Sunday, May 23, 2010

In defense of argument from authority, and when it goes too far

I could probably just post a link to the wiki page on why the Argument from Authority can be a legitimately useful crutch for non-experts, but doing that wouldn't waste enough of your time. So let's look at a dustup based on what David Roberts had to say about my seconding Matt Yglesias' criticism of an incomprehensible post by Sean Casten opposing cap-and-dividend.  I guessed that Casten is confusing the effect of cap-and-dividend with the effect of a carbon tax, and that was too much for Roberts to take:

I’ve been trying to stay out of this, but this is just hilarious. Brian, Sean’s been working for years deploying capital to reduce CO2. He knows electricity markets better than 99% of people in the world, certainly better than you or Matt. Along with his father Tom, he’s written peer-reviewed papers on the subject.
So when he comes along and suggests that the Economics 101 cliches that substitute for energy policy knowledge among most bloggers might be off-base, their reaction is … to assume he’s making remedial errors about policy?

So let's unpack the appeals to authority.  Roberts is arguably an expert on policy, but while he's been cagey about it, I think he quietly disagrees with Casten but doesn't like the criticism that his fellow Grist blogger has received.  I'm just guessing that from the several comments he made that defended Casten as an authority but not Casten's particular arguments.

So leaving aside Roberts' expertise, we've got Casten.  Had Casten written only, "I don't have time to spell out the problems with cap-and-dividend, but I want people to realize that my renewable energy business that I've put years into will not benefit from cap-and-dividend," then I would've put that into my "this gets some weight" category and moved on.  Instead, however, Casten makes arguments that he apparently considers easily understood, but make no sense:
a tax on your competitor does not make you wealthier. This is so obvious it shouldn’t need repeating.
and in a comment:
The theory that all power prices will rise giving the CO2-free source a higher margin is predicated on an absolutely perfect transfer of cost through the system
At the margin, a tax on your competitor and not on you will help you out.  Yes, sometimes it won't, but if Casten's going to argue that it never helps in the renewables biz, then he needs to supply a missing argument.  And the theory that power prices will rise doesn't depend on 100% perfect pass-through, it just depends on more than 0% pass-through.


I feel competent to judge these arguments, especially the second one, and by making them Casten has decreased my willingness to defer to his authority.  


That's the way I'd handle it - climate denialists seem to think "appeal from authority" is a sufficient response to the mountains of science academies screaming bloody murder about AGW, but I think there's more work to be done than simply that.  I think they have to rely on conspiracy arguments for why almost all experts could be so wrong, and that's a pretty slender reed.

1 comment:

  1. Anonymous2:34 AM

    Skeptico makes a similar argument (http://skeptico.blogs.com/skeptico/2009/02/global-warming-denial.html ):

    "It isn’t necessarily fallacious to consider that thousands of climate scientists writing in peer reviewed journals might know more than you do about such a complex subject."

    quoting Nizkor:
    "…a person who is a legitimate expert is more likely to be right than wrong when making considered claims within her area of expertise. In a sense, the claim is being accepted because it is reasonable to believe that the expert has tested the claim and found it to be reliable. So, if the expert has found it to be reliable, then it is reasonable to accept it as being true."

    "What we have here is trust in the scientific method. And we trust it because we have reason to believe it works – just look around you."

    Bart

    ReplyDelete

Note: Only a member of this blog may post a comment.