Friday, March 23, 2007

How Not to Bet on Global Warming

Brian at Backseat Driving has been challenging people to bets about global warming, and it looks like I’m the latest to be challenged.

In general, I like the idea of betting on propositions, because it forces people to put their money where their mouth is, as well as to think carefully about what they really believe. That said, I’m not sure Brian’s proposed bets really get to the heart of the matter.

You’ll notice that many of his proposed bets are about whether the world is getting warmer, without any reference to causes. As I’ve said before, I’m reasonably convinced that the world is getting warmer; I’d say it’s about 95% likely. That means I’d require at least 19-to-1 odds in order to bet that global temperatures will stay the same or get cooler. If Brian is willing to offer those odds, I’ll take them. But I would fully expect to lose the bet, having agreed only because of the high payout in the unlikely event of winning. I hardly think any loss-of-face would be on the line.

As far as I can tell, only one of Brian’s proposed bets raises the issue of the human contribution to global warming, as opposed to other hypotheses such as increased solar activity. Brian is willing to bet that the IPCCC report of ~2016 will conclude that anthropogenic global warming “likely intensified tropical storm damage in the 1995-2005 period.” So in other words, this is a bet about what a politically charged committee will conclude about the causes of events that have already occurred. Again, is this really the issue? I thought the point of bets like these was to arbitrate among hypotheses by reference to objectively measurable future facts of the world. Even a sincere global-warming-denier (and I’m not one) could reasonably reject this bet because he strongly suspects the IPCCC will continue to espouse the conventional wisdom, whether that wisdom is correct or not.

And, as far as I can tell, none of Brian’s proposed bets raises the issue of the extent of global warming, which is the main issue that separates alarmists like Al Gore from the scientific consensus. None of them relate to the extent of resulting harms. None of them relate to the existence or extent of benefits from global warming. None of them relate to the effectiveness of policies aimed at limiting global warming. In short, none of the proposed bets really addresses the main points being raised by those who admit that some global warming is taking place but doubt the desirability of enacting major policy initiatives to deal with it.

5 comments:

Brian said...

Thanks for looking at the bets - I've got two bets challenging the "warming is only natural" dodge, and maybe you missed them. The simplest one is that the warming rate will accelerate substantially compared to the overall rate in the last century:

http://backseatdriving.blogspot.com/2006/09/not-betting-with-glen-raphael.html

The 19-1 odds you offer are interesting. I will also take the cool side of the bet if given 20-1 odds - natural variability makes that a completely reasonable bet for someone, like me and Al Gore, who accepts the scientific consensus.

As for the purpose of the bets, they have multiple purposes. One thing they've done successfully is constrain wild claims. Maybe you've said elsewhere that that you expect long term warming to continue 19 times out of 20, but the betting scenario is at least useful in getting that fact emphasized. It doesn't sound to me like you really do have much room for pure denial, at least when it comes to betting. I would argue for equal non-reliance on denialism in that case when determining the best policy on climate change, then.

Glen Whitman said...

I've got two bets challenging the "warming is only natural" dodge, and maybe you missed them. The simplest one is that the warming rate will accelerate substantially compared to the overall rate in the last century

For this to address the anthrophogenic claim, you have to think non-anthropogenic warming would not accelerate in the next century. That's not necessarily so. The main alternative hypothesis is increased solar activity, and no one knows how long that will last. It could intensify, stay the same, or diminish.

...a completely reasonable bet for someone, like me and Al Gore, who accepts the scientific consensus.

Al Gore's position is at odds with the scientific consensus. He continually claims there will be dramatic increases in both temperature and sea level, whereas the consensus is much smaller increases. See the recent IPCCC report, for instance, which predicts increases in sea level in the range of 1 to 2 feet.

Brian said...

What's the basis for believing the allegedly "natural" warming will accelerate, other than a pathetic attempt to deny the implications of the evidence? After 150 years of warming, how could anyone credibly expect the warming rate from a natural cycle to increase over 50%, which is what I'm willing to bet? Even the denialists who dispute whether now is the warmest time in centuries will acknowledge it's one of the warmest times, so there's no reason to expect the acceleration.

And by the way, you misunderstand the time scales that the IPCC and Gore are talking about for sea level rise. I agree that Gore should have been more specific, but it's not a direct conflict. I'm not aware of Gore talking about warming exceeding the standard 1.5-6.0C that's part of the consensus.

Glen Whitman said...

And by the way, you misunderstand the time scales that the IPCC and Gore are talking about for sea level rise. I agree that Gore should have been more specific, but it's not a direct conflict. I'm not aware of Gore talking about warming exceeding the standard 1.5-6.0C that's part of the consensus.

I'm not sure how much temperature change Gore has predicted, but he's definitely been talking about sea level rises of 20 feet, whereas the IPCC's predicted range has an upper limit of about 3 feet. To get the 20 foot increase, you need to assume the melting of all the glaciers in Greenland -- and that is not the consensus. At a minimum, Gore has repeatedly emphasized the upper end of possible outcomes, while failing to mention that those are not the most likely outcomes.

What's the basis for believing the allegedly "natural" warming will accelerate, other than a pathetic attempt to deny the implications of the evidence?

I think you're confusing two different hypotheses, natural cycles versus solar activity. If I understand correctly, solar activity does not seem to follow any regular cycle -- it just happens to be at a relatively high level right now.

Just to be clear, I'm not saying the anthropogenic theory is wrong and the solar activity theory is right. I think the anthropogenic theory is probably right; I think it's also possible both are right.

After 150 years of warming, how could anyone credibly expect the warming rate from a natural cycle to increase over 50%, which is what I'm willing to bet?

Hmmm. You think the rate of warming will actually go up by that much? In that case, we might be able to find a betting proposition after all -- although the outcome still won't really get at the causes of the warming, since warming is consistent with more than one theory. I might be willing to bet you, however, because neither theory (in its most likely version) predicts such large increases.

But we'll have to be more precise. Rather than looking at acceleration, which requires establishing a baseline level of warming, let's talk about how much warming will occur in degrees Centigrade. How much warming is predicted by year X (2020, 2050, whatever) if you're correct that the rate of warming will go up by more than 50%? If your prediction is greater than that resulting from a constant increase of 2 degrees Centigrade per decade (based on the 1.8 degrees per decade mentioned here), then I'll seriously consider taking your bet.

For instance, say we pick the year 2027 (two decades from now). If you think temperature will rise by more than 4 degrees Centigrade, I'll bet that it will increase by less than that.

(To be clear: nothing I say here should be taken as agreeing to a bet until the exact terms and means of determining the winner have been finalized.)

Brian said...

Glen, you're off by a factor of 10, and I'm predicting the temperature will go up 50% faster than when it was in the peak rebound from 1850 temperatures, according to the natural warming theory. From 1900 to 2000, temps warmed 0.6C/decade.

Since no one disputes that we're now either at or near peak temps compared to previous warm periods in the last 1300 years, the natural warming theory gives no reason to assume accelerating warming.

(sorry about the late comment, didn't see yours for quite a while)