In the emerging digital age, companies are increasingly using and developing pricing algorithms to help them determine optimal prices on a near real-time basis. When used lawfully, pricing algorithms can lead to greater competition, reduced cost, and reduced barriers to market entry. In this way, there is potential for greater consumer efficiencies with access to “real-time” prices for consumers based on prevailing market conditions.
However, pricing algorithms cannot be used to engage in unlawful practices that would breach competition law if carried out by a real person. There is a risk that using an algorithm provider could result in price-fixing behaviour between competitors. Price-fixing is a form of collusion that happens when two or more competitors act together to raise, lower, maintain or stabilise prices. When a small number of algorithm providers can influence a large segment in a particular market, there is a greater chance that the algorithm provider can be used to facilitate such collusion. In those circumstances, any consumer efficiencies that might otherwise be brought about by using pricing algorithms are undermined by the fact that consumers are limited in their ability to “shop” around for the best prices.
A lesson from across the pond
The impact that algorithmic pricing can have on competitive outcomes in markets has been closely followed over the last decade in the UK and beyond, particularly with the rise in the use of artificial intelligence. Last month, the Federal Trade Commission and the Antitrust Division of the Department of Justice filed a statement of interest (accessible here) outlining that hotels cannot use algorithms to collude on room pricing.
The statement relates to the case of Cornish-Adebiyi, et al. v. Caesars Entertainment Inc., et al., a class action filed by a group of New Jersey residents that allege several hotels engaged in an illegal price-fixing conspiracy using a common pricing algorithm to set hotel room rates. Price-fixing is unlawful in the US under Section 1 of the Sherman Act. In the statement, two aspects of competition law are drawn out:
- First, there need not be direct communication between competitors to allege a competition law infringement under Section 1 of the Sherman Act. Fundamentally, competitors cannot cooperate to set their prices, whether via staff or an algorithm.
- Second, an agreement to use shared pricing recommendations, list prices or pricing algorithms is still unlawful even when those competitors retain some pricing discretion. Setting or recommending initial starting prices can still breach competition law even if those prices are not the prices that consumers ultimately pay.
Whilst the US example is not precedent for UK courts, there is clear potential for a similar case to arise in the UK. And, as we discuss below, the hotel online booking sector is not a new target for the Competition and Markets Authority (CMA).
Algorithms and UK competition law
The US case draws on the fact that ‘horizontal’ price fixing between competitors is illegal under US antitrust law. Price-fixing is also unlawful in terms of UK competition law. Chapter I of the Competition Act 1998 in the UK is broadly similar in its effect to Section 1 of the Sherman Act in the US. Chapter I prohibits agreements between undertakings, decisions by associations of undertakings, or concerted practices which may affect trade within the UK and have as their object or effect the prevention, restriction, or distortion of competition within the UK.
Importantly, concerted practices are within the scope of Chapter I, which means that, like the US case, there need not be evidence of direct communication between competitors to allege an infringement of Chapter I. A form of coordination between undertakings without an agreement necessarily having been reached will be sufficient to breach Chapter I where the other elements of the Chapter I prohibition are met. Similarly, the impact on competition need not have been the “intention” between a set of competitors. It is sufficient that the “effect” of the behaviour impacts competition in the UK.
This means broadly that, where competitors use the same algorithm provider to set their prices in the UK, there is a risk that they could fall foul of competition law (whether in the hotel online bookings sector, or another sector).
When used appropriately, pricing algorithms can be employed by businesses independently to monitor competitors’ prices and inform their own price setting. This can lead to more intense competition, reduced costs and/or barriers to entry. Naturally, if a business is aware that its competitors are charging lower prices for similar goods or services, they are encouraged to lower their prices to remain competitive. This has efficiencies for consumers who can, in turn, have access to a wider range of goods or services at lower prices.
That said, it has been long understood in UK competition law terms that algorithmic pricing can also be used unlawfully as a tool to facilitate collusion amongst competitors. Where third parties offer tools and algorithms to businesses to help them to set and manage their prices, or even offer to autonomously set prices on behalf of those businesses, there is a risk that this could create a ‘hub-and-spoke’ structure, i.e., a structure where information is not exchanged directly between competitors but instead disclosed to a third party in order to reduce competitive uncertainty. Pricing algorithms can also more broadly allow businesses to collect detailed data relating to their competitors and can in turn be used to increase the predictability of competitors’ pricing behaviour.
On this, the CMA is clear that, at a minimum, if a pricing practice is illegal when implemented offline, there is a high probability that it will also be illegal when implemented online. In addition, the CMA is clear that businesses involved in illegal pricing practices cannot avoid liability on the grounds that their prices were determined by algorithms. In this way, an algorithm remains under the control of the business using it, and therefore that business is liable even if its actions were informed by algorithms.
Potential wider harms resulting from the use of pricing algorithms
The CMA has also historically looked at how algorithms can reduce competition and harm consumers, noting that potential harms from the use of algorithms are “wide ranging” and include search algorithms as well as pricing algorithms (see the CMA’s research and analysis piece from January 2021 here).
These include personalised pricing and rankings harms, which occur when businesses use algorithms to advertise different prices to different consumers, and to display results in a different order depending on an individual user’s location, previous queries, and search/purchase history. In those instances, consumers cannot always observe exactly how or why search results are presented in a particular order. There is, therefore, potential for businesses to manipulate consumers into making different decisions that are more profitable for the business but which the consumer would not have made under more objective and/or neutral conditions.
Previous CMA scrutiny of pricing practices in hotel online booking sector
The hotel online booking industry is also not a new target for the CMA. In 2019, the CMA secured undertakings from several online accommodation booking sites, each of whom agreed to clarify how hotels are ranked in search results, including by explaining how the results have been impacted by the amount of commission a hotel pays them. They also pledged not to give a false impression of the availability or popularity of a hotel, or to rush consumers into making a booking decision based on misleading information. These undertakings also signed up to the CMA's principles for businesses offering online accommodation booking services (accessible here). It includes, amongst other things, an expectation that where the receipt of commission affects how search results are ranked, the hotel booking website will explain this to the consumer.
With the CMA set to obtain new consumer law powers under the Digital Markets, Competition and Consumers Bill (the DMCC Bill) expected to come into force later this year, businesses engaged in hotel online booking activities (and other sectors more widely) can expect increased scrutiny around the use of algorithms from a consumer law perspective.
Further thoughts
Going forward, a challenge for businesses will be the extent to which algorithmic pricing is becoming increasingly complex, particularly those systems involving machine learning algorithms. There is a risk in those circumstances that harms may not be perfectly anticipated by the businesses that use them.
Digital markets are dynamic and algorithmic systems change at a fast pace, so new harms can manifest quickly. The CMA is clear that where a business uses an algorithm, it remains responsible for how it is used from a competition law perspective. There is, therefore, a challenge for businesses to ensure they understand and use pricing algorithms in a way that keeps them on the right side of competition law.
In the context of new consumer law powers for the CMA contained in the DMCC Bill, it will also be important for businesses to maintain a strong focus on ensuring transparency and efficient outcomes for consumers in the face of increasingly complex algorithms.
If you would like to discuss the contents of this article in more detail, please get in touch with a member of our Regulation and Markets team.
This article was co-authored by Trainee Emma Hill