Colin’s canal, again

Apologies to readers for the rather odd post below, which was meant for my own blog. Since people have made comments, I’ll leave it in place, and add a few notes of explanation. The post refers to a state election campaign in Western Australia, which the incumbent Labor government was, at the outset, expected to lose. A major issue in the campaign is the water supply problems facing Perth, the capital city where the great majority of the population lives. Issues include the traditional Australian ideology of developmentalism, and the role of public-private partnerships

My piece in today’s Australian Financial Review, over the fold, brings together arguments about the Kimberley canal project, which has been debated here on the blog on my blog. As usual, I got a lot out of all the comments, whether or not this is obvious in the published article. Thanks to everyone who contributed to the debate.

Economic rationalists are often criticised, not entirely without reason, for lacking vision. Still, when confronted with a visionary project like the canal being advocated by WA Opposition leader Colin Barnett to bring water 3700km from the Kimberleys to Perth, we are reminded of that mirages are a kind of vision.

There aren’t many issues on which I agree with the Institute of Public Affairs. But I’m happy to join IPA Director Mike Nahan in suggesting that this is a project fraught with danger for WA taxpayers. It could easily join the Ord River fiasco as an illustration of what happens when developmentalist rhetoric trumps economic rationality.

Campaigning on the slogan ‘decisions not delays’, Barnett has promised to go ahead with the scheme without waiting for feasibility studies, environmental assessments or any of the other impediments of bureaucratic rationality. The evidence supporting the proposal appears to consist of some very slim documents proposed by project proponent Tenix.

Even a moment’s scrutiny of these documents indicates some obvious problems. For example, the feasibility of the project is supported by reference to the “1,065 km Californian Aqueduct’, built in the early 1900s. The Aqueduct’s own website states its length at 223 miles (less than 400 km).

But the real problems are with the economics. Barnett has promised to deliver water at $1 a kilolitre, but this can’t be done, even on the most optimistic assumptions. Taking Tenix’s claimed construction cost of $2 billion, and a very conservative EBITDA ratio of 10, the capital costs for the project would be $200 million a year. Even at the claimed maximum flow of 200 gigalitres a year, this would use up the whole $1/kl without allowing anything for pumping, maintenance, treatment or reticulation.

More plausible estimates for construction costs, including staging costs, are up to $4 billion, and a more likely estimate for total capital costs (return on capital, depreciation and amortisation) is 12.5 per cent, putting the capital costs alone at $2.50/kl. In addition, each kl of water is a tonne of matter that has to be pumped 3700km, with little help from gravity. That’s not going to be cheap. The WA Treasury has estimated the delivered costs at $6.50/kl, compared to current prices ranging from $0.41 to $1.50.

Then there’s the fact that the proposal is a BOOT scheme, with the illusory benefit of handing the project back to the public after 40 years. Of course, there is no free lunch here. The cost is built into prices, usually with a handsome profit margin built in.

It gets worse. The Tenix proposal is based on the assumption that all the water is sold to residential users, but Barnett has talked of diverting up to 80GL each year to irrigators. This will no doubt do wonders for the politics of the proposal but it will be awful for the economics. There’s no way irrigators will be able to pay any more than the pumping costs (if that). Even $1/kl is $1000 a megalitre, which would render most irrigated crops uneconomic. That means that all the capital and maintenance costs will have to be spread over, at most, 120 GL of residential use, so the costs for residential users could be as high as $10/kl. At this price, fanciful options like transporting icebergs from Antarctica start to look attractive.

But even at much lower prices, market forces would resolve the problem. A mere doubling of prices would induce a reduction in water use, and would bring forth alternative sources of supply, such as repurchase of water currently used for irrigation. Then there’s the backstop option of desalinisation, already under way on a relatively small scale.

It’s just possible that careful study could show that the kinds of problems raised above can be overcome, and that the canal option is a feasible one. But there’s no time for such a study before the election.

Unless he can present a better case than he has done so far, it seems quite likely that this proposal will cost Barnett an election he seemed to have in the bag. This episode, and Labor’s difficulties with Medicare Gold, provide a harsh lesson for Opposition parties. If you have a complex and innovative policy, it’s better to put it out for scrutiny well in advance of the election, and risk having the government steal it, than to try and defend it, perhaps unsuccessfully, in the heat of an election campaign.