Scott Armstrong of the Wharton School at the University of Pennsylvania is co-author of a paper which was presented to the International Symposium on Forecasting in June. The conclusion: “The paper explained the need for simple methods and conservative forecasts in the face of uncertainty and complexity and pointed out that simple no-change benchmark forecasts are sufficiently accurate for policy decisions. In contrast, simple causal models with CO2 as the policy variable are not credible.”
Prediction markets for temperatures in three and ten years time agree that the no-change forecast is the more likely outcome than the IPCC 0.03C per annum forecast. Finally, similar (analogous) alarms in the past identified by the authors and others turned out to be false alarms. The slides for the talk are available as a PowerPoint file and as a PDF file.
In layman’s terms, Armstrong et al conclude that if you had to bet what the temperature will be in ten years, “the same as today” is a more accurate than any scientific model. And the no-change model is the one they recommend using for climate change policy.
This sounds absurd! Why, there is a well-known Scientific Consensus on man-made global warming! Do you think they’re wrong? Then here’s your chance to make some easy money: The Climate Bet.
Scott Armstrong of the Wharton School challenges Al Gore $20,000 that he will be able to make more accurate forecasts of annual mean temperatures than those that can be produced by climate models. Scott Armstrong’s forecasts will be based on the naive (no-change) model; that is, the forecasts would be the same as the most recent year prior to the forecasts. The money will be placed in a Charitable Trust to be established at a brokerage house. The charity designated by the winner will receive the total value in the fund when the official award is made at the annual International Symposium on Forecasting in 2018.
So far, Al Gore has proffered several reasons for declining the bet. Among the more laughable – “Mr. Gore does not make financial bets”. No, Mr. Gore likes a financial “sure thing’, like being a partner in Kleiner Perkins.
But, dear reader, don’t think that this bet is only available to Nobel laureates like Mr. Gore. You, too, can have a financial stake in the outcome! There is an Intrade market for the proposition “Gore vs Armstrong”. Betting Armstrong to win costs you $62 right now (to win $100); taking the Gore side of the bet would win $100 on only $38 at risk – over 150% profit! Given the overwhelming scientific consensus behind anthropogenic global warming, how can you possibly say “no” to this proposition?
Paradoxically, over the last 150 or so years, a 10-year forecast of “no change” has been more accurate 68% of the time, despite the fact that the underlying temperature trend is increasing.
[I encourage anyone with a curiosity about climate statistics to examine the linked Powerpoint presentation. It is full of interesting factoids like the following table, which should be titled “Correlation Does Not Imply Causation”. The authors examined several time-series data sets and evaluated each in terms of its correlation coefficient with respect to global temperature.]
Atmospheric CO2 1850-2008
U.S. Postal rates 1885-2009
U.S. Price Index 1850-2009
NOAA* expenditure 1970-2006
Books published in U.S. 1881-2008
No change (naïve model)