Lawyers and judge mull how to allege Uber’s rating system is racist

A lawsuit that a federal judge recently dismissed for being too "sparse and poorly drafted" came under review again Thursday, when attorneys argued over what information the plaintiff needed to prove that Uber's ratings tool is discriminatory.

The proposed class action was filed by noted Uber foe Shannon Liss-Riordan of Lichten & Liss-Riordan PC, who alleged the ride-share company racially discriminated against minority drivers by basing its firing decisions on customer ratings. Liss-Riordan alleged Uber upheld this firing policy without accounting for customers' racial biases, and that Uber was therefore wrong for deactivating the account of the Asian plaintiff, Michael Liu, for his low ratings.

In March, U.S. District Judge Vince Chhabria dismissed the case. Liss-Riordan filed an amended complaint, adding citations of studies and articles about technology and racial bias by the Boston Globe, Bloomberg, and New York-based think tank Data & Society. Liu v. Uber, 3:20-cv-07499-VC (N.D. Cal., filed Oct. 26, 2020).

At a hearing Thursday, Uber's attorney, Littler Mendelson PC shareholder Andrew M. Spurchise argued the amended complaint still did not contain plausible allegations.

"So what do you think Liu would need to allege to connect his deactivation termination to the racially biased system of terminating?" Chhabria asked.

Spurchise responded, "I would think that he would need to contend ... that he was performing his job in a way that would rule out at least some of the causes, some of the reasons for low ratings."

Chhabria pushed back. "One point of disparate impact [cases] is you have unconscious bias, right? So why does he need to include any allegations about explicit overt bias that he suffered at the hands of passengers? Why isn't it enough to say, look, we have this system for terminating drivers that accommodates subconscious bias? We have strong reason to believe that there is in fact subconscious racial bias in these ratings and Uber is terminating people based on the ratings, and I fell below the rating cutoff, and I got terminated, and I'm a racial minority. Why isn't that enough?"

Chhabria also grilled Liss-Riordan. "There's the stuff that Mr. Spurchise was talking about, which was a good point that I hadn't thought of, is more facts to bolster the idea that ... Liu was a victim of this discriminatory system and not just some terrible driver."

"I understand there's a question of whether you need to do -- whether you're required to do more at this stage or not," he added. "But couldn't you have done more pre-suit investigation on the effect of this customer rating system on Uber drivers? How hard would it have been to do some survey of former or current Uber drivers? ... Because it would be difficult and expensive?"

Liss-Riordan said, "We don't have access the way Uber has access. We've alleged a plausible complaint, we've stated the claim, and in order to do that kind of study ... we'd need discovery."

"And when you do disparate impact cases ... you don't get that data until after you've filed suit. You get the data in discovery, and then you have experts work with it," she argued.