One root pattern in the set of issues (race, gender, religion) is of between-group variance attracting more attention than within-group variance.
I suspect this pattern has deeper roots than a simple neglect of variance: At least some participants seem to fully accept that a model of suffering based only on group membership may involve too much noise to apply to individuals, but still feel very concerned about the predicted group differences, and don’t feel a pressing need to develop better models of individual suffering.
(BTW, this is the heart of my critique of Jonathon Haidt’s claim that left-leaning people think predominately along the care/harm axis.)
In any case, we probably care more about group differences for political reasons, and because our group definitions may correspond to “levers” that are easier to pull. Another theory is that group inequality really is a leading cause of suffering, because we’ve evolved to feel the stigma of belonging to less well-off group as much more painful than the raw difference in non-group-related suffering would predict. Or perhaps we fear that group differences can diverge much more rapidly, and so must be monitored very closely (there is some historical evidence for this).
Finally, we might feel that group membership (in something like race or gender) is not morally assignable to a single member, while the error terms (what I’d call the contributing factors to within-group variance) are morally assignable to individuals. But this is just a restatement of the idea that “fairness” concerns are involved.
This perspective also frames another perennial puzzle of these debates, which is the appearance of both of a critique of stereotypical thinking (which is typically a neglect of within-group variance), and an approach that focuses so heavily on group differences!
One root pattern in the set of issues (race, gender, religion) is of between-group variance attracting more attention than within-group variance.
I suspect this pattern has deeper roots than a simple neglect of variance: At least some participants seem to fully accept that a model of suffering based only on group membership may involve too much noise to apply to individuals, but still feel very concerned about the predicted group differences, and don’t feel a pressing need to develop better models of individual suffering.
(BTW, this is the heart of my critique of Jonathon Haidt’s claim that left-leaning people think predominately along the care/harm axis.)
In any case, we probably care more about group differences for political reasons, and because our group definitions may correspond to “levers” that are easier to pull. Another theory is that group inequality really is a leading cause of suffering, because we’ve evolved to feel the stigma of belonging to less well-off group as much more painful than the raw difference in non-group-related suffering would predict. Or perhaps we fear that group differences can diverge much more rapidly, and so must be monitored very closely (there is some historical evidence for this).
Finally, we might feel that group membership (in something like race or gender) is not morally assignable to a single member, while the error terms (what I’d call the contributing factors to within-group variance) are morally assignable to individuals. But this is just a restatement of the idea that “fairness” concerns are involved.
This perspective also frames another perennial puzzle of these debates, which is the appearance of both of a critique of stereotypical thinking (which is typically a neglect of within-group variance), and an approach that focuses so heavily on group differences!