Wednesday, January 9, 2008

Economist's Notebook: Self-Fufilling Expectations

As I was standing in line at security at PDX on my way out to New Orleans, I happened to see Steve Novick, Democratic candidate for the US Senate seat currently held by Gordon Smith, in line ahead of me. I immediately noticed him because of the fact that he is very, very short. Now, we all know intuitively that height has absolutely nothing to do with the ability to be a good senator - or does it? Steve is obviously extremely intelligent, personable and hard working, so I would argue there is nothing innate about his height that lessens his ability to be an effective senator. But suppose I was worried about his ability to be elected not because of any prejudice I had about his height, but because I thought others might worry about the effect of his height. If everyone in Oregon were like me, personally not worried about his height but worried about others holding his height against him, notice what happens: my initial worry would be confirmed. No one would vote for him because of his height, not because they believed it made him less able but because they believed it would make him unelectable, and they would all have been proven right ex post. This is what in economics are known as self-fulfilling expectations: you expect his height to be a problem, so you act that way, and by such action you cause it to be true.

You can make a similar argument about his effectiveness once in the senate. A senator who is considering supporting him on some legislation he is proposing, might worry about his ability to drum up support among other senators, so decides not to - not because of his own ability, but because of how this senator believes others will act towards Novick. This worry causes the senator not to support Novick and thus causes his own worry to be shown to be correct.

So the punch line is that you can quite legitimately be worried about his height without believing that it has any effect on innate ability. But should you be? Notice that if everyone chose not to worry about what others might do, another equilibrium would result. An equilibrium where height played no role either in the election or in the senate. So what is a voter to do with such economic insight? Well, which equilibrium results really depends on how much you think other voters and senators will be influenced by height. Any one voter's marginal contribution to this is so small it is virtually meaningless, so an individual decision to disregard these concerns is not that meaningful in terms of outcomes, but may be meaningful in terms of one's own utility.

11 comments:

Unknown said...

This phenomenon happens often in politics, especially with African American candidates. Those polled will say that they support an African American. But when asked if their neighbors will do so, a disproportionate number will say that they wouldn't. What is clear is that usually they're revealing something about themselves.

I don't know if this will play out when it comes to height, but it's conceivable.

Portlandia said...

It's hard for me to believe that Novick's height will be any kind of a factor. If he were 5'6" (i.e. short "normal" height) it would probably be more of a factor that his more extreme shortness, arising from a physical disability.

Portlandia said...

argh - I meant to say

it would probably be more of a factor THAN [not that] his more extreme shortness

Pat Malach said...

Interesting perspective. But there's an obvious flip side to that coin.

Novick, by virtue of his physical "unusualities," instantly has the thing that every cookie-cutter 6’2" traditional politician with perfect teeth and hair pays millions for: voters asking themselves, "who is this guy?"

It certainly didn't hurt his ability to get this kind of invaluable news coverage.

(I also posted the gist of this comment at BlueOregon before that discussion was closed and directed here)

Robert G. Gourley said...

Let's face it, life's not fair - Merkley's too tall, and he lacks a hard left hook.

Chris Lowe said...

This of course extends to ideological positions and identities too, and even specific ideas. The resulting intellectual sclerosis why some sort of ranked preference voting would be better than what we've got.

But more directly relevant, this is the whole "electability" game, isn't it? We Dems got the unelectable John Kerry in 2004 because so many people convinced themselves that better candidates would not be electable and that Kerry would be.

Is this a variant of the Prisoners' Dilemma?

Patrick Emerson said...

Chris,

No this is not the same as a Prisoner's dilemma where we all act in our self-interest and end up at a sub-optimal group outcome. Here, simply because I believe height is a problem, I act as if it is, and by doing so I (and everyone else) make height a real rather than just perceived problem. In a game-theoretic sense, if I am a Novick supporter and in a 2 person game, and I know you will vote for Steve, then my best response is to vote for Steve, however if you vote for Jeff, and if by voting for Steve I open a door for a third best candidate, I want to vote for Steve as well. So it is really a coordination game.

In this sense, Steve is doing exactly the right thing in addressing the height issue head on. The more he can convince voters not to be so worried about height the more likely he is to avoid the self-fulfilling prophecy.

Chris Lowe said...

OK. I guess the similarity I see is that both cases work through a process of assuming the worst actions on the part of others we don't know. But the self-interest in the electoral cases is partly a self-interest as part of a collective interest so that differs from a pure self-interest game as well I suppose.

I haven't exactly thought this through, but it's also striking me as something like an perverse inversion of John Rawls' veil. That may not really work either.

Interesting problem, and I agree with your final comment about Steve's strategy, which reduces the information vacuum about what others might think. Thanks for putting this up.

Patrick Emerson said...

Not to get too technical and off topic, but in the prisoner's dilemma, in fact, I don't need to assume anything about the other player. If I act solely in my own self-interest without any regard to what I think the other person is likely to do, and the other person does similarly, we end up in a sub-optimal equilibrium.

Perhaps this would be a nice thing to write a post on...stay tuned.

Portlandia said...

"sub-optimal equilibrium" sounds like the world I live in.

Jeff Alworth said...

It's probably really important to separate these phenomena as they occur in the political and economic spheres. Voters have a far more diffuse sense of their own benefit when they vote than when they, say, buy a stock.

I would argue that self-fulfilling expectations are a more profound element of elections than they are in economics and, mechanistically, different.