Since Lina Khan – famously an Amazon critic – was appointed chair of the Federal Trade Commission (FTC), a significant lawsuit against Amazon has been expected. The lawsuit was finally launched on Tuesday, brought jointly by the FTC and 17 state attorneys general.
The FTC says that Amazon has raised its fees and commissions, cluttered its webpages with more and more ads, showed consumers worse results in the Buy Box and results rankings, and generally degraded its platform. Rival e-commerce platforms should be able to take advantage of this situation, but cannot due to Amazon’s “anti-discounting” policies. Rival logistics firms cannot compete because sellers are effectively pushed into using Fulfillment by Amazon (FBA).
An ambitious case
I was interested to see whether the FTC would bring a wide case that tries to put together Amazon’s various strategies into an overall exclusionary and exploitative narrative, or a narrower case that tries to pinpoint a small number of specific strategies. The latter approach tends to be adopted by the UK and EU agencies who worry that their cases take too long and become undeliverable if they are scoped too widely, but some would argue these narrow cases do less to improve competition and consumer outcomes overall.
As we’ve discussed previously on this blog, Amazon’s business model involves a large number of interlinked strategies that revolve around the Prime membership programme. Chipping away at individual issues such as Amazon’s use of third-party sellers’ data is a worthy issue that the European Commission and UK’s Competition and Markets Authority (CMA) have both investigated, but it will not kick-start competition in e-commerce.
It is clear that the FTC’s case is towards the wider end of the spectrum, and is therefore ambitious. It aims at the more strategic issue of competition in e-commerce, and it deals with both Amazon’s first-party and third-party platforms. Time will tell whether that was the right decision.
The FTC’s allegations
The FTC argues that Amazon is dominant in two separate markets: online superstores in the US (i.e. the first-party platform), and online marketplace services in the US (i.e. the third-party platform). Notably, these markets do not include: (1) bricks and mortar stores, (2) more focused online stores (e.g. a sports equipment website hosted by Shopify), or (3) groceries (which are stored and sold in a different way). Its conduct is said to enhance its market power in both markets.
The case is primarily about Amazon protecting its position and harming inter-platform competition, and so does not cover things like self-preferencing that would only harm intra-platform competition. The case describes a most-favoured nation (MFN) policy, which has the practical effect of preventing third-party sellers from discounting their products on competing marketplaces, combined with other “anti-discounting” tactics. It also describes how Amazon effectively forces third-party sellers to use its logistics services because otherwise they will not win the Buy Box and their sales will “tank” (in the words of an internal Amazon email).
The FTC describes Amazon’s strategies as interconnected, and says they are aimed at denying e-commerce and logistics competitors the ability to gain the necessary scale to compete effectively. It states that the “cumulative impact of Amazon’s unlawful conduct is greater than the harm caused by any particular element”.
The lawsuit does not address all the possible competition issues raised by Amazon. For example, it does not directly allege that excessive prices infringe antitrust law (not surprising considering the different approach of US law compared to European law). It does not allege that other unfair terms are imposed on sellers and wholesalers. It does not address Amazon’s dual role as a platform and a seller, or its use of seller data, which are the issues that have been the focus of cases in the EU, Germany, Italy and the UK. It does not directly deal with the cross-subsidy of Prime. However, it is undeniably ambitious, and the bundle of practices it describes has a coherent narrative.
Most-favoured nation (MFN) and other anti-discounting policies
The FTC alleges that Amazon enforces an MFN policy in practice, despite having dropped its explicit contractual policy years ago (due to a European Commission investigation and political pressure in the US). Amazon is said to employ web crawlers to monitor sellers’ prices on other websites, and then to punish sellers who offer cheaper prices elsewhere.
For example, Amazon is said to enforce the policy by removing the ability to win the Buy Box from sellers it wishes to punish. All sellers know how important the Buy Box is – a huge majority of sales go to the Buy Box winner. The relevant figure is redacted from the FTC’s lawsuit, but a recent CMA document revealed that more than 75% of sales go to the winner of the Buy Box in the UK, and the European Commission has published a similar number for France and Germany.
Other forms of punishment are also alleged, including downgrading a seller’s position in the search rankings, and even being excluded from using the marketplace (of which, according to an internal Amazon email, some sellers “live in constant fear”).
If Amazon’s rivals such as Walmart and Target, or new entrants such as TikTok and Shein, cannot compete on price, they stand little chance of dislodging Amazon from its market leading position. The MFN policies may not matter so much if Amazon was charging sellers very low fees and commissions. Over the past few years, we have seen various estimates of how much third-party sellers pay Amazon as a percentage of their sales. The FTC quotes a figure of “close to half” of their revenues for sellers who use Amazon’s logistics services, which is said to have increased alarmingly over the past few years. It therefore alleges that sellers’ prices are high on Amazon’s platform and those high prices are mirrored across to other websites. The high Amazon price serves as the internet’s price floor. There is little point in rivals like Walmart and Target offering lower fees and commissions to sellers since, due to Amazon’s policies, it would not result in the sellers lowering consumer prices or an increase consumer sales.
The FTC further alleges that Amazon uses an algorithm on its first-party site that discourages price competition from other websites. The details are redacted, but perhaps the algorithm aggressively discounts products on Amazon’s platform in the short-term when it sees another website cut its price. It would therefore discipline the competitor website over time as the operation of this algorithm becomes known and the competitor realises that price cutting will not result in increased sales. The FTC quotes an internal document that predicts that “prices will go up” as a result of the practice, and that this is in practice what happened.
The Loch Ness Monster of pricing algorithms?
Nothing makes a corporate initiative sound more exciting than giving it a project name and making it secretive (c.f. Google and Meta’s Project Jedi Blue relating to header bidding, which was investigated in various jurisdictions, and for which they possibly should have chosen a less memorable name). This is even more the case when the project name refers to a mythical monster.
In this case, the FTC refers more than a dozen times to a Project Nessie. It may be named after the legendary Loch Ness Monster, but it is also the name of one of the buildings on Amazon’s Seattle campus, and it is mentioned on Amazon’s own website, so perhaps it is not as secret as you might think. The FTC says that Nessie is an “algorithm”, a “pricing system” that has “extracted” something (presumably money and data) from “American households”, and it is an “unfair method of competition”. It goes “a step further” than the anti-discounting measures discussed above. Unfortunately, the lawsuit redacts any further details, so we don’t know if the Nessie algorithm is as exciting as it sounds.
So many ads
The increase in Amazon’s advertising revenues has been a major story over the last few years. The FTC describes the rise in Amazon’s ad revenues and the increased space on its platform that is devoted to ads. These revenues are derived from “pay-to-play” fees that sellers nowadays cannot avoid without significantly harming their sales. The FTC says that these are effectively price hikes for sellers, and that the result for consumers is a lower-quality platform which favours sellers who pay over the sellers who might be better suited to the consumer’s requirements.
Forced use of logistics services
The FTC alleges that sellers are effectively forced to use Amazon’s expensive logistics services, FBA, if they want to be Prime-eligible (and therefore want to avoid their sales “tanking”). This raises their costs and prevents them from multi-homing, therefore excluding other logistics firms from competing effectively at scale.
Amazon will presumably argue that its delivery guarantee is core to its business model and they have made immense investments in logistics. They should not be forced by a government agency to allow third parties to piggyback on that investment, and it is Amazon’s reputation that would suffer if delivery standards were inadequate. The response to that seems to be that Amazon did in fact open up to third parties through the Seller Fulfilled Prime programme. Amazon has offered to the CMA and European Commission a remedy whereby it will allow third party logistics firms to deliver Amazon orders as long as they meet objective standards, which are set by Amazon.
Some commentators will say that Amazon generally makes little profit overall, and indeed that it is only Amazon Web Services rather than e-commerce that brings them close to break-even. Therefore, the prices charged by Amazon for its e-commerce services cannot be excessive and Amazon must be keeping consumer prices low. This is wrong, and competition policy practitioners should move past that misleading narrative.
The Prime bundle is at the core of Amazon’s flywheel business model. Data is hard to come by, but we have estimated well over $100 billion of annual losses – it is expensive to offer fast delivery for free to Prime members throughout the US and other countries and to commission large amounts of television content. The high fees and commissions (including advertising fees) that are imposed on sellers and brands are subsidising the Prime bundle, and are therefore subsidising Amazon’s business model as a whole. The more attractive the Prime bundle is, the more consumers join the programme. Prime members are more locked into Amazon’s ecosystem, and spend multiples more money on its e-commerce platform, than non-members (in the UK, for example, Prime members represent more than 80% of sales, according to the CMA). Sellers will pass on part of their inflated costs to consumers, so it is therefore consumers who ultimately subsidise the Prime bundle and therefore Amazon’s market power. A consumer may not want or use Amazon’s television, music or photo storage services, but she is paying for them. In short, everyone except Amazon appears to be paying to help Amazon lock in its customers and build a monopoly that competitors cannot match.
Other e-commerce platforms cannot afford such huge losses and so cannot compete on a level playing field. Companies who offer services that compete with a subset of the Prime bundle cannot compete with services that a consumer is already getting for “free”. And if the FTC is right about the anti-discounting policies, even consumers who do not buy from Amazon face higher prices when they shop online.
The scoping of the FTC’s case is very effective. It describes some of the most exclusionary conduct under a coherent umbrella narrative, and in doing so it ignores other types of conduct that may or may not be exclusionary or exploitative, but do not add to the chosen narrative. It aims at the most strategic issue of all – Amazon’s leading position in e-commerce – rather than addressing secondary issues such as its use of third-party sellers’ data. It seems to have some good evidence, although much of it is redacted. We will have to wait and see if the FTC pursues separation remedies or if it adopts a more behavioural approach like the UK and EU agencies. Of course, Amazon will have some good arguments in response and the FTC has not found it easy to persuade the US courts to adopt progressive competition policy.
From a UK and European standpoint, we do not know to what extent the same conduct exists on this side of the Atlantic. Certainly, the MFN policies described by the FTC would be much riskier in these jurisdictions where the substance of a policy takes precedence over its legal form, and it is not therefore relevant that a policy is not explicitly stated in a contract. Amazon has previously promised the European Commission that it would stop using MFN policies in the e-books sector, in commitments that expired in May 2022.
To some extent the FTC’s allegations have already been covered by the European Commission, for example withholding Prime-eligibility from sellers who do not purchase FBA services is already part of the European Commission’s case (which was settled through commitments last December) and the CMA’s case (which has recently consulted on some commitments).
I raised an eyebrow at the hyperbole in the FTC’s press release where John Newman, FTC Deputy Director, said: “[s]eldom in the history of US antitrust law has one case had the potential to do so much good for so many people.” This is just some typical American overstatement that is good fodder for the media, right? But actually, when we are talking about a company responsible for half a trillion dollars in consumer sales, and the allegation that it has stitched up the whole of e-commerce so that others cannot compete effectively and consumers overpay for products that are less well suited to their needs, perhaps he’s right?