Julian Jonker, Wharton
DISCRIMINATION AT A DISTANCE:
THE PLACE OF PROTECTED CLASSES IN ANTI-DISCRIMINATION LAW
Abstract
While there is good reason to be pessimistic about the state of anti-discrimination law, some recent jurisprudential trends also give reason for optimism. Federal courts, in recent decisions concerning race and sexual orientation discrimination such as Holcomb v Iona College, 521 F.3d 120 (2d Cir. 2006), Hively v Ivy Tech Community College, 853 F.3d 339 (7th Cir. 2017) (en banc), and Zarda v Altitude Express, 883 F.3d 100 (2d Cir. 2018) (en banc), have effectively enlarged the category of classes protected by Title VII by deploying a theory of associational discrimination. This approach builds upon the Supreme Court’s willingness to depart from a strict reading of Title VII in order to impose liability for harassment and stereotype policing in decisions such as Meritor Savings Bank, FSB v Vinson, 447 U.S. 57 (1986) and Oncale v Sundowner Offshore Services, 523 U.S. 75 (1998). Surprisingly little has been done to understand and elaborate the theory immanent in these developments, perhaps because they are underwhelming from the perspective of longstanding critiques of the structure of anti- discrimination law. Yet room must be made for the former approach, since it is more likely that the courts will respond to the opportunities for extension suggested by a rational reconstruction of their current practice than that they will depart radically from that practice on the basis of an a priori theory of inequality.
Judicial developments favoring liability for stereotype policing and associational discrimination are surprising given that they do not appear to fit within the comparativist understanding of discrimination that we find in the statutory text and judicial dicta. Comparativism, to which nearly all philosophers, legal scholars, and judges subscribe, is the view that discrimination necessarily involves differential treatment on the basis of some socially salient trait. It is reflected in statutory requirements that discrimination be ‘because of’ a plaintiff’s membership in a protected class, as well as in judicial reliance upon a comparative test requiring that a plaintiff show that they were treated differently than they would have had they not been a member of a protected class. In contrast, stereotype policing of the type involved in Oncale involves treatment that need not vary with possession of a protected trait. And associational discrimination essentially involves a plaintiff being treated unfavorably because of whom they associate with—so that the plaintiff’s treatment may not in fact vary with their possession of a protected trait. From the viewpoint of comparativism, these sorts of cases seem to rest on a fiction of the sort involved in the doctrine of transferred intent. Some judges have therefore resisted the extension of liability in these cases.
But the consensus about comparativism rests on an ambiguity in logical scope. Narrow-scope comparativism requires that the differentiation involved in discrimination be on the basis of a protected trait, so that it treats the plaintiff unfavorably relative to how they would have been treated had they lacked the trait. Wide-scope comparativism requires that the differentiation involved in discrimination merely have some causal connection to the plaintiff’s possession of a protected trait, so that the treatment may vary with possession of some non-protected trait, though the variation arises because the plaintiff possesses the protected trait. Title VII’s requirement that actionable discrimination be ‘because of’ the plaintiff’s possession of a protected trait is capable of both a narrow-scope and a wide-scope reading. A wide-scope reading is preferable given the historic purpose of Title VII. It also makes sense of the recent decisions regarding stereotype policing and associational discrimination. Also, it would allow courts to respond to the cases involving intersectionality and structural discrimination that prompt other scholars to urge far more radical critiques of existing anti-discrimination law.
Finally, wide-scope comparativism sheds new light on debates about the mental state required for disparate treatment. Scholars have worried about forms of discrimination that do not fit easily into the taxonomy of disparate treatment and disparate impact. An example is ‘masking,’ in which decisions that treat protected classes unfavorably are deliberately made on the basis of traits that are statistically correlated with protected traits, thereby excluding disparate treatment liability. Setting aside difficulties of proof, such cases can properly be seen to involve disparate treatment if we take a wide-scope view of discrimination. Developing this interpretation anti-discrimination law will be a crucial analytical tool for making sense of discrimination in a world of opaque algorithms and black box decision-making.