|
| The destructive beauty contest |
|
navigational aids: News ticker:
|
11 January 05. I don't want to imply that it's just about my kitchen floor. [The sudden start should indicate to you that this one won't make sense unless you've read the last item.] There are much more significant things at stake here. For example, black folk lower housing values. Why? Because, academic studies show, people really are fucking racist. Emerson et al [1] called 1,663 white folk, and asked them about their choices of home. Their focus was race, but they tried to bury race under a long list of other issues: school, race, recent changes in property values, crime, the relative value of your house vs the neighborhood average. It hurts how unsurprising the results are: controlling for those other factors, a primarily Asian neighborhood had no significant effect on choice (a one SD downward shift), a primarily Hispanic neighborhood had no significant effect (half an SD upward), and a primarily black neighborhood had a way significant negative effect (five SDs down). People who currently live in a hispanic neighborhood are more likely to be OK with living with hispanics, and similarly with blacks. I could find you more citations about how people are fucking racist, but this is one of those deals where academics just confirm the obvious. [In non-confirmatory news, I'm told that the white flight thing is a bit mis-stated. You don't have to believe me on this one, `cause I don't have a citation and am not in the mood to look it up, but whites are not more likely to move out of a neighborhood that's become racially mixed. However, new residents are more likely to be minority, meaning that in a few decades, the natural turnover makes it look like white flight.] Back to the beauty contest: say you're not a racist. You were bit by a radioactive chameleon and are no longer capable of perceiving people's race in any way. But you're still familiar with Emerson et al, and you know that 51% of the country voted for Dubya. As you go house shopping, you do the calculations on the value of a proposed house in a black neighborhood, and you have no choice but to take into account the fact that you're likely to get a lower value for the house when you sell. As long as there are some racists in the market, everyone will price in a racist manner. Now say that radioactive chameleons secretly sweep the nation, and one day everybody wakes up color-blind. You don't know this, so you still believe that others are racist, and so you will still value the house less. Maybe I overstated the level of racism in Emerson et al's survey: maybe the respondents are not fucking racist; so long as they believe that others are, they may still value a house in a black neighborhood less. The same is true everywhere that people can agree on dumb assumptions. If we read in books that there are people who think that women are less productive because they're always birthing babies, then it makes sense to offer a lower salary to women. If you know others are offering a lower salary to women (even if you've entirely forgotten the reason why), it makes no sense for you to offer an above-market salary. Value-by-fiat, or the destruction of value by fiat, is a coordinated equilibrium: when everybody sees the female signal, everybody knows to offer a lower wage. This is indeed a proper Nash equilibrium, where nobody has an incentive to defect. I'm frankly not sure about the policy implications for housing. But it makes for a tough row to hoe for a black person: if you buy the same house as a white guy, it'll have less value because you're you. What is likely to be the number one largest and most significant investment you ever make just won't get as high a return. For employment, this is a solid justification for overhiring those who have traditionally been discriminated against---affirmative action. As long as there are racists or sexists anywhere, minorities and women will be paid less, and the market will not correct this. Shifting people out of a bad equilibrium is one of the roles of government; in this case it does so by breaking the significance of the signal used for coordination. Under a proper AA regime, people won't know if others are offering a lower wage when they get the `girl' signal, so it is not necessarily optimal for them to also offer a low wage. In the mean time, there are still racists and sexists out there who will continue to offer a lower wage just on principle, meaning that even if the affirmative action regime is correctly implemented, there will still be pressure to shift back to the discriminatory coordinated equilibrium.
@journal{citation1,
[link] [A comment] Replies: A comment
|
