On “Climategate”

Before I begin my monologue, I’d like to remind readers of my underlying position on the topic of global warming:

• I believe that climate change happens, and that humans may have a role.
• I believe that the AGW have overhyped their message in an effort to attract attention, thereby risking doing more harm than good by losing public trust if their worst predictions don’t occur as forecasted.
• I am not a climatologist, but my own experience in finding a signal in noisy data has had me wondering if the AGW crowd may be drawing too much meaning from “noise” rather than real “signal”.
• I’ve suspected that the climate change message may be magnified and/or distorted as it percolates up from scientists through activists and the media to public attention.
• I think that some of the changes called for by AGW are not necessarily bad  ones, on conservation/ecological/anti-pollution grounds, and I worry that if/when the current AGW hype ebbs, some good ideas may fall by the wayside, due to the public backlash.

With that background aside: I’m sure that you’re aware of events from about a week ago, when some enterprising characters cracked the computers of some climatologists, and acquired for distribution many emails and raw datasets presumably in an effort to shine additional light on, and potentially discredit, the AGW movement.

There are certainly emails that look bad.  Some folks are performing forensics on pro-AGW papers and finding errors/discrepancies in the data.  And at least one Senator is calling for an official investigation into alleged fraud committed by the Intergovernmental Panel on Climate Change.

I have to admit – it is incredibly tempting to sit here saying, “told you so!”  However, I find the “no it’s not” and “neener-neener” style of debate engaged in by some of the anti-AGW crowd to be childish…and there are lessons to be learned from this mess.  Some of these lessons should be of particular interest to actuaries and others who make a career of studying and drawing meaning from data.

First: Especially in this day and age, when drafting a document or composing an email, it is vitally important to ask yourself: “How would this look if it appeared on the front page of the New York Times?”  It’s rare that words die any more, and it’s surprising what can be dug up when you least expect it.  It would be a shame if a perfectly correct message were destroyed or discredited because of a flip comment inserted into an email, and taken the wrong way.

We’re all guilty of not thinking defensively in this way.  I know that work wouldn’t be as much fun if I couldn’t banter with a few folks in email.  I don’t think that written communications necessarily should be reduced to dry, lawyer-approved recitations of facts (or replaced with unrecorded verbal communications).  But engaging brain before clicking “send” is probably a good rule to follow.

Second: To do good analysis, one must be open to the possibility that you might be wrong…and that you may have been wrong for some time.   Admitting error can be embarrassing, but sometimes the greatest lessons can be learned from exploring why an experiment or a computation did not turn out as expected.

In the case of Climategate, I wonder how much of the AGW crowd’s sin was in the refusal to accept the possibility they were wrong.  While it may be extremely tempting to reject junk science and ridiculous claims emitted by critics who refuse to accept the possibility that they may be wrong, a climatologist cloistering him/herself in a cabal of scientists bound together by near religious certainty in their findings is engaging in faith, not science.   Criticism isn’t pretty, and the public does generally lack appreciation for the messiness of good analysis…but good science requires that you always be open to the possibility that you might be wrong, and that you honestly test for that possibility.

Third: Once of the biggest challenges in engaging in analysis can be in communicating the results.  When your audience is similarly geeky/nerdy, communication is relatively easy – there should be some common understanding and tolerance in the audience of noisy data, confidence intervals, and the like.  However, when the audience is less analytically-inclined, communicating results involves walking a fine line between accurately presenting the outcome and not clouding the message with too much background information (like the nature of “noise”).

So, when communicating the results, it is important that the perceived need to distill the message doesn’t also distort the message, or paint yourself into a corner such that all credibility is lost when reality doesn’t quite conform to your prediction. 

I can’t help but wonder if, when the dust settles on Climategate, we find that the AGW crowd, in their efforts to find a way to best communicate the seriousness of their fears, they settled on a message that was worse than the data supported, and from which they could not back away lest they be banished by others of their faith.