This article assesses the development of EU anti-discrimination against a typology of models of equality. Under a first model, discrimination is prohibited, but there is no obligation to ensure a proportionate representation of the diverse social groups whose members are protected from discrimination. Nor is there an effort to monitor the situation of these groups with respect to the global allocation of social goods in order, if necessary, to take remedial action where imbalances are found to exist. Such imbalances as such are not seen as problematic, as long as each individual has not been discriminated against in identifiable ways, by particular agents. Under a second model, the prohibition of discrimination extends to the prohibition of disparate impact discrimination: any measure which disproportionately and negatively impacts upon certain groups which are already underrepresented (or which already receive a less-than-proportionate share of the social good to be allocated), should be revised, unless it can be demonstrated that such measure, although presumptively suspect, aims to realize a legitimate objective by means which are both appropriate and necessary. Under a third model, that of affirmative equality, the aim of ensuring equal treatment is not only to avoid instances of discrimination, but also to make progress towards a fair share of social goods among the different segments of the population. Under this model, affirmative policies are pursued which seek to improve the representation of certain groups in the areas or at the levels where they are underrepresented, and to arrive not only at a situation where discriminatory rules, policies or practices are outlawed, but where, moreover, social goods are distributed more equitably between the diverse groups composing society. The article shows that although European anti-discrimination law belongs to the first model of equality described, the Member States are not prohibited from espousing the second model – and, indeed, they should be incentivized to do so. It also highlights the obstacles that explain why the third model of equality has been slow to emerge in European law
