Do the Ends Justify the Means? : On the Justifiability of Statistical Discrimination

While statistical discrimination comes with similar objections as other types of discrimination it may also lead to a number of beneficial and important consequences, thereby leaving us with conflicting intuitions regarding the right course of action. This study investigates whether there is any superior account of statistical discrimination that can explain when and why such a practice can be defended; i.e., when it is justified. Employing four desiderata — i.e., requirements — for a successful account of statistical discrimination, this study sorts out a number of popular but inadequate accounts and is left with three main contenders: the demeaning account, the consequentialist accounts, and the ex-ante contractualist account. The study notes several challenges with each of these main contenders and tries its best to answer them given the tools available for each account. This study concludes that the ex-ante contractualist account is superior to the other two main contenders given these four desiderata. In contrast to consequentialist accounts, it can more fully account for the reasons we typically hold against discrimination in a plausible way, and in contrast to the demeaning account, it can avoid the indeterminacy that seems to follow from relying, mainly or solely, upon demeaning treatment. Furthermore, it is argued that ex-ante contractualism’s emphasis on the need to discount uncertain burdens leads to intuitively appealing distinctions and implications. This conclusion gives new fuel to the idea that demeaning treatment is an important wrong of discrimination and offers a refined idea of what it means to take the consequences of statistical discrimination seriously. ; Medan statistisk diskriminering kommer med liknande invändningar som andra typer av diskriminering så kan den också leda till ett antal fördelaktiga och betydande konsekvenser, och lämnar oss således med motstridiga intuitioner gällande rätt väg framåt. Den här studien undersöker om det finns någon överlägsen förklaringsmodell av ...