Megan McArdle

« Curious questions | Main | Development economics »

Decision science

September 8, 2007

Back by popular demand

In the comments to the post on assuming that your opponents don't really believe what they're saying, a commenter requests a post I did a while back on egotistical bias and the fundamental attribution error. Ask, and ye shall receive:

As usual, blogging about abortion is producing a lot of heated and very emotional debate. There are a few categories of this: some very good posts about philosophy, which make me sad, because they are so interesting, and yet so pointless, because the main figures in the debate seem to me so clearly to be looking not for a well-thought out position, but for a philosophical premise which will tickle the crude moral intuitions of a public which does not and will not read moral philosophy. Nor can I say that I've found a well-reasoned answer to the abortion question; my only defense is that I'm well aware of it.

Then there are the personal anecdotes. Pro-choicers waving around the abortions they've had to say "See! I needed that abortion!" and (currently) pro-lifers waving around the abortions they've had to say "See! It was the worst thing I ever did!" Pro-choicers talking about their children to prove that they do too like kids, and pro-lifers showing cute pictures of Jimmy and asking "What if he'd been aborted?" Paul Campos has pointed out the futility of these demonstrations: everyone's seen the coathanger pictures and the fetuses with the faces and the fingers and toes, and everyone's still disagreeing. The premise that the other side would agree with you if you could just use provocative images to call forth some emotion on the topic seems particularly odd in the case of abortion; if there is one thing that the debate is not suffering from, it is a lack of emotion.

My favourite, though, are the posts where everyone speculates on the motives of the other side. You see, pro-lifers don't care about babies at all, because that would make their points something you might have to listen to and we can't have that, can we? So what they obviously really care about is screwing up women's lives so that they'll have to spend the rest of them barefoot and pregnant and in the kitchen making lemonade for Pa and his friends when they come in from a hard day of plowing and oppressing colored people. And pro-choicers don't actually care about women; all they're really interested is enforcing a radical feminist agenda on the rest of us so girls won't be able to wear dresses and lipstick any more and boys will have to have their genitalia surgically removed at puberty and replaced with a copy of The Feminine Mystique. Also, while we can't be totally sure, it's reasonable to assume that many of them enjoy baby-killing, and would sacrifice live infants if not restrained by the hard work of good, Christian folk.

I was recently at the University of Chicago graduate school of business interviewing professors about their work, and got a chance to spend half an hour talking to Nick Epley, who is a psychologist. He's done a lot of fascinating work, but one of the first things we talked about is a simple concept he didn't invent: egotistical bias. We tend to use ourselves as a model for everyone else's behaviour, even when doing so is wildly inappropriate. This is the source of fundamental attribution error, among other things.

We actually spoke specifically about abortion in reference to this, and he pointed out something I hadn't noticed up until then, but which now leaps out screaming at me from these posts and comments: people involved in a debate tend to assume that their opponents are disagreeing with them about the aspect of the debate they themselves care most about.

Pregnancy is a fundamentally unique situation in which there is no possibility of Coasean bargain; the rights of the mother to things like control over her own body conflict directly and irreconcileably with the rights of the fetus-as-potential-human. There is no good metaphor for it, no parallel where we can try to work out the ancillary issues. It's a toughy. The pro-choice end up thinking the rights of the mother are more important; the pro-life think that the rights of the baby that will be here in nine months are more important. Neither of these positions is obviously wrong, as far as I can see. As I say, it's a toughy.

But the really committed pro-choicers I've seen commenting act as if what pro-lifers were doing were considering the rights of the mother over her own body, and rejecting that principle, rather than curtailing it after weighing it against other principles; hence the accusations that pro-lifers hate women, or women with careers, or freedom, or whatever. And pro-lifers respond by calling pro-choicers baby killers. Nice.

Admittedly, there are some crazy people on both sides who really do cherish a vision in which [women leave the workforce and get back to having like, a zillion babies/men and women merge into a single androgenous species]. This is not, however, the majority. It is fair to say that there is quite a lot of reasoning from the result to the premise on both sides: people who think that it is of surpassing importance for women to take their place at the helm of half the world's institutions notice that this would be much more difficult in a world containing both frequent sex with people you don't intend to spawn with, and restricted access to abortion; they therefore reason that abortion must be moral. Conversely, the pro-life side notes that if you deny the obligation of a woman to carry her pregnancy to term, all sorts of other family obligations become harder to logically support, and therefore conclude that abortion must be wrong. Since the pro-life side generally does not care so much about total workplace parity, and the pro-choice side generally does not care so much about preserving traditional family structures, they conclude with some truth that there is a somewhat questionable moral discounting going on across the divide. But the sin is sufficiently equally distributed that this is not enough reason to dismiss the moral heft of the other side's arguments. Pregnancy and parenthood are HUGE DEALS. So is not existing. That's why so many people are so frightened of doing either.

This practice is, obviously, not limited to abortion. Hence the number of liberals who accuse libertarians of hating poor people, because they simply cannot believe that anyone could frame questions of property and economic liberty in any other way than "What do these do to the poor?"; and of course, the conservatives and libertarians who accuse liberals of hating freedom, for the same reason.

August 20, 2007

Love shack

Mark Kleiman and I have been discussing the theory that the disaster recovery bunker in WTC 7 was placed so as to make the best love nest for Rudy Giuliani and his then-extramarital-girlfriend (now wife). Mark has just responded to a post I made on my old blog:

She makes two points:

1. Traffic near City Hall was tied up on 9/11, vindicating the decision to place the command center should be within walking distance.

2. Giuliani and Nathan had other places to canoodle, including "her apartment and the City's many fine luxury hotels."

As to point #2, as long as Rudy and Judi were engaging in discreet adultery rather than flagrant adultery, having the Mayor visit her apartment, or taking a room in a luxury hotel, would have created certain ... security risks. Much better to have the city supply the love nest. As noted above, Barrett's flat assertion that the couple repeatedly visited the command center has gone uncontradicted, so far as I know. What do you think they were doing there: fire drills?

As to point #1, I don't pretend to be an expert on emergency management, but as far as I can tell no one but Giuliani supported the decision to place the command center at WTC7, or the "walking distance" criterion that justified that decision. That ought to raise red flags. This is not a case where "common sense" deserves serious consideration when it conflicts with expert opinion.

Yes, traffic near City Hall was tied up. But traffic near any disaster site was likely to be tied up. If the disaster wasn't near City Hall, the Mayor and his staff could have been easily transported to whatever command center had been chosen. (Apparently the favored location was actually in Brooklyn.) In a pinch, NYPD has helicopters.

On the point about their illicit liasons I am under the impression that, whatever the danger, they did in fact routinely visit hotels and her apartment, which is how the thing became common knowlege; certainly, showing up at the bunker together wasn't exactly a good way to keep it secret. I'm sure they did go to the bunker . . . and if I had such a place, I think I'd be tempted to at least neck on the desk near the secret button. But that doesn't mean it was therefore purpose-built for illicit assignations.

On the second point, it wasn't just traffic near the buildings that was snarled; that traffic may have been better than uptown. Remember, the island was sealed shut in the middle of rush hour. Police cars and ambulances can only clear a route to get somewhere if there is somewhere for the other cars to turn into. In Midtown and the financial district at rush hours, the streets are literally a solid block of cars. They would have to drive up onto the sidewalk to give a mayoral entourage a clear path--except that there are cars parked along the side of the street, so they can't do that.

In retrospect, of course, it was dumb. But hindsight bias distorts our perceptions. People who were against an idea remember having been against it much more strongly (and for better reasons) than they actually were at the time; people who were for it forget or play down their support; and people who didn't take part in the decision vastly overestimate the likelihood that they would have correctly predicted the actual outcome. Thus, it starts to make sense to look for conspiracies . . . a mole inside the organisation, or a venal politician who placed a bunker to use it as a love nest. It is not impossible that this was the case, but it doesn't strike me as particularly likely.