I recommend reading the UK Competition Appeal Tribunal’s (CAT) decision in Flynn/Pfizer vs UK Competition and Markets Authority (CMA) as both lawyers and economists can find a fair number of details to note for later cases related to excessive pricing. In this short article I do not wish to argue whether the CMA’s methodology of assessing the excessiveness and fairness of pricing (and the derived fine) was correct. Rather I question the concept of unfair pricing and dispute that the CAT cannot dismiss the estimation of competitive prices and, at the same time, accept a product market definition.

In summary, the CAT quashed the fines the CMA imposed on Pfizer (GBP 84.2 million) and Flynn (GBP 5.2 million) due to excessive pricing on the epilepsy drug Phenytoin Sodium Flynn, that is, over-pricing consumers as dominant drug-suppliers. Since 2012 Pfizer is the manufacturer, and Flynn is the supplier of the product. The CAT’s main allegation is that the CMA incorrectly applied the legal test for excessive pricing: (i) the estimation of Phenytoin Sodium Flynn’s economic value was not substantiated enough and (ii) there was insufficient assessment of prices of comparative products. At the same time the CAT accepted the CMA’s product market definition, according to which the drug Phenytoin Sodium Flynn is a product market in itself. The CAT also accepted that Pfizer/Flynn hold a dominant position in that market.


Unfair pricing

In the context of pricing practices of dominant companies, the CMA and the CAT claim – based on existing case law[1] – that a product is unfairly priced if the price “bears no reasonable relation to the economic value”.[2] Unfair prices is a legal term derived from the special responsibility a dominant company has. However, it lacks economic meaning.

One of the issues with economic value is that it is not obvious how to estimate it. The CMA’s assessment implies that the economic value is a function of production costs (cost-plus approach). However, even for very simple products it is difficult to calculate economic value. Let’s take the example of a potato producer. Is the economic value of potatoes the sum of the values of input materials, land rental and human workforce? How about the knowledge of when and how to plant, treat and harvest? Furthermore, how could the producer allocate common cost elements to individual products (for example, if the same infrastructure is used for producing potatoes, onions, and sunflowers)?

Therefore there are at least two problems with the estimation of economic value:

  • What is the cost estimation for intellectual property and other intangible assets?
  • How can we allocate these costs and other common costs to individual products (or services)?

The difficulty with measuring certain cost elements and cost allocation may lead to substantial underestimates of economic value, hence, an underestimation of fair prices.

The other, and probably more fundamental, issue is that the concept of “fair prices” is defective. Rationally behaving companies will always seek to maximize profits. This objective has nothing to do with setting prices that bear a reasonable relation to the economic value of the product. Turning back to the example of the potato producer: even if (s)he knew the exact economic value of their product (which is unlikely as set out above) why would (s)he sell it at a price that has a relation to this economic value? (S)he would rather set the price at a level based on supply conditions (other potato sellers, substitute products) and demand for potatoes. Rationally behaving profit-maximizing companies (including dominant firms), that have a fair knowledge about the market in which they are active, would never follow a cost-plus approach.

It is important to understand the implications of the above: prices can be competitive or excessive (that is, excessing competitive levels) depending on the level of competition in the market, but it is difficult to grasp the economic meaning of unfairness. Motta is also sharing this view.[3] In any case, fair prices based on a cost-plus approach with a reasonable margin is likely meant to be a close approximation of competitive prices. Hence, I presume that the CMA’s interpretation of fair prices is closely aligned with competitive prices. This is supported by the CAT decision which sets the requirements of analysing prices of comparative products as part of a comprehensive excessive prices assessment.


Product market definition

One piece of evidence given by the CMA/CAT for defining the product market narrowly is that Flynn did not consider NRIM as a significant competitive threat.[4] During the cross-examination, Flynn’s witness Mr Walters said that “[…] NRIM’s commercial strategy is not generally to start a “race to the bottom” on price but rather to build up a 30-50% share of the market.[5] This could be seen as evidence of the lack of competitive pressure exerted by NRIM, but aiming to build up a 30-50% share of the market is, in my view, a fierce enough competitive threat. NRIM also provided the information that they saw a market entry opportunity to “establish NRIM as the only generic competitor to Pfizer and to carve out a market share of around 35-40% in total sales of Phenytoin Sodium capsules in the UK.[6] This is a clear indication that NRIM’s objective was to compete for Pfizer/Flynn epilepsy patients.

The objective of NRIM to carve out 35-40% share indicates that, according to NRIM, there is a market for Phenytoin Sodium capsules in the UK including both Pfizer/Flynn and NRIM Phenytoin Sodium capsules. In contrast, the CMA defined the relevant product market to be Pfizer-manufactured phenytoin sodium capsules only (that is, Phenytoin Sodium Flynn).

To understand the CMA’s stance on product market definition we need to look into the details of the case. First, the CMA revealed that the average selling price of Phenytoin Sodium Flynn was significantly higher than the price of NRIM’s product throughout the whole infringement period. Note that the price difference between products that potentially belong to the same product market does not reveal anything about the competitive interaction between them. Second, Pfizer/Flynn did not decrease the price of Phenytoin Sodium Flynn after the market entrance of NRIM in any material extent and later on (from April 2013) the price decrease of Phenytoin Sodium Flynn was not motivated by competition from NRIM. Third, and probably of most significance, in November 2013 the Medicines and Healthcare Products Regulatory Agency (MHRA) raised concerns about switching between anti-epileptic drugs of different manufacturers, which influenced NRIM’s initial success of gaining volumes.

According to the CMA, Pfizer/Flynn were able to set the price of Phenytoin Sodium Flynn independently from the pricing decisions of NRIM. The fact that NRIM managed to gain 20-30% share of the 100mg phenytoin sodium capsules sales by November 2013,[7] that is, within eight months after launching the product, contradicts this statement. The significant volume-switch within a short period of time reveals a competitive interaction between the products of Pfizer/Flynn and NRIM – at least until the publication of the MHRA guidance.

Whether this competitive interaction is strong enough to place them within the same relevant product market is dependent on the results of the SSNIP-test (Small but Significant and Non-transitory Increase in Price): would a hypothetical monopolist find it profitable to increase prices of a product (or products) by 5-10% from its competitive level? If the answer is yes, the product (or products) of the hypothetical monopolist is the relevant product market. The test was, unfortunately, sidelined by the CMA.

It is rightly pointed out by the CMA that there is a risk that the observed prices in the market are already significantly higher than the prices would be at competitive level[8], hence, they cannot be used as the basis for a SSNIP-test. This is an obvious point to make but the main criticism of the CAT, in the context of excessive pricing, is that the CMA’s estimation of the “but-for” competitive price was not substantiated enough (insufficient assessment of prices of comparative products). This criticism also holds in the framework of product market definition – which also requires the estimation of competitive prices.

Now, if the CAT requires a proper assessment of the “but-for” competitive price with respect to excessiveness and unfairness (see above), then it is difficult to understand why this “but-for” competitive price, once properly estimated, cannot be used for product market definition as well. In particular, the difference between the “but-for” price and the price that Pfizer/Flynn set for Phenytoin Sodium Flynn may be less than the SSNIP threshold 5-10%. This would disprove the conclusion of the CMA that the relevant product market includes Phenytoin Sodium Flynn and nothing else and would turn the case on its head.


Final thoughts

If the CMA and/or the NHS are concerned about the dominant position of Pfizer/Flynn, and they can show that prices are excessive, the relevant authority should interfere, and the market should be regulated. This can be achieved by direct market intervention like price caps but, in certain circumstances, promoting market entry may also be beneficial (and probably the first-best solution).

The CMA did not opt for this path but, instead, asked Pfizer/Flynn to set prices at a fair level from now on and fined them for their past behaviour. Authorities should be extremely careful in fining companies for excessive pricing for two reasons. First, as shown above, the dominant company is simply following a profit-maximizing motive. Second, to impose a proportional fine the authority should have a precise estimate of the price that would have emerged under competitive conditions. This is an extremely difficult exercise (as one should note after reading the CAT’s decision).

[1] Judgment of the Court of 14 February 1978, United Brands Company and United Brands Continentaal BV v Commission of the European Communities, Case 27/76, EU:C:1978:22.

[2] CAT Judgment, para 443.

[3] M. Motta, Competition Policy, Theory and Practice, Cambridge University Press, 2004, 12th printing, p69.

[4] See e.g. CAT Judgment, para 169.

[5] CAT Judgment, para 185.

[6] CAT Judgment, para 188.

[7] CMA Decision, 4.11.

[8] CMA Decision, 4.34.


To make sure you do not miss out on regular updates of the Kluwer Competition Law Blog, please subscribe to this Blog.


Leave a Reply

Your email address will not be published. Required fields are marked *