Betting on Climate Change

(based on a poster presented at the EGU 2005 - but with extra material about Richard Lindzen)

Back to my home page

More on this can be found on my blog., including my comments on the "Reason" magazine article for those readers who came here from there. I've just had an article published on which covers similar ground.


Probabilistic climate forecasting is not easy, and each forecast inevitably depends on a number of assumptions made by the researchers who build the models and explore their behaviour. This makes the construction of a probabilistic forecasting somewhat subjective. (It may be worth pointing out that weather prediction - and just about everything else you could think of - suffers from the same theoretical problem, but short-term forecasters can validate and fine-tune their methods on a daily basis, whereas we get one shot at a 100 year forecast and therefore cannot use the same iterative approach to improve methods and build confidence). Nevertheless, we must make policy decisions and should do so based on our best estimates of the future. How can we form these estimates in such a way as to reduce the risk of researcher bias? This page presents one possibility, based on a futures market for the Earth's climate.

Let the Market decide

That's all there is to it really. All we need is a market in climate futures such as "The amount of global warming from 2000 to 2030". The market price would then be the consensus of all players (currently around 0.6C for that claim in the link, but that is only an internet game with no real money involved). More sophisticated futures could represent a range of outcomes in a probabilistic distribution (eg what are the odds on warming > 0.1, 0.2, 0.3C etc).

Weather futures markets are widely used for hedging in the energy markets, and can be shown to have skill at predicting temperature (and therefore energy consumption) at the monthly time scale.  However, this market only extends to the seasonal scale at longest. Agricultural commodity markets also relate to expectations of future weather, but are again too short-term to relate to anthropogenic climate change. The concept of an "Ideas Futures Market" is not new, but the inherent intractability of the climate prediction problem, together with the importance of the outcome, adds new impetus to the case for a climate futures market. There are many reasons why I believe a climate futures market could be useful. Firstly, it would enable us to find the real consensus view. Commentators who act in bad faith (either through exaggerating the chances of climate change, or understating it) might find the likelihood of financial loss would motivate more honesty. Interestingly, Julian Simon thought that such a market "would thus no doubt often serve as a corrective to alarmist stories about impending doom" but I have also found that climate change sceptics are remarkably reluctant to back their proclaimed beliefs by betting against significant climate change. See below on my exchanges with prominent sceptic Richard Lindzen.
Of course no market is infallible, but anyone who genuinely thinks the current prices are badly wrong can go and make a lot of money, so at least they won't complain too loudly :-)

A climate futures market would enable those who feel vulnerable to climate change to hedge against their perceived risks - betting on sea level rise would have a very real relevance to Pacific Islanders. By betting on rapid sea level rise, they would either be able to stay in their homes at the cost of losing the bet (if sea level rise was slow), or would win the bet and have money to pay for sea defences or relocation if sea level rise was rapid. If they couldn't afford the bet, at least they would know they are already in trouble.

Richard Lindzen's "Climate Bet"

Professor Richard Lindzen is a prominent sceptic who claims that the anthropogenic influence on our climate is close to negligible and that in fact cooling is about as likely as warming over at least the next 20 years. Now, although there must be some non-zero chance of cooling on this time scale (due to the natural decadal-scale variability of the climate, and the possibility of large volcanic eruptions), this opinion certainly seems to be well to one extreme of what most climate scientists think is likely.

Recently, my attention was drawn to some comments attributed to Lindzen: "Richard Lindzen says he's willing to take bets that global average temperatures in 20 years will in fact be lower than they are now." (thanks to William Connolley for the tip). Given his widely-promulgated views, I took this quote at face value and contacted him to arrange a wager. A payoff at retirement age would be a nice top-up to my pension.

Now here's the kicker. Richard Lindzen will indeed accept a bet - but only if offered odds of 50:1 in his favour! He actually started out quoting 100:1 - but came down to 50:1 in what he described as a "special favor" to me. If the temperatures went down, I was to hand over $10,000, but in the event of a rise, I'd get a whopping $200. That's worth around $8 per year on my pension. Whoop-de-doo. That's not really quite what I had in mind. In fact, not only is $200 too small a win to be worth bothering with, but moreover I think that his side of the bet is probably more attractive than mine. Note that I certainly do not consider myself to be a sceptic, but on the contrary am just a bit-part player in climate research who thinks that the IPCC TAR (by which I really mean Working Group 1, "The Scientific Basis" which is the only bit I know much about) has basically got it right. Yet here is one of the most prominent sceptics who is apparently less confident about the chances of medium-term cooling than I am myself!

As a ballpark estimate, I reckon something in the region of 10:1 or 5:1 odds on the bet would probably be fair on the 20 year time scale - I haven't tried to pin it down with any great precision, as there is such a gulf between my viewpoint and Richard Lindzen's, that it should be easy to find a mutually agreeable wager in the region of 2:1 or 3:1 odds. As I explained to him, 3:1 would be a reasonable middle ground between an honest sceptic - Mr 50-50, who believes cooling is as likely as warming and would therefore expect to win $1 on average from a $3 v $1 bet - and a (mythical?) "true believer" who believes that a rise in temperatures is certain and would therefore also expect the same bet to be worth $1 to him. I'm not claiming here that 3:1 is necessarily the correct odds to arrange a bet - but it certainly provides a meaningful and realistic starting point for negotiations, in contrast to Lindzen's offer. I hadn't reckoned that the gulf in opinions was that he is more insistent on warming taking place, and less confident about the prospects of cooling, than I am! A 1 in 50 chance of cooling is hardly a consensus-busting viewpoint.

Richard Lindzen's words say that there is about a 50% chance of cooling. His wallet thinks it is a 2% shot. Which do you believe? He says he was misrepresented in the quote on the site - can we expect to see a correction, and headlines in the sceptic press along the lines of "Richard Lindzen thinks there is at least a 2% chance of global cooling"? I'd probably take bets against that happening, too - and 50:1 odds against it seems reasonable in this case!

Is there any sceptic out there who is prepared to place a reasonable bet against significant climate change?

If any other sceptics put more faith in Lindzen's predictions than he does himself, I'd be very interested in hearing from you. There isn't much money in climate science, and I'm still looking for that gold watch at retirement. I'm flexible about what sort of bet I would be willing to accept - for example, it could be a scaled claim, where the payoff depends on the magnitude of the temperature change (eg $100 per 0.01C above/below 0.2C in 20 years), and anything in the 20-30 year time frame would be ok.

Recently, Reason magazine ran a short article on this story. Disappointingly, Ron Bailey chose to misrepresent my opinions (and indeed the IPCC TAR). I've briefly commented on his article. As for why he chose to misrepresent me in this way (quite deliberately, as we had exchanged emails immediately prior to him writing the article), I will refrain from speculating here.

There are many more examples of sceptics ducking the challenge on my blog.

Related articles and sites:

My article on

"Using weather futures as weather forecasts", Vishu Kulkarni 2003, published by Chicago Mercantile Exchange

Foresight Exchange Market - an internet-based game
Idea Futures by Robin Hanson - some pioneering work on Idea Futures, including:

The Policy Analysis Market ("Terrorism Futures"), cancelled in a media frenzy - not because it wouldn't have worked, but because it was politically incorrect.

The story of the bet between economist Julian Simon and Paul Ehrlich ("The Population Bomb"). Specifically, Simon bet in 1980 that commodity prices would drop over the coming decade, and Ehrlich that they would rise. Simon won.

However, Simon lost a second bet on pine sawtimber prices.

Information on weather derivatives etc.

Wanna Bet by Jim Giles - Nature article concerning the Simon Market etc

The Simon Market is not yet operational, but Tom Bell is hoping to set up a real ideas market.

The Longbets Organisation takes bets with real money, but the proceeds go to the charity of the winner's choice (due to betting/licensing issues). Money is invested in a long-term investment portfolio, so the payoff will be worthwhile. But since neither the winner nor the loser can collect, it is really nothing more than a charitable donation and a soapbox and no-one has any incentive to actually bet honestly.

Chris F. Masse's pages have a lot of links and information about prediction markets in general.

Back to my home page