I can see how you were confused, but I have a habit of making statements like “we should insist people believe X” without meaning that I favor enforcing actual laws mandating belief in X (because, as you suggest, it’s icky and impractical). What I meant is: I encourage people to behave rationally by believing X, and would like them to further encourage others.
Also, I suppose that if done well, such an instruction to jurors would be good. It’s not “blacks are X% more likely to commit”, it’s “of blacks brought to jury trial, they’re X% likely to be guilty on average” (or “they’re X% more likely guilty than all people brought to jury trial”, if you prefer, but that’s less informative.)
What I meant is: I encourage people to behave rationally by believing X, and would like them to further encourage others.
Either you have a way to actually get most people to believe in X, or the jury system fails in that respect.
There’s also room for bias in the selection of the reference class used. (All the numbers that follow are random.) A long-haired blonde man is brought to trial. On the one hand, men are 10% more likely than average to commit violent crime. But on the other hand blonde men are 13% less likely than average. But then, this is a truck driver, and they’re 25% more likely than average. But then, he grew up in Boston, and only 37% of accused Bostonites are convicted. But then...
“of blacks brought to jury trial, they’re X% likely to be guilty on average”
To be precise, they’re X% more likely to be found guilty on average. Which is not the same, in a relevant way.
Suppose you think there’s unfair bias against long-haired people in the courts today. If you use today’s statistics for conviction % of long-hairs accused, then you may perpetuate that bias. But where else do you get your prior?
I definitely don’t think the method of picking a single reference class is a good one. Many overlapping big and small classes aren’t a problem if you reason correctly.
Re: what if conviction rates are drastically unfair? That’s a good point. It’s definitely possible to prove that some people who were convicted are likely innocent (but not all of them, just those with e.g. exonerating DNA evidence). I suppose I assumed that the false-guilty rate in the US is less than 10%, but knowing that the false guilty for some privileged class (let’s say attractive white folk) is significantly lower, would mean that “X% likely to be guilty when tried” could only come from “X% found guilty when tried” after adjusting as much as possible for the difference in false-guilties and false-innocents.
I agree with you that it’s difficult to reason correctly. I’ll still encourage people to do so.
I can see how you were confused, but I have a habit of making statements like “we should insist people believe X” without meaning that I favor enforcing actual laws mandating belief in X (because, as you suggest, it’s icky and impractical). What I meant is: I encourage people to behave rationally by believing X, and would like them to further encourage others.
Also, I suppose that if done well, such an instruction to jurors would be good. It’s not “blacks are X% more likely to commit”, it’s “of blacks brought to jury trial, they’re X% likely to be guilty on average” (or “they’re X% more likely guilty than all people brought to jury trial”, if you prefer, but that’s less informative.)
Either you have a way to actually get most people to believe in X, or the jury system fails in that respect.
There’s also room for bias in the selection of the reference class used. (All the numbers that follow are random.) A long-haired blonde man is brought to trial. On the one hand, men are 10% more likely than average to commit violent crime. But on the other hand blonde men are 13% less likely than average. But then, this is a truck driver, and they’re 25% more likely than average. But then, he grew up in Boston, and only 37% of accused Bostonites are convicted. But then...
To be precise, they’re X% more likely to be found guilty on average. Which is not the same, in a relevant way.
Suppose you think there’s unfair bias against long-haired people in the courts today. If you use today’s statistics for conviction % of long-hairs accused, then you may perpetuate that bias. But where else do you get your prior?
I definitely don’t think the method of picking a single reference class is a good one. Many overlapping big and small classes aren’t a problem if you reason correctly.
Re: what if conviction rates are drastically unfair? That’s a good point. It’s definitely possible to prove that some people who were convicted are likely innocent (but not all of them, just those with e.g. exonerating DNA evidence). I suppose I assumed that the false-guilty rate in the US is less than 10%, but knowing that the false guilty for some privileged class (let’s say attractive white folk) is significantly lower, would mean that “X% likely to be guilty when tried” could only come from “X% found guilty when tried” after adjusting as much as possible for the difference in false-guilties and false-innocents.
I agree with you that it’s difficult to reason correctly. I’ll still encourage people to do so.