US DOJ deputy: algorithmic cartel requires agreement

Pallavi Guniganti

05 February 2018

US DOJ deputy: algorithmic cartel requires agreement

Ethan Litwin and Barry Nigro at GCR Live Miami 2018 (Credit: Freedom Film)

Senior US Department of Justice official Barry Nigro has said concerns about price fixing through algorithms stem from a lack of understanding of the technology, and that tacit collusion through such mechanisms is not illegal without an agreement among participants. Pallavi Guniganti at GCR Live Miami

When analysing whether conduct constitutes collusion, an observer should “take out” the fact that an algorithm was involved and also disregard the effect on prices, the deputy assistant attorney general in the DOJ’s antitrust division said on Saturday. Nigro spoke in his individual capacity at the GCR Live 7th Antitrust Annual Law Leaders Forum.

Conscious parallelism and interdependent pricing are not illegal under US antitrust law, and both criminal and civil enforcement require an agreement among the cartelists, Nigro said. He acknowledged the anxiety surrounding algorithms and how they could help cartelists prevent other members from cheating on an agreement, but said the technology’s quick response can also help sellers compete aggressively by lowering prices.

While an agreement can be inferred depending on the circumstances, Nigro said he was sceptical of finding a hub-and-spoke arrangement simply based on the “spokes” all using a single price-setting mechanism. “Agreement to me is the key,” he said.

He cited the example of reselling sports tickets on StubHub’s website, which gives sellers data on the prices others have set for tickets to the same event, so they can choose whether to price higher, lower or somewhere in between. Having an algorithm advising a seller allows them to offer a more competitive price because they are eager to sell quickly, Nigro said, adding that he is unsure whether an algorithm by itself shows a conspiracy.

Even if the use of algorithms causes higher prices than would otherwise exist, that would be legal conscious parallelism so long as there is no communication between sellers or their algorithms that rises to the level of an agreement, he said, comparing the classic antitrust hypothetical of four gas stations at an intersection whose owners can see and respond to each other’s posted prices.

Nigro said he is unsure if new guidelines are necessary to deal with algorithms, following the US antitrust agencies’ 2017 joint statement to the Organisation for Economic Co-operation and Development, which he said did a “good job” outlining the application of antitrust law to algorithms.

Describing algorithms as “a bunch of if/then statements” or equations written by humans that can be run by machines much more quickly than a human could calculate, Nigro said, “I don’t know that the rules need to change just because we have a vehicle, a computer that can do for humans what the humans tell it to do, faster than the humans can do it.”

Nigro spoke on a panel alongside Kai-Uwe Kühn, a professor of economics at the University of East Anglia; Simpson Thacher & Bartlett partner John Terzaken; Google senior competition counsel Kevin Yingling; and Susanne Zuehlke, a partner at Willkie Farr & Gallagher. The panel was moderated by Dechert partner Ethan Litwin.

The GCR Live 7th Annual Antitrust Law Leaders Forum ended on Saturday.

Comments Add your comment

Add Your Comment