Coalmine approvals based on flawed models
In this story
Let’s not suggest former prime minister Tony Abbott was wilfully misleading the parliament in question time on August 18 last year. Let’s give him the benefit of the doubt and suggest he was but a naive and willing dupe.
And an angry dupe at that. Abbott was thundering on about the pesky conservationists who were mounting a legal challenge – “lawfare”, in his parlance – to Australia’s biggest coalmine, the Carmichael mine proposed by India’s Adani group in Queensland’s Galilee Basin.
On this occasion, Abbott was not arguing that coal was good for humanity, rather that it was good for the economy and that the mine opponents were saboteurs.
“This,” he bellowed, “is a project that will create 10,000 jobs.”
But that was not true. It was not even remotely close to being true, according to sworn evidence given to the Queensland Land Court by Adani’s own expert witness, Jerome Fahrer, of ACIL Allen Consulting.
By Fahrer’s calculations, made public as part of an export report many months before Abbott’s question time tirade, the number of jobs created would be but a sliver of that: “Over the life of the project it is projected that on average around 1464 employee years of full-time equivalent direct and indirect jobs will be created.”
Now, 1464 new jobs is not a figure to be sneezed at. But it is also about 85 per cent less than the number claimed by our former prime minister.
So how did he get it so wrong?
The short answer is that Abbott, like many others in politics, the media and elsewhere – including both Labor and conservative governments in Queensland – was simply regurgitating propaganda fed to him by Adani. Mining companies fudge the numbers, and those numbers are unquestioningly accepted by people who should know better. More deeply, though, there are the political pressures that drive the whole deceptive process.
A few other case studies usefully illustrate these various points, starting with Rio Tinto’s huge Mount Thorley Warkworth open-cut coalmine in the Hunter Valley.
Rio Tinto began digging at its Warkworth site in 1981. Originally, the mine was to operate until 2021. In 2003, Rio reached an agreement with the New South Wales Labor government to expand the mine. This was allowed on the condition that a ridge and the native vegetation on it, which separated the mine from the small historic village of Bulga, would be conserved. Thus the town would be spared much of the noise, dust and visual pollution of the massive open-cut mine.
But the company quickly recanted on its promise. The price of coal had gone up dramatically and Rio decided it could profitably mine an extra 230 million tonnes of coal from areas it previously deemed uneconomic. In 2010, it sought to increase the life of the mine by 10 years, to 2031, and expand it through Saddleback Ridge to within a couple of kilometres of the town – inter alia clearing a total of 766 hectares of four types of ecological communities listed as endangered under the state’s Threatened Species Conservation Act.
Its consultants set about compiling the required information on the economic and environmental impacts – the benefit and cost analysis – to justify its proposal.
On February 3, 2012, the NSW planning authorities waved the proposal through – as they always did – subject to some conditions, the most significant of which was that Rio would provide “biodiversity offsets”. That is to say, they would locate and protect areas of endangered ecologies somewhere else, to substitute for the ones they had previously promised to protect before they decide to knock them down.
But the Bulga Milbrodale Progress Association did not give up. It assembled its own team of experts, including several serious economists of leftish persuasion – Richard Denniss of The Australia Institute, Professor John Quiggin of Queensland University, and Roderick Campbell, then a principal of a small outfit called Economists at Large – who had a history of working for various NGOs, including some opposing mines.
The association took the company and the NSW planning department to the Land and Environment Court. And it won.
In a decision handed down on April 15, 2013, the chief judge of the court, Brian Preston, savaged the company’s case. He found that there would be significant impacts on the endangered ecologies, and that the environmental offsets were inadequate. He also found that there would be serious negative social consequences for the 350 residents of Bulga, particularly significant noise impacts, and that economic analyses of the project – input-output analysis and benefit-cost analysis – were inadequate.
Rio Tinto and the mining industry in general were aghast at this unaccustomed defeat. As Richard Denniss notes, they are so used to winning that they are apt to talk about the “approvals” process rather than the “assessment” process.
Rio immediately set about lobbying the then Liberal premier of NSW, Barry O’Farrell, to change the rules so such a thing could not happen again, which he obediently did. Then they reapplied for their extension. Late last year, the planning department again approved the mine extension.
It doesn’t look good for the people of Bulga, but at the time of writing a further legal challenge was being mooted. By the time you read this, it may have been initiated.
Meanwhile, the coal price bounces along at about one-third of what it was at the peak of the resources boom, when Rio began the process. At this price, opponents argue, it is not worth it for Rio to go ahead and mine its extension.
Either the mining giant is taking a long view to when prices pick up again – maybe in five years, maybe never – or it sees it as a process of fattening the asset for sale to some smaller chancer. All the major resource companies are now looking to unload projects.
We’ll see how it plays out. But regardless of that outcome, the court decision on the Warkworth challenge did yield one lasting victory for the critics of mining, particularly coalmine developments. It blew a big, permanent hole in claims mining companies could make for the economic benefits of their pits, specifically in relation to employment.
“Rio argued the Warkworth mine would create 44,000 jobs,” Denniss says. “I told the court the true number was closer to zero.”
The problem, as Denniss explained to the court, was with the kind of modelling the miners’ consultants habitually used to inflate job numbers, called input-output modelling. It is a useful economic tool for some purposes, but not fit for the purpose for which the miners were using it.
“The Productivity Commission, the Treasury, the ABS, all know it’s crap [in that context],” Denniss says.
“Yet for at least 10 years, the planning departments in every state accepted this kind of modelling, uncritically.”
And that fact illustrates the degree to which state planning departments were captive to the industry they were supposed to be assessing.
“The departments,” Denniss says, “were happy to basically take whatever was fed to them, put a megaphone in front of it and broadcast it.”
In the case of Warkworth, Rio was proudly asserting the 44,000 jobs figure, produced by its dodgy modelling. But in its revived application based on other modelling, it claims a tiny fraction of that number – 1300.
“The judgement in the Warkworth case destroyed input-output modelling for them. Ever since that case, no one has taken input-output modelling to court again,” says Denniss.
Case in point: another appeal in the NSW Land and Environment Court against a mine approval by a community group, against the Ashton mine proposal, from the Chinese mining giant, Yancoal.
“It was already in train when the Warkworth decision came down,” says Roderick Campbell. “It was different in many respects, but exactly the same as far as the economics was concerned.”
The economic analysis was done by Gillespie Economics, run by a former planning department bureaucrat, Rob Gillespie, and the input-output modelling was done by Hunter Valley Research, the same team that did the work on Warkworth and a number of other mine proposals.
But when Yancoal saw the Warkworth decision, they opted to get new consultants, and brought in Jerome Fahrer from ACIL Allen, a bigger company that throws more money at more complicated modelling, which they would put to the court after agreeing that the input-output model had been debunked.
“Fahrer basically threw it under the bus,” says Campbell of the previous consultants’ modelling.
The court transcript bears him out.
“Well, your honour,” Fahrer said in evidence, “you will recall from the Warkworth case that input-output modelling was criticised by the chief judge and was criticised in evidence and I think for good reason.”
He went on to offer essentially the same critique as the anti-mining economists had – except he was appearing for a mining company, and was setting up the presentation of his superior model on their behalf.
The big consulting houses have a vested interest in nailing down the lid on the methods previously used by their smaller competitors, say the anti-coal economists, because they are making big money by selling their more complex, expensive and somewhat better models.
“The industry has now switched to something called computable general equilibrium modelling,” Denniss says, “and their cases are based on the effect of proposals on GDP, not jobs.”
The econometrics here are impenetrable to non-economists. Which does not matter too much in this case, as the process of making mining assessments is largely a political one anyway.
Campbell cites the example of a proposed development at Cobbora, near the little town of Dunedoo in NSW.
“It was dreamed up by the former Labor government who were very keen to privatise electricity assets. The problem was they were trying to sell old generators no one was prepared to pay for. They also had some poor coal assets nobody wanted to dig up.
“So they proposed to develop a state-owned mine that would sell its coal to the generators at cost, hoping somebody might buy them, because they had access to cheap fuel,” Campbell says.
“Rob Gillespie was brought in and found the project to be worth $2 billion to NSW. The NSW planning department trumpeted the benefits.”
The hitch was that the government changed. The incoming Liberals got treasury to reassess the mine and it found the project would waste $1.5 million and never make money.
The new government abandoned the idea of a state-run mine. By then, though, the state had bought up more than 100 farms and houses.
The new government decided to sell out. No buyer has yet been found. Meanwhile, the farms are going to weed and the district has lost about 10 per cent of its population.
There is not much Gillespie the mine modeller and Campbell the community-advocate economist agree on, except that the mine assessment process is pretty farcical.
The process varies a bit from state to state. Gillespie explains the system in NSW thus:
“In the old days, the department of planning would assess the merits of a proposal and make a recommendation to the minister. Occasionally things would go to a commission of inquiry. And then the minister would make the decision.
“Now they have a Planning Assessment Commission. So it goes from the experts in the department of planning to the PAC and they do another technical assessment – although I’m not sure they have the technical capability – and what they come up with is often very different from what the planning department came up with.
“Then it’s bounced back and forth a bit and the PAC makes a recommendation to a second PAC. But the second Planning Assessment Commission is the same as the first. So they make a recommendation to themselves. So all the arguments had in between about whether they got it right or not are a bit moot.
“Plus there’s political machinations there.”
Campbell concurs, but adds more detail: the community consultation process gives opponents only a few weeks to prepare their responses to company proposals that typically run to a couple of thousand pages of detail, and usually takes no notice of their objections. Various parts of this process are heavily loaded with former planning department people and the odd former politician. Some members have clear conflicts of interest through their associations with companies that consult to the industry.
There are other flaws in the assessment system, too. The big one – in these days of rock-bottom resource prices – is that it does not take into account the likely profitability of a proposal, as was pointed out by Fairfax Media’s Michael West this week.
He noted that when the Chinese company Shenhua proposed its huge Watermark mine, in the middle of prime agricultural land in 2012, the company’s consultant, Gillespie Economics, found it would bring a net benefit to the country of $3 billion.
But that was predicated on coal prices twice as high as they are today.
“Gillespie’s input prices were provided by none other than its client Shenhua,” West wrote. “Anyway, the department happily signed off on the Gillespie analysis of the Shenhua mine in 2014.”
Rob Gillespie confirmed to The Saturday Paper that he simply worked with the price estimates given by the companies, which in turn relied on market predictions by big international advisers such as the firm Wood Mackenzie, which self-describes as the “global leader in commercial intelligence for the energy, metals and mining industries”.
No matter, Gillespie suggests.
“The premise that the government should intervene in the market by trying to determine profitability is very odd in a capitalist market-driven economy,” he says.“It’s anathema. Let these people invest, and if they fall over, they fall over. It’s like the government saying to a restaurant, ‘No, no, you won’t make any money, so we won’t allow you to invest in it.’”
The difference, however, is that when a restaurant opens it does not usually drive its neighbours off their properties. And when it closes, it just leaves an empty shop. It does not usually leave taxpayers out of pocket for white elephant infrastructure or leave a big hole in the ground or large-scale social or environmental damage.
Wood Mackenzie reported last December that two-thirds of the world’s coalmines, and half of Australia’s, were not turning a profit. Citi predicts coal prices will continue to fall through 2016 and 2017, and Bloomberg foresees more mine closures. Even the optimists don’t see a pick-up in prices for five years or so. No one suggests coal will ever boom again.
But the coal cargo-cultists continue to berate those who have delayed projects such as Warkworth, Carmichael and Shenhua.
They should probably be thanking their opponents for saving them from a terrible financial mistake. But that is no concern to those who just keep on approving more coalmines. Not yet, anyway.
This article was first published in the print edition of The Saturday Paper on Feb 27, 2016 as "Poor coal models". Subscribe here.