Amazon is a ripoff
If you’d like an essay-formatted version of this post to read or share, here’s a link to it on pluralistic.net, my surveillance-free, ad-free, tracker-free blog:
There’s a cheat-code in US antitrust law, one that’s been increasingly used since the Reagan administration, when the “consumer welfare” theory (“monopolies are fine, so long as the lower prices”) shoved aside the long-established idea that antitrust law existed to prevent monopolies from forming at all.
The idea that a company can do anything to create or perpetuate a monopoly so long as its prices go down and/or its quality goes up is directly to blame for the rise of Big Tech. These companies burned through their investors’ cash for years, selling goods and services below cost, or even giving stuff away for free. Think of Uber, who lost $0.41 on every dollar they brought in for their first 13 years of existence, a move that cost their investors (mostly Saudi royals) $31 billion.
The monopoly cheerleaders in the consumer welfare camp understood that these money-losing orgies could not go on forever, and that the investors who financed them weren’t doing so for charitable purposes. But they dismissed the possibility that would-be monopolists could raise prices after attaining dominance, because these prices hikes would bring new competitors into the market, starting the process over again.
Well, Uber has doubled the price of a ride and halved the wages of its drivers (not that consumer welfare theorists care about workers’ wages — they care about consumer welfare, not worker welfare). And not just Uber: companies that captured whole markets have jacked up prices and lowered quality across the board, a Great Enshittening whose playbook has been dubbed “venture predation”:
Not only was this turn predictable — it was predicted. Back in 2017, Lina Khan — then a law student — published a earthshaking Yale Law Journal paper, “Amazon’s Antitrust Paradox,” laying out how monopolists would trap their customers and block new competitors as they raised prices and lowered quality:
Today, Khan is the chair of the FTC, and has brought a case against Amazon that turns her legal theories into practice, backed by a cheering chorus of Amazon customers, workers, suppliers and competitors who’ve been cheated by the e-commerce giant:
Khan’s case argues that Amazon is not the house of bargains that it’s widely billed as. She points to the sky-high fees that Amazon extracts from its sellers (45–51% of every dollar!) and the company’s use of “most favored nation” deals that force sellers who raise their Amazon prices to pay those rents to raise their prices everywhere else, too:
Now, a new Amazon Paradox has dropped, and it drills into another way that Amazon overcharges most of us by as much as 29% on nearly every purchase, disqualifying it from invoking that consumer welfare cheat code. The new paper is “Amazon’s Pricing Paradox,” from law professors Rory Van Loo and Nikita Aggarwal, for The Harvard Journal of Law and Technology:
The authors concede that while Amazon does have some great bargains, it goes to enormous lengths to make it nearly impossible to get those bargains. Drawing from the literature on behavioral economics, the authors make the reasonable (and experimentally verified) assumption that shoppers generally assume that the top results in an Amazon search are the best results, and click on those.
But Amazon’s search-ordering is enshittified: it shifts value from sellers and shoppers (you!) to the company. A combination of self-preferencing (upranking Amazon’s own knock-offs), pay-for-placement (Amazon ads), other forms of payola (whether a merchant is paying for Prime), and “junk ads” (that don’t match your search) turn Amazon’s search-ordering into a rigged casino game.
The ability to manipulate customers and sellers and get more money from both is why Amazon has so many incentives to use Amazon’s internal search tool, rather than, say, searching Amazon via Google, which can yield far superior results. For years, Amazon ran a program called Amazon Smile, where a share of every purchase you made would be given to a charity of your choice — but only if you found that item by searching for it on Amazon, and not via Google or a direct link:
In their new paper, the authors extract and analyze a large dataset of common items you might buy on Amazon, determining which result is best — the lowest price at the highest rating — and then calculating how much more you’ll pay for that item if you click the first relevant (non-ad) item on the search results.
If you trust Amazon search to find you the best product and click that first link, you will pay a 29% premium for that item. If you expand your selection to the “headline” — the first four items, which are often all that’s visible without scrolling — you’ll pay an average of 25% more. That top row accounts for 64% of Amazon’s clicks.
On average, the best deal on Amazon is found in the seventeenth slot in the search results. Seventeen!
Amazon argues that none of this matters, because it allows users to refine their searches to get the best bargains, but Amazon’s search won’t let you factor in “unit pricing” — that is, the price per unit. So if you order your search by price, the seller who’s offering a single pencil for $10 will show up above a seller who’s offering ten pencils for $10.01.
Here is an iron law of cons: any time someone adds complexity to a proposition bet, the complexity exists solely to make it hard for you to figure out if you’re getting a good deal. Whether that’s the payout lines on a craps table, the complex interplay of deductibles and co-pays on your health insurance, the menu of fees your bank charges, or the add-ons for your cell-phone plan, the complexity exists to confound your intuition and overwhelm your reason:
And Amazon certainly knows how to pile on the complexity! First, there’s the irrelevant results — AAA batteries that show up in a search for AA batteries, or dog accessories that show up in a search for cat accessories:
Then there’s the “drip pricing”: extra charges that get tacked on at checkout, like shipping fees. I once found an item on Amazon that advertised “free shipping” — but at checkout, that “free shipping” came with a delivery date that was three months in the future. Upgrading to shipping in the current quarter doubled the price.
Drip pricing makes it hard to figure out if Prime is a good deal, too. Recall that Amazon already comps shipping on orders over $25, so a potential Prime purchaser has to evaluate whether they’ll place enough sub-$25 orders in the coming year to justify the price — and also factor in the fact that Prime items are often more expensive on a per-unit basis than their non-Prime equivalents. Yes, Prime comes with other perks — music and videos — but valuing these just adds complexity to your calculations about whether Prime is a good buy for you, and requires that you factor in the possibility that Amazon will enshittify those services and reduce their value in the coming year, say, by taking away the ability to turn off shuffle when listening to music:
Or stuffing ads into your videos:
Finally, there’s the nonsense labels that Amazon pastes onto its search results: “Best Seller,” “Climate Pledge Friendly,” “Highly Rated,” “Top Rated From Our Brands” and other gibberish that doesn’t necessarily mean what it seems like it means. Is an item a “best seller” because it was briefly price-dropped, or elevated in search results, or both, or because other shoppers genuinely liked it better?
The authors conclude that getting the best price on Amazon requires that you “first spend considerable time searching through pages of results and then utilize, at a minimum, spreadsheet algebraic capabilities to determine the product’s full price…[and] somehow de-bias from the psychological effects of anchoring, and labels such as ‘limited time deal’ and ‘Best Seller,’ as well as many other subtle psychological influences.”
Amazon says it’s entitled to use the consumer welfare cheat-code to get out of antitrust enforcement because it has so many bargains. But to get those bargains, you have to pay such minutely detailed attention — literally spreadsheeting your options and hand-coding mathematical formulas to compare them — that you’ll almost certainly fail. The price of failure is incredibly high — a 25–29% overcharge on every purchase.
Amazon’s burying of this vital information will be familiar to Douglas Adams readers, as the “Beware of the Leopard” tactic. It’s not even the first time Amazon’s deployed it:
Another group of scholars recently coined a useful term to describe this ripoff: in a paper published last week, Tim O’Reilly, Mariana Mazzucato and Ilan Strauss dubbed the costs of all this complexity “attention rents”:
It’s fascinating to see these two different groups of scholars, coming at this problem from multiple disciplines, all converging on the same analysis! When technologists, trad economists, behavioral economists, and antitrust lawyers all study Amazon and come away pointing at the same sleazy tactic as being at the heart of the scam, it feels like maybe we’re having A Moment. What’s more, all of this is so thoroughly presaged by Khan’s 2018 paper that it suggests that she’s a bona fide prophet.
The authors of this new paper are pretty confident that this gimmick violates antitrust law. They point out that it doesn’t matter if Amazon customers feel like they’re getting a good deal — just as it doesn’t matter if don’t know that you got charged a higher rate for your mortgage because you’re Black, that’s still illegal.
What’s more, consumer protection law doesn’t require that the merchant intends to rip you off. There’s plenty of laws requiring supermarkets to post unit prices on their shelves. These laws don’t start from the assumption that supermarkets who don’t use unit pricing are trying to scam you! Rather, they start from the assumption that you will make better-informed purchases if you have that information, and so you should get it.
Regulating the presentation of prices is firmly in the purview of antitrust law, especially consumer welfare antitrust, which fetishizes low prices above all else. The less competitive a market is, the less pressure a company will feel to offer clear price information to customers, because those customers will have fewer places to go if they don’t like the company’s business practices.
All of this militates for antitrust intervention: rules for how Amazon must do its business. The authors propose three different kinds of rules:
I. Force Amazon to halt its most deceptive practices, like hiding the true price including shipping or chaffing search results with confusing junk ads. One fascinating tidbit: just a few days after this paper was published, the FTC revealed that Amazon had been deliberately cluttering its results with junk ads in order to juice revenue:
II. Mandate interoperability between Amazon and comparison shopping sites by forcing the company to publish its pricing data in machine-readable format, and allowing customers to authorize shopping bots to access their purchasing data to help them figure out how to get a better deal. Another fascinating turn — the same week this paper came out, the CFPB proposed a rule that would force your bank to do the same thing — let you forward your data to comparison shopping sites that would tell you which bank you’d get the best deal from:
The CFPB rule goes one step further, strictly limiting how those comparison sites can use your data, banning them from retaining, selling or sharing it or using it to target ads to you. This is the approach that my EFF colleague Bennett Cyphers and I proposed in our “Privacy Without Monopoly” paper:
III. Create legal safe harbors for scraping. Scraping is a form of “adversarial interoperability,” the self-help measures that technologists use to modify and adapt existing services without their owners’ consent. Think of reverse-engineering, bots, etc:
Comparison shopping sites have historically relied on scraping to help their users get better deals. Amazon almost certainly scrapes its competitors’ sites to figure out if a merchant is selling more cheaply elsewhere (these merchants are punished by being banished to screen eleventy-million of the search results, which has the same effect of just kicking them off of Amazon).
Scraping was once the norm online, then it dwindled, as monopolists used their cash reserves and market power to get governments to punish rivals that used it. But scraping is a very important backstop to any kind of price-analysis. Though Mario Zechner used grocery stores’ own official APIs to prove that they were colluding to rig prices, he has gone on record to say that he would also use scraping if they shut down those gateways or denied him access to them:
In my latest nonfiction book, The Internet Con, I lay out virtually the same program for addressing monopoly power in every tech industry:
I. Start with traditional antitrust remedies (breakups, bans on unfair or deceptive practices)
II. Mandatory APIs that allow tinkerers, co-ops, nonprofits, and startups to interface with dominant platforms and offer their users and suppliers better services and deals;
III. Safe harbors for adversarial interoperability, so that when companies cheat on their mandatory APIs by blocking or degrading them, those rival services can keep things going while they wait for fact-intensive regulatory proceedings to force the big companies back into compliance.
Reading this new paper, I was struck by how much convergence there is among different kinds of practitioners, working against the digital sins of very different kinds of businesses. From the CFPB using mandates and privacy rules to fight bank ripoffs to behavioral economists thinking about Amazon’s manipulative search results.
This kind of convergence is exciting as hell. After years of pretending that Big Tech was good for “consumers,” we’ve not only woken up to how destructive these companies are, but we’re also all increasingly in accord about what to do about it. Hot damn!