Cheap Talk

by Henry Farrell on March 19, 2009

Tyler Cowen has a “couple”: of “posts”: arguing that columnists (and others) should not be required to bet their views. I’ll confess to having mixed feelings on the underlying question – which is whether you should make cheap talk on politics etc more expensive. On the one hand, if you view public opinionators as political actors (as they surely are), then they ought to be held accountable for their screw-ups. A certain degree of epistemological caution is entirely warranted, for example, when advocating major wars and the like, and I’d dearly love to see people who screwed this up (for what were at the time entirely obvious reasons) held properly accountable. On the other hand, if you view public opinionators as providers of novel heuristics etc, there may still be some value to their arguments even if they are wrong much or most of the time, as long as they are wrong because of ways of viewing the world that are (a) under-represented in public debate, and (b) help one grasp features of situations that are not immediately obvious given other heuristics. In other words, even if these ways of viewing the world are usually wrong, they can potentially supplement and improve other ways of viewing the world that are more usually correct. How to calibrate the balance between these two desiderata is not immediately obvious to me. One can still say that in an ideal world, one would see people who are _both_ more likely to be right than not _and_ provide relatively novel and interesting ways of viewing situations enjoy a prominent place in public debate. The continued prominence of e.g. Thomas Friedman (who is both often wrong _and_ the Davos Consensus Made Flesh, and Dwelling Amongst Us) suggests that the real world incentives don’t run in this happy direction.



joel turnipseed 03.19.09 at 4:13 pm

Funny you should mention this, Henry. I have just started exchanging e-mails with some friends/acquaintances on how to implement bets regarding my posts at the Minneapolis Star-Tribune.

I think that if there were some kind of instant link that made it easy to include bets (and report on current markets in) Op-Eds, blog posts, and the like we could have it both ways (especially since the prescient would make [some] money and people like to make money as well as watch people who make money [sad, but true–esp. in the U.S.]).


joel turnipseed 03.19.09 at 4:21 pm

… it could also be a breakthrough in micro-payments: I may not be willing to pay you a dollar for a post: but I might bet a dollar one way or the other on the outcome. If structured with a small cut (both ways) for the market-making firms and opinion jockeys, you could see real money on a well-trafficked post.


Lee 03.19.09 at 4:51 pm

“Tyler Cowen has a couple of posts arguing that columnists (and others) should be required to bet their views. ”

He argues against that view in the post you link to. (Did you intend to link to Bryan Caplan?)


Henry 03.19.09 at 5:19 pm

Urgh – I meant to say “posts on the argument that …” and somehow screwed up. Corrected.


PG 03.19.09 at 7:55 pm

Why not have one place for the arguments whose substance we should take as the expression of the columnist’s sincere beliefs, on which he should be willing to bet, and another place for playing out wacky novel arguments that might lead to new heuristics? Certainly the fact that so many columnists now also have blogs seems like we have a tidy binary set up, and the existence of JournoList and similar “off the record” debating ventures for prominent people to work out kinks in their ideas means that the columnists themselves probably would like the idea.

So if William Kristol is saying in the print version of the Weekly Standard that we ought to invade Iraq because of the danger of WMDs, he should be betting on our finding the WMDs; but if he says only in a blog that we should invade Iraq because it’ll be so cheap due to our being welcomed as liberators, then he oughtn’t get too much grief if it turns out otherwise.

I think it’s reasonably for a publication to pay opinionators only for the opinions they’re willing to back with their own money, but to provide online space, research and editorial staff for more uncertain claims. As with what gets knocked about on JournoList, the dubious online argument of today might become the fairly well-reasoned dead-tree printed column of tomorrow.


Zamfir 03.20.09 at 10:31 am

PG, I would say bets alone would suffice for that purpose, people could show the seriousness of a particular text by the size of the bet they are willing to take. There would still be room for offcial columns with interesting-but-I wouldn’t-bet-on-it arguments. In principle such a system could grow to a situation where columnists are expected, by the public, only to publish articles of high certainty, but I suspect that people would also be happy with uncertain-but-interesting views.

Still, I don’t like the idea at all. it would just encourage people to write columns like politicians, with weasel words and hidden terms. That’s unavoidable in real politics, but opinion pieces are mostly entertainment, and they should stay that way.


jackd 03.21.09 at 4:20 am

I’d prefer a system in which a pundit’s record could be evaluated and displayed along with any opinion he or she puts out in public. The devil is in the implementation, of course, as I’m sure there would be a noisy contingent supporting virtually any absurdity as being correct. Personally I’m convinced that for a glaring example Bill Kristol has been so egregiously wrong that no responsible organization would publish his idiocy, but there are people with actual money who disagree.

Comments on this entry are closed.