Tuesday, October 17, 2006

The Power of Incentives -- Political Polemic Edition

The standard Republican response to various scandals the past few weeks has been to "question the timing" of the revelations. E.g.,

Conservative religious leaders described themselves as shocked yesterday by a new book's charge that Bush administration staffers privately dismissed evangelical Christian political activists as "nuts" and "goofy." James Dobson, Charles W. Colson and other stalwarts of the conservative Christian movement defended the Bush administration and questioned the timing of the book's publication, a month before the midterm elections.
I don't really understand this defense. The logic for this argument seems to be that the revelation of any new information near an election is politically motivated and therefore not credible. I think this is wrong. The timing of such revelations are far more likely to be economically motivated than politically motivated. If you were a supplier of new politically relevant information (particularly a book publisher) wouldn't you wait and enter the market when demand for your product was high?

Regardless of whether or not revelations are politically or economically motivated, the fact that information suppliers respond to incentives does not tell us anything about the quality of the information. Certainly, people face political and economic incentives to lie near elections; however, suppliers of good information also face strong incentives near elections. So, while the quantity of low quality information increases near elections, the quality of high quality information also increases. The fact that the this information was revealed near an election tells us very little about its expected quality.

People may believe that the information is less credible because, since it comes out at a moment where the information can be most politically damaging, that implies that the author is politically motivated rather than an "objective observer." That, of course, assumes away profit considerations. But that said, I actually think this is a reasonable hypothesis TO A DEGREE. The timing may be a signal as to the alignment of the author, and the skeptics argue that the alignment of the author really affects the credibility fo the information.

Your point that both liars and honest people face strong incentives in an election season is a good one, though. It's not obvious if those incentives are distinct.
Think about in terms of relative prices. Honest people have strong incentives to reveal the truth always. (That's my definition of honest.) Liars only have strong incentives when their lies can make a difference. (Again, I guess that is my definition of a liar in this context.)

So, knowing this, it's not unreasonable to infer that damaging information that comes out in January of an odd-numbered year is more likely to be from an honest source than damaging information that comes out in October of an even-numbered year.

You're right that the profit motive is stronger in even-numbered years, but that's not going to make a big difference in the relative incentives, as both the honest and the dishonest have the economic motive-- it mainly means that we are more likely to see damaging info in October of an even-numbered year.
I disagree. First, I don't think the change in incentives facing honest and dishonest people are symmetric. For the dishonest, there is some chance of being discovered and disgraced (or worse) (like the guy who gave 60 minutes the fake Bush National Guard records -- of course the swift boat liars were thoroughly discredited and nothing happened to them). Truth-tellers just get a bump up in their own glory or revenge or whatever. On net, I think this creates a bigger positive change in incentives for the truth-tellers than for non-truth tellers.

Second, and more importantly, the informants themselves are only one input into the information production process. The other inputs have very strong and clear incentives to find, gather, and release information when it will sell the most copies (of books or newspapers). So even if truth-tellers tell the truth uniformly across time, we should expect to see an increase in the supply of their truth near elections simply because this is when they are asked to tell it. This increase in the supply of truth certainly could offset the increase in dishonesty.

There is a tendency to, as Dave argues, treat all information released near elections with a skepticism that I am not sure is justifed. Certainly, the stuff from the parties or the campaigns is full of spin and lies, but that's not really what I am talking about. I am talking about stuff contained in new books or new reporting of "facts". Ultimately, I think that these sources need to be critiqued on their merits. To defend yourself (or your party) by questioning the timing of the release is not, in my mind, a defense. It refutes nothing and is an attempt simply to change the topic. This debate judge gives no points for such efforts and thinks people should stick to the topic.
Post a Comment

Subscribe to Post Comments [Atom]

<< Home

This page is powered by Blogger. Isn't yours?

Subscribe to Posts [Atom]