In the set of (relatively) liberal recent pronouncements from the United States Supreme Court features its judgment in Texas Department of Housing v Inclusive Communities Project (2015). The Court, by a slender majority, held that the Fair Housing Act 1968 prohibited not just disparate treatment (direct discrimination in UK law), but also disparate impact (indirect discrimination), based on race. It did so whilst insisting upon certain stringent qualifications within which disparate impact prohibition must operate in order to avoid ‘constitutional questions’.
Disparate impact occurs when a facially neutral rule or policy which nevertheless, and without adequate justification, has a disproportionate adverse impact on a protected group. The US SC held in Griggs v Duke Power Company (1971) that disparate impact was prohibited in the context of employment discrimination under Civil Rights Act 1964. The concept was imported into the British Sex Discrimination Act 1975 and Race Relations Act 1976 as ‘indirect discrimination’ (Hepple, p 608-9). Since then, the evolution of the concept of ‘discrimination’ has taken divergent routes in the two jurisdictions: US law insists on the primacy of discriminator-oriented and reasons-based direct discrimination prohibition, and (just about) tolerates the prohibition of indirect discrimination. British law has come to cast direct discrimination in the image of indirect discrimination, recognising that it is the impact on the victim, and not the reasoning of the discriminator, that is key to either form of discrimination (Khaitan, pp 1-4, ch 6).
The US story:
In 1976, the US SC refused to extend the logic of Griggs to the constitutional prohibition on discrimination under the 14th Amendment: this provision continues to prohibit only disparate treatment, not disparate impact (Washington v Davis). The next landmark is Wards Cove v Antonio (USSC 1989), where the Court tried to defang disparate impact law through a series of procedural and doctrinal innovations. Most of these changes were undone by Congressional statute. But the spark from Washington v Davis was fanned by Justice Scalia’s concurring judgment in Ricci v Destefano (2009). Given that voluntary compliance with disparate impact prohibition usually requires race-conscious action, Scalia wondered whether such prohibition did not therefore violate the 14th Amendment’s prohibition on the state undertaking or mandating disparate treatment. In her dissent, Justice Ginsburg argued that while disparate treatment law demanded race-neutrality in form, it did not prohibit race-consciousness of purpose (p 619-20, a distinction defended by Siegel). Note, in contrast to British position, that even this approach to disparate treatment concerns the reasons and actions of the defendant, rather than the impact on the plaintiff.
The majority’s recent decision in Inclusive Communities affirmed what all nine Courts of Appeal had held all along: that the Fair Housing Act, like the Civil Rights Act, prohibits disparate treatment as well as disparate impact. Against the context of Scalia’s threat in Ricci, however, Justice Kennedy held that ‘serious constitutional questions’ can be avoided only if the disparate impact prohibition is subjected to strict limitations. The extent to which these limitations bring the ghost of Wards Cove back to life remains to be seen. But some of the rhetoric suggests this could be a pyrrhic victory for the advocates of disparate impact law.
In Britain, the law has followed a different trajectory. In a string of cases, the House of Lords and then the UK Supreme Court have given primacy to indirect discrimination rather than direct discrimination [see generally, James v Eastleigh Borough Council (1990), R (E) v JFS (2010), Bull v Hall (2013)]. Citing Advocate General Sharpston, Lady Hale affirmed in Bull that ‘I take there to be direct discrimination when the category of those receiving a certain advantage and the category of those suffering a correlative disadvantage coincide exactly with the respective categories of persons distinguished only by applying a prohibited classification’ . This approach discards the viewpoint of the discriminator, even when assessing direct discrimination, and looks only for the impact on the victim. If the coincidence is not exact, but the disadvantage is nonetheless disproportionate on the victim group, the discrimination is indirect rather than direct. In fact, British Equality Act 2010 effectively mandates race-consciousness of sorts for public authorities under the public section equality duty (for the connection between indirect discrimination and the public sector equality duty, see Fredman).
This turn from reasons towards effects in British law has been criticised not only by conservatives (Finnis, Law Quarterly Review, 2010), but also some liberals (Gardner, 182). Ultimately, the choice is moral, rather than linguistic: whose point of view is key to our understanding of the concept of discrimination: the discriminator’s or the victim’s.