The elephant in the room.
In 2018, following a super-complaint by the Citizens Advice Bureau, the CMA investigated the existence of a loyalty penalty for UK consumers across a broad range of markets. They described the concern as follows:
«In some markets, longstanding customers can pay much more than new customers for the same services – this is known as the ‘loyalty penalty’. This happens when suppliers charge higher prices to their existing customers, who they think are unlikely to switch to get a better deal. The result is that in many cases, people who stay with their supplier end up paying significantly more.
Many services are paid for through automatically renewed or rolled over contracts. While this can be convenient for customers, it also increases the risk that customers who get rolled over year after year will pay a loyalty penalty […] Overall, we have found that the loyalty penalty is significant and impacts many people, including those who can least afford it. Customers rightly feel ripped off, let down and frustrated. They should not have to be constantly ‘on guard’ or spend hours negotiating to get a good deal. This erodes people’s trust in markets and the system as a whole.»
The CMA concluded that practices such as “price walking” are unfair,2 and made a series of recommendations to regulators on remedies that might help to address the issues it identified but did not take any enforcement action.
Fideres believes that, while the CMA stopped short of denouncing this practice as a breach of competition law, these findings provide a real opportunity to take a class action on behalf of loyal consumers against the firms in question for abusing their dominance.
Recent work on inclusivity and antitrust has emphasised that to enforce competition law in an inclusive fashion requires public and private enforcement to explore whether the relevant antitrust market needs to be defined with reference to a specific group of consumers (OECD, 2018). This allows the investigation to ensure that harm to marginalised groups is identified and not mistakenly balanced out against possible benefits to other consumer groups that in fact participate in other markets. Any such benefits are therefore strictly out-of-market efficiencies that should not be used to justify the abuse of market power to exploit the marginalised group of consumers.
While this work has thus far focused on harm to women and ethnic minority groups,3 we note that the same logic can often be applied to groups of loyal consumers.4 In each case the question is whether a hypothetical monopolist could profitably apply a SSNIP to the product(s) that it sells to a specific group of consumers. If a firm can, and indeed does profitably price discriminate towards those customers then this price discrimination can indicate the existence of a separate market for the sale of the product to these customers. For example, a razor marketed as a women’s razor that is priced at a higher level may indicate the existence of a market for the sale of razors to women. After all the higher price level can itself be interpreted as a SSNIP that has evidently proved to be profitable. The same logic can be applied to credit that is offered to women at a higher cost based on otherwise identical information (see apple card).5
In the case of the loyalty penalty, it is evident that the incumbent firm knows the purchasing history of the customer and is therefore able to identify those that have been loyal, and for how many years.6 Arbitrage is often impossible in the services in question and so it is unsurprising that we see price discrimination across these groups.
The CMA, Ofcom and other regulators have in their analysis treated this as price discrimination within a market, which can, as is well-known, have ambiguous effects on aggregate consumer welfare. However, in our view a more rigorous market definition exercise would have clarified that instead these are in fact distinct markets. This makes the impact of excessive pricing, mergers and exclusionary conduct within these distinct markets much less ambiguous.
In particular, with a more specific market definition it becomes easier to see whether a firm is abusing its dominance within the market for loyal consumers by setting excessive and exploitative prices. Moreover, it can be seen that where it does so this unambiguously harms the consumer welfare of those consumers within this narrower market. Any out-of-market efficiencies that it creates by enabling the cross-subsidisation of products for other consumers in other markets becomes irrelevant (as it should be).
In some cases, it will make sense to cluster together markets for the sale of products to different customer groups, and to consider market shares across the cluster. This however should only be done, where the competitive constraints in those markets are the same and the firms each price discriminate. An analysis of competition between cinema’s for example will often group together the sale of cinema tickets to students, children, adults, and pensioners rather than considering them separately since the competitive constraints across cinema’s are likely to be common across those groups (unless some chains do not offer the same discounts to pensioner as others, in which case they will pose a weaker constraint on the price that rivals’ set for pensioners, though not their price for students and other groups).
We therefore believe there are a series of products for which the relevant antitrust market is the sale of the product not to all consumers, but only to those consumers that have been previously loyal to a specific firm. The firm in question therefore holds a clear dominant position in each market. Moreover, we would argue that many of these firms have taken advantage and abused that dominant position by setting exploitative and excessive prices.
In each case we would therefore propose work to assess whether these cases meet the ECJ ruling in United Brands that a price can be considered excessive when it has “no reasonable relation to the economic value of the product supplied”.7 In particular, we suggest applying the ECJ’s two-limbed test for the excessive nature of a price:
In our view, the large difference between the price charged to loyal customers and the average price for the same product demonstrates the excessive and unfair nature of the price. We note for example that while they have not taken enforcement action, the CMA and other regulators have identified these prices as unfair. We believe that any reasonable benchmark on cost, economic value, a normal profit margin, or the price of the same product in other markets would demonstrate the excessive nature of these prices. Therefore, we see significant scope for a series of abuse of dominance cases that can and should be taken in order to address the loyalty penalty issue and deter firms from continuing to target loyal and often vulnerable and marginalised groups through such practices in future. Such cases would exemplify an inclusive approach to private enforcement of competition law in the UK.
It is unfortunate that competition agencies have not as yet examined this angle during their work on the topic, and that other economics consultancies that represent defendants as well as plaintiffs have not advised the law firms and litigation funders of this possibility (indeed the firm that the CMA itself commissioned did not even mention market definition in its work on the topic).8 However this is perhaps unsurprising since the conflict of interest that such firms hold will inevitably mean that such arguments for narrow markets will prove inconvenient for them in other cases in which they are defending well-paid clients.
It is important to emphasise that, while a focus on marginalised groups is to some extent a new development in competition enforcement, the definition of markets with reference to a specific group of consumers is not. Instead, this has long been set out in best practice guidance in the US, the UK, and the EU, each of which are set out below.
In the FTC/DOJ merger guidelines (2010), section 4.1.4. sets out the US agencies approach to identifying price discrimination markets:
«If a hypothetical monopolist could profitably target a subset of customers for price increases, the Agencies may identify relevant markets defined around those targeted customers, to whom a hypothetical monopolist would profitably and separately impose at least a SSNIP. Markets to serve targeted customers are also known as price discrimination markets. In practice, the Agencies identify price discrimination markets only where they believe there is a realistic prospect of an adverse competitive effect on a group of targeted customers.»
The 2010 UK merger guidelines were, like the US guidelines, clear that a market could be defined in relation to a set of customers:
«The relevant customer market: when suppliers can target higher prices at those customers willing to pay more than others (ie price discriminate), the market may be defined by customer group, with the customers in each market being offered different terms.»
In contrast, the new 2021 UK merger guidelines place little emphasis on market definition. This reflects a sensible move towards focusing on competitive effects of mergers or conduct rather than market definition.9 It therefore identifies the SSNIP test as a framework and says it may aggregate markets if the main parameters of competition are set uniformly across those markets. However it does not set out any guidance on the circumstances in which it will identify price discrimination markets.
The EU’s Market Definition Notice is currently being updated, however until revisions are released, the 1997 version continues to apply. The 1997 Notice says in summary that the Commission will take account of: «different categories of customer and price discrimination: a distinct group of customers may constitute a narrower, distinct market when such a group could be subject to price discrimination.»
It then specifies that: «The extent of the product market might be narrowed in the presence of distinct groups of customers. A distinct group of customers for the relevant product may constitute a narrower, distinct market when such a group could be subject to price discrimination. This will usually be the case when two conditions are met: (a) it is possible to identify clearly which group an individual customer belongs to at the moment of selling the relevant products to him, and (b) trade among customers or arbitrage by third parties should not be feasible.»
We note that firms are of course able to quickly identify those customers that have in the past been loyal and hence less likely to switch in response to a higher price.
The CMA report the estimated size of the loyalty penalty across a range of different markets.
Citizens Advice estimate that 12.4 million loyal customers were paying £57 on average more than new customers (£709million each year). Then in 2020 the FCA went further and identified that in 2018, 6 million loyal policy holders would have saved £1.2 billion had they paid the average price for their actual risk. Since the average price cannot sensibly be below cost, this serves as a more cautious benchmark on the possible overcharge. In May 2021 the FCA confirmed that new rules would be established to eliminate these pricing practices by prohibiting price discrimination on the basis of loyalty.
Citizens Advice estimate that 11 million loyal customers were paying £113 more than new customers (£1.13billion each year). Ofcom estimated this at approximately £675million each year. It then in July 2020 identified that out-of-contract customers paid just under £500 million more each year, compared to the average price for the same service from the same provider. Again, since the average price cannot sensibly be below cost, this serves as a more cautious benchmark on the possible overcharge.
Ofcom choose not to make a regulatory intervention and instead sought voluntary commitments from firms to reduce the difference in price. It hopes that these voluntary steps might reduce the overpayment by loyal customers by approximately £270 million per annum. However, even if this appeal to the firms to end the exploitation were successful, this would still leave them paying approximately £230 million each year more than the average price for new customers.
The FCA estimate that 800,000 loyal customers were paying £1000 for the first two years, and £100 more in subsequent years than new customers (£800 million each year). It amended regulations to make it easier to switch in October 2020 though it remains unclear whether this will address the problem.
The CMA conducted a detailed market investigation into the energy market in 2016 which found that 70% of domestic customers of the six largest energy suppliers were on expensive ‘default’ standard variable tariffs (SVT) and could potentially save over £300 by switching to a cheaper deal. This worked out at £2billion in 2015. A market wide price cap was implemented in January 2019 following legislation introduced by the government. However, unlike the FCA ban on loyalty based discrimination, the price cap permits firms to continue to price discriminate between loyal and new customers. It is unclear how much more a loyal customers will pay compared to a new customer.
The CMA and regulators have emphasised that the difference between the price for new customers and loyal customers is not necessarily a reliable estimate of the overpayment. For example, a firm might be setting prices at a level that is below-cost for new customers in the expectation that some will turn out to be loyal, and hence will in future choose to pay higher prices in later years.
Such below-cost selling can make it difficult for entrants to compete in this market for new customers if they do not have the same ability to cross-subsidise (and hence recoup) from excessive prices charged in the market for sales to loyal customers (at least until a later date).
To deal with the possibility of below cost pricing the FCA take a more cautious approach and estimate the difference between the price paid by loyal customers and the average price. Since the firms should not set an average price that is below cost, it should be safe to assume that this covers costs. However, it does of course include the excessive prices themselves and so may still represent an underestimate (or a lower bound) on the overcharge to loyal consumers.
In cases where a large loyalty penalty is set, the application of well-established, internationally recognised best practice principles of market definition would lead to the definition of a narrow market for the provision of the service to loyal consumers alone. This narrow market definition clarifies that in many of these services the incumbent firm is in a dominant position in the market for sales of the service to consumers who have previously been loyal towards them. We argue that by setting a substantial loyalty penalty, an incumbent is therefore exploiting its dominance by charging an excessive price. We believe that this practice in a series of markets may therefore constitute an exploitative abuse of a dominant position. We propose that these are therefore assessed against the relevant two-limbed test for excessive pricing.
1 ‘Price Walking’ is the practice of applying successive price rises each year, as customers demonstrate their loyalty.
2 See CMA (2018), paragraph 8.41,
4 As noted, these groups may often be marginalized or vulnerable in some sense. Indeed, as the CMA note, the most vulnerable in society can have particular challenges engaging in markets, such as those on low incomes, people who struggle to use online services, or people with poor mental health who may avoid or fear change. This means they may be more at risk of paying the loyalty penalty and may be least able to afford it.
5 See allegations made in relation to Apple Card and Goldman Sachs: Apple’s ‘sexist’ credit card investigated by US regulator – BBC News
6 Indeed this ability to profile and identify loyal consumers and target higher prices at them is only likely to increase in future in digital markets (see OECD, Price discrimination, 2016). In this context establishing a framework for assessing market definition will be an increasingly important part of putting inclusive competition enforcement into practice.
7 United Brands, ECJ, Case 27/76, paragraph 250.
9 OECD, Rethinking the Use of Traditional Antitrust Enforcement Tools in Multi-sided Markets, 2018
Chris joined Fideres in 2021. Chris holds a PhD, an MA and BA in Economics from the University of East Anglia. At Fideres Chris has provided expert economic advice on class action complaints against Amazon, Facebook and Apple. He has written expert reports, developed models to quantify damages, and developed analysis of market definition and abuse of dominance (monopolisation) in digital aftermarkets and multi-sided platforms.
Before joining Fideres, Chris spent 7 years as a Competition Expert for the OECD where he led the economic thinking on antitrust in digital markets, as well the role for competition law in delivering inclusivity. He published numerous papers and led a working party of the OECD Competition Committee in developing new international standards on competitive neutrality and competitive assessment in light of the digitalisation of the economy.
Chris has advised the UK Government’s Department of Trade & Industry on the benefits of competition policy, and the UK Competition Commission (predecessor to the Competition and Markets Authority) on digital mergers, retail market investigations and competition cases. He was an advisor to the Co-operation and Competition Panel on mergers, market studies and antitrust in publicly-funded healthcare markets, and later became Director of Competition Economics at the UK Healthcare Regulator. Chris is a founding member of the Centre for Competition Policy of the University of East Anglia. He remains an associate of the Centre, a member of various advisory boards at non-profit making organisations, and peer reviews papers for the Journal of Competition Law and Economics & the Journal of Antitrust Enforcement.
The elephant in the room.
Evidence of collusive behaviour In academic publishing.
Fideres investigates algorithmic mortgage discrimination in the US.