Note: Where available, the PDF/Word icon below is provided to view the complete and fully formatted document
Third-party effects of water trading and potential policy responses.



Download PDFDownload PDF

A key feature of water policy reform in Australia has been the separation of water access entitlements from land titles and the establishment of markets for water. However, the separation of water entitlements from land entitlement is not a suffi cient condition to ensure that water markets are complete. In the absence of fully defi ned property rights, water markets will be incomplete and trade has the potential to create third-party effects. However, property rights over natural resources are seldom completely specifi ed, because of the existence of transactions costs. The benefi ts from eliminating third-party effects need to be suffi cient to warrant intervention.

The objectives in this paper are to identify key potential third-party effects of water trade under existing property rights structures in Australia, and to ex-amine policy responses to address these effects. The discussion draws on the concepts of exclusiveness and rivalry as key characteristics that determine the applicability of property rights solutions to third-party effects of trade.

It is likely that many of the third-party effects of trade discussed in this paper do not warrant policy intervention at the national or state level. In some instances, effects are likely to be relatively minor, although some may be signifi cant at the local level. The costs of addressing some third-party effects may outweigh the benefi ts. Where there are signifi cant gains from trade, the existence of these third-party effects should not been seen as a reason to impede trade. There are fi rst-best policy instruments to address these effects at an appropriate scale.

1 Australian Bureau of Agricultural and Resource Economics 2 Productivity Commission* * The views expressed in this paper are those of the authors and do not necessarily refl ect those of the Productivity Commission.

CONFERENCE PAPER

American Agricultural Economics Association Providence, Rhode Island, 25-27 July 2005

Anna Heaney1, Gavan Dwyer2, Stephen Beare1, Deborah Peterson2, Lili Pechey1

Third-party effects of water trading AND POTENTIAL POLICY RESPONSES

2

water trading

Introduction

The focus of water policy in Australia over recent decades has changed from assisting ongoing resource development for consumptive use, to managing competing demands for a fully allocated resource. A turning point in Australia’s water policy occurred in 1994, when the Council of Australian Governments (COAG) agreed to a water reform framework that recognised that management of Australia’s water resources as an issue of national impor-tance.

A key feature of the 1994 reforms included the introduction of a cap on diverting water from the Murray‐Darling Basin, the area with the largest agricultural water use in Australia. The objective of the cap is to achieve a balance between the economic and social benefi ts of water resource development and the provision of water for ecological purposes.

The importance of water resources for ecological purposes was highlighted in the 1994 COAG reform framework through explicit provision for environmental water. Accordingly, there has been a ‘clawing’ back of entitlements to water in river systems that are currently considered overallocated and overused.

The cap on diversions and the provision of water for the environment have been accom-panied by the separation of water access entitlements from land titles and the establish-ment of a market for water trade. Trade can ameliorate the economic and social effects of entitlement reductions by allowing water resources to move from water uses that generate relatively low economic returns to those that generate greater returns. In this way, trade can minimise the economic cost of providing additional water for the environment. Water property rights will, therefore, play a pivotal role in ensuring effi cient water use.

However, the introduction of tradable water rights has been problematic and trade has been constrained by a number of institutional issues. The 2004 National Water Initiative (NWI) was designed to complement and extend the reform agenda commenced in 1994 and seeks to further expand water trading. It contains actions to be implemented as priorities by the Commonwealth, state and territory governments over the next ten years. The objectives of the NWI are to achieve:

A nationally compatible market, regulatory and planning-based system of managing surface and groundwater resources for rural and urban use, that optimises economic, social and environmental outcomes, and is able to adapt to future changes in the supply of, and demand for, water. (COAG 2004, p. 1)

The separation of water entitlements from land ownership is a necessary but not suffi cient condition to ensure that a water market is complete. In the absence of fully defi ned property rights, trade has the potential to create adverse third-party effects that prevent the benefi ts of trade from being fully realised, or have distributional effects that can have impacts on the wealth of other water users. The existence of some third-party effects has been raised as a reason to impede trade.

The objectives in this paper are to identify key potential third-party effects of water trade under existing property rights structures in Australia and, where possible, to assess the

water trading

3

relative signifi cance of these effects. Examples are drawn from the southern Murray‐ Darling Basin system.

In the fi rst section of this paper, water property rights, water use and trade are described. It is observed that the separation of water entitlements from land failed to take into account a number of characteristics that were implicit in the joint right. In the second section, a description is provided of four third-party effects of water trade that are related to reliability of supply and delivery, storage and delivery charges and water quality, which have been raised as potential impediments to trade (Goesch and Beare 2004). Available policy instru-ments to address these effects are considered in the third section, drawing on a framework using the concepts of exclusiveness and rivalry (Randall 1983). Most effects are found to be amenable to property rights solutions, with the exception of effects on water quality. The relative importance of each effect is then discussed in the fi nal section, to help policy makers determine whether potential reforms are worthwhile and to help prioritise reform efforts. Some of the effects, such as those on delivery reliability, are likely to be relatively localised and small in terms of scale, and the costs associated with policy intervention are likely to outweigh the potential benefi ts.

1. Water property rights and use Irrigated agriculture contributes just over a quarter of the value of agricultural production in Australia, or around $9.6 billion a year (ABS 2004). The focus in this paper is on the southern Murray‐Darling Basin region, located in the south east of Australia and extending across the jurisdictional boundaries of Victoria, New South Wales, the Australian Capital Territory and South Australia. The region:

■ accounts for a large proportion (around 70 per cent) of irrigated agriculture in Australia

■ is characterised by the provision of large scale public and private infrastructure to regu-late water delivery on a district basis to farms and

■ is hydrologically linked, enabling intraregional (within valley) and interregional (between valleys or states) water trading.

The region covers eight irrigation districts supplied by two main river systems: the River Murray and the Murrumbidgee River. A schematic diagram of the area under consideration is shown in fi gure A. Most of the irrigation water in these districts comes from large storages, with diversions and delivery infrastructure managed by private and public utilities (Appels, Douglas and Dwyer 2004). Three irrigation companies account for about 75 per cent of water supply in the southern Murray‐Darling Basin (table 1).

1 Major irrigation scheme entitlements, southern Murray-Darling Basin, 2003-04

Irrigation company Entitlement a

GL

New South Wales Murray Irrigation 1 479

Murrumbidgee Irrigation 1 193

Victoria Goulburn-Murray Water 1 767

Other b (including South Australia) 1 326

Total 6 497

a 1 GL (gigalitre) equals 8107 acre feet. b 2001-02 estimate. Sources: ANCID (2005, p. 13); Appels et al. (2004, p. 5).

4

water trading

Trade in water entitlements and allocations Most irrigators in the southern Murray‐Darling Basin hold a nominal volumetric entitle-ment that is delivered as a share of an annual resource pool. The share is determined by the total nominal volume and entitlements. The yield of the entitlement depends on the size of the resource pool that, in turn, depends on surface water runoff, storage capacity and the way that storages are managed. Storage management is important as most irrigators have little or no on-farm storage capacity. Irrigators receive an annual volumetric allocation from these rights that they can call on (in full or in part) during the irrigation season.

Entitlement specifi cations and yields can differ between jurisdictions. South Australia and Victoria have the most conservative supply arrangements and nominal entitlements have a yield of around 95‐100 per cent. In New South Wales there are two classes of entitlement. Around 10 per cent of entitlements are high security — similar to those in Victoria and South Australia. The remainder has a yield of around 80 per cent of entitlement. Further, there is a difference in the storage to allocation ratio in New South Wales and Victoria and, as a result, supply reliabilities can be affected signifi cantly by climatic conditions.

Irrigators in the southern Murray‐Darling Basin can trade both water entitlements and seasonal water allocations:

■ Trade in water entitlements (‘permanent trade’) is the transfer of the ongoing right to access water for the term of the right.

■ Trade in seasonal water allocations (‘temporary trade’) is the transfer of some or all of the water allocated in accordance with the entitlement for the current irrigation season or for an agreed number of seasons.

Schematic diagram of the southern Murray-Darling Basin A

Queensland

Victoria

South Australia

NewSouth Wales

Inflow Irrigation area Confluence Storage infrastructure supply constraint

Hume Dam

Dartmouth Dam

Burrinjuck Dam

Blowering Dam

Eildon Dam

Murrumbidgee River

Murrumbidgee Irrigation Area

Murray Irrigation

Goulburn ‐Murray Water

River Murray

Darling River

Tumut River

Mitta Mitta River

Goulburn River

Loddon River Barmah Choke

Waranga channel

Snowy Mountain Scheme

On-river

water trading

5

Most trade occurs in seasonal allocations, and it is common for 10‐20 per cent of alloca-tions to be traded within an irrigation district in an irrigation season (Peterson et al. 2004). The volume of net interregional trade (exports minus imports) in seasonal allocations is relatively small, however, and varies across irrigation districts from year to year. Irrigators in South Australia are net sellers of seasonal allocations to irrigators in New South Wales and Victoria in most years, refl ecting the higher reliability of allocations in South Australia. Net trade between New South Wales and Victoria differs across years, but irrigators in New South Wales are usually net buyers of seasonal allocations from irrigators in Victoria, refl ecting the lower reliability of allocations in New South Wales.

While trade in allocations has increased over recent years, trade in water entitlements remains small, at less than 1 per cent of diversions in 2001-02. Further, around 1 per cent of the volume traded in both temporary and permanent entitlements was interregional (MDBC 2003). Low levels of interregional trade may be partly explained by administrative impedi-ments such as trade quotas imposed by irrigation authorities that restrict out of district trade (Goesch 2001). There are also a number of physical and environmental factors that impede interregional trade.

Two recent studies examined the effects of removing administrative impediments to trade in the southern Murray‐Darling Basin (box 1). Both studies suggest that additional trade may be relatively small, even if impediments are removed.

Implicit water rights and trade The development of irrigation in the southern Murray‐Darling Basin created a link between access to water, infrastructure and land. Prior to the cap on diversions being imposed, it

Box 1: Expansion of trade in the southern Murray‐Darling Basin

Heaney et al. (2004) used a competitive partial equilibrium model of water markets in the southern Murray‐Darling Basin to assess the economic impacts of water trade under freer administrative arrangements and alternative charging options for water delivery. This work suggests that removal of administrative impediments to trade will result in around 600 giga-litres of additional trade in permanent water entitlements. This represents a relatively small share of total water use and partly refl ects the large sunk investment in on- and off-farm infra-structure. While water may become more mobile as these investments reach the end of their economic life, the demand for water for environmental purposes is likely to be an important driver of future trading patterns.

Peterson et al. (2004) used a computable general equilibrium model to estimate the regional impacts of expanding trade in the southern Murray‐Darling Basin under a number of scenarios where water availability was reduced by 10, 20 and 30 per cent. With a 10 per cent reduction in water availability, total net water trade in the southern Murray‐Darling Basin was found to be a relatively small proportion of total allocations, with only 2.3 per cent of allocations traded among regions. Similarly, under the same scenario, net water exports or imports from a region was a small percentage of total water allocations in that region. For example, the Murrum-bidgee region was projected to be the largest net exporter of water with around 4 per cent of allocations traded out of the region.

6

water trading

was access to infrastructure rather than water resources that limited surface water use. The link between water, infrastructure and land allowed the creation of water property rights that were largely implicit. These implicit rights included access to a location-specifi c pool of resources, storage facilities and delivery channels, together with rights over the quantity and quality of return fl ows and conveyance losses (Goesch and Beare 2004).

Trade can alter the implicit rights attached to those entitlements. When a water entitlement that is defi ned at the point of delivery is traded, for example, it may change the location from which it is sourced. This can have implications for the infrastructure used to store and transport this water; therefore the specifi cation of resource access rights and hence the water market is incomplete. Implicit changes in resource access rights arising from trade may not always affect other water users. Altering the location of storage and delivery infra-structure from where water is sourced, for example, will not affect other users if capacity constraints are not binding, but when and where these constraints are binding, trade can impose third-party effects. In general, utilities currently only approve trades if the trade is not likely to cause congestion.

2. Third-party effects of trade Some of the key third-party effects of trade in entitlements and allocations in the southern Murray‐Darling Basin are reviewed in this section. The effect of trade on supply and delivery reliability is discussed in the following subsections. The effect of water trade on the costs of providing and maintaining storage and delivery infrastructure is then discussed. The many dimensions of the effects trade on water quality are then considered. In the southern Murray‐Darling Basin, for example, the main water quality issue is salinity and its potential effect on agricultural yields, urban and industrial use and stream habitat.

Reliability of supply The reliability of supply can be defi ned in terms of the probability that an entitlement holder will receive a volume of water in a given season — that is, the expected level and variation in physical water allocations that are realised from holding an entitlement. Supply reliability is determined by the natural variability of the resource pool and by the institu-tional arrangements that determine the share of that resource over which the entitlement is granted. The reliability of supply can also be affected by institutional arrangements that govern physical water trade in allocations, including conveyance losses, access to return fl ows and the introduction of tradable water entitlements.1

The signifi cance of the factors affecting supply reliability varies across the major resource pools of the southern Murray‐Darling Basin. Importantly these factors are not explicitly

1 Prior to the introduction of tradable entitlements, unused allocations from irrigators who did not exercise all or part of their entitlements were returned to the resource pool and reallocated. This increased the yield of the entitlement of active water users. The introduction of trade created an opportunity cost to irrigators who did not fully exercise their entitlements, which effectively led to an increase in the number of shareholders in the resource pool. While not causing problems of effi ciency, the distributional effects of this reallocation of resources induced by the potential for trade slowed the reform process.

water trading

7

specifi ed in the irrigator’s entitlements. As a result, each irrigator’s share of the resource pool is not fully specifi ed, and third-party effects of trade can arise if:

■ the source of the entitlement is not specifi ed, and the number of irrigators and the total volume of entitlements allocated varies by source

■ entitlement specifi cations differ between jurisdictions

■ evaporative and conveyance losses are not specifi ed separately from the allocation or entitlement and

■ the allocation or entitlement is specifi ed as a right to extract (rather than to use) a speci-fi ed volume of water.

In the southern Murray‐Darling Basin, entitlements are defi ned at the point of delivery. The number of potential sources from which these delivery points can be supplied increases moving downstream, as the effective catchment area increases with tributory infl ows. Unless traded entitlements retain the features of the reliability of supply from the exporting catchment, net trade that spans one or more tributories can affect the reliability of entitle-ment in both the source and the destination regions. Upstsream of a tributory, a given pool of resources may be spread over a greater number of users. At the same time, there is an increase in the share of resources that is potentially available to users below the tributory.

Loss of water through evaporation and, in some circumstances, seepage, during storage and transport can reduce the availability of water for irrigation. For example, changing patterns of trade can alter the timing of required irrigation releases from the dam and, in turn, the period of time water is held in storage. Conveyance or transport losses can be either non-fl ow dependent or fl ow dependent. Non-fl ow dependent conveyance losses are those asso-ciated with, for example, saturating earthen channels and occur regardless of how much water is transported.

Total volumetric conveyance losses associated with each megalitre of water delivered can differ within and between irrigation districts because of the natural physical and engineering features of the delivery system. Differences in conveyance losses between regions may also refl ect the hydrological characteristics of a river or channel system and climatic condi-tions. Flows from the upper tributaries of the River Murray to South Australia, for example, will incur greater evaporative losses during transport owing to the expanded width of the river and warmer and drier weather conditions. However, trade does not signifi cantly alter these non-fl ow dependent losses. Conversely, fl ow dependent losses vary depending on how much water is transported and may be affected by trade. They may occur, for example, if additional water caused the river to breach its banks.

Under traditional irrigation techniques, such as fl ood and furrow, irrigation runoff from farms is recycled via surface water runoff, drainage schemes, and accession to groundwater tables that eventually reach the river system. These fl ows are included in calculations of entitlements that are allocated to irrigators downstream in the basin. However, irrigators presently hold an implicit right to the return fl ows in that they can trade or save water without considering the downstream effects on other water uses associated with changes in water volume and quality. Water traded to an irrigator who employs more effi cient on-farm

8

water trading

water application techniques than the seller, for example, may lead to a reduction in water available to downstream users and thus affect the reliability of supply for some irrigators (see Appels et al. 2004). Under current institutional arrangements, reductions in the volume of return fl ows are simply absorbed as an additional diversion above the cap that may be at the expense of desired environmental fl ows.

Reliability of delivery The reliability of the delivery of an allocation is the timeliness of delivery of the water to the farm gate and depends on access to storage and delivery infrastructure. Congestion can occur in the river system as a result of the physical features of the river or within the irriga-tion system in near farm delivery infrastructure. Congestion costs are those associated with high levels of delivery infrastructure utilisation and occur when it is physically infeasible to move a certain volume of water. Increasing marginal congestion costs are borne by irrigators and other water users through reduced timeliness of delivery on water orders. Water trade can have an effect on third-party access to infrastructure if, for example, water is traded from a scheme where channel capacity is rarely reached into a scheme where capacity is often reached. Under current institutional arrangements this trade could reduce the reliability of delivery for all irrigators in the destination region.

Further costs will be imposed if, as commonly occurs, access to delivery capacity for irri-gators within a district is rationed on the basis of historical allocation during times of congestion rather than on the basis of delivering to those who face the greatest costs asso-ciated with a shortfall in deliveries. Trade into an irrigation area that worsens congestion can cause further ineffi ciency if access is not rationed according to costs. An irrigator on a local channel delivery system in the Goulburn‐Murray Irrigation District, for example, can only trade water to the farm if there is suffi cient capacity in the channel system for the delivery not to affect the supply reliability of other irrigators on the channel. Congestion may also have environmental impacts if, for example, it impedes the delivery of water to meet environmental outcomes. The ecology of natural watercourses can be compro-mised if using these watercourses as delivery channels overrides the natural fl ow variations (Hillman 2004).

Congestion in the river system itself may impose costs. Water utilities avoid on-river constraints, to some extent, by using a number of rerouting mechanisms, including moving water to downstream storages before the start of the irrigation season. Losses from these storages through evaporation and seepage are high and this, in turn, may have an effect on users’ security of supply. Trade that exacerbates channel constrictions at the Barmah Choke is currently forbidden. As mentioned above, nonmarket rationing does not ensure that the irrigators with the greatest net return gain access to the infrastructure.

Storage and delivery charges Trade in entitlements can result in a net trade of water permanently out of an irrigation district. If utility costs are apportioned to a smaller number of entitlements, charges for remaining irrigators may be higher. This may lead to their trading water out of the region and, ultimately, the utility may no longer provide irrigation water to the assets, so the

water trading

9

infrastructure may cease to be used. This is sometimes referred to as the ‘stranded asset’ problem. Stranded irrigation assets can include:

■ major infrastructure including dam and diversion works

■ major channels and diversion infrastructure

■ local channel and delivery works

■ on-farm irrigation delivery systems

■ other on-farm infrastructure assets associated with the irrigation activity.

It is important to distinguish between ‘ex ante’ and ‘ex post’ policies — in other words, whether the policy decision is made before or after an infrastructure investment decision. Ex ante, the requirement for undertaking and charging for an infrastructure investment (including infrastructure replacement) is for the total benefi ts of its creation and operation to be greater than the costs. Ex post, the infrastructure is a sunk asset which may have little or no salvage value and may not be replaced. It may be economic to operate that asset even if only covering variable costs. Ineffi ciency will arise if the pricing rule does not allow for this possibility. Remaining debt on the infrastructure does not have economic effi ciency implications although there may be equity issues.

If a utility ex ante allocates fi xed and variable costs in an appropriate two part charging scheme, such as part of a long term contract with irrigators, the stranded assets problem seems to fall into the class of third-party effects where there are no deadweight social losses. These effects are sometimes known as a pecuniary externality. They are distinct from physical externalities, which occur when water transfer affects the quality or quantity of water use. Pecuniary externalities arise when the external effects are transmitted through higher prices. The stranding of assets that result from the exit of entitlements from an irrigation district can result in pecuniary externalities for the remaining irrigators. To the extent that these third-party effects do not create deadweight social losses, their removal does not improve economic effi ciency (Katz and Shapiro 1994; Hanak 2003).

Where a utility adopts an inappropriate pricing model, such as one that allocates fi xed costs to a variable charge, trade can have effi ciency implications (Goesch 2001; Heaney et al. 2004). The average cost of delivery may rise in source regions, while in the destina-tion region, average costs may fall. These artifi cial conditions of decreasing and increasing costs can distort the spatial pattern of trade and result in movement of water into lower returning activities.

Water quality Changes in water quality because of trade arise through changes in the volume and quality of return fl ows (including runoff, drainage and groundwater discharge) and (to a lesser extent) the movement of traded water through the river system. Water quality may be affected if water is traded to an area or industry that has different agronomic practices from the source area. If water is traded to a use that relies more heavily on agrochemicals that may reach waterways, for example, third-party costs may be imposed on downstream water users.

10

water trading

The water quality issues in the River Murray system include the increasing river salinity, and its effect on productive uses for water as well as river health. Much of the increase in salinity can be attributed to subterranean return fl ows through the mobilisation of saline groundwater to the river system, a consequence of high levels of groundwater recharge from excess irrigation water. Within the southern Murray‐Darling Basin, there is consid-erable variation in the salinity of groundwater underlying irrigation areas. In the northern areas (such as Goulburn‐Broken), groundwater is generally fresh, between 3000 to 5000 milligrams of salt per litre. In parts of southern Victoria and South Australia, however, groundwater salinity can be as high as 33 000 milligrams of salt per litre, comparable with that of sea water (Heaney and Beare 2001). The highly location specifi c nature of the underlying hydrology means that the third-party effects of trade on water quality depend on the source and the destination of the water traded.

Return fl ows can either improve or reduce water quality depending on the location of water use after trade, thus having either positive or negative effects on users not directly engaged in the trade. Within the southern Murray‐Darling Basin system, for example, relatively fresh return fl ows from areas characterised by fl ood irrigation technologies can reduce the river salinity concentration. Conversely, return fl ows from irrigation areas located above high saline groundwater deposits can increase salinity in the River Murray.

The location specifi c nature of these third-party effects is important for at least two reasons. First, trade may lead to improvements in water quality, and policy instruments that provide incentives to trade water to ‘low impact’ areas can generate positive environmental and economic outcomes. Second, as the effects of trade vary with the source and the destination of the trade, it is infeasible to internalise fully, through a system of private property rights, the effects of return fl ows on others. Potential policy initiatives for internalising the impacts of negative third-party effects will be discussed in more detail in the following section.

3. Policy options Water entitlements are access rights to the stream of benefi ts (or costs) derived from using water for irrigation. The property right is a claim over some or all of the returns from water as a productive resource. Water trade is an instrument whereby irrigators can enhance the value of that right. Third-party effects from water trade arise if some of the benefi ts (or costs) of that action are not exclusive and not captured by the holder of the property right. The set of markets for this right is incomplete, and the true value of that asset will not be accurately refl ected in its price.

The Randall framework Developing effective policy instruments that will improve the management of natural resources requires an understanding of why the market is incomplete. Randall (1983) argued that the concepts of exclusiveness and rivalry represent the characteristics of goods and resources that matter in a public policy context. Randall arranged goods into nine groups based on their exclusivity (nonexclusive, exclusive and hyperexclusive) and rivalry (nonrival, congestible and rival). Randall’s classifi cation of goods is presented in table 2, along with some illustrative examples.

water trading

11

Under an exclusive property right, an individual bears all the benefi ts and liabilities associ-ated with consuming or producing a good or service. An exclusive property right is complete if it coveys all the costs and liabilities of either production or consumption. If a property right does not convey sole rights and liabilities, it is nonexclusive. If access to this right is restricted to a subset of individuals, such as a club, it is hyperexclusive; the limiting cases are monopoly or monopsony access.

A good or service is rival if its consumption or production by one individual has an impact on the consumption or production of others. Consumption and production of nonrival goods does not alter the choice set or incentives faced by other producers and consumers. Congest-ible goods are nonrival up to some point but as consumption or productions increases, delays begin to occur that impose costs.

Properties of exclusivity and rivalry can refl ect the institutional arrangements that defi ne property rights that exist over goods and services. Governments may grant resource access rights that are open, exclusive or even hyperexclusive. For example, water entitlements are exclusive in nature but access to the resource pool is capped by government — a form of hyperexclusion. The properties of exclusivity and rivalry may also be intrinsic to a commodity or service in that they limit the nature of the property right that can cost effectively be placed over that good or service. For example, existence values are intrinsi-cally nonrival and the costs of excluding individuals from the amenity benefi ts of natural resources is often prohibitive.

Beare and Newby (2005) note that exclusivity and rivalry can exist in both production and consumption of goods and services and that this can have implications for the design of an appropriate policy instrument. For example, unregulated emissions may generate nonex-clusive damages in consumption and are nonrival in production in that one fi rm’s emissions does not limit another’s. User or benefi ciary charges may not lead to an effi cient solution to the problem due to the transactions costs of compulsory charges and the incentive for benefi ciaries to underinvest in abatement. However, a cap on emissions, another form of hyperexclusion, creates rivalry in production, allowing the introduction of a tradable permit scheme.

2 Randall classifi cation of goods based on exclusivity and rivalry Category Exclusive Hyperexclusive Nonexclusive

Rival Water entitlement or Allocation of water In-stream conveyance losses

allocation resource pool between

consumptive and

environmental uses

Nonrival On-farm saving Improved water quality

of evaporative losses

Ecosystem services

Congestible Tradable infrastructure Congestion charge Open access access right by a delivery utility delivery channel

12

water trading

Randall (1983) argues that the joint classifi cation of exclusivity and rivalry characterise a good or service and determines the most effi cient means of provision and trade. Beare and Newby (2005) suggest that the policy options available to complete a market are largely determined by characteristics of exclusion and rivalry in the missing primary good or goods. The Randall classifi cation provides a framework to examine policy instruments to address third-party effects of trade that can arise in incomplete water markets.

Reliability of supply Water entitlements are not fully exclusive (even though they are rival) because they are not defi ned according to the location from which they are sourced. Consequently, trade in water entitlements fails to account for storage and conveyance losses and return fl ows. The lack of exclusion gives rise to third-party effects on the reliability of supply of an entitlement. Trade in water entitlements that do not take into account jurisdictional differences in the reliability of supply specifi cations can also generate third-party impacts.

The third-party effects of trade on reliability of an entitlement can be addressed using a source based, property rights solution, which adds a further component to the current water right. One option is to redefi ne the entitlement from the point of delivery to the source of extraction. Alternatively, Beare et al. (2005) suggest that a system of administered exchange rates could be used to account for the differences in the yield of an entitlement. The exchange rate would convert the volume of a traded entitlement so that the yield was similar to other entitlements in the destination region. Where there were no conveyance losses, the exchange rate determined would result in no net change in demand for water from the system on average. Exchange rates, however, can be diffi cult to specify correctly so that effi ciency losses do not occur. Appropriate exchange rates would need to be loca-tion specifi c and suffi ciently fl exible (including being adjustable retrospectively) to allow for changes in factors affecting supply security, such as changes to water sharing plans and long term climate change.

An exchange rate could also be used to implicitly account for conveyance losses associ-ated with physical water trade. An exchange rate may be used, for example, to account for trade resulting in a 10 per cent increase in conveyance losses, by converting the volume purchased to 90 per cent of the volume sold. Alternatively, water property rights could be explicitly defi ned at the source to make conveyance losses exclusive. The key issue is accu-rately identifying the conveyance losses and effi ciently attaching a property right given the transactions costs.

Water trading has been shown to have an effect on the reliability of supply of third parties, through changes in patterns of return fl ows that alter the quantity of water available for irri-gation downstream. This problem arises because water property rights are currently defi ned in terms of water diversions rather than the volume of water that is consumed. While these third-party effects are nonexclusive, they may be amenable to a property right solution if the transaction costs of defi ning and measuring return fl ows are not prohibitive. This solu-tion will be particularly complex when accounting for changes in water quality, because a fraction of nonconsumed water will return to the river system in an altered state, possibly generating downstream costs or benefi ts.

water trading

13

Trade in allocations, on the other hand, only affects supply reliability where there are differ-ences in the conveyance losses between the source of supply and destination of the water after the trade. If the extra fl ow created by the trade does not result in water breeching the banks of the distribution network and the trade occurs when distribution networks are full (as is usually the case with trade during peak irrigation period) the potential for further loss is greatly diminished since the distribution network is usually fully saturated and evapo-ration losses are not altered since the surface area of the channel is unchanged. Overall, conveyance losses are likely to be small relative to overall releases from storages (see, for example Pratt Water 2004).

Reliability of delivery Irrigation delivery channels are generally referred to as congestible goods because they exhibit nonrival characteristics for a limited number of users or levels of use. Rivalry sets in once this limit is exceeded, and intensifi es as the number of users increases.

Because demand for delivery capacity is highly seasonal and subject to periods of expan-sion and contraction, it is seldom optimal to invest in delivery capacity to meet periods of peak demand. While congestion does impose costs, it does not necessarily follow that there is a need to ration access to minimise congestion costs. Where all irrigators seeking access to delivery infrastructure face the same marginal cost of delay, delivering services on a ‘fi rst come, fi rst served’ basis will lead to optimal allocation. Where irrigators face different marginal costs associated with a delay, however, delivery services must be rationed in some way to ensure that those irrigators with the greatest net return gain access.

It may be possible to reduce congestion costs by allocating access rights to delivery infra-structure and by allowing trade in those rights if the right to access water and delivery infra-structure were separated and made explicit. Alternatively, congestion charging may be used during periods of congestion when services are rival, and where irrigators face differing marginal costs of congestion. By increasing peak period access charges during conges-tion, those irrigators facing higher congestion costs will be most prepared to pay the extra charges. Congestion pricing will have no rationing effect during continual congestion and will result in economic rents accruing to the infrastructure supplier, signalling a possible need for increases in capacity.

Further, rationing access to delivery infrastructure will only lead to a more effi cient alloca-tion of resources once rivalry sets in. Tradable access rights or congestion charging will create an ineffi cient exclusivity (rent seeking) if imposed when access to the infrastructure is nonrival.

With small net trade (see box 1), third-party effects are most likely to be limited to near farm delivery infrastructure where trade causes or exacerbates peak period congestion. Even then, trade may only cause problems during a relatively small number of peak demand days. Net trade into an irrigation area is more likely to exacerbate peak period congestion in systems with similar agricultural enterprises (for example, rice production) because the timing of demand for water is likely to be similar across the region. Areas characterised by large variation in agricultural production may be less likely to have peaks in water demand.

14

water trading

Storage and delivery charges Problems related to storage and delivery charges were identifi ed in section 2 as potentially having both equity and effi ciency aspects. The policy challenge posed by pecuniary exter-nalities, similar to the challenge arising from agricultural trade liberalisation, is an equity issue (Hanak 2005). If governments wish to intervene, however, constraining trade arrange-ments is unlikely to be the most effi cient and effective mechanism to use.

The effi ciency aspect of the problem may have arisen for two reasons. The fi rst reason for ineffi ciency could stem from missing options markets for access to storage and delivery infrastructure. Purchase of such options would be akin to joining a club. Club goods are a class of public goods that are exclusive in consumption but nonrivalrous, at least to the point of congestion. Creation of such markets ex ante would allow long term contracts to be established such that the actions of club members (such as exit) may not harm other members of the club.

The second reason is inappropriate pricing regimes. This could be addressed ex ante, for example, through adopting a two part charging scheme for allocating fi xed and variable costs. Ex post, where the asset might not be replaced, this could be addressed through more fl exible pricing agreements between the utility and the irrigator, such as marginal cost pricing. However, under current institutional arrangements, there are restrictions to utilities adopting such a regime.

When weighing up the benefi ts and costs of government intervention related to stranded assets (or for any other reason), it is important to consider the positive as well as negative third-party effects that can result from trade. Positive effects associated with permanent water trade include the alleviation of congestion and pressure on groundwater tables in the exporting district, and greater economies of scale in the importing region. If governments wish to assist affected irrigators, they should choose instruments that are targeted and that do not impose unnecessary cost on other parties.

Inappropriate intervention to address stranded asset concerns can reduce effi ciency. Restric-tions on trade are the most common example in Australia. The imposition of ex post exit fees, for example, can lock water into low productivity enterprises and regions. There is opportunity for the utility to rationalise its delivery system with asset redundancy, for example, by decommissioning or ‘mothballing’ redundant infrastructure. Some parts of the local distribution network may no longer be needed because water is no longer diverted from the main distribution network to smaller feeder channels to the irrigator’s farm. The utility could then reduce charges to refl ect the new patterns of infrastructure use. There may also be opportunities for the utilities to negotiate exits with irrigators.

In the absence of trade constraints, the incidence of stranded assets will be highest in regions where the marginal value product of water is lowest (for example, where a signifi -cant proportion of water is used for lower value activities), or in areas facing environmental degradation problems (for example, rising saline water tables). Localised channel and diversion infrastructure have been the main utility assets affected by trade in entitlements in the Goulburn‐Murray Irrigation District to date. The net exit of entitlements in Goul-

water trading

15

burn‐Murray Water subdistricts has tended to be geographically concentrated, with some subdistricts frequently reaching their quota. This may, in part, refl ect commodity prices and salinity (see, for example, Barr 1999). Other factors, such as the size of the farm, the age of the irrigator and their off-farm income are also likely to be important infl uences.

If administrative constraints to trade were removed, modeling suggests small net exit from irrigation regions (box 1), indicating that the rise in costs to remaining irrigators will be correspondingly small. Even where some subdistricts lost a third of irrigators over the decade, utility charges would only increase by about $11 per megalitre in nominal terms.2 With allocations trading around $67 per megalitre in 2004-05 (during the 2002-03 drought they reached $500 per megalitre), in the Goulburn Murray region such increases in charges are unlikely to bridge the gap between utility charges and the traded prices for allocations and are, therefore, unlikely to infl uence investment and production decisions of irrigators.

Major infrastructure assets such as dam and diversion infrastructure are unlikely to be affected. The cost of major dam infrastructure is passed on through trade, and entitlements are traded to meet the water needs of the purchaser. Just as the seller has relied on major infrastructure to store and deliver the water allocated to the entitlement in the past, so too will the purchaser in the future, regardless of whether the trade is intradistrict or inter-district, or whether the purchaser is an irrigator or an environmental manager.

Water quality The effect of trade on water quality is intrinsically nonexclusive. For example, trade out of a high impact area that reduces saline discharge to the river would benefi t all downstream users. Further, because the benefi ts accruing to water users differ according to their loca-tion, high transaction costs may prevent downstream users from collaborating to encourage investment upstream to improve water quality. Both these factors limit the usefulness of property right solutions that can capture the benefi ts of trade between parties. As a conse-quence, polices need to be directed to those activities that are the source or abatement of pollution.

Water use rights may also be used to impose specifi c conditions of use on irrigators. These rights may be regulations that apply to the location and intensity of water use, for example, in high impact zones. Limits may be placed on the type of soil that can be used for partic-ular activities, and maximum water application rates and standards may be imposed on water use effi ciency. Regulations can, however, be infl exible. Imposing a maximum appli-cation rate, for example, will prevent irrigators from applying additional water in dry years, even if the benefi ts vastly outweigh the social costs. The conditions of water use rights can include the use of fl exible economic instruments including taxes, subsidies or exchange

2 Goulburn-Murray allocates 220 000 megalitres of water to entitlements on 1260 properties (including stock and domestic supplies) in the Pyramid Hill subdistrict. Annual fi xed and variable charges are approximately $22 per megalitre (not including bulk water charges) plus an administration fee of $100 per property, generating around $4.96 million in revenue annually. If a third of entitlements and properties left the subdistrict, there would be a reduction in annual revenue of $1.65 million. To recover this loss, from the remaining 840 irrigators, annual allocation charges would need to rise to $33 and the administra-tion charge to $150 per property.

16

water trading

rates; for example, an irrigator might be required to purchase a salinity credit from a salt extraction scheme to offset the impact of saline drainage water discharge.

A form of water use right is currently implemented in the Murrumbidgee Valley, where rice production is limited to areas of specifi c soil types, and where maximum water application rates are imposed. These use rights are effectively tied to the land. Use rights can also be applied specifi cally to deal with the change in third-party effects associated with use when water is traded. Where the transfer of water results in an increase in external costs, for example, it may be possible to impose a tax on the use of water traded into that region. The tax revenue could be used to provide an incentive to trade water from regions with high external costs to regions with lower external costs.

Because the effects of water use vary continuously according to location in the river system, a tax or subsidy must be source and destination specifi c. A system of exchange rates could also be used to deal with third-party effects, for example, if water trade between regions results in an increase in salinity downstream of the recipient region, an irrigator may be required to purchase water in excess of requirements. This additional water would be used as a dilution fl ow to offset the increased salinity arising from this trade (Beare and Heaney 2002; Goesch and Beare 2004).

From a producer’s point of view, unregulated pollution is nonrival — that is, the discharge of pollution by one producer does not affect the ability other producers to pollute. It is possible, however, to make water quality rival by creating tradable use rights and estab-lishing a market for pollution caused by water use that meets some ‘target’ level of pollu-tion at least cost. A tradable salinity credit scheme could, for example, be used to control the level of saline emissions from an irrigation scheme. Salinity credits equivalent to the desired level of emissions from the scheme would be initially allocated to irrigators. Once allocated, these credits represent an asset that can be traded. Irrigators with a lower marginal cost of abatement will have an incentive to sell credits to irrigators with a higher marginal cost of abatement.

An irrigator wishing to expand water use by trading water into the scheme will either have to purchase salinity credits from other irrigators or invest in salt mitigation to generate addi-tional credits. The ability to sell salinity credits creates an incentive to invest in mitigation. If the market was expanded to allow entities other than irrigators to provide mitigation (for example, private companies providing mitigation through engineering works such as salt interception schemes), it may be possible to provide mitigation at a lower cost than if the market is restricted to irrigators (Goesch and Beare 2004).

The effects of trade on River Murray water quality have been shown to be considerable, depending on the source and destination of the trade (Heaney and Beare 2001). Net trade into the highly saline regions of South Australia and Victoria, for example, imposes costs on downstream water users through higher salt concentration of water used for productive purposes and also affects the riverine environment more generally. In contrast, regional trade between irrigation areas with similar agronomic and hydrological characteristics may not warrant policy intervention.

water trading

17

4. Conclusions

The separation of water entitlements from land failed to account for the spatial character-istics of water supply, demand and use that were implicit in the joint right. Trade in water entitlements and allocations have therefore given rise to third-party effects. The existence of third-party effects has been cited as a reason to restrict or prohibit intra- and interregional trade in the southern Murray‐Darling Basin.

Third-party effects on delivery reliability are likely to be relatively localised and small in terms of scale and cost, but nonetheless amenable to property right solutions. Other effects have the potential to be more substantial, such as some of the effects on the security of supply. Accounting for differences in entitlement specifi cation between jurisdictions, for example, would generate considerable benefi ts if there were large volumes of water traded between states.

While for the most part, these third-party effects can be addressed through the introduction of more completely specifi ed water rights, the creation, implementation and enforcement of property rights regimes is not costless. In some instances, the costs of property right solu-tions may be higher than the benefi ts they generate. The regionally specifi c nature of the third-party impacts of trade examined in this paper highlight the need to recognise regional characteristics of surface and groundwater systems, soil and climate as well as investments in fi xed infrastructure when considering policy interventions. Adding a further component to existing property rights to account for the water quality effects of water traded within the Goulburn‐Broken region, for example, may generate costs that exceed the benefi ts. The same action attached to water used in the highly saline regions of South Australia, on the other hand, may generate considerable benefi ts and allow the benefi ts of trade to be fully realised. Similarly, if the costs imposed by water losses and trade restrictions as a result of the current management of capacity constraints such as the Barmah Choke are such that they warrant intervention, property rights or pricing regimes to ration access may be considered as appropriate policy interventions.

It is likely that many of the third-party effects of trade discussed in this paper do not warrant policy intervention at the national or state level. In some instances, effects are likely to be relatively minor although some may be signifi cant at the local level. The costs of addressing some third-party effects may outweigh the benefi ts. Where there are signifi cant gains from trade, the existence of these third-party effects should not been seen as a reason to impede trade. There are fi rst-best policy instruments to address these effects at an appropriate scale.

18

water trading

References

ABS (Australian Bureau of Statistics) 2004, Water Account Australia 2000-01, cat. no. 4610.0, Canberra.

Appels, D., Douglas, R. and Dwyer, G. 2004, Responsiveness of Water Demand: A Focus on the Southern Murray-Darling Basin, Productivity Commission staff working paper, Melbourne.

ANCID (Australian National Committee in Irrigation and Drainage) 2005, Benchmarking Data Report for 2003-4, Key Irrigation Industry Statistics and Performance, Canberra.

Barr, N.F. 1999, Salinity Control, Water Reform, and Structural Adjustment: The Tragowel Plains Irrigation District, PhD thesis, University of Melbourne.

Beare, S. and Heaney, A. 2002, Water Trade and the Externalities of Water Use in Australia, ABARE report prepared for Natural Resource Management Unit, Department of Agricul-ture, Fisheries and Forestry ‐ Australia, Canberra.

Beare, S. and Newby, J. 2005, Incomplete markets, excluded goods and natural resource management, ABARE conference paper 05.5, presented at the 49th Australian Agricul-tural and Resource Economics Society Conference, Coffs Harbour, New South Wales, 9‐11 February.

Beare, S., Goesch, T., Pechey, L. and Bosch, C. 2005, Water Rights and Trade: The Role of Exchange Rates and Tagging, ABARE report prepared for the Natural Resource Manage-ment Unit, Department of Agriculture, Fisheries and Forestry, Australia, Canberra.

COAG (Council of Australian Governments) 2004, Attachment A: Further Information on the National Water Initiative, Meeting outcomes from 25 June, Canberra.

Goesch, T. 2001, ‘Delivery charges for water: their impact on interregional trade in water rights’, Australian Commodities, vol. 8, no. 4, December quarter, pp. 626‐34.

Goesch, T. and Beare, S. 2004, ‘Water rights and trade: meeting the water reform agenda’, Australian Commodities, vol. 11, no. 1, March quarter, pp. 180‐9.

Hanak, E. 2003, ‘Who should be allowed to sell water California? Third-party issues and the water market’, Public Policy Institute of California, Los Angeles.

Hanak, E. 2005, ‘Stopping the drain: third-party responses to California’s water market’, Contemporary Economic Policy, vol. 23, no. 1, pp. 59‐77.

Heaney, A. and Beare, S. 2001, ‘Water trade and irrigation’, Australian Commodities, vol. 8, no. 2, June quarter, pp. 339‐49.

water trading

19

GPO Box 1563 • Canberra • 2601 • Australia • Fax +61 2 6272 2001 • Tel +61 2 6272 2000 www.abareconomics.com

abare

Productivity Commission

Australian Government

PO Box 80 • Belconnen ACT 2616 • Australia Telephone: (02) 6240 3200 • Facsimile: (02) 6240 3399

Heaney, A., Thorpe, S., Klijn, N., Beare, S. and Want, S. 2004, Water charges and inter-regional trade in the southern Murray Darling Basin, ABARE conference paper 04.14 presented at the Establishing Water Markets Symposium, Melbourne, 9 August.

Hillman, T. 2004, Murrumbidgee Valley Ecological Assessment, Working paper for the Pratt Water Project, Griffi th, New South Wales.

Katz, M.L. and Shapiro, C. 1994, ‘Systems competition and network effects’, Journal of Economics Perspectives, vol. 8, pp. 93‐115.

MDBC (Murray‐Darling Basin Commission) 2003, Water Audit Monitoring Report 2001-02, Canberra.

Peterson, D., Dwyer, G., Appels, D. and Fry, J. 2004, Modelling Water Trade in the Southern Murray-Darling Basin, Productivity Commission staff working paper, Melbourne.

Pratt Water 2004, The Business of Saving Water, Campbellfi eld, New South Wales.

Randall, A. 1983, ‘The problem of market failure’, Natural Resources Journal, vol. 23, pp. 131‐48.