What does “disparate impact” mean?
RSS 最後由 編輯
Mar 3rd 2021
A LANDLORD WHO posts a sign in front of his apartment building that says “Disabled people unwelcome” is clearly engaged in unlawful discrimination. So is a nightclub owner who refuses to hire women. Federal law bars these acts under the rubric of “disparate treatment”: intentionally treating people differently because of their race or sex (two of several federally protected classes; among the others are religion, national origin, age and disability status).
But what about a landlord who only rents apartments to people with full-time jobs, or a nightclub owner who only hires people above a certain height? The first policy excludes disabled veterans, who may be unable to work but can still afford their rent. Because the average American woman is shorter than the average American man, the second policy makes it harder for women than men to be hired. Both policies are facially neutral; neither overtly discriminates. But both adversely affect members of a protected class—they discriminate in practice, if not in intent. In American law, this phenomenon is known as “disparate impact”.
A disparate-treatment claim relies on evidence of a discriminatory motive behind a policy—which means that, in the absence of a smoking-gun racist or sexist remark, or an explicit policy of unequal treatment, that claim can be hard to prove. This is where disparate-impact lawsuits, cognisable under several civil-rights statutes, come in. Claims of disparate impact look to the effects rather than the intent of a rule or policy. Policies that harm vulnerable people without a valid reason are illegal, the theory says, even if the discrimination was inadvertent.
The first step in advancing such a claim is to identify a specific policy that produces a pattern of adverse results for people who are members of protected groups—lower salaries for female employees, for example, or higher mortgage rates for African-American homebuyers. If a pattern of inequality from the policy can be shown, employers or mortgage-brokers could defend themselves by citing a legitimate business reason for the differential that has nothing to do with racial animus or sexism. They could also argue that the plaintiffs have not fulfilled their duty of presenting an alternative policy that could reach the same legitimate goal with less adverse results.
This framework for disparate-impact claims comes from Griggs v Duke Power Company, a 1971 Supreme Court decision. In that case, Duke Power Company stipulated that employees who wanted to transfer internally to higher-paid jobs must hold a high-school diploma and register a minimum score on two separate aptitude tests. Willie Griggs, who worked for Duke, sued his employer, alleging that because those policies barred a disproportionate number of African-American employees from advancing, they fell foul of the 1964 Civil Rights Act, which among other things forbids racial discrimination in employment. The court found that neither of the two requirements directly related to an employee’s professional aptitude, and, more broadly, that employment rules that are facially neutral but “discriminatory in operation” may, unless they are appropriately justified, be illegal.
Many conservatives have had disparate-impact analysis in their crosshairs for decades. At the confirmation hearing for Merrick Garland, Joe Biden’s nominee for attorney-general, John Kennedy, a Republican senator from Louisiana, asked sceptically whether disparate impact is “just a product of the numbers”. Racism, Mr Kennedy implied, cannot be proven by showing statistical patterns; individuals and institutions should not be held liable for discrimination just because the chips fell where they did.
Mr Kennedy’s question was a red herring, as disparate-impact claims must identify a specific policy causing the alleged harm; pointing to raw data alone is a non-starter. But in line with the sentiment behind Mr Kennedy’s comment, in September 2020 the Trump administration finalised a rule making it harder for the Department of Housing and Urban Development (HUD) to bring disparate-impact claims against mortgage-lenders, insurance companies and developers. A federal judge in Massachusetts promptly blocked this change, finding the new rule made it “easier for offending defendants to dodge liability and more difficult for plaintiffs to succeed.” Mr Trump also oversaw a weakening of enforcement and fines at the Consumer Financial Protection Bureau (CFPB), a federal agency charged with guarding against misleading and discriminatory lending by banks, credit unions, retailers and credit-card companies. Under his administration, the CFPB turned away from disparate-impact analysis in its enforcement of the Equal Credit Opportunity Act, a 1974 law protecting borrowers from discrimination.
In an executive action days after taking office, Mr Biden asked HUD to re-evaluate Mr Trump’s weakened fair-housing rules and to come up with new ways of promoting racial justice in homeownership. The Biden administration is also expected to combat systemic racism in the terrain overseen by the CFPB by using stiffer penalties to hold lenders accountable for racially discriminatory practices. This new tone may transform federal civil-rights enforcement, but it could ultimately be hemmed in by the Supreme Court. A 2015 decision, Texas Department of Housing and Community Affairs v Inclusive Communities Project, rejected a bid to scrap disparate-impact lawsuits under the Fair Housing Act. But Inclusive Communities was a 5-4 decision written by the now-retired Justice Anthony Kennedy; his seat and that of another member of the majority, the late Ruth Bader Ginsburg, are occupied by successors whose views are to the right of theirs. Should Mr Biden’s policies come under review, the disparate-impact understanding of equality may be vulnerable to the court’s new 6-3 conservative majority.