April 17, 2006
A New Article
Posted to Author: Pielke Jr., R. | Science Policy: General
I have an invited article just out in the magazine Regulation, published by the Cato Institute. The article is titled "When Scientists Politicize Science" (here in PDF). The first part of the article retells the story of debate over the Skeptical Environmentalist, and my views of the role of science in that debate, which I first presented in a peer-reviewed paper in 2004 (here in PDF). The second part gets into the the broader context of science and politics, and in this essay I am more explicit that I have before about the notion of "honest brokers of policy alternatives". Here is a short excerpt:
Instead of the futile effort to keep science and politics separate, it may make more sense to ask scientists to engage more substantively in policy debate, not by taking sides but instead by serving as “honest brokers of policy options.” Such honest brokers might distinguish themselves from policy advocates (who work to reduce available options) by furnishing policymakers with a broad set of policy alternatives and their relative pluses and minuses. The policymakers would then decide what course of action to take.
I welcome comments and reactions. Thanks.Posted on April 17, 2006 06:32 AM
You wrote, "In thinking about how things might be different, it is absolutely critical to differentiate scientific results from their policy significance. To illustrate the distinction, consider the central conclusion
Ah, but the statements of "projection" are NOT statements of projection in the sense that 9 out of 10 readers of the IPCC TAR would assume.
MOST readers associate the word "projection" as being synonymous with "prediction." For example, "The 5-day projected path of hurricane Katrina is so-and-so."
The word "projection" as used in the IPCC TAR is synonymous with "scary scenarios that have no basis in reality."
I've already challenged William Connolley, Kevin Vranes, and Andrew Dessler to label these eight assertions at "true," "false," or "don't know," based on what's in the IPCC TAR. I invite you to do so. Invite your father to help you. Invite anyone.
Unless you or anyone else can come back with anything but eight "don't know" answers, the IPCC TAR projections are completely meaningless, as a matter of science.
1) The IPCC thinks that there is an approximately 50/50 chance that the warming will be less than 3.6 degrees Celsius.
2) The IPCC thinks that there is an approximately 50/50 chance that the warming will be less than 3.1 degrees Celsius.
3) The IPCC thinks that there is less than a 10 percent chance that the warming will be less than 1.4 degrees Celsius.
4) The IPCC thinks that there is less than a 10 percent chance that the warming will be more than 5.8 degrees Celsius.
5) The IPCC thinks that there is more than a 50/50 chance that the warming will be less than 1.4 degrees Celsius,
6) The IPCC thinks that there is more than a 50/50 chance that the warming will be more than 5.8 degrees Celsius.
7) The IPCC thinks that there is more than a 99 percent chance that the warming will be less than 1.4 degrees Celsius.
8) The IPCC thinks that there is more than a 99 percent chance that the warming will be more than 5.8 degrees Celsius.
Posted by: Mark Bahner at April 17, 2006 10:50 AM
Actually, the IPCC (at least WG I) does explicitly associate probabilistic degrees of confidence with words such as "likely" etc. Offhand I don't have the ranges and thresholds for which percentages go with which qualifiers, but my guess is that, just going by what is actually in the IPCC report, you would label #s 3 and 4 as "true" (and, by extension, #s 5, 6, 7, and 8 as "false").
You could be trying to make one of two points here (I'm not sure which, or neither): One is that you might object to assigning quantitative percentages without an associated empirical quantification of the impacts of all possible uncertainties on the final "answer" (e.g., a giant Monte Carlo exercise). However, assigning numerical probabilities is accepted practice in expert elicitations and other Bayesian-type analyses where such experimentally based quantification of uncertainty is not present.
Alternatively, you might think the IPCC's probabilities are wrong because they're based on emissions scenarios and model projections with which you disagree.
Posted by: Chris Weaver at April 17, 2006 12:31 PM
Though now I've sidetracked myself from commenting on Roger's article, the original task ...
Posted by: Chris Weaver at April 17, 2006 01:51 PM
The paper is very good and a logical extension of your past writings that I have read.
The only issue I would raise is that while do accept that scientists who receive money from industry may have been influenced by it, you do not acknowledge that government and non-for-profit funding is not free of influence on the all-too-human scientists who receive it.
Anyway isn't the whole long-term process of science an adversarial review aimed at weeding out error whether it is inadvertent or a conscious attempt at propaganda? After all the peer-reviewed publication of a paper does not establish its truth. It may only establish that is is on the ladder to truth.
No more than we can find Madison's angels to run our governments can we find angels to man our laboratories.
Maybe part of the problem we face with the uses of science in our society is that we have lost sight of the evolving nature of scientific truth. We are much more in the mode of publication on Monday, laws on Tuesday, and guillotine by Friday.
Posted by: D. F. Linton at April 17, 2006 02:16 PM
You write, "Actually, the IPCC (at least WG I) does explicitly associate probabilistic degrees of confidence with words such as 'likely' etc."
Yes, there are indeed qualifiers. See Footnote 7 on this page:
"...virtually certain (greater than 99% chance that a result is true); very likely (90-99% chance); likely (66-90% chance); medium likelihood (33-66% chance); unlikely (10-33% chance); very unlikely (1-10% chance); exceptionally unlikely (less than 1% chance)."
1) I can find ***no*** such qualifier for the "projected" temperature rise:
"The globally averaged surface temperature is projected to increase by 1.4 to 5.8°C (Figure 5d) over the period 1990 to 2100."
2) Even if the projected temperature rise did have such a qualifier, unless the qualifier for the projected temperatures was "virtually certain" or "very likely," it would potentially leave a wide margin outside of the projections.
For example, SUPPOSE the qualifier was "likely" (again, there is no qualifier that I know of). That would mean that there could be up to a 34% chance that the warming will be LESS than 1.4 degrees Celsius, or up to a 34% chance that the warming will be MORE than 5.8 degrees Celsius.
3) Even if there WAS a qualifier (which there is not, to my knowledge) and even if that qualifier WAS "very likely" or "virtually certain," it would tell the reader virtually nothing about the probabilities INSIDE the projected range.
For example, suppose the IPCC knew that there was a 80% chance that the warming will be between 1.4 and 2.4 degrees Celsius (with a 10% chance above and 10% chance below that range). Well, that would be "compatible with" a projection of:
"It is very likely that the warming will be between 1.4 and 5.8 degrees Celsius."
Or conversely, suppose the IPCC knew there was a 90% chance of the warming being between 4.8 and 5.8 degrees Celsius (also with a 10 percent chance of below and a 10 percent chance of above that range). That ALSO would be "compatible with" a projection of:
"It is very likely that the warming will be between 1.4 and 5.8 degrees Celsius."
You continued, "You could be trying to make one of two points here (I'm not sure which, or neither):..."
I'm trying to make the point that the IPCC TAR "projections" are basically worthless, as a matter of science. (Let alone being useful for policy guidance.)
You have answered my questions based on your ***recollection*** of the IPCC TAR. You didn't actually go to the IPCC TAR to find out whether your answers were supported by the actual wording in the IPCC TAR. I don't think any of the qualifiers you recollect ARE ever explicitly applied to the temperature "projections."
And I’m “virtually certain” there is no discussion of probabilities within the range...which is extremely important, given the very large range that is “projected.”
P.S. A reasonable scientific analysis of present trends in methane atmospheric concentrations (and methane emissions), and CO2 emissions and atmospheric concentrations, would lead one to the conclusion that the odds are about 50/50 that average lower tropospheric warming in the 21st century will be approximately 1.4 degrees Celsius…i.e., at the very BOTTOM of the IPCC TAR's "projected" temperature range:
Posted by: Mark Bahner at April 17, 2006 03:21 PM
Mark and Chris- I'm not one to upset a good conversation on the IPCC, but that's pretty far off topic here. I'd ask that you either exchange emails or follow this up on a more relevant thread, like:
Posted by: Roger Pielke Jr. at April 17, 2006 03:37 PM
Yes, it was based on my recollection - I thought I had already said that in my comment (down boy with the *** and the all caps, by the way). You may very well be correct that the given range is not directly associated with those qualifiers - I believe there are a few studies that came out post-TAR that tried to put explicit confidence bounds on the range, and that may have been because this was a perceived gap in the TAR.
But, digging a little deeper (dammit, man, you've forced me to actually look at the TAR, the horror), I find the following in the Summary for Policymakers:
p.9: "Projections using the SRES emissions scenarios in a range of climate models result in an increase in globally averaged surface temperature of 1.4 to 5.8°C over the period 1990 to 2100. This is about two to ten times larger than the central value of observed warming over the 20th century and the projected rate of warming is very likely to be without precedent during at least the last 10,000 years, based on
So, in the sentence after they mention the projected temperature range, they use the words "very likely" to qualify "without precedent during at least the last 10,000 years."
Also, p. 10: "Nearly all land areas will very likely warm more than these global averages, particularly those at northern high latitudes in winter."
Again, "very likely" is associated with temperatures explicitly and with the 1.4-5.8 degree range implicitly.
So, unless you want to argue that putting "very likely" right before or right after "1.4-5.8," as opposed to in close but not exact association with it, changes the document from something valueless to something of value, I think we're arguing about a different point.
You do make an excellent (and underappreciated) point, however, about the relative probability of values within the range. I agree that it would be better for assessment reports to provide actual PDFs - people who study characterizing and communicating scientific uncertainty definitely endorse this approach.
Posted by: Chris Weaver at April 17, 2006 04:06 PM
Oops, sorry Roger - I hit send before reading your calling out our "hijacking."
Posted by: Chris Weaver at April 17, 2006 04:08 PM
Chris, Mark- No worries, thanks for commenting ;-)
Posted by: Roger Pielke Jr. at April 17, 2006 05:08 PM
I have one more question, that I think does relate to your article, even though it's an extension of what Chris and I were discussing:
Suppose the IPCC TAR's central conclusion had been, "The number of Category 5 hurricanes landfalling in the United States between 2020 and 2050 will be between one and six."
Based on your knowledge of hurricanes, would you have mentioned that conclusion in your paper?
If not, why not?
Posted by: Mark Bahner at April 17, 2006 06:54 PM
Hi Mark- Thanks for your comments. I don't think I understand your question. My paper was not at all about summarizing the IPCC, but rather using the IPCC as an illustrative example for explaining how he science-policy connection might be different if we focused on the significance of science for action, rather than just stopping at the science itself. Hope his makes sense, Thanks.
Posted by: Roger Pielke Jr. at April 17, 2006 08:09 PM
"I don't think I understand your question. My paper was not at all about summarizing the IPCC, but rather using the IPCC as an illustrative example for explaining how he science-policy connection might be different if we focused on the significance of science for action, rather than just stopping at the science itself."
My question was based on the fact that the science you quoted isn't really science.
I assume that, if the IPCC TAR had a conclusion that, "The number of Category 5 hurricanes landfalling in the United States between 2020 and 2050 will be between one and six"...you wouldn't have used that as your illustration of an example of "science."
For one thing, you'd know that there is a huge difference between 1 and 6 Category 5 landfalling hurricanes in 30 years. So simply saying, "between 1 and 6" wouldn't be sufficient science.
But perhaps more important, you'd know that six landfalling Category 5 hurricanes in the U.S. in 30 years is simply not credible. One, definitely; two, possibly; three, pretty unlikely; four, amazing...and six in 30 years is simply not credible.
When you reference the IPCC TAR's "1.4 to 5.8 degrees Celsius," you're doing the same thing. The "science" you are using for your example simply isn't defensible as science.
Posted by: Mark Bahner at April 17, 2006 08:49 PM
Would that red herrings were on the endangered species list, with catch limits and all that. Lots of Roger's bandwidth could be saved that way.
Posted by: Dano at April 17, 2006 11:12 PM
I was just about to depart the climate science blogosphere when I saw this article on the Cato website. It gives me the opportunity to get on the soap box one last time even if sounds like a rehash. So here it is.
Your point about scientists proposing alternative policy options should be digested well. The science and the policy makers sorely needs this. I have been hearing about this abstract thing called a warming globe for a very long time. I have also been hearing about the carbon dioxide monster till my eardrums are about to go numb. Then I have heard the laments about not unshackling the Kyota superman who will solve everything. All of this has gone beyond paradigm to dogma and every new scientific paper somehow supports it even when it doesn't
Its amazing, absolutelly amazing that such an overwhelmingly complex field, which the most authoritatuve sources say is loaded with uncertainties, somehow has its central problem reduced to one simple cause and that cause has a self evident solution!
Are there not other causes of climate change? Black carbon, methane, ozone, other green house gases and other aerosols. I am sorry, they are significant, very significant. How about land use change? In addition to the forcings, might not feedback mechanisms be addressed? How about feedback loops?
Go after these forcings instead of CO2 and you have the advantage of going after a true pollutant instead of a gas essential to life that has been distorted into a widespread belief that it too is a pollutant. Going after these other forcings would have the double-barreled effect of getting rid of a pollutant, probably be cheaper, and have a more immediate effect.
The lagged effects of CO2, the impossibility of asking an underdeveloped world to give up aspiring to a modern standard of living, means to me that chasing Kyota-like ideals will not have an effect for decades, if ever. Dealing with other forcings are alternatives that must be discussed.
Throwing away that abstract globe and looking at the real world also needs attention. Jim Hansen goes to Greenland and is amazed that melting is going on that his models never told him about. Ha! His models did so tell him as did papers he co-authored. Darn right that the abstract models fixated on CO2 did not tell him. But NASA papers on carbon black and ozone say that over half of that Arctic warming may be due to these causes and also that the polar winds are contributing. Looks to me like the effects of CO2 up there may be fairly minor.
Of course everythng in print uses the goings on in the arctic as testament to the ravages of CO2. Doesn't that figure? Instead of looking at the Arctic as a problem that may have a solution, we use it instead as further proof that CO2 is heating up the world. Other forcings are not even mentioned. You can bet that those Inuits and polar bears will continue to be inconvienced. But that ancient Inuit culture looks pretty adaptable to me as they drive about in their snowmobiles. They are probably savvy enough to charge tolls when the Nortwest Passage opens up. In such an event, they might join midwestern farmers, with their increased crop yields, in praising "global warming".
That leads to another set of options... regions in the real world. Look at them. Some will affect climate change and some will be affected by it. Does anyone expect that midwest farmer to give up his best interests so that the Maldives won't flood? Maybe if we had a set of options we could address both. Regions are where the climate changes not in a model!
I am a pragmatist whose profession is advisory. To me the world of climate science reeks of political correctness not progress.
This is my last shot guys so fire away. I bet that the first nasty shot comes from Dano, high priest of the dogma!
Posted by: Paul Dougherty at April 19, 2006 02:40 PM
"To me the world of climate science reeks of political correctness not progress. "
I'm very dogmatic about pointing out how the current scientific understanding and community is mischaracterized, cherry-picked and mendacicized by rubes, ideologues and shills for their agenda, yes.
Thanks for allowing me to point that out.
Posted by: Dano at April 19, 2006 03:21 PM
I think the point is that the "current scientific understanding and community is" ...
Although, the point is certainly made, as long as one includes the sheep bleating and the donkeys braying from the bandwagon on the other side...
Posted by: Steve Hemphill at April 19, 2006 03:49 PM
"I think the point is that the "current scientific understanding and community is" ...overrated. "
Yup. Marginalizing science certainly is the new wish, isn't it.
Some hide it better than others.
Posted by: Dano at April 19, 2006 05:25 PM
Yep. Sometimes real science comes through:
Posted by: Steve Hemphill at April 19, 2006 06:31 PM
I found your article both reasonable and thought-provoking. As such, I forwarded on the article link to my MP and asked her to consider the argument and whether it would be worth raising a discussion in Parliament re science and public policy in the UK.
Posted by: John Lish at April 20, 2006 09:29 AM