The FTC’s Amazon Antitrust Case Centered on Wildly Misleading Statistic

The FTC’s Amazon Antitrust Case Centered on Wildly Misleading Statistic


One of the central arguments in the Federal Trade Commission’s (FTC) antitrust lawsuit against Amazon is that the online retailing giant has stunted the growth of potential competition by forcing small businesses and other independent sellers to funnel their products through Amazon’s own in-house distribution system.

Much of that argument seems to hinge on a single statistic—one that top officials at the FTC have cited in interviews and on Twitter and that pops up in a newly unredacted part of the FTC’s lawsuit. There’s just one problem: that stat doesn’t say what the FTC keeps claiming it does.

To understand what it does say, first you need a bit of background.

Amazon has allegedly deprived potential competitors of the “ability to gain the scale and momentum needed to effectively compete online,” as FTC Chairwoman Lina Khan told Bloomberg TV in an interview on September 26.

In fact, the argument goes, Amazon has been so determined to squash that potential competition that in 2019 it shut down the so-called “Seller Fulfilled Prime” (SFP) program—an arrangement in which independent sellers could offer free shipping to buyers with Amazon Prime subscriptions but where the sellers remained responsible for getting the orders out the door accurately and on-time. Since the SFP program was shuttered, all independent sellers using Amazon Prime have been forced to go through Amazon’s own distribution network (known as the “Fulfillment by Amazon,” or FBA, system).

That’s evidence of anti-competitive monopoly power, according to Khan, who called the arrangement a “coercive scheme” during that same Bloomberg TV interview.

“At various points, Amazon did experiment with giving sellers more leeway to use Seller Fulfilled Prime,” she explained. “But once Amazon recognized that that would threaten its monopoly power, it switched that off, even though sellers were effectively meeting the same standards that FBA does.”

In the newly redacted part of the lawsuit, the FTC reiterates this claim—and attaches a figure to it: 95 percent. This is the key statistic.

“Amazon shut SFP down because they said deliveries weren’t on time. But new info today shows sellers using SFP met the delivery requirement set up by Amazon more than 95% of the time,” Douglass Farrar, director of public affairs for the FTC, tweeted on Thursday, along with a screenshot from the lawsuit.

In a nutshell, the FTC’s argument is that sellers using the SFP program were meeting their delivery targets 95 percent of the time, but Amazon shut the program down anyway in order to consolidate power and limit competition (even competition that was already coming through its own front door).

But, importantly, that’s not what the 95 percent statistic actually says.

Just read the line in the FTC’s own lawsuit, as helpfully screenshotted by Farrar: “Sellers enrolled in SFP met their promised ‘delivery estimate’ requirement set by Amazon more than 95% of the time in 2018.” (Emphasis added.)

That doesn’t mean those sellers were meeting the requirements of Amazon’s Prime program—it means they were meeting whatever shipping standards they set for themselves when setting up their Amazon seller account.

“It was correct to say sellers met their own estimates, but those estimates for delivery could be days, weeks or even months,” Carl Szabo, vice president at NetChoice, points out. “Customers who bought from these sellers would get their stuff at the ‘promised delivery date,’ but that date could be several days, weeks, or months later—not the two days promised under a ‘Prime’ badge.”

How many of those sellers were actually meeting the standards for Prime-level shipping? According to Amazon spokesman Tim Doyle, it was about 16 percent.

“The misleading figures the FTC points to in the complaint falsely portray how we work with sellers to meet our customers’ high expectations,” Doyle tells Reason. He says Amazon made the decision to pause new enrollment in the SFP program in 2019 because fulfillment rates were “far below the high standards and expectations our customers have for Prime.” Since then, Amazon has restructured the program and reopened it.

Rather than trying to squeeze competition and harm consumers, Amazon seems to have taken proactive steps to make sure its customers were getting the Prime-level service they had paid for. This isn’t evidence of a monopoly; it’s a demonstration of how Amazon has become so successful: by putting customers first.

This is in some ways similar to the bizarre claim the FTC made about Amazon making it too difficult for users to cancel their Prime subscriptions—a process that takes six clicks, one fewer than what’s required to submit a complaint to the FTC, as Reason previously reported.

That was at least an accurate stat—albeit a facetious one, and a pretty silly thing to base an antitrust lawsuit upon. By comparison, the inaccurate way that the FTC has used this statistic about Amazon’s SFP program cuts to the core of the agency’s lawsuit. It suggests that Khan either fundamentally misunderstands the claim she (and the FTC at large) is making, or that she is lying about what it means.

“By presenting this as some sort of ‘gotcha’ moment, the FTC is trying to suggest that SFP outperforms FBA, which is simply not true,” wrote Carl Holshouser, senior vice president of TechNet, a nonprofit that advocates for the so-called “innovation economy,” on Twitter. “The FTC, by mischaracterizing the facts of this case and continuing their efforts to undermine consumer preferences, is once again undermining their own credibility.”

Amazon didn’t kill an alternative that was working for consumers as a way to entrench its monopoly further. It shut down a program that was clearly flawed and that threatened to undermine customer confidence in the Amazon brand, reconfigured it, and now has re-opened it with better controls in place.

Then again, it’s probably no surprise that the federal government is unfamiliar with that process.