Journal of Environmental Management (2001) 62, 113–120 doi:10.1006/jema.2001.0423, available online at http://www.idealibrary.com on

The environmental impact of water markets: An Australian case-study

J. G. Tisdell

Water markets are developing as part of a Council of Australian Governments initiative to promote an efficient use of ’s water resources. The consequences of these policies on river health is yet to be fully understood, but recognised as having significant interrelationships which need to be explored. This paper examines the consequences of introducing trade and allocating water for environmental use in the Border Rivers region of . The results of this study suggest that: (1) trade in water entitlements is likely to increase the differential between extractive demand and historical flow regimes as extractive water-use concentrates on the most profitable crops, and (2) water markets are likely to limit the effectiveness of water policies aimed at restoring natural flow regimes. As a result, trade-offs between environmental needs and income from extractive use will need to be determined. This work is important and timely in water-policy development demonstrating the linkages and trade-offs between ecological and economic objectives.  2001 Academic Press

Keywords: environmental flows, water policy, water allocation, water trading.

Introduction understood or adequately accounted for (Allan and Lovett, 1997). Uncertainty sur- rounding the needs and dynamics of riverine In 1992 the Council of Australian Govern- ecosystems makes it difficult to give appropri- ments (COAG) put in place principles to ate consideration for environmental needs. promote efficient and sustainable use of water Methodologies to model the relationship in resources in Australia (ARMCANZ, 1995, specific catchments in Australia are evolving NCC, 1998). Markets for entitlements to (Black, et al., 1996, DNRQ, 1999a). Following extract water have been introduced in many the precautionary principle, policy options states of Australia as a mechanism for redis- are being developed to account for the per- tributing available water to its most efficient ceived needs of riverine environments (ARM- use (Brennan and Scoccimarro, 1999; Dra- CANZ, 1996, Young, 1997, DNRQ, 1999a). gun and Gleeson, 1989; Topp, 1998; Topp The work in this paper will contribute to that and McClintock, 1998). To achieve an effi- development. cient distribution of the resource market transaction costs must be minimal and exter- This paper explores the consequences of: nalities accounted for (Chan, 1989; Colby (1) introducing transferable water entitle- et al., 1993, Russell, 1995). Water markets ments on flow regimes; and (2) reducing the may lead to significant environmental exter- level of extractive allocations to allow for nalities, as they are dependent on hydro- water to be used for environmental use. Lin- logical constraints and have the potential of ear programming models have been used to adversely impacting on flow regimes and as a model trade between 112 irrigation farms on the Queensland side of the Border Rivers. result riverine environments (Bjornlund and Faculty of Environmental McKay, 1995). The results of the modelling suggest that: Sciences, Griffith How trade in aggregate will impact on (1) free trade has the potential to significantly University, Brisbane, health of Australia’s rivers is yet to be fully alter flow regimes; and (2) improvements in Australia, 4111. flow regimes can be achieved by reducing Received 9 March 2000; Email of author: [email protected] the announced allocation by 5 to 10% and accepted 22 January 2001

0301–4797/01/050113C08 $35.00/0  2001 Academic Press 114 J. G. Tisdell

releasing the water in accordance with natu- The convention has been to re-establish ral requirements. natural flows in some form—be they in total, as first—flush flows, or as minimum flow regimes (DNRQ, 1999b). In Australia the dominant idea is that the natural or pre- Water allocation, development flow pattern should be mim- environmental flows and trade icked (Hart and Campbell, 1991, BGATF, 1992). The Queensland State Government in Australia, for example, is developing water While there are variations in the structure allocation management plans for each of the of water allocations, most regulated water catchments in the State (DNRQ, 1999b). Mea- in Australia is allocated to irrigators under surement of the impacts of extractive uses a doctrine of non-priority (Randall, 1981). of water on environmental flows is judged Irrigators are issued with permits to a nomi- against such natural or predevelopment flow nal allocation of water. Each water year, the regimes, taking into account key biological water authority, dependent on the availabil- trigger processes (DNRQ, 1999b). The rela- ity of supply, declares an announced alloca- tionship between flows and aquatic benefits tion; a percentage of the nominal allocations is site specific and gaining an understand- of water that each irrigator may extract. The ing of it requires familiarity with biological announced allocation has been dependent on models beyond the scope of this study. demand, the hydrological characteristics of the catchment and evaporation transmission losses. A case-study of the border Over the last 30 years demand for rural water has changed significantly as a result of rivers region of queensland changing farm practices and greater areas of land under irrigation (Edwards et al., 1996). Trade in water entitlements in Queens- In order to redistribute existing water enti- land formally first started in the Border tlements to their most efficient use trade Rivers region. The Border Rivers separate in water entitlements has been introduced. the states of and Queens- At the same time, increasing pressure on land. The catchment consists of three main the riverine ecosystem to service irrigated rivers—the Dumaresq, Macintyre and Bar- agriculture has resulted in significant out- won (see Figure 1). Extractive use of water is breaks of blue-green algae (BGATF, 1992). regulated between Bonshaw and , While environmental water requirements a distance of approximately 130 km (Brown have historically been considered in deter- et al., 1983). In total 231 000 ml and 62 900 ml mining announced allocations, they have not are allocated to individual water users in been given sufficient weight. Where water Queensland and New South Wales, respec- trading is introduced, formal consideration of tively. In this study the extractive use of environmental flows will be necessary to give water was modelled using data on 112 status to instream water uses. Water flows, licensees on the Queensland side of the Bor- previously considered unused, will need to der Rivers region. Historical flow records of be recognised as a legitimate use of water. the catchment till 1979 are used to calculate In fully allocated systems announced allo- natural flow regimes. cations may need to be reduced to provide Data was collected on crops grown, irri- water for environmental use. Understand- gatable land area and the water allocation of ing the economic/environmental interactions each farm. Data was also collected on regional arising from with the introduction of trade crop factors, rainfall and evaporation and in water entitlements and consideration of crop gross margins. Data from the 1985/1986 environmental flow regimes is critical to water year was used because it was the first water policy development in Australia and official year of trade in water entitlements in in other countries developing water alloca- the region. In that water year the announced tion and trading rules (Rosegrant and Bin- allocation of water was 60%, which resulted swanger, 1994, Shatanawi and Al Jayousi, in a total of 37 472 ml of water being allocated 1995, Bauer, 1997). to the 112 licensees in this study. The environmental impact of water markets 115

Border Rivers Region N

Great

Maranoa Roma Condemine River QUEENSLAND

River

River Toowoomba Beardmore River Dividing Dam Brisbane St. George RiverGoondiwindi Balonne Boomi Weir Weir Weir Coolmunda Regulators Dam Warwick Moonie e River ntyr Dumareso aci Glenlyon Stanthorpe Little Weir M Weir 'Keetah' Dam River Diversion River River River Mungindi Glenarbon Weir Texas River Cunningham Weir Mingoola Gauging Station River River Mungindi Weir

Bonshaw Tenterfield Range Moree Weir BirrieBokharaNarran Pindari

Culgoa Dam Glen Innes Barwon River Bourke Walgett

NEW SOUTH WALES

Scale 0 50 100 150 200

Kilometres

Figure 1. Map of the Border Rivers Region. Source: DPIWR, Toowoomba.

The flow regime prior to major human in 1975/1976. It was assumed that the aim intervention was assumed to represent the is to mimic the environmental flow regime of environmental flow requirement. The flow of that water used for extractive use. This sim- water over Bonshaw Weir in the recorded ple flow regime based on monthly averages years prior to commencement of construc- is recognised as having a number of limita- tion of was be taken as tions. By averaging the monthly flows, the representative of the natural flow regime extreme variation in flow and environmen- of the river (QWRC, 1980). Glenlyon Dam tal conditions inherent in Australia’s riverine is the main dam regulating the flow of systems are damped. Modelling of irregu- water in the river basin. Bonshaw Weir lar events such as floods usually involve was selected because it is the first recording probabilistic models over time. The mod- station below the dam and is largely unaf- elling in this study is static in that it uses fected by ancillary rivers and streams enter- data from a single water year only. Using ing the main river system. Monthly stream alternative environmental measures, such flow records over Bonshaw Weir were first as meeting minimum flow requirements or collected in 1965. For this study, recorded extreme flow events, may well produce dif- monthly streamflow averages for the water ferent results and is an important area for years up to the construction of Glenlyon Dam further research. 116 J. G. Tisdell

Estimated extractive demand for estimated by: water X3 Xn Xm X12 Maximising GM x .2/ Extractive demand for water in the region kijq kijq kD1 iD1 jD1 qD1 stems predominately from irrigation. Farm- ing in the Border Rivers is broad acre agricul- Subject to: ture with pasture, lucerne and cereal crops such as sorghum upstream and cotton domi- X3 Xn Xm X12 X3 Xm nating downstream agriculture. Due to the wkijqxkijq Ä akj length of the river system the basin was kD1 iD1 jD1 qD1 kD1 jD1 divided into three climatic zones: Glenlyon Water constraint Dam to Macintyre Brook, Macintyre Brook X3 Xn Xm X12 X3 Xm to Boomi Weir, and Boomi Weir to Mungindi. wkijq Ä lkj The water requirements of crops grown in the kD1 iD1 jD1 qD1 kD1 jD1 three zones were estimated using crop fac- Land constraint tors, and zone rainfall and evaporation data on a monthly basis. The aggregate extrac- tive demand for water without trade was Evaluating the differences between estimated by: extractive and environmental flow X3 Xn Xm X12 regimes Maximising GMkijxkijq .1/ kD1 iD1 jD1 qD1 Here, the environmental objective was defi- ned in terms of minimising the average sum subject to: of squared differences between the actual and natural flow regimes, subject to the extrac- Xn X12 tive use of water and available water for wkijqxkijq Äakj for all k and j environmental use. This was estimated by: iD1 qD1   Water constraints X3 Xn Xm 2 12   Xn X12 X    kD1 iD1 jD1  xkijq Älkj for all k and j   qD1 C iD1 qD1 .wkijqxkijq/ eq hq Minimising MSDD Land constraints 12 .2/ where: subject to:   GMkijq is the gross margin in zone k X12 X3 Xn Xm X3 Xm   for crop i on farm j in month q .wkijqxkijq/Ceq Ä akj xkijq is the area of irrigated land qD1 kD1 iD1 jD1 kD1 jD1 planted in zone k with crop i on farm j in month q where: akj is the allocation of water in zone k to farm j eq release of water for environmental use in month m lkj is the total area of irrigable land in zone k on farm j hq natural flow of water prior to the construction of Glenylon dam in wkijq is the water requirement in zone k of crop i on farm month m j in month q Environmental flow performance measures When water entitlements are tradeable in most countries are based on some mea- the farmers are no longer constrained by sure of flow deviation. The annual propor- their individual water allocation but theo- tional flow deviation (APFD), for example, retically by the aggregate supply of water. is based on the difference between exist- Extractive demand for water with trade was ing and natural flow regimes (Gehrke et al., The environmental impact of water markets 117

1995) and is used to assess environmental and December with the final watering of flows in Queensland (DNRQ, 1999b). Such crops such as cotton. The water demand of models are used in conjunction with hydro- crops during the winter months of June and logical models (such as the Integrated Water July mean that the flow during these months Quality and Quantity Model (IQQM) (Black is less than the natural flow regime. The et al., 1996)) and ecological indicators (such as effect of trade estimated using Equation (2) fish species diversity) to assess environmen- is to concentrate water use on the more tal flow requirements. Data requirements for profitable crops, thereby concentrating the such integrated environmental evaluation is demand for water to particular months and significant and to date have occurs in a select locations. This is demonstrated clearly with number of catchments. As such integrated the demand for water during November and modelling and assessment of environmen- December far exceeding the environmental tal flow requirements develops it is hoped flow requirement. Demand for water during that accurate measures of environmental December is estimated at over 8000 ml after improvement as a result of a specific reduc- trade, compared to an estimated pre-trade tion in extractive water use will be possible. extractive demand of 4563 ml and an environ- mental water requirement of only 3976 ml. Results of the modelling: These results suggest that the introduction of transferable water entitlements have the environmental and extractive flow potential to further differentiate extractive regimes compared from natural flow regimes. The modelling empirically suggests that trade will signif- Figure 2 shows the monthly water use prior icantly change in flow regime of the rivers to trade, the monthly water use after trade away from the natural flow regime that and the environmental water requirement. existed prior to the construction of the dam. The water year for this region operates from Empirically, the cost on the environment of October (Month 1) to September (Month 12). introducing trade is reflected in the increase The historical flow regime peaks during Octo- in the MSD. At 60% announced allocation ber through to February then declines during the without trade MSD is 30 000 compared to the winter months and finally rises dur- over 1Ð8 m with trade. ing August and September. The results of the modelling suggest that the extractive demand for water is less during the winter Consequences of decreasing the months of May and June and is high dur- announced allocation ing November and December. The allocative demand for water prior to trade estimated Decreasing the announced allocation will using Equation (1) peaks during November make water available for improving the

9000

8000

7000

6000 ) l m

( 5000

r e t

a 4000 W 3000

2000

1000

0 Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Month

Figure 2. Extractive and environmental demand for water. Historical flows ( .....); pre-trade extractive use ( ); at equilibrium after trade ( ). 118 J. G. Tisdell

2 300 000

60% 2 250 000 59% ) $

58%

S 57% U 2 200 000 A

( 56%

e 55%

m 2 150 000 o 54% c n

I 53% 2 100 000 52% 51% 2 050 000 0 5000 10 000 15 000 20 000 25 000 30 000 35 000 MSD

Figure 3. MSD and income as announced allocation declines without trade.

flow regime. Figure 3 shows that without reductions in announced allocation if trade trade significant improvements in environ- occurs; and (2) the cost of reducing announced mental flows can be achieved by reducing allocations by 5% from 60 to 55% is greater the announced allocation from 60 to 50% and if trade occurs. In absolute terms the cost is dedicating 10% of available water to envi- approximately AUS $113 000 without trade ronmental use. At 60% announced allocation compared to AUS $161 000 with trade. In the MSD is approximately 30 000. Reduc- relative terms however the costs are simi- ing the announced allocation to 50% and lar; 4Ð9% without trade and 5% with trade, dedicating 10% of available water to the envi- as trade results in greater aggregate farm ronment reduces the MSD to below 40 with income. A 5% reduction in announced alloca- an associated decrease in aggregate income tion reduces the MSD from over 33 million to from AUS $2Ð3mtoAUS$2Ð06 m, a decrease 1Ð88 million, a relative decrease of 43%. of AUS $240 000. As expected, the shape of These results suggest that trade will have the curve suggests that the marginal cost of a significant impact on the flow regime of reducing announced allocations increases as rivers and with or without trade, reducing the the announced allocation declines, especially announced allocation by 5% during median below 55% announced allocation. Reducing climatic years from 60 to 55% could make the announced allocation from 60 to 55% significant improvements in the flow regime without trade reduces the MSD from 30 000 at a similar relative cost to extractive users to 7600, a relative decrease of 75%. of water. Figure 4 presents the results of modelling Throughout this study, the cost of water reductions in announced allocations with available for environmental use has been trade. Comparing Figure 3 with Figure 4 measured by the income loss to extrac- suggests that: (1) improving the flow regime tive users from an associated decrease to comparable levels requires much greater in announced allocations. It should be

4 000 000 3 500 000 60% 55% ) 3 000 000 50% $ 45% 40% S 2 250 000

U 35% A (

2 000 000 e m

o 1 500 000 c n I 1 000 000 500 000

0 500 000 1 000 000 1 500 000 2 000 000 MSD

Figure 4. MSD and extractive income after trade for various levels of announced allocation. The environmental impact of water markets 119 remembered that increasing such flows also challenges. Australian Journal of Environmental creates non-market benefits, such as recre- Management 4, 200–210. ARMCANZ (1995). Water Allocations and Entitle- ational and aesthetic values; increased water ments: A National Framework for the Implemen- quality which may reduce treatment costs tation of Property Rights in Water. Canberra: and potentially increase extractive options Task Force on COAG Water Reform. downstream. Introducing tradeable environ- ARMCANZ (1996). National Principles for the mental entitlements is also possible. Mod- Provision of Water for Ecosystems. Canberra: elling these values and trade options in Aus- Occasional paper. Bauer, C. J. (1997). Bringing water markets tralian catchments is the challenge ahead. down to Earth: the political economy of water rights in Chile, 1976–95. World Development 25, 639–656. Conclusions BGATF (1992). Final Report of the New South Wales Blue-Green Algae Task Force. Sydney: This paper explored the environmental con- Department of Water Resources. Bjornlund, H. and McKay, J. M. (1995). Can water sequences of two water policy options for an trading achieve environmental goals? Water 22, Australian River system. The results of this 31–34. study suggest that: (1) trade in water entitle- Black, D. C., Sharma, P. K. and Podger, G. M. ments is likely to increase the differential (1996). Simulation modelling for water manage- between extractive demand and historical ment planning in New South Wales. Water and the Environment 01/02, 1–8. flow regimes as extractive water use con- Brennan, D. and Scoccimarro, M. (1999). Issues centrates on the most profitable crops; and in defining property rights to improve Aus- (2) water markets are likely to limit the effec- tralian water markets. Australian Journal tiveness of water policies aimed at restor- of Agricultural and Resource Economics 43, ing natural flow regimes. Trade-offs between 69–89. Brown, J. A. H., Harrison, R. D. and Jacob- environmental needs and income from extrac- son, G. (1983). Water demand and availabil- tive use will need to be determined. ity with reference to particular regions, Water The flow regimes resulting from extrac- 2000. Canberra: Print- tive demand were compared to historical flow ers. records prior to the construction of major Chan, A. H. (1989). To market or not to market: allocating water rights in New Mexico. Natural dams on the river system. The needs of the Resources Journal 29, 629–643. riverine ecosystem are still being discovered Colby, B. G., Crandall, K. and Bush, D. B. (1993). and as they become more transparent so the Water right transactions: market values and results of this study can be modified. Never- price dispersion. Water Resources Research 29, theless, it was clear that the trade of water 1565–1572. entitlements distorts the flow regime further DNRQ (1999a). Methodology for Assessing Envi- ronmental Flows. Brisbane: Queensland Gov- away from the needs of the environment. ernment Printers. While this study does not capture the full DNRQ (1999b). Water Allocation Management dynamic nature of flow regimes; due to its Plan (Fitzroy Basin). Brisbane: Queensland deterministic structure, it does contribute Government Printers. to the important debate on water policy in Dragun, A. K. and Gleeson, V. (1989). From water law to transferability in New South Wales. Australia. The issues expressed and method- Natural Resources Journal 29, 645–661. ologies used could be modified for other insti- Edwards, G., Dumsday, R. G. and Chisholm, A. tutional water structures and policies. (1996). Efficient use of Australia’s land and The question still remains whether trade in water. Search 27, 188–191. water entitlements is a long-term ecologically Gehrke, P. C., Brown, P., Schillerm, C. B., Mottatt, D. B. and Bruce, A. M. (1995). River sustainable mechanism for supplying water regulation and fish communities in the Murray: with quality to extractive users. As knowl- System, Australia. Regulated edge of riverine ecosystems improves, this Rivers: Research and Management 11, 363–375. analysis may provide a platform for greater Hart, B. T. and Campbell, I. C. (1991). Water analysis. Quality Guidelines and the Maintenance of Australian River Ecosystems. Armidale: CWPR. NCC (1998). Compendium of National Compe- References tition Policy Agreements. Canberra: National Competition Council. QWRC (1980). Queensland steam flow records Allan, J. and Lovett, S. E. (1997). Managing to 1979. Brisbane: water for the environment: impediments and printers. 120 J. G. Tisdell

Randall, A. (1981). Property entitlements and Shatanawi, M. R. and Al Jayousi, O. (1995). pricing policies for a maturing water economy. Evaluating market-oriented water policies in Australian Journal of Agricultural Economics Jordan: a comparative study. Water Interna- 25, 195–220. tional 20, 88–96. Rosegrant, M. W. and Binswanger, H. P. (1994). Topp, V. (1998). Efficiency and Equity in Water Markets in tradable water rights: potential for Markets. Canberra: ABARE. efficiency gains in developing country water Topp, V. and McClintock, A. (1998). Water allo- resource allocation. World Development 22, cations: efficiency and equity issues. Australian 1613–1625. Commodities 5, 504–510. Russell, C. S. (1995). Review of: Economics of Young, M. (1997). Water Rights: An Ecolog- water resources: from regulation to privati- ical Economic Perspective. Canberra: Cen- zation. Journal of Economic Literature 33, tre for Resource and Environmental Studies, 284–285. Australian National university.