I think this is probably a phenomenon that happens in all disciplines, but I know economics so that’s where I see it. Anyway, while helping some of the Texas Tech grad students study for their qualifiers, I noticed that it’s hard to answer economics questions, when you don’t know what assumptions the professor has in mind.
For example, take Bryan Caplan’s recent lament:
A recurring final exam question for my undergraduate Public Choice class:
T, F, and Explain: Democracies would spend a higher share of their budgets on genuine public goods.”
Almost no one gets it right – and we cover the most relevant material just one week before the final exam!
Now at the time I read that, I thought, “I could give any answer you want, and justify it by making the appropriate assumptions.” For what it’s worth, here’s Bryan’s suggested answer. And I’m sure if someone sat in his class, it would be “obvious” what Bryan wanted. But really, that means the metric is, “What does Bryan Caplan think the answer to this question is?”
Now what I worried about, when I first read the question, was the motivation to vote in the first place. After all, if people are rational in the game theoretic sense, and they’re self-interested, then they shouldn’t bother studying political platforms and casting votes if they think there’s a better chance of dying in traffic than their vote influencing the outcome of the election.
However, you can’t just say, “Nobody would vote!” because then, there would be an incentive for one guy to show up and write himself in as president.
I hope you see now what I mean, when I say I could justify either True or False, and craft the assumptions accordingly. Or more specifically, the things I chose to worry about in my analysis, would decide whether the answer turned out to be True or False. (Incidentally, this is how Krugman manages to almost-always justify the political conclusion his fans want, while always pulling mainstream economics off the shelf to do so.)
While I’m venting, here’s an anecdote from grad school that contributing to my disillusionment:
At the weekly Austrian colloquium, a guy presented a paper showing that there were two strands in the voting literature. On the one hand, you had “median voter theory” kind of stuff. But on the other hand, you couldn’t explain why anybody voted in the first place.
So I’m sitting there thinking, “Wait a minute… I bet I could build models with purely instrumental voting, where I get large numbers of people to vote in a Nash equilibrium.” And sure enough, I did that. (The intuition: Build a 3-person voting game involving a mixed strategy. Then add voters to each side, to either abstain or vote with certainty, so that they cancel each other out. You can build up families of equilibria.)
The NYU expert on this stuff at first told me my idea was impossible. So I sat in his office and after about 15 minutes I got him to realize he was wrong, and that my result *was* possible. Then he changed his tune to, “Nobody would be interested in this result.”
I thought, “Well since you had earlier said something false, surely my result is *moderately* interesting?” But I just submitted my result on my own to a journal. A month later the report came back: “We are rejecting your submission because So and So in 2001 published a more general result including yours as a special case.”
And now you have more insight into my personality…