On the basis of the preceding discussion it would appear that there is some merit in taking a national approach when water resources are shared across competing jurisdictions. For example, without a national approach, a full appreciation of the downstream impacts of investments in ‘water-use efficiency’ in one upstream jurisdiction may not emerge. Regrettably, there is no evidence that this broader, integrated view has arisen from national intervention in Australian water policy. To illustrate this problem we use this section to describe and analyse the Food Bowl Modernisation Project (FBMP), which recently won the support of the Commonwealth government. Similar arguments may well apply to other projects of this genre although the absence of data on many, particularly those that are ‘currently at the conceptual planning stage’ (CoAG 2008b) makes scrutiny problematic.

The euphemistically named FBMP forms only one component of the Victorian Government’s ambitious water policies assembled under the Our Water Our Future framework in 2007. This framework also comprises the construction of a 150GL desalination plant in the Wonthaggi region, expansion of the Victorian water grid by establishing additional pipelines between major centres, increased capital expenditure on water-recycling projects and additional support for water-recycling programs.

The FBMP has six key elements.

First, manual structures for managing the supply of water via channels are to be replaced with automatic channel-control technologies.

Second, some sections of the open-channel network are to be replaced with pipes and/or remodelled.

Third, Dethridge wheels (which measure water use) are to be replaced with more accurate metering devices. ^[6]

Fourth, changes to water charging to reflect the additional investment base are foreshadowed. ^[7]

Fifth, some farm system adjustments, such as a reduction in the number of off-takes, ^[8] are anticipated as part of the project.

And sixth, a sequence of consultations and communication to adjust to different service demands are predicted (DSE 2008).

The FBMP reportedly aims to ‘save’ 225GL of water per year by improving distribution efficiency with the resulting ‘savings’ to be shared equally — one-third being allocated to irrigators, one-third being exported to Melbourne via the Sugarloaf pipeline, and another third assigned to environmental uses. ^[9]

In total, the Our Water Our Future initiatives are estimated to cost $4.9 billion, with 90 per cent of the cost being borne by water consumers via increased charges (Victorian Auditor-General 2008: 19). Like the Howard government’s National Plan, the Victorian initiatives were hurriedly assembled over a six-month period, largely in response to the unprecedented low inflows in 2006. Whilst the Victorian Auditor-General concedes that ‘the speed of the response’ may account for some of the deficiencies in planning, he nevertheless observed that ‘for some of the key projects the rigour was inadequate’ (p.v). In the case of the FBMP, the Auditor General specifically noted that ‘the upgrade costs (reported in the plan) represent the lowest level of rigour and were, at the time, based on a preliminary study by a stakeholder group (the Food Bowl Alliance)’ (p.31). Importantly in the context of ‘water savings’, the Auditor-General also sourced earlier work used to develop the business case for the FBMP and found that in these earlier documents ‘the estimated water losses were more refined and lower than those published in the food bowl steering committee’s final report in November 2007’ (p.35; original emphasis).

In order to shed additional light on the magnitude of this problem it is worth considering some of the earlier work undertaken on water-use efficiency in this setting. Of particular interest is a pre-feasibility assessment undertaken by Marsden Jacob on behalf of the Murray-Darling Basin Commission in 2006. This work was undertaken primarily to assess the quantum of water that might be ‘recovered’ from the Shepparton Irrigation Area (SIA). The project had three main goals: to reduce irrigation outfalls through channel automation; to improve the detection of losses in the channel system; and to undertake investments in seepage and leakage reduction. The SIA is only one of six districts covered by the FBMP but the empirical approach and findings are instructive on several grounds.

To estimate the quantum of water that might be ‘saved’ by this project Marsden Jacob and Associates (2006) categorise the various forms of distribution losses within the irrigation network. In the context of the proposed automated channel technologies that forms a core part of the FBMP, the greatest potential for ‘savings’ is attributed to the water ‘lost’ through channel outfalls. A channel outfall is the mechanism by which excess water and return flows pass from the irrigation network to the surrounds. In some instances, this will be a structure that links to a river or, in other cases, water might pass to a swamp, creek or depression. ^[10] Marsden Jacob and Associates (2006: ESiii) concede that ‘the destination or final use of this return flow is unknown but could conceivably include extractions by diverters in the Goulburn and Broken system or discharge into the River Murray where it becomes part of the tributary contribution to Victoria’s share of the River Murray water resource. Notwithstanding this caveat and numerous instances where data were unavailable or embodied significant measurement error (see, for instance Marsden Jacob and Associates 2006: 16; 21) the study arbitrarily assumed that differing percentages of the water that entered outfall drains constituted return flows. These range from an assumed 10 per cent return flow for four large drains to 100 per cent return flow where the outfall was directly to a river. An accompanying assumption was that 50 per cent of all outfalls came about as a result of rainfall rejection flows. These flows arise when irrigators close their receiving infrastructure because of heavy rainfall during an irrigation event. Put differently, this study assumed that a significant portion of the water rejected by farmers had no other use, including maintaining in-stream flows. Whilst it is encouraging that the study at least acknowledged the existence of return flows, there is considerable conjecture about the actual volume of water involved, its present uses and possible end destination. In addition, whilst the FBMP forecasts a change in distribution efficiency from 70 per cent to 85 per cent, Marsden Jacob and Associates (2006) offer a more cautious outcome suggesting a change from 70 to 80 per cent as being plausible. The upshot of the work by Marsden Jacob and Associates is that there is considerable conjecture about where the water purportedly ‘lost’ from the Goulburn Valley is presently going. Moreover, redistributing this water under the guise of irrigation efficiency runs the risk of depriving existing users with low-ranking claims, including environmental beneficiaries.

There are two key issues here. First, water is invariably fugitive and measuring it with precision is costly and difficult. This is not unique to Australia and, as we noted in the earlier section, there are numerous projects around the world where the purported ‘water savings’ turned out to be much less than the original estimates used to justify the project. Second, the scale of analysis and the incentive to focus on local water use invariably leaves downstream users worse off. Without a clear view of the quantum of return flows before embarking on a ‘modernisation project’ it will always be difficult to assess the actual detriment to downstream users/uses. Moreover, once the project is completed it will be costly and difficult to redistribute the resource in its original configuration. This is further complicated in the context of the FBMP since one-third of the water ‘saved’ is purportedly to be used for environmental benefit. Since there is uncertainty about the quantum of water presently accruing to ‘the environment’ under the status quo, it will not be possible to test whether the 75GL assigned for ‘the environment’ represents an increase or decrease in environmental amenity.

Notwithstanding these uncertainties, inconsistencies and potentially flawed logic, the Federal government announced its support for the FBMP in March 2008, following the twenty-first meeting of the CoAG. In reaching this decision, the Commonwealth ‘agreed in principle to fund 90 per cent of the project costs, up to $1 billion of the Stage Two Food Bowl Project in Victoria, subject to a due diligence assessment and delivery of half the gains in additional flows to the Murray River’ (CoAG 2008a: 7). The Commonwealth also signalled to other states that it intended to continue down the path of ‘modernising irrigation’ and funding additional ‘water saving’ projects. The largesse of the Commonwealth at the July meeting of CoAG gave effect to this commitment.

Regrettably, the mythology that attends ‘water-use efficiency’ projects seems likely to be perpetuated. Even within the academic profession there are signs that the fiscal suasion of the Federal government can override water distribution logic. In June 2008, the Federal government announced $8.6 million of funding to two universities that have long and distinguished histories in hydrology. The project reportedly aims to ‘find ways to make better use of the water we have, creating benefits for both farmers and the environment’ and to ‘provide farmers with practical ways to make the most of available irrigation water supplies — including rainfall and recycled water — through better planning, technology and predictive tools’ (Wong 2008b: 1). Unfortunately, there is no indication that the project will assess the impact of these measures at a wider and more appropriate scale.

On a more cynical note, the political allure of the water-use efficiency chimera shows no signs of weakening. By definition, water politics is hard work and orchestrating the genuine redistribution of a tightly held resource in favour of broader environmental interests was always going to be viewed by the polity as a zero-sum game, at best. Convincing the electorate that more water can be ‘created’ in order to satisfy environmental interests whilst maintaining the existing distribution of rents remains far more politically palatable, even if this approach results in long-run negative outcomes.