More water for irrigation and the environment? Some problems and prospects for worthwhile investments.

More water for irrigation and the environment?
Some problems and prospects for worthwhile investments.

Because of a growing concern about the riverine environment, there are calls to increase environmental flows in the Murray-Darling Basin (WWF Australia, 2002). Allocations for consumptive use in the connected Murray River system would fall under a series of proposed scenarios by 350 gigalitres (GL), 750 GL or 1500 GL (MDBC, 2002); and by 750 GL, 1630 GL or 3350 GL (Young et al, 2002) as shown in Figure 1.

Figure 1: Schedule of increased environmental flows proposed for the Murray System (after Young et al, 2002)

Increasing environmental flows on this scale is a big idea. While there may be some complementary outputs in river management, environmental flows and consumption are ultimately competitive uses. On an area basis, the increased environmental flow scenarios contemplated by Young et al have the potential to reduce the area of irrigated agriculture[1] by 95,000 hectares, 200,000 hectares or 420,000 hectares. This is equivalent to wiping out irrigation in Northern Victoria.

Increasing the efficiency of irrigation water use is seen as a way to offset reduced allocations. Indeed some see increasing water use efficiency as the next quantum leap in water resource development. Options such as reducing water storage and transmission losses, improving irrigation efficiency and improving plant water use efficiency can help maintain production under reduced water availability. And switching from production of “low value” to “high value” commodities can increase gross value of returns. However the costs of implementing these options must constitute a critical economic constraint to the adoption of these solutions.

To provide a basis for analysis, inefficiencies in water use are defined, the illusory nature of some proposed savings is explained and a method for valuation of real savings in comparison to costs of proposals is described. The simple treatment of these issues here is not complicated by the unique attributes of local situations. This is not a major difficulty if real options are examined in detail using benefit:cost analysis principles before policy changes are made or investment is sunk.

The limited prospect for obtaining a significant volume of real savings is discussed. This highlights the need for a sound policy for achieving the best allocation of limited water resources to competing uses.

Flows exceeding demand spill over the end of the channel or drain off the end of the irrigated paddock. Estimates of combined gross losses range from 25-50% of stream diversions. Figure 2 shows a hypothetical irrigation system where paddock tail water and channel outfalls do not return to the river. Of gross diversions of 100 GL only 60 GL are used for crop production. The remaining 40 GL comprising channel outfalls and paddock tail water is lost from the system. Net diversions are 100 GL.

Figure 2: Schematic illustration of water flows for an irrigation system with 40% gross outfall and paddock tail water losses. Arrows show flow volume and direction, star symbols indicate consumptive use and dotted circles show volume of losses

The magnitude of real or net losses depends on the ability to recycle within the irrigation system or return excess flows to the river. Returned flows contribute to environmental flows.

Figure 3 shows the same system where diversions exceeding irrigation demand flow back to the river via the farm and district drainage network. In this example excess flows of 40 GL return to the river. Net diversions are 60 GL

Water that seeps below the channel bottom or the root-zone in the irrigated paddock supplements existing groundwater resources. Gross surface system losses depend on channel/pipe materials, length of irrigation season, soil type, irrigation technology and management. Magnitude of real or net losses depends on the proportion of groundwater returning to the river and the ability of sub-surface drainage systems to recycle groundwater accessions.

Gross losses are in the order of 15-20 ML/ha of water surface depending on climate. These losses are not recoverable, except that within irrigation areas increased humidity from evaporation may moderate plant water demand.

There are diminishing returns to increasing water use intensity (irrigation or rainfall) as other factors of production become limiting.

The assumption here is that, given the market for produce, water resources are irrationally allocated to low value enterprises.

Since returned flows already contribute to downstream allocations there are no system savings obtained from reducing return flows. This simple algebraic reality obliterates the major forlorn hope of increasing catchment water resources. Figure 4 shows that eliminating tail water losses and channel outfalls and supplying only crop irrigation demand does not create new water. Net diversions are still 60 GL and downstream flows are not increased above 140 GL.

Conceptual difficulties occur when only parts of a system are considered. Outfalls are in fact spillovers. They may be negative spillovers as losses from one part of the system. But they are also positive spillovers providing inflows for the downstream component.

At the basin scale there is basically only one outfall, through the barrages at Goolwa, close to the mouth of the Murray. Calling transfers between jurisdictions “losses” and then aggregating “losses” from each of the n jurisdictions introduces an iterative process of nonsensical double counting between jurisdictions all the way down the system.

Given the interconnectedness of surface and groundwater systems, seepage losses are also spillovers. The prospects for real savings depend on the extent to which seepage is used as a water resource and the time lag between accessions and groundwater pumping.

If seepage is already being recycled by existing groundwater pumps, the only real savings from seepage reduction are reduced operating and maintenance costs for the groundwater pumps.

Prospects for real savings depend on opportunities to decrease specific exposure by reducing the surface area exposed to evaporation and/or increasing the water depth of storages. Options include piping open channels and changing system operating rules and decommissioning shallow storages such as Lake Mokoan and Lake Alexandrina (Anon, 2001).

Given a reasonable standard of management, increased production per unit of water can only be obtained by investing in developing and adopting new production technology. The adoption of higher harvest index semi-dwarf wheats in the 1980s is an outstanding example. Other options include regulated deficit irrigation of peaches and partial root zone drying of winegrapes using drip irrigation technology, amelioration of physical and chemical constraints to soil fertility and development and/or introduction of plant types more suited to the climatic conditions experienced. An example of the latter option would be the replacement of temperate C3 photosynthetic pathway species with more water use efficient sub tropical C4 plants for summer production.

It is often suggested that because horticulture has high gross margins per megalitre, and modern horticulture can deliver high water use efficiency, that the best policy solution for increasing water use efficiency is to mandate or subsidise horticultural use.

Unfortunately the market reality does not support this policy option (if the objective of policy is to increase net social welfare). Commodity composition is in loose equilibrium with capital markets because the mobility of capital in market economies leads to equal rates of adjusted[2] net return in all activities. For commodity composition to change dramatically, extensive changes in demand for irrigated produce is necessary. This may be engendered by trends in global demand (Hooke, 1997) and development of new production technology conferring a comparative advantage to local production. Until then, too rapid expansion into horticulture is a recipe for financial ruin.

Water markets have been operating for more than a decade (Simon and Anderson, 1990). Average prices for permanent transfer of water right in recent years in a number of irrigation areas is shown in Figure 5. The price dispersion can largely be explained by the expected mid to long run average allocation on different systems, by immediate seasonal allocations prevailing and by other factors such as locational variability in terms of institutional arrangements, prices for inputs and commodities and climate (Colby et al, 1993). When these factors are taken into account a price of $500-$600 per megalitre of permanent entitlement to annual delivery seems a reasonable estimate of the recent market price of water.

Figure 5: Average recent prices for permanent water right. Because of different allocation policies on different irrigation systems the figure does not indicate the price of permanent entitlement to annual delivery of one megalitre. (Data after Marsden Jacob in ACIL (2002))

Given the existence of contestable water markets, and land and water management plans to manage or tax the external impacts of irrigation, market prices should represent the social value of water at the margin of resources.

Markets facilitate the transfer of rights between willing buyers and willing sellers. Trade occurs when willingness to pay (WTP) at least equals willingness to accept (WTA). Provided buyers and sellers are equally well informed, the equilibrium market price of water will represent the net present value (NPV) of the future stream of benefits flowing from the water entitlement in either use. Buyers and sellers will base their estimate of the value of water on the expected timing and magnitude of the additional production from irrigation using the entitlement, the expected market value of the additional produce, the magnitude and timing of additional costs and the required rate of return on marginal or core capital, whichever is appropriate.

There seems to be some underlying policy apprehension that reluctant sellers are seeking inordinately high rents from speculation. Despite the fact that the use of futures trading to manage risk in agricultural markets relies purely on speculation, some consider it inappropriate to speculate on the value of water. Yet, given the uncertainty inherent in the estimation outlined above, a non-speculative valuation is impossible.

A large part of the commonly perceived gap between the NPV of water in “high” and “low value” uses is due to the inappropriate use of unadjusted gross margins as a means of comparison. The annualised additional capital development costs should first be deducted from the gross margin of the expanding enterprise. This substantially reduces the annual net margin for the “high value” use. The relative present value of the “high value” net margin will be further reduced when discounted at the desired rate of return on marginal capital rather than the low discount rates used for sustainability of core capital advocated by Quiggin (1992).

Figure 6: Present value of continuing existing irrigated grazing enterprise or developing new irrigated horticulture or irrigated dairy activities.

Figure 6 shows how inclusion of development costs and risk adjusted discount rates reconciles a large disparity in gross margins between enterprises. In this example, the NPV of irrigated development in horticulture and dairy generating gross margins of $600/ML and $163/ML respectively is much the same as that of an existing irrigated grazing enterprise with a gross margin of $30/ML.

As the long run agricultural development costs are already sunk, the present value of the future loss of gross margin should be used to estimate the agricultural opportunity cost of heightened environmental demands. Using recent market prices, the cumulative cost of purchasing water entitlement for the full implementation of the scenarios outlined in Young et al (2002) is $1.8 billion. The present value of the cost of the scheduled program of acquisition is $940 million. Given that the market price of water will rise as the supply for consumptive use is restricted, this must be very much an underestimate. Yet this very underestimate is roughly double the estimate made by Young et al of $450 million using an economic model for a scenario where there is no adjustment through investment in increased water use efficiency.

What is the reason for this extreme discrepancy? Some increases in future environmental flows may be released from storages during seasons of high inflows and low irrigation demand. Depending on inflows and demand in following seasons, this approach may moderate the impact on agricultural output at the margin of regional water resources. The potential for this moderation would tend to disappear at the higher levels of proposed increases in environmental flows. Higher environmental flow regimes may bring some benefits to downstream users through lower salinity levels. But the value of these benefits is relatively minor and comparatively low cost engineering options for salt interception are available. Further Quiggin (1988) has shown the rational national adjustment to salinity is to move salt sensitive uses upstream.

Surely, an if economic model is to be effective in guiding profitable investment, its structure must entertain all feasible options and its output must reconcile with the reality of market prices.

Private and public investment should yield increased profit and net social welfare. The corollary of this is that it is foolish to promote a state of higher technical efficiency if the benefits of being there don’t exceed the costs of getting there. Thus the appropriate evaluation of proposed intervention should be based on a conventional financial or benefit: cost analysis and its implementation should be driven by cost sharing arrangements recognizing private and public net beneficiaries (Mishan, 1976).

While there is a growing realisation that investment in unprofitable efficiency gains is nonsensical, there is a continued clamour by vested interests for funding of unprofitable projects. In some instances there may be complimentary benefits or other trade-offs to bear in mind which may complicate decision making. These aspects may be made explicit in the benefit:cost analysis but are not central to the issue of identifying real water savings considered here.

The most complicated proposals are for the funding by government of water authorities’ projects to reduce outfalls in exchange for increased environmental flows. These arrangements must attenuate the property rights of water entitlement holders. This is so because the net effect on environmental flows is zero as shown in Figure 4. Hence additional water must be released from storage to keep the bargain to increase environmental flows. The additional releases mean allocations to irrigators are reduced. It can be seen as a scheme by water authorities to appropriate and sell part of irrigators’ bulk water entitlements. Such schemes promote an opposite view to that of Randall (1981) who advocated that “The simplest solution, it seems, would be to vest ownership of all tailwaters with the original water title holder”.

Another scheme to reduce water losses is the proposal to improve the accuracy of measurement of water deliveries to farms. The major assumption here is that water deliveries are significantly underestimated. Be that as it may, very little if any real water savings will result from improved measurement of deliveries per se because crop water demand will remain unchanged. Given that farm practices and technology remain the same, either the same real volume of water will be delivered to satisfy crop demand or a reduced area of crop will be grown under a limitation imposed by a cap on diversions. In the former case there is no increase in environmental flows and in the latter case increased environmental flows will come at an agricultural opportunity cost in addition to the cost of improved metering.

As Adam Smith said “It is the maxim of every prudent master of a family never to attempt to make at home what it will cost him more to make than to buy.” On this basis it would seem difficult to justify investment in water savings projects that cost more than the market price. An estimated supply curve for water savings is shown in Figure 7. A market price of $500/ML is also indicated. The fact that there are no savings identified below the market price and very limited volume is available at the market price indicates the market is well informed and operating efficiently.

Figure 7: Estimated cost and possible yield of water savings projects. Note that white circles indicate projects where savings are at best dubious, illusory or non-existent. (Data after Marsden Jacob in ACIL (2002) and Anon (2002))

After considering the dubious, illusory or non-existent nature of the water savings claimed for many of the proposed projects (indicated by white circles) Figure 7 shows that the prospects for obtaining high volumes of real water savings at any cost are very limited.

Figure 8: Volume and cost of potential real water savings identified in the connected Murray system

In comparison to the proposed increased volumes for the environment, Figure 8 shows only a couple of projects with significant potential real savings identified in the interconnected Murray system. These are 123 GL for on-farm options and channel sealing in the Murrumbidgee irrigation area (ABARE, 2001) and 800 GL for reduced evaporation losses from Lake Alexandrina and Lake Albert (Anon, 2001). These savings may come at a cost of $1000/ML and $1250/ML respectively.

Taking $500/ML as the market price for permanent entitlement to delivery of irrigation water, Figure 7 shows that there are no economical technical solutions to the problem of overuse of water resources by competing uses. Because catchment yield is limited by biophysical factors and the efficiency of use is limited by economic constraints to the adoption of technical solutions, a system of rational allocation is needed if unacceptable levels of degradation are to be avoided (Hardin, 1968).

“The economist can imagine circumstances in which, for example, organised groups of recreationists and wildlife enthusiasts would purchase water entitlements and leave them unused to augment, at their own expense, in-stream flows beyond the required minima. Realistically, one would not expect such behaviour to be especially prevalent. But it is hard to conceive of any resource misallocation which would result from its occurrence” Randall (1981).

Indeed, the ACF recently indicated it would not support property rights for water for the environment while it could obtain increased environmental flows more cheaply through the political process (Moss, 2002).

This examination of the nature of water losses due to inefficiency has outlined basic principles and a detailed analysis should be carried out to evaluate major prospects. But notwithstanding this caveat, the majority of anticipated savings from most projects promoting increased water use efficiency are illusory due to errors in logic and the inability or reluctance of the promoters to view water flows in a systems context.

The indisputable conclusion is that the economical opportunities for real water savings in the connected Murray system can only be measured in hundreds, rather than thousands of gigalitres. Thus increasing environmental flows beyond some hundreds of gigalitres will have nationally significant opportunity costs measured in billions of dollars rather than millions.

To the extent that LWMPs tax and manage the external impacts of irrigation, market prices indicate the social cost of moving water out of agriculture. Little is known of the demand curve for environmental flows but institutional reform properly defining water rights and allowing wider access to the water market would make the derivation of environmental demand an academic exercise.

Market prices indicate the net present value of existing and new irrigated agricultural development opportunities at the margin of regional resources. That governments have indicated willingness to pay double the market price for water savings projects (ACIL, 2002) may indicate either a reluctance to allow an adjustment to policy decisions through the market or an anticipation that market prices will rise dramatically in response to increasing scarcity. This further underscores the prevailing gross underestimation of the agricultural impact of reduced allocations based on some economic modelling.

While there are no currently economical options for greatly increasing water resources in the connected Murray system some may become so as market prices rise in response to reduced allocations for consumptive use. A promising prospect for real increases in effective water resources from reduced evaporation is the decommissioning of Lakes Alexandrina and Albert as irrigation storages (Anon, 2001).

Some very high cost proposals such as pipelining are being promoted on the basis that water savings will be transformed into expertly marketed produce of “high value” far exceeding the cost of water savings. Yet a moment’s reflection will show that, however financially successful such developments may be, the economic value of the water savings can not exceed the least cost alternative source of supply.

Well defined property rights and soundly constructed markets can value and provide that source of supply.

ABARE (2001) Benefits of increasing water use efficiency: A case study within the Murrumbidgee Irrigation Area, report prepared for Land and Water Australia, Canberra

ACIL (2002 unpublished draft) Scope for water use efficiency savings as a source of water to meet increased environmental flows – Independent review, Murray-Darling Basin Commission, Canberra

Anon (2001) Options for water savings from the lower lakes for improved flows in the Coorong and the Murray Mouth, MDBC Technical report 2001/11, Murray-Darling Basin Commission, Canberra

Colby B G, Crandall K and D B Bush (1993) Water right transactions: market values and price dispersion, Water Resources. Res. 29-6, 1565-1572.

Hooke A (1997) Long term trends for irrigated agricultural commodities in Final Report for Project I-6604, Murray-Darling Basin Commission, Canberra

Marsden Jacob et al (2002) Improving water use efficiency in irrigation conveyance systems: A study of investment strategies, Final report to Land and Water Australia, Canberra. 13-15

Moss W (2002) Why the property rights debate is holding back reforms: A case for a focus on Structural Adjustment, A WWF Working Paper, Retrieved January 10, 2003 from the World Wide Web: http://www.agric.uwa.edu.au/ARE/AARES/

Quiggin, J. (1988), 'Murray River salinity - an illustrative model', American Journal of Agricultural Economics 70(3), 635-45.

Quiggin J (1992) Discount rates and sustainability, Proc 36th Ann. Conf. Aust. Ag. Econ. Soc., Canberra.

Randall A (1981) Property Entitlements and Pricing Policies for a Maturing water Economy, Australian Journal of Agricultural Economics, 25:195-220

Simon P and Anderson D (1990) Water auctions as an allocation mechanism in Victoria, Australia. Water Res. Bull. 26 3, 387-395.

Smith, Adam. An Inquiry into the Nature and Causes of the Wealth of Nations. Library of Economics and Liberty. Retrieved January 8, 2003 from the World Wide Web: http://www.econlib.org/library/Smith/smWN13.html

WWF (2002) Blueprint for a living continent: A way forward from the Wentworth group of concerned scientists, World Wide Fund for Nature, Sydney. www.wwf.org.au

[1] Irrigation intensity of 8 ML/ha

[2] Adjusted for market risk, existence of sunk capital, production uncertainty etc.