BeskrivelseMost people agree that discrimination based on “forbidden traits”such as gender, ethnicity, religion, and age is unacceptable. That is until they are informed that the discriminatory practice relies on statistically sound differences between groups and that the discriminatory practice furthers an otherwise legitimate goal. Then people’s assessments seem to change: Experimental research finds that people accept statistical discrimination to a higher degree than non-statistical discrimination. But whereas the rationale of statistical discrimination seems to persuade people to accept discriminatory practices, it is ambiguous which specific components of statistical discrimination drive the higher degree of acceptance, and whether such components are morally significant. There could be several reasons why people are more willing to accept and engage in statistical rather than non statistical discrimination. They might have normative principles and arguments justifying why statistical discrimination is more acceptable than non-statistical discrimination. Or on the contrary, the assessments might reflect morally insignificant factors such as socialized cues to perceive statistical and economic rationales as good and emotional or instinctive rationales as bad. Based on the current studies, we cannot conclude which reason is more plausible.
I take an experimental-philosophical approach to shed light on the role of 1) sound statistics and 2) a non-discriminatory goal in the moral assessment of discrimination. I analyze data from an original survey experiment testing the acceptability of four different rationales of discrimination and discuss the experimental insights’ relevance to existing normative theories of wrongful discrimination. The results have implications for our theories and policies on discrimination as they will help us identify 1) when statistical and non-statistical types of discrimination are morally problematic and 2) whether the wrong-making features used for our assessments differ between statistical and non-statistical discrimination. It is central to make sure that we assess statistically discriminatory actions and policies on the grounds that people find relevant to the specific type of discrimination. What makes non-statistical discrimination problematic might simply differ from what makes statistical discrimination problematic. It is equally central to make sure that introducing statistics will not automatically lead people to skip the process of evaluating whether a policy or action is morally acceptable or not. In the end, most policies and actions (should) derive from more than just numbers.
|27 okt. 2022
|DPSA (Annual Meeting of the Danish Political Science Association)