Discounting Identity
I recently remarked: “Appealing to your identity is a reason to discount what you say, not a reason to pay extra attention.” Why do I believe this?
Well, suppose you want to make people agree with you. You’ve got two main routes.
Offer arguments in favor of your view to change what listeners sincerely think.
Make continued disagreement feel uncomfortable so listeners pay you lip service.
Only #1 is epistemically respectable. But it has a major downside: Good arguments are scarce. #2 is epistemically sleazy, but it has a major upside: It is open to everyone, regardless of the merits of their views.
So what rhetorical route should we expect speakers to take? Well, if you’ve got at least halfway-decent arguments, you’ll probably make them – hoping to change your listeners’ minds as well as their words. In contrast, if your arguments are flimsy, you’ll probably play on people’s emotions. You won’t really convince them, but at least they might act convinced.*
What does this have to do with identity? Simple: When someone you disagree with appeals to their identity, it is usually uncomfortable! If I want to discuss the prevalence of sexual abuse among Catholic clergy, and a listener responds, “As a Catholic…” they’re not-so-subtly telling me, “You better tread lightly, lest you insult my faith!” If I want to discuss the right to burn flags, and a listener responds, “As an American…” they’re not-so-subtly telling me, “You better tread lightly, lest you insult my country!” The same goes for all of the standard appeals to identity – religious, national, ethnic, gender, etc. When you invoke them, you are undermining the truth-seeking mission of the conversation – and the reasonable response is to discount what you say.
But can’t identity provide extra information? Once in a while, yes. But again, the truth-seeking route is normally to simply share your extra information without making identity an issue. In special cases, admittedly, you can’t certify your credibility without mentioning your identity. For example: “I’ve attended Catholic Church for 32 years, and never seen the slightest sign of clerical sexual abuse.” Even here, though, truth-seekers will acknowledge their identity casually to keep information flowing freely.
In any case, your identity provides far less information than you think. For two reasons:
Belonging to a group lets you learn lots of details about the group, but this depth comes at the expense of breadth. Being Danish teaches you a lot about what Danes are like. But the more energy you invest in your Danish identity, the less you learn about non-Danes.
The more you identify with a group, the worse your myside bias normally becomes. When you invest energy in your Danish identity, you grow more likely to overestimate the wonder of Danes and underestimate Danish shortcomings.
By analogy: Each of us knows more about our own lives than anyone else on Earth. But the more you dwell on your own life history, the less you’re likely to know about what life is like for anyone else. Furthermore, the fact that you know lots of details about your own life does not make you a reliable judge of your own merits and failings. Quite the opposite.
The knowledge that identity provides is cut from the same cloth as self-knowledge. Indeed, it’s probably worse, because the social sanctions for personal arrogance are far stronger than the social sanctions for group arrogance. Even your parents and closest friends will roll their eyes if you say, “I am the greatest.” But among people who share your identity, declaring “We are the greatest” might even make you friends.
The “discount appeals to identity” maxim can plainly be abused. An Indian nationalist could selectively use it against Pakistani nationalists – and Pakistani nationalists could return the favor. But the same goes for any de-biasing rule. You can’t make the whole world reasonable. But you can still be the change you want to see in the world.
* Needless to say, both claims are only tendencies; people with good arguments occasionally appeal to emotions, and people with bad arguments occasionally make them anyway.
The post appeared first on Econlib.
As a person with an identity, I find this article wrong.
There's a category you missed, where people say "As a ____, I have a strong incentive to oppose this policy; yet I believe this policy is so just, I support it anyway. So you can trust that this policy is just".
For example, you might see "As a police officer, I believe body cams are a good policy that will keep cops honest", and to me at least that has a lot more weight than proposals to impose responsibility on police from people who hate cops. Because from someone who hates cops, they might be proposing a policy that really is a good idea, or maybe they're proposing it just to make cops lives harder. From a police officer, I can be pretty confident they're not just proposing to make the cops' life harder.
And it's useful to have shorthand heuristics on who to trust, so I can read summaries of the arguments/research instead of doing it myself. Because on a lot of issues, which policy is better ultimately comes down to empirical research on which is better, not just moral arguments; and reading and judging the quality of empirical research takes a while. It's much better to have trusted people who can tell you what the research says than expecting every individual to judge the literature themselves. And someone's identity can tell you if they're worth trusting.
Other examples in this sort of category: "As someone currently applying to immigrate, I think immigration is too easy to get permission for", "As a soldier, I think the government funds the military too much", "As a white man, I think racism and sexism against minorities and women are major problems", etc. Of course just because you're arguing against your self-interest doesn't mean you're actually correct or honest. Especially if you're a "token" talking head who gets paid a lot by one political side or the other to espouse views against your self-interest, and get paid more than you'd lose from those policies being implemented. But I think it is a fair justification to put some extra weight on the likelihood someone's being honest.