Saturday, January 1, 2011

Revealed Preferences

I was looking at the "Spousonomics" blog tonight, which Steve Horwitz turned me on to a short while ago and I came across a great little explanation of the idea of stated vs. revealed preferences. The issue centers around the all-too-common problem most couples (and friends/roommates) face when discussing what to have for dinner:
"Here’s how the conversation might go in my house:

Him: “What do you want for dinner tonight?” Me: “Whatever you want is fine.” Him: “Whatever I want?” (tone: skeptical) Me: “Yeah, sure, I’m easy.” Him: “You’re easy? (tone: skeptical-plus) Ok then, how about tacos?” Me: “Tacos…not so sure I’m in the mood for tacos.”

Economists would describe this kind of jiu-jitsu as a problem of revealed vs. stated preference. My stated preference is that I don’t care what we eat. But my revealed preference–which I reveal when push comes to shove–is that I do care, and quite a bit. The problem is, the more my revealed and stated preferences contradict each other, the less my husband believes a word I say, like, “I’ll be home by 6, I swear,” or “No thanks, I already ate.” Fortunately, he’s a fast learner.  He never waits up, and always makes enough for two."
I suspect that most of us have experienced this time & time again, but I also suspect that few people have ever taken the time to stop and consider why this feature of humanity is so incredibly devastating to the idea that economies can be planned.

When governments and others try to "design" economies, they are stuck using information provided through various statistical information gathering methods. This tends to mean asking people supposedly relevant questions what they think or what they want. A classic example of this is the silly Census Bureau bus advertisement that made appearances around Los Angeles (and presumably other cities) last year:

If you don't understand how amazingly stupid this statement is, just stop and consider this... The owner of your local bakery has no ability to perform a census or learn how many people live in his area. So how does he know how much bread to bake?

And moreover, how does he know how much of his time he should devote to baking bread versus cakes, versus cupcakes, versus croissants, etc.?

The answer is of course: Prices, profits & losses.

This is the essence of the socialist calculation problem, however. The thing is, most of our actual preferences are only knowable at the very moment which they are "revealed' through direct action... And this means that for all the collected census data, all the public opinion polls, all the questionaires that central planners send out in their attempts to acquire information about what people need & want is almost entirely useless!

People invariably say they will act a certain way, but in reality they act another. Like how people always claim they would love to have a new light-rail or other forms of public transit but once billions of dollars are spent on a brand new monorail or whatever, they don't bother to use it. Thus it is that big, top-down, centrally planned "solutions" to problems perceived by politicians and other top men wind up failing spectacularly - while typically costing tax-payers gobs of money.

Revealed preferences are real. Stated preferences are... typically... not. And that makes all the difference.

So the next time some smart chap tries to explain that the only way to provide some service (like education) or solve some problem (like public transportation) is to use some centrally planned bureaucracy, just remember that as little as you know about your own true preferences for the food you'll be eating in a few hours - central planners know even less about people's true preferences for all of the more complicated things that they are busy trying to manage.

Their knowledge is limited... but they frequently act on the premise that it is not.

No comments: