Sexism can mean a whole bunch of different things. It’s not a simple binary predicate: this is sexist, that isn’t. In general, I mean a cluster of attitudes and actions that harm people based on their sex. Usually, its women being harmed, but definitely not always.
Affirmative action is, of course, an interesting case. On its face, it involves advantaging one group, which naturally comes at the expense of all other groups. So, of course it’s sexism in one sense of the word. So why does anyone think it’s fair? Because there are believed to be cognitive biases in play that prevent people from (for instance) selecting an equally qualified woman for a job (one day, I would like to write up a post on the evidence for this). The theory is that an explicit adjustment for these biases will result in treatment more like what there would have been if employers were unbiased. If this theory is correct, then in cases where we believe that there is such discrimination, maintaining the status quo would be sexism. Naturally, not all cases of affirmative action qualify for this.
As the discussion on The Bedrock Of Fairness shows, fairness can have many meanings. They frequently correspond almost exactly to meta-ethical stances (consequentialist, deontological, virtue ethics). I’m a consequentialist with regards to fairness (since I view it as merely a part of the whole system of ethics). And affirmative action is only justifiable under a consequentialist (or perhaps virtue ethics) framework of fairness—and then only sometimes. I guess that is, as you say, one particular interpretation of fairness, but it’s one that I would imagine is relatively common here, since consequentialist ethics are relatively popular on Less Wrong.
The theory is that an explicit adjustment for these biases will result in treatment more like what there would have been if employers were unbiased.
This is a deontological stance, namely immoral act X was performed so we must bring the world as close as possible to the state it would have been in had X not happened.
I’m a consequentialist with regards to fairness
I have no idea what this means. That is, I have no idea how to incorporate ‘fairness’ into a utility function that won’t produce absurd things (like saying life extension research is immoral because it’s not fair to those who will die before it gets implemented).
The theory is that an explicit adjustment for these biases will result in treatment more like what there would have been if employers were unbiased.
This is a deontological stance, namely immoral act X was performed so we must bring the world as close as possible to the state it would have been in had X not happened.
I don’t think it’s necessarily the case that an immoral act was performed in these cases; often people just make mistakes. And trying to stop an ongoing harm is entirely compatible with consequentialism.
I’m a consequentialist with regards to fairness
I have no idea what this means. That is, I have no idea how to incorporate ‘fairness’ into a utility function that won’t produce absurd things (like saying life extension research is immoral because it’s not fair to those who will dies before it gets implemented).
First, unfair situations make people unhappy in and of themselves. That is, in some sense, absolutely absurd, but no more so than boredom is absurd. Nonetheless, it is the way humans seem to be (research on apes also shows this effect as well). Gwern’s post on the psychology of power discusses some of the less obvious effects of this on e.g. cortisol.
Second, when talking about money, utilities are non-linear in dollars. If A has $1 million, and B has $100, and utility is the square root of money, then, ceteris parabus, redistributing money from A to B would be the utilitarian thing to do. Of course, this ignores the incentive and precedent effects of this (why should B bother to work if they can just get A’s money?), as well as A’s unhappiness at losing the money, so of course in the real world the computation is considerably more complex.
Third, if everyone benefits from having the better person doing any given job, then correcting for biases that prevent this will make society better off.
[Edit] Fourth, when a group of people is treated as abnormal or subordinate, their desires are not given full weight (and thus, they are less likely to be happy). An example of this in the US is that only one non-Christian group has ever won a Free Exercise Clause case.
Sexism can mean a whole bunch of different things. It’s not a simple binary predicate: this is sexist, that isn’t. In general, I mean a cluster of attitudes and actions that harm people based on their sex. Usually, its women being harmed, but definitely not always.
Affirmative action is, of course, an interesting case. On its face, it involves advantaging one group, which naturally comes at the expense of all other groups. So, of course it’s sexism in one sense of the word. So why does anyone think it’s fair? Because there are believed to be cognitive biases in play that prevent people from (for instance) selecting an equally qualified woman for a job (one day, I would like to write up a post on the evidence for this). The theory is that an explicit adjustment for these biases will result in treatment more like what there would have been if employers were unbiased. If this theory is correct, then in cases where we believe that there is such discrimination, maintaining the status quo would be sexism. Naturally, not all cases of affirmative action qualify for this.
As the discussion on The Bedrock Of Fairness shows, fairness can have many meanings. They frequently correspond almost exactly to meta-ethical stances (consequentialist, deontological, virtue ethics). I’m a consequentialist with regards to fairness (since I view it as merely a part of the whole system of ethics). And affirmative action is only justifiable under a consequentialist (or perhaps virtue ethics) framework of fairness—and then only sometimes. I guess that is, as you say, one particular interpretation of fairness, but it’s one that I would imagine is relatively common here, since consequentialist ethics are relatively popular on Less Wrong.
This is a deontological stance, namely immoral act X was performed so we must bring the world as close as possible to the state it would have been in had X not happened.
I have no idea what this means. That is, I have no idea how to incorporate ‘fairness’ into a utility function that won’t produce absurd things (like saying life extension research is immoral because it’s not fair to those who will die before it gets implemented).
I don’t think it’s necessarily the case that an immoral act was performed in these cases; often people just make mistakes. And trying to stop an ongoing harm is entirely compatible with consequentialism.
First, unfair situations make people unhappy in and of themselves. That is, in some sense, absolutely absurd, but no more so than boredom is absurd. Nonetheless, it is the way humans seem to be (research on apes also shows this effect as well). Gwern’s post on the psychology of power discusses some of the less obvious effects of this on e.g. cortisol.
Second, when talking about money, utilities are non-linear in dollars. If A has $1 million, and B has $100, and utility is the square root of money, then, ceteris parabus, redistributing money from A to B would be the utilitarian thing to do. Of course, this ignores the incentive and precedent effects of this (why should B bother to work if they can just get A’s money?), as well as A’s unhappiness at losing the money, so of course in the real world the computation is considerably more complex.
Third, if everyone benefits from having the better person doing any given job, then correcting for biases that prevent this will make society better off.
[Edit] Fourth, when a group of people is treated as abnormal or subordinate, their desires are not given full weight (and thus, they are less likely to be happy). An example of this in the US is that only one non-Christian group has ever won a Free Exercise Clause case.