998/1999 and from irrigators who have never participated in any kind of water trading provided some interesting insight into how different sectors of the irrigation community use water markets to cope with adjustment pressures [Bjornlund, 2002c]. Three main groups emerged representing the socioeconomic structure of the community: the nonadjusters or strugglers (57%), the adjusters or defenders (35%), and the comfortable farmers (8%) (Figure 6).
 The nonadjusters or strugglers make up 57% of all farm businesses. They have given up developing their properties to be viable in the long term; they are simply trying to maintain the farming life style and postpone exit adjustment to the time of intergenerational change and stay within the community. The existence of this group is clearly illustrated by the fact that only 39% of the sellers, 48% of the buyers, and 26% of the nontraders expected family continuity of the property, while ∼15% expressed uncertainty. The buyers in this category use the informal market to replace what they used to get as sales, that is, to maintain their production. They have low farm incomes and a high dependence on off-farm work. The sellers are in two subgroups: (1) The first is “professional water sellers;” 15% of all sellers on the informal market are selling all their water every year. They are selling to generate an income, which combined with off-farm work and possibly some dry-land farming, secures their annual income and makes it possible for them to stay in the community. (2) The second group is struggling farmers, who are selling a large proportion of their water every year. They have given up developing their properties to be viable in the long term but have not given up irrigated farming all together. They are able to stay on the property by some low-value production, annual water sales, and a large dependence on off-farm work.
 The adjusting irrigators are using the informal market to assist them in the adjustment process and are generally higher-value users. The buyers do not have enough water for this process, and they cannot afford to buy water in the formal market because all their available finance is tied up in other parts of the adjustment process. This group was clearly reflected when the buyers during 1998/1999 were interviewed: 62% said that they actually need the water every year but could not afford to buy it. The sellers have water enough for the development of their property and are able to use the informal market to generate some income to assist them in the adjustment process.
 The comfortable irrigators generally have larger properties with higher farm incomes and low off-farm work dependence. They are opportunistic traders, buying water in the informal market if commodity prices relative to the price of water are good and selling if prices are low. This discussion indicates that the informal market has several important ways in which it drives the process toward economically and socially sustainable rural communities and maintain community cohesion.
6. Working Together
 Analyses of water rights and water trading registers indicate that farm businesses cannot simply be classified as buyers or sellers of water. In total, 23% of all farm businesses have both bought and sold water in the informal market. This reflects the findings that farmers respond to annual fluctuations in prices and supply as discussed above. Looking at the informal market during 1998/1999, it was found that 16% both bought and sold water during that season, with half of them net trading nothing, that is, ending the season with the same volume as they started. They shifted their risk position during the season, buying early when allocations were low, and selling again as allocations increased or they speculated in price developments during the season.
 The analysis also indicated that a large section of farm businesses trade both in the formal and informal market: 26.2% within the PHB and 17.4% within the Torrumbarry area. These farmers are changing their risk position and adjusting the capital assets that they have tied up in water rights. Cross tabulating the activities in the formal and informal markets revealed the following: (1) Some 39.3% within Torrumbarry and 65.6% within PHB selling water in the formal market also buy water in the informal market. This indicates that there is a group of irrigators shifting their risk position to rely more heavily on annual purchases in exchange for a cash injection. This clearly was more predominant within PHB. This reflects that 61% of the sellers in the formal market said that an important reason for selling was that they needed the money. (2) Some 74.7% within Torrumbarry and 88.5% within PHB of those selling water in the formal market also sold water in the informal market. This indicates that there is a group of irrigators who have sold some of the water that they traditionally have not used and who sell the remaining excess in the informal market; in effect, they have not shifted their risk position but cashed in on an unused asset. This reflects that 69% of the sellers in the formal market said that an important reason for selling was that they did not need the water. (3) Some 84.8% within Torrumbarry and 79.7% within PHB of those buying water in the formal market also bought water in the informal market. These irrigators have shifted their risk position to rely less on the informal market. They have, in effect, bought some drought insurance but could not afford full coverage. This reflects that 60% of the buyers in the informal market said that an important reason for using the informal market was that they could not afford the prices in the formal market, despite the fact that they needed the water every year. (4) Some 53.2% within Torrumbarry and 60.9% within PHB of those buying in the formal market also sold water in the informal market. These irrigators have, in effect, bought full drought insurance and now sell water annually when they do not need it. The above clearly shows that irrigators use both formal and informal markets to shift their risk position during individual seasons, between seasons, and on a long-term basis, which has driven the process toward socially and economically sustainable rural communities.
7. What are the Pitfalls?
Sections 4, 5, and 6 have illustrated how water markets have helped irrigators in managing changes in policy, economic, and climatic conditions and, consequently, have assisted communities in the process toward economic, social, and environmental sustainability. However, the introduction of water markets is not without risk to the sustainable future of irrigation communities. The focus groups and workshops showed a very high level of concern with the potential negative impacts of the formal market. The problem is not that irrigators fail to see the benefits from this trade but, rather, that they can see some potentially serious negative community impacts as well. The following are some of the identified concerns, which provide important lessons for policy makers, researchers, irrigators, and water managers to consider before introducing water markets in other regions.
 1. If large volumes of unused entitlements exist at the time trade is introduced and use is not capped, then trade will activate unused water, resulting in environmental degradation.
 2. If large volumes of unused entitlements exist at the time trade is introduced and use is capped, then trade is likely to activate such unused entitlements, which will reduce water availability for existing users. This has been the case within the MDB and has caused hardship and resentment within irrigation communities [Bjornlund, 2002a]. Many farmers see it as wealth transfer when they have to pay their neighbors for water they have traditionally received via the allocation process.
 3. If trading moves large volumes of water out of an irrigation area, it will have significant community implications: (1) Economic activity in the area will be reduced as farm output declines, which in turn results in fewer jobs both on and off farm and a loss of local businesses services as well as a decline in the population base. (2) Farms will be left dry, which can have a negative impact on neighboring farms as weeds and pests infest the abandoned land and spread to other farms. (3) Property values decline when water is sold off; this will reduce the revenue from rates and taxes for local councils, resulting either in a reduction in services or an increase in the rate burden for other farmers. Such developments will result in unsustainable communities and could eventually result in community disintegration.
 4. If trading moves large volumes of water out of an irrigation area, the volume of water used within the system will be reduced and thereby increase the burden of maintaining the infrastructure for the remaining irrigators. Eventually, the supply channels may become unviable, and the authority will be forced to stop supply; this will leave remaining properties dry. This is a major impediment to formal markets in Australia and has been dealt with differently in each state [Bjornlund, 2002d]. It is also one of the main problems with the wider implementation of interstate trading [DNRE, 2001a].
 5. If trading moves water into certain channels, the supply pressure will increase, with the result that supply reliability is reduced for existing irrigators. Within the study region, channel capacity is one of the assessment criteria that the authority uses when approving trade. Trade will not be allowed into a channel if the supply reliability of existing irrigators will be affected. However, many irrigators now relying on purchases on the informal market are concerned that other irrigators on their channel will buy water on the formal market and thereby commit the full supply capacity of the channel. This would effectively prevent them from buying water on the informal market in the future.
 6. If the point of extraction is moved, it can have third-party effects. If water has to be transported longer in the river, more evaporation losses or seepage may occur before it reaches the new user. The question is, who is going to carry that loss? If water is traded upriver, sources of supply can be cut off, which reduces supply options and thereby the supply reliability of existing irrigators.
 7. If water is traded upstream, dilution flow from the new point to the old point of extraction will be reduced, with potential impact on river salinity. Further, some locations produce more saline drainage back into the river because of higher natural salinity levels of the groundwater. If trade moves water from areas with low salinity return flow to areas with high salinity return flow, it will have an impact on river salinity levels. Both of these impacts have been identified in South Australia [Bjornlund and McKay, 2000].
 8. If trade moves water to more efficient users, return flow will be reduced; this has some environmental benefits but also decreases supply for downstream users and reduces total river flow with negative environmental impacts.
 This paper identifies several ways in which Australian water markets have facilitated the process toward more sustainable rural communities guided by strong regulatory frameworks. It also identifies that formal and informal water markets play separate but equally important roles in this process and that the practice of using both markets has been widely adopted to shift irrigators' risk position and to manage increased supply uncertainty. Irrigators clearly acknowledge that without water markets they could not have maintained production during the last 5 years of drought and communities would have suffered much more than they have.
 Water markets have achieved these outcomes by assisting irrigators in managing resource constraints, increased supply risk, and the general adjustment pressures within agriculture. Markets have allowed many farm families to remain in the community while partly or fully exiting irrigation. This has significantly helped individuals and their communities to adjust over a long time period, rather than forcing a large number of irrigators out of business and out of the communities.
 These findings clearly illustrate the potential positive outcomes that can be expected if trade is introduced into other regions of the world. However, it should be noted that several of these positive outcomes are contingent on proper regulatory instruments.
 The introduction of markets therefore holds significant promises for regions of the world with a semiarid climate like that of Australia's. However, the Australian experiences also show that water markets can have a negative impact on the sustainable future of rural communities if regulatory instruments and policies do not properly guide trade and that several pitfalls exist during the process of introducing water trading. These experiences and pitfalls should be taken into account by water managers and policy makers planning to introduce water markets: the water market can be a very good servant to move water around between competing uses and drive the process toward sustainable rural communities, but if left to its own forces, it could prove a very unforgiving master.
 This research project is funded by the Australian Research Council and 10 industry partners: Department of Water Land and Biodiversity Conservation, Department of Primary Industries, SA−Water, Central Irrigation Trust and River Murray Catchment Water Management Board in SA, Goulburn−Murray Water, and Department of Natural Resources and Environment in Victoria, Murray Irrigation Limited, and Department of Land and Water Conservation in NSW, and the Australian National Committee on Irrigation and Drainage. The interviews were conducted as part of a project funded by Land and Water Australia held jointly with Jennifer McKay. I would also like to acknowledge the advice provided by Brian Cheers, associate professor or Rural Sociology.