‘If rain don’t come this month,’ said Dan,

And cleared his throat to speak —

‘We’ll all be rooned,’ said Hanrahan,

‘If rain don’t come this week.’

‘If we don’t get three inches, man,

Or four to break this drought,

We’ll be rooned,’ said Hanrahan,

‘Before this year is out.’

Said Hanrahan, J O’Brien, 1921

O’Brien’s poem tells the story of a very pessimistic farmer whose obsession with the weather was probably understandable, given that his cropping fortune depended on getting the right quantity of seasonal rainfall at the right time. For managers of public water utilities and their responsible ministers, the stakes of the rainfall gamble are also very high. They take the blame when a city runs short of water in a country where urban water supply seems to be a matter of politics rather than of economics. The difference, however, between Hanrahan’s situation and that faced in the public water sector is that utilities can now take action to protect themselves from the vagaries of the weather by investing in desalination, a climate-independent technology. Unfortunately, there is little to protect society from the Hanrahanian pessimism that seems to be ingrained in those who have a stake in the Australian weather. As Hanrahan found out soon enough, too much rainfall can lead to ruin too:

…In God’s good time down came the rain…

… and every creek a banker ran

And dams filled over top.

‘We’ll all be rooned,’ said Hanrahan

‘If this rain doesn’t stop.’

The rainfall probability distribution also poses substantial risk to investors in desalination technology. Cooley et al. (2006) report the case of a desalination plant built in Santa Barbara, California, in response to a prolonged drought, that was shut down shortly after it began operation because the drought broke and it was deemed too expensive to run compared to conventional supplies. In the two weeks prior to the announcement of the successful contractor for construction of Sydney’s 80 gigalitre (GL) per annum desalination plant, heavy rains had recharged Sydney’s dams by 328GL (two-thirds of annual consumption). In this case, no change was made to the construction schedule on the grounds that the desalination plant was still justified to compensate for a reduction in rainfall associated with climate change. However, it remains to be seen how palatable the high costs of running the desalination plant will be if Warrangamba dam is full when the plant is completed.

The trade-off between the cost of supply failure if it doesn’t rain, compared to the financial burden of redundancy if it does, has economic and political dimensions. A large proportion of urban water demand in Australia is used outdoors and the strategy for addressing short-term supply failure has been to impose restrictions on outdoor use. Water restrictions impose a welfare cost on the consumer, such as a loss of utility from outdoor gardens and the time cost of hand-watering. Based on decisions made in Western Australia in recent years, the perceived political costs of urban water restrictions are substantially higher than this. Whilst there is an argument for reserve capacity as an insurance policy where there is a perceived high cost of failure, there has been no scrutiny of supply-augmentation decisions made by planners regarding the economic trade-off of augmentation vs. supply failure.

In days past it would have been possible to measure the perceived value of supply failure by comparing expenditure on system capacity against the probability that failure will occur; but this is no longer possible because the probability of rainfall outcomes can no longer be securely based on the historical climatic record. Thus the ‘climate change’ issue seems to present an opportunity for Hanrahan’s descendents to assume whatever they like about the future catchment yield, so making it is possible to justify substantial investment in desalination technology that may have questionable economic foundations. In Victoria, the calculations for the desalination plant there are based on future catchment yield being defined by the mean of the previous three years (Department of Sustainability and the Environment 2007). In Perth, the Water Corporation recently changed the historical period used to assess future system yield from a nine-year (1997–2005) average to a six-year (2001–06) average, arbitrarily dropping the wettest year in recent history (2000) and adding the driest (2006), thus ensuring an additional 23GL ‘shortfall’ in the yield of the conventional sources.

Even if political processes deem that an ultra-conservative, ruler-and-pen approach to forecasting the climate is acceptable, there is an economic-efficiency problem if the assumption is not applied consistently across water management. If it is reasonable to assume that the current drought will last forever, then the implications for ‘long-run marginal cost’, and thus efficient water pricing, are substantial: for even though urban water demand is relatively inelastic, the extent of the price increase rationally necessitated by climate change may reduce demand sufficiently to make the rush of current desalination construction planning across the country surplus to requirements. If only the economic regulators would adopt the same climate forecasts as the water utilities!

This paper presents an analysis of these issues using a case study of the proposed second desalination plant to be built in Western Australia, at Binningup. The aim of the paper is to highlight the need for more rigour and more consistency in the planning and pricing of urban water, especially given the prospect of climate change. The next section provides a brief overview of the climate situation in Western Australia. The magnitude of uncertainty regarding climate and catchment yield and its implications for urban water-supply planning are then demonstrated using a simple balance sheet of water supply and demand under current institutional arrangements. A more rigorous assessment of system yield and risk of supply failure is then presented, which demonstrates the relationship between assumptions about the probability-distribution of catchment yield, the timing of supply augmentation, and the cost of supplying urban water over the medium term. There is no attempt in the paper to provide a complete economic analysis of the problem, because of lack of information on the economic cost of supply failure. However, the long-run marginal-cost implications associated with these climate and supply-augmentation assumptions are calculated and compared to prices currently paid by consumers.