Lounge readers, particularly those working in the fields of discrimination and employment law, are probably already familiar with the research on “implicit” or unconscious bias. Many will recall, for example, the heated debates a few years back over Implicit Association Testing, featuring, on one side, Mahzarin Banaji and her colleagues, and Philip Tetlock and Hal Arkes, on the other. Given the implications for discrimination law, the dispute naturally made its way into the legal academy quickly, with prominent contributions from Jerry Kang and Banaji, Greg Mitchell and Tetlock, Christine Jolls and Cass Sunstein, Sam Bagenstos, and Ian Ayres, among others.
In a provocative piece recently posted to SSRN, Pat Shin (Suffolk) sets aside the practical and empirical disputes that largely characterized these prior exchanges in order to explore what he terms the “deep problem:” as a normative matter, should employment discrimination laws be understood to encompass liability for actions tainted by unconscious bias? In order to do so, Shin makes several simplifying assumptions and develops two intriguing hypotheticals, in the hopes of gaining a keener understanding of our currently prevailing notions of actionable discrimination.
Shin assumes, first, that unconscious bias is real and, second, that it is provable in individual cases. Shin then develops two hypotheticals:
Work Experience I: Black Applicant performed slightly better in the job interview than White Applicant, who has more work experience. Employer decides that work experience is the most important factor, and hires White Applicant. Employer’s decision was influenced by unconscious bias in favor of whites, yet honestly believes that his choice was based on White Applicant’s superior work experience.
Work Experience II: Same facts as above, but the Employer's decision was not affected by unconscious bias and the reason that Black Applicant has less work experience than White Applicant is because he had great difficulty obtaining employment in the early years of his career because of intentional discrimination by other potential employers.
Should the employer in Work Experience I be liable for employment discrimination? Shin argues that liability in Work Experience I (much less II) is implausible under current case law, but is not foreclosed by the literal language of Title VII. But Shin contends that liability for unconscious discrimination would be controversial, even if there were convincing evidence that it would reduce workplace inequality. The interesting question is why this is so. He argues that it is because the justificatory conception of discrimination (under which courts judge Employer’s rationales for acting) is deeply engrained in current law and a move to the causal conception (under which courts engage in the quasi-scientific business of identifying the psychological and causal antecedents of Employer’s actions) would represent a radical shift in our understanding of what discrimination is, not just an incremental expansion of the scope of the employment discrimination laws. In other words, it moves our conception of discrimination away from the paradigm of individual blame, and more toward a tax on the societal factors that contribute to workplace inequality.
But Shin argues that, although he doubts anyone would say that Employer is liable for discrimination in Work Experience II, it is deeply plausible to believe that there is no meaningful difference between Work Experience I and II. In other words, both should be actionable (as a theoretical matter) under current discrimination law or neither should be.
Though the paper remains somewhat agnostic on the ultimate question of whether discrimination law should reach Work Experience I and II, Shin’s conclusion strikes me as willing to contemplate that possibility. Such an understanding of discrimination, Shin correctly concludes, “must be understood not as a practice of enforcing norms of individual responsibility, but about effecting social change and reform.”
Unlike my colleague, Kate Bartlett, whose recent Virginia Law Review article defends the importance of “good intentions,” I’m untroubled by the potential demoralization of discrimination law that such a move would entail, for reasons that Shin largely addresses in the paper. (Interested readers can stop by Duke almost any day to watch Kate and me dispute the importance that the legal regime should attach to motive -- particularly, altruism versus profit-seeking -- in any number of settings. The paradigm of a great senior colleague, she’s always game for these exercises).
The questions that I’m left with, though, are largely the ones that Shin purposely sets aside for purposes of what, admittedly, is a real challenge to my prior conceptions about both the doctrine and purposes of discrimination law. Assuming that Work Experience I is really no different than Work Experience II, and assuming that the law should appropriately address such problems, then why is a case-by-case liability regime the appropriate means for doing so? In other words, if discrimination law is really a tax on harmful conduct or a redistributive regime of some sort, then what is the benefit of administering that regime through what – in real life – is likely to be a haphazard and imperfect case-by-case analysis fraught with problems of proof, evidence, and the like?
That more than exhausts my limited expertise on discrimination, which is primarily limited to organizational issues, such as compliance and business rationales for diversity. I’ll have to leave the heavy lifting on issues like this to Pat and the others who have been engaging these problems, including Susan Sturm, Tristin Green, Amy Wax and numerous others (all of which are reviewed and summarized in the Bartlett piece referenced above).
It sounds a little like Shin's paper is an exercise in futility. Why bother bringing up such an important question and all its various parts if you're not going to offer some sort of solution?
Posted by: Joe | May 19, 2010 at 07:50 PM
I don’t think that it is, Joe. Admittedly, the exercise is a fairly abstract one, but an abstraction worth considering in some depth in the case of liability for unconscious bias. The reason for that, I think, is that debates about liability for unconscious bias typically become bogged down with empirical disputes about whether such bias exists, the extent to which it actually impacts behavior, whether it could be proven in any given case, whether people can be “de-biased” or the behavior corrected, and so on. By assuming away these currently intractable problems, Shin lasers in on a point that has been given relatively little serious consideration – assuming it existed, and could be proven, and that liability would increase equality, should liability for implicit bias attach? The answer to that, as Shin points out, is surprisingly difficult. But if one thinks the answer is “yes,” then Shin’s argument is that this is a complete reconceptualization of discrimination law – not just an incremental evolution as some have argued – and would logically imply the possibility of liability for structural inequality not of the prospective employer’s own making. The paper is not an easy read and I’ve skipped some nuance here, as the issues Shin is tackling are quite difficult -- but I think it’s a worthwhile endeavor. I would not have blogged about it in such depth if I didn’t think so.
Posted by: Kim Krawiec | May 19, 2010 at 08:54 PM