Was the Garnaut Report based on a statistically dubious science paper?

Here’s an eye opener………

I read a paper linked  by ClimateAudit.com which describes a finding by a statistician, David Stockwell about a widely circulated paper suggesting AGW was getting worse. Stockwell strongly asserts that the paper applied “wrong” statistical smoothing in a non linear data series.

What’s the problem with that?

Well, Garnaut used that paper called Rahmstorf  et al (2007) to basically anchor  his entire economic analysis. So it could very well be that Garnaut’s advice to the government was actually premised on wrong statistical evidence that AGW was getting worse than originally thought, particularly as it compares to what the IPCC said.

Abstract from the Stockwell paper says:

The non-linear trend in Rahmstorf et al. [2007] is updated with recent global temperature data.  The evidence does not support the basis for their claim that the sensitivity of the climate system has been underestimated.

And here is what Stockwell says about Garnaut.

Despite the lapse in statistical rigor, Rahmstorf et al. [2007] has been widely cited in support of more urgent action on emission controls [e.g. Garnaut 2008].

So from what I can gather Garnaut moved away from the science applied in the IPCC report to a much more gloomier report and based his economic study on a paper, which has now been possibly found to be of dubious quality and reliability. Further, if is found to be dubious is there a legitimate claim for a refund?

See here for the paper. Sourced from ClimateAudit.

Here’s my question: are we going to see Garnaut and the Government come back and revise their strategy and economic analysis if the paper is found to be a unreliable?

Who is David Stockwell?

After receiving his Ph.D. degree in Ecosystem Dynamics from the Australian National University in 1992, Dr Stockwell worked as a consultant until moving to the San Diego Supercomputer Center at University of California San Diego in 1997.

[Formatting edited 06/07/09]

23 thoughts on “Was the Garnaut Report based on a statistically dubious science paper?

  1. In my own areas of interest statistical analysis is a huge problem. There are many problems here that have been longstanding. I referred a recent paper entitled “WHy most research findings are false” to a friend of mine in New York. He replied with a 1962 reference from his statistics professor which found that in abnormal psychology studies it was statistically absurd to find that many instances of statistical significance. Additionally, and this is how sad it really gets, one source of error was that if significance was attained the bods did not check their calculations, whereas if it was not achieved … .

    The geneticist Steve Jones once wrote: There are two kinds of scientists: those who use mathematics and those who understand mathematics.

    In my view relying on statistical analysis alone, as so often happens in biomedicine and health related matters, is not good enough.

  2. Rahmstorf is based at the Potsdam Institute for Climate Change. Their whole reason for existence is to produce mitigation modelling for the non-existant problem of global warming. I’m sure Rahmstorf is a committed scientist but the emphasis at Potsdam, led by Schellnhuber the founding director, is one of impending disaster and the temptation is always there to draw conclusions that may not exist. After all, funding depends on the belief in global warming by politicians and they have to be kept scared.

    Also at Potsdam there are two very influential scientists who produce many of the scary scenarios, but are actually Greenpeace activists, one of them Australian. See the details and the proof here:

  3. It seems to me that pretty much ALL global warming alarmists are relying on computer models. So the accuracy of computer modelling is the key issue and the point of contention in the global warming debate.
    The “scientific” consensus that people care so much about seems to be predominantly based on computer models which all assume the same hypothesis.
    Does anyone know, how accurate is this statement?

    I certainly think that computer models could be inherently flawed by assuming CO2 is a fundamental driver of climate – this is the hypothesis, so scientists attempt to make the empirical data fit the hypothesis. Absolutely nothing wrong with that approach to science unless the empirical data doesn’t actually fit. eg/ The earth saw cooling in the last 10 years and from the 1940s-1960s too just as man made CO2 levels were significantly increased.

    And I don’t think a universal computer model that works for data from different time periods, different locations and different temperature reading locations (eg/ stratosphere/troposphere/land/sea etc) exists?
    But maybe someone knows more about the limits of modelling? or some good links? How are researchers validating the limits of their models?

    Anyway nice post JC.

  4. I’m bemused by those that are convinced that AGW is definitely happening and is definitely dangereous. However I’m equally bemused by those that say AGW does not exist (eg Harbinger). How do people end up so damn certain?

  5. Except the one’s saying that it isn’t happening aren’t going to end up sticking us in useless policies that will end up costing serious money, lower the potential rate of GDP, turn things into a renting seeking sludge match and end up doing next to nothing to mitigate.

    You think they’re still equal, Terje, or is that a way of gaining a senate seat (by looking even handed)?

  6. I agree with Terje. Having a strong belief that AGW is not happening shows a lack of skepticism.

    TimR — the models are long term models and aren’t designed to pick up short-term fluctuations. Same with most economic models.

    No economic model can predict when a recession is going to happen. But an economic model will tell you that a tax hurts the economy eventually. You shouldn’t take the models as gospel, but you also shouldn’t entirely ignore them.

  7. People can and will argue over the Garnaut reports and the rest. There are now other signs emerging. Ocean acidity is rising and some suggest this has severe implications for ocean life. Some studies are claiming that we are already seeing thinning in the shells of various organisms as calcium is leached out through the acidification. Then there is the recent claim re shrinking sheep.

    My position is identical to Terje and John Humphreys. I don’t how bad it will be but I’m prepared to invoke Pascal’s wager on this one. For those who care about the future of our civilisation we should at least have some plans in mind to address what might happen. As for myself, no kids so don’t care.

  8. Having a strong belief that AGW is not happening shows a lack of skepticism.


    1. believing it is and then crafting a stupid policy to mitigate that simply won’t work.

    2. seeing recent ” studies” present even worse case that IPCC worse case and then finding out they’re padding the data treatment.

    What do you think of these issue?

  9. JohnH:

    Your comment is pretty meaningless if you think Garnaut’s report isn’t that important. In a sense it’s actually more important than the scientific papers as he basically advised the government on what to do economically.

    He’s based, it seems, the science on the most pessimistic scenario rather the simply relying on the IPCC.

    In other words it’s a junk report.

  10. JC,

    As I have repeatedly stated, I have bugger all faith in modeling long term trends in complex systems. I don’t think we are capable of that. Damn it, we can’t even model the dynamics of a simple cell. We can see trends, we can catch the overall direction but we cannot nail these things down with the precision some climatologists and economists suggest. In any event criticality may intervene and make all our ideas redundant. For example, it was only several years it was believed that an ice age took a long time to end, new data now suggests it can happen in a geological instant(30-40 years). Actually I wonder if the confusion is emerging because we are driving a warming but at the same time the planet is heading towards another ice age. This is a long inter glacial period.

    Thus the economic models are precarious and I don’t think we should be going overboard. I would much rather we just start building lots of nuke plants and do it real quick. We may not know what is going to happen but lots of energy on tap is a good safeguard against the possible consequences.

    So tell me JC, just what should we be doing?

  11. Possibly just watching as the warming is actually quite benign over the past 100 years or so… .65 degrees which is hardly a huge concern.

    IPCC mid range is 2% in 2100 which is really quite manageable.

    Do nothing and watch or perhaps even tax carbon and reduce income taxes.

  12. JC — I agree that a lot of people are fear-mongering and exaggerating well past the point of reasonable science. And I agree that the science does not justify bad public policy (including the current ETS proposal).

  13. Possibly just watching as the warming is actually quite benign over the past 100 years or so… .65 degrees which is hardly a huge concern.

    IPCC mid range is 2% in 2100 which is really quite manageable.

    Do nothing and watch or perhaps even tax carbon and reduce income taxes.

    That pretty much summarises it for me. I’d probably add one extra point that there is no definitive science which identifies what is going on, at least not on the balance of probabilities – so any action we take is actually pretty close to arbitrary in nature. The carbon tax plus offset by reductions of other taxes just makes it in as a rational piece of ‘insurance’.

  14. JC – your proposed action plan sounds perfectly reasonable to me. Taxing carbon and cutting income tax would mitigate some political risks. Risks that are currently being realised by way of dud alternatives.

  15. Terje

    and I’m sorry if I sounded terse earlier. You’re a “brother” in the sense that disagreements are tiny if any.

    And if you ever decided to run and win a senate seat the country would he better off by multiples.

  16. JC – thanks. Our differences are mostly about form rather than substance. And half the reason I blog is to encounter different ways of looking at things. I don’t expect to make another serious tilt (long shot or otherwise) at trying to win a seat until 2016 or later. In the interum I’m hoping that some other libertarians can break through for the LDP. In any case my earlier statement had nothing to do with political posturing and was simply a thought I shared.

  17. PS:

    If John Humphreys runs again in Qld I hope it is mentioned here. I voted for him last time and will vote for him again.

  18. Regarding models………Having emailed Garnaut last year, asking him whether the Garnaut report would include assessments of the economic impacts of carbon trading on Small and Medium Manufacturers, I got the following reply, “Many of your questions could not be definitively answered by economic modeling as there are many compounding sources of uncertainty in the climate change science and impacts analysis, as well as uncertainties inherent in undertaking economic modelling over long time periods.” Say no more.

Comments are closed.