Perverse Incentives

Posted on


I stopped checking the traffic stats on my independent blog a couple of years ago, and the Prospect looks at Tapped posts more in terms of generating “impact” that can be pitched as important to donors than to traffic as such. Now that I’m working for a genuine for-profit business corporation, however, I’m more aware of traffic spikes and so forth. Yesterday, for example, my post about David Petraeus’ dull dissertation got an Instapundit link. It also prompted James Fallows to do a post defending Petraeus from charges of unusual banality. And that post got an Instalink as well.

Advantage: Yglesias, valuable and productive employee. Except, of course, the incentives here seem terrible since the premise of all this traffic is that I was being dumb.

Allow me, however, to engage in some post hoc defense of my dumbness. The point was that I had my hands on a copy of Petraeus’ dissertation. It seemed like a document worth checking out. Maybe it would say something staggeringly stupid, and I could write “aha! this is dumb! we shouldn’t listen to this guy!” Alternatively, like the COIN Field Manual it might say smart things that, being smart, could be used jujitsu-style as arguments against the surge. In truth, though, the dissertation just turned out to be really, really boring. Given that all that happened, it seems like I might as well report my findings to the world: the dude’s dissertation doesn’t say anything interesting. I know that traditional journalism doesn’t work this way, but maybe it should. We know that publication bias (basically, journals only publishing interesting results, rather than “failed” experiments) is a real problem in academic research and it probably is in journalism as well.