An Economist’s Perspective on Exploitative Abuse of Dominant Position and Excluding Competitors

Introduction

This chapter addresses the following competition issues, from an economics perspective, as they relate to the pharmaceutical industry:

• exploitative abuse of a dominant position through excessive pricing; and
• excluding actual or potential/nascent competitors via:
  • rebates and discounts; and
  • killer acquisitions.

Excessive pricing

The European Union, the UK and many other jurisdictions can use their powers against abuse of dominance to act against the exploitation of a situation of weak competition, even if the firm under investigation did nothing to create that situation. These investigations of ‘exploitative abuse’ are almost always focused on excessive pricing.

Many economists have expressed concern that prosecuting excessive pricing can be economically harmful, particularly because it can create uncertainty for investment incentives. Perhaps for that reason, enforcement of the excessive pricing prohibition has been rare, particularly at EU level. However, there has been a flurry of cases in the pharmaceutical sector recently.^[2] To illustrate the economic principles involved in assessing excessive pricing, we focus on (1) the UK’s Phenytoin case concerning Pfizer and Flynn’s pricing of the branded off-patent drug Epanutin,^[3] and (2) the European Commission’s (EC) investigation of Aspen’s ‘Cosmos’ portfolio: six off-patent oncology drugs, which was settled with commitments in 2021.^[4]

Most recent cases in the pharmaceutical sector involve large increases in price, often accompanying a change of owner. In the UK, the Competition and Markets Authority (CMA) found that the price of Epanutin increased by around 2000 per cent after Flynn took over distribution from Pfizer. Price increases in the Aspen case varied by product and country but the EC described them as ‘often by several hundred percent’. Even in the US, although federal competition law cannot deal with such cases, similar concerns have led to the use of state and consumer protection laws to control prices, especially in emergencies.

The covid-19 crisis highlighted such concerns, particularly in the early days when masks and other personal protection equipment (PPE) were in short supply. It might be thought competition law should deal with this, but in fact Article 102 of the Treaty on the Functioning of the European Union (TFEU) and its equivalents would have been slow and to show dominance would have been difficult. Typically, concerns involved a small independent supplier buying up a stock of PPE and selling it on online marketplaces at very high prices. Instead, consumer protection or bespoke legal and regulatory action was typically used to control these price spikes, for example, by executive order in the US^[5] and by additional legislation in France (controlling prices of hydroalcoholic gel) or Portugal.^[6]

Indeed, excessive pricing prohibitions under Article 102 TFEU and its equivalents are not ‘quick fixes’. The EC settled its Aspen case with commitments after two to three years rather than taking it through to a decision. The CMA in the UK faced (and essentially lost) a series of legal appeals to its initial Phenytoin decision. This might seem surprising – as one of the UK Appeal Court judges noted:

The CAT’s decision is in my view somewhat repetitive. It was quite easy to lose sight of a stark reality, which was that, literally overnight, Pfizer and Flynn increased their prices for phenytoin sodium capsules by factors of between approximately 7 and 27, when they were in a dominant position in each of their markets. That did not, of course, abrogate the need for a rigorous reasoned approach to the legal and factual questions before the CAT, but it was important to keep in mind.^[7]

Why is it so difficult? The economic analysis in an excessive pricing case can be surprisingly complex. In the EU’s United Brands framework (also relevant in the UK), a competition authority must show:^[8]

• a relevant market;
• dominance within that market;
• an ‘excess’ of prices over costs; and
• that this excess be considered ‘unfair’, either:
  • in comparison to the price of ‘competing products’; or
  • ‘in itself’.

All five of these require economic analysis.

Market definition and dominance

Defining the market and establishing dominance are not inherently different in excessive pricing cases from other abuse of dominance cases. However, features of pharmaceutical markets can introduce added complexity, for example because of differing approaches to clinical substitutability between jurisdictions and regulatory intervention. This can be particularly difficult for the EC because health policy is a national competence and markets are generally national. The EC’s Preliminary Assessment found Aspen dominant in ‘most or all’ of the relevant markets and did not find prices above cost in all of them, for example.

United Brands Limb 1: identifying the excess of price over costs actually incurred

Prices must be compared with all the costs actually incurred by the company to bring the product to market, including (1) the costs directly incurred in supplying the product or service; (2) an apportionment of indirect costs; as well as (3) a reasonable return.

What proportion of fixed and common costs should be allocated to the products under investigation? Economic principles do not provide a simple answer to this question. In a market, a product’s price might cover any proportion from none to all of a company’s fixed costs.

Costs allocated for the purpose of assessing excess must at least include all of the incremental costs specific to the product under investigation: if Flynn did not supply phenytoin, how much lower would its costs be in the long run? However, that benchmark is surely too low as it could find any contribution to overheads to be excessive.

At the other extreme, one could construct a stand-alone business supplying only the products under investigation, which would therefore have to cover all the necessary overheads just from the sale of the products. Such a stand-alone supplier’s unit costs could be very high in some cases, so using this as a benchmark might make it difficult to find an excessive price. This might strike competition authorities as an unreasonable hurdle. However, such an approach cannot be ruled out on economic principles (indeed, some pharmaceutical firms do make almost all their revenue – and all of their profits – from one blockbuster drug).

Suppliers themselves often simply do not allocate these costs. All businesses routinely price some products to make little or no contribution to fixed costs, while pricing others for which the elasticity is lower at a higher price. Economists call this ‘Ramsey pricing’; businesspeople might call it charging what the market will bear. However, competition authorities and courts have often rejected this principle in excessive pricing cases.^[9]

In practice, therefore, any approach adopted by a competition authority to cost allocation is likely to be arbitrary. In Phenytoin, for example, the CMA allocated cost across product types according to packs sold.^[10] This is not only arbitrary in the sense that it has no economic justification, it is arbitrary even as a unit because a ‘pack’ of tablets is not the same as a ‘pack’ in the form of a bottle. In Aspen, the EC allocated different kinds of costs on the basis of different ‘keys’ across Aspen’s multinational business. There is no ‘right’ way of doing it that has economic meaning. A prudent competition authority will therefore use several methods – and perhaps take action only when the price-cost gap is sufficiently large for all of them.

As for return on capital, there are well-established techniques from utility regulation to estimate this, but in practice these techniques are harder to apply without the ‘regulatory accounts’ required by utility regulators. The CMA in Phenytoin used a 6 per cent return on sales (ROS) reasonable rate of return measure instead, based on the average rate in the UK’s scheme for regulating drugs prices, having rejected Pfizer’s submissions that a relevant benchmark is provided by other generic pharmaceutical companies. In its remittal decision, the CMA increased the reasonable ROS on Pfizer’s products to 10 per cent based on the average ROS of the business units managing supply of Pfizer’s products.^[11] For Flynn, a distributor rather than a manufacturer of phenytoin, the CMA found that a return on capital employed (ROCE) approach was feasible and applied that in preference to an ROS measure.^[12] In contrast, the EC assessed Aspen’s prices using a 23 per cent ROS margin,^[13] based on the median of comparator pharmaceutical companies.

Both of these measures suffer from the problem that they apply average, portfolio-level profit rates to profitability calculated for a single product or smaller group of products. The CMA recognised this in Phenytoin and accepted that generic pharmaceuticals could in principle earn margins above 6 per cent, for example to recover significant investment or if there were significant commercial risks, but it decided that in this case no such factors applied.

United Brands Limb 2: is that price-cost gap ‘excessive’?

How large a gap between prices and a sufficiently inclusive measure of cost should be considered ‘excessive’? High prices act as important signals. They signal to consumers the value of looking elsewhere for alternatives and they signal to producers the value of launching competing products. An incorrect decision is likely to deter investment and innovation. There must therefore be a gap: a price just above the cost measures set out above is unlikely to be considered excessive.

The United Brands judgment itself defines excessiveness as a price that bears no reasonable relation to a product’s ‘economic value’. However, this phrase has little meaning to economists. The normal meaning of ‘economic value’ would be the price the supplier could charge, but this would then imply no price can ever be excessive provided that purchasers pay it. The notion of a ‘fundamental value’ exercised early economists (from the French physiocrats who thought all value derived from land, to Marx and his ‘labour theory of value’) but more recently economists see value as contingent on supply and demand, rather than fixed. A litre of water has no value in a flood but it has almost infinite value to a stranded traveller in the desert.

This has not stopped competition authorities considering a range of possible factors that they assess under ‘economic value’, under two headings.^[14]

Is the price unfair ‘in itself’?

EU authorities and courts have assessed unfairness by considering whether prices were unfair ‘in themselves’. This is not a well-defined test and in practice seems to involve considering whether there were any good reasons for the price to be as high as it was, and if not, concluding it was unfair ‘in itself’. In Phenytoin, the CMA noted in its first decision:

• it had found no additional factors that might justify the excess of prices over cost;
• the substantial gap between prices and costs made the prices unfair in themselves;
• the prices could only be sustained because the companies were shielded from competition; and
• the excessive prices had adverse effects on the National Health Service and its resources (a finding with many parallels, such as the Italian Competition Authority’s Aspen decision in Italy).

Of these four, only the first criterion seems to provide an additional test to the analysis the CMA had already conducted. The second follows from the finding of an excess of prices over cost, the third from a finding of dominance and the fourth seems to reflect public policy concerns.

On remittal, the CMA amended its assessment and noted that:

• the fact that old, generic drugs are essential for some patients is of limited relevance to their economic value, given that these drugs are at a stage of their life cycle where the degree of competition should be the primary driver of prices;
• there was no identifiable enhancement to the product that might have justified a price increase;
• there were no additional factors that might justify the excess of prices over cost; and
• the excessive prices had adverse effects on the National Health Service and its resources.

Similarly, the EC provided something of a laundry list of ‘in itself’ factors in Aspen, including:

• the nature of the products (old and off-patent);
• no improvements in quality or other reasons for the price increase;
• the disproportionality of the price increase, compared to cost;
• the magnitude of Aspen’s profits; and
• Aspen’s strategy to increase prices.

This approach is similar to the CMA’s in its first Phenytoin decision, with the addition of the last criterion reflecting the EC’s concern that Aspen had negotiated unfairly with health regulators (a concern mirrored – and more prominent – in the earlier Italian investigation of Aspen).

Is the price unfair in comparison to ‘competing products’?

As mentioned above, modern economics generally does not recognise any concept of ‘inherent value’, so instead it makes sense to assess excess by comparison with the price of similar products in reasonably competitive markets. The United Brands test allows for this, although rather confusingly it refers to ‘prices of competing products’ which cannot mean the price of products directly competing with the product under investigation (because the supplier must be dominant) but instead prices in other markets – usually other geographical markets or markets for similar products. When investigating price increases, competition authorities have sometimes taken the pre-increase price as a comparator, as the CMA did in Phenytoin.^[15] On remittal, the CMA did not put forward any comparator evidence such as the pre-increase price, relying only on the ‘in itself’ test.^[16] However, to address criticisms from the court of appeal, the CMA addressed the evidence advanced by Pfizer and Flynn in relation to phenytoin tablets and other anti-epilepsy drugs as comparators, finding they did not undermine the CMA’s conclusion under the ‘in itself’ test.^[17]

As economists, we suggest that competition authorities could be more creative in this area. There is no obvious logic to require the comparator to perform a similar clinical function, as the CMA and EC have required. A drug with similar costs of production, supplied under reasonably competitive conditions would seem to provide a reasonable comparator, no matter what condition it treats. However, if anything the authorities seem to prefer to avoid comparators, perhaps because the ‘in itself’ test provides a more discretionary alternative. Both the Phenytoin and Aspen decisions considered that there were no suitable competing products to take as comparators. The Competition Appeal Tribunal strongly criticised the CMA’s reasoning here: the CMA had not considered the price of phenytoin capsules, as it must do before falling back on the ‘in itself’ test.

Conclusions

Like many other economists, we are sceptical about the value of penalising excessive pricing with fines.^[18] The tests applied in practice by competition authorities seem to involve a lot of discretion, creating uncertainty. We are not aware of any competition authority that has published guidelines on excessive pricing: the CMA’s published advice for business makes no reference to it, even though the fines it imposed in Phenytoin were the largest in its history.^[19]

We do see a role for assessing excessive pricing and measuring profitability to identify constraints on competition, whether leading to enforcement action (if the constraints arise from conduct) or perhaps regulatory change. Indeed, the CMA’s investigations led to a change in UK law to eliminate a pricing loophole,^[20] which seems a valuable result.

However, the law is as it is and competition authorities can and do enforce it, sometimes feeling pressure from public opinion. As with any other competition investigation, good economic analysis can help companies under investigation – or third parties – bring relevant evidence to the authorities’ attention and if need be, challenge the authorities’ own analysis.^[21]

Rebates and discounts

Discounts and rebates are both means of reducing the net prices paid by purchasers. Discounts are typically deducted from gross prices at the time of discounting. Rebates are typically deducted at a future date, after invoicing, and may be applied on top of discounts. Both can often be pro-competitive or competitively neutral, as we discuss here before addressing situations in which discounts and rebates have the potential to harm competition, illustrated with reference to three specific recent cases.

Discounts and rebates in the pharma sector

Discounts and rebates are a means of reducing the net prices paid by purchasers. Manufacturers may discount when supplying wholesalers, hospitals, pharmacies or other purchasing bodies. Wholesalers may also discount when supplying their customers (hospitals or pharmacies). While prices paid by purchasers could be reduced by lowering list or gross prices, discounts and rebates are prevalent in the pharmaceutical sector because in many situations manufacturers might have to reduce other prices if they reduce list prices. For example:

• in many European countries (e.g., France, Portugal and Ireland), the amount reimbursed by payors to manufacturers is based on ‘reference pricing’; and
• the list price of existing drugs may also be a reference point when new innovative drugs are introduced.^[22]

Because discounts and rebates have the effect of lowering the prices paid by those who pay for medicines (which, in most cases, are not directly the patients that receive the medicines), in many circumstances this will be pro-competitive. They may, for instance, pass on cost savings by manufacturers or distributors when selling greater volumes.

When can discounts and rebates harm competition?

The primary competition concern associated with discounts and rebates is whether they are used to limit competition, for example, by excluding rivals, by foreclosing entry or expansion by new generic or biosimilar drugs after patents on an original drug expire.

The EC’s 2008 ‘Guidance on the Commission’s Enforcement Priorities in Applying Article 82 of the EC Treaty to Abusive Exclusionary Conduct by Dominant Undertakings’^[23] discusses anticompetitive foreclosure and price-based exclusionary conduct. On 27 March 2023, the EC amended the 2008 Guidance and announced plans to issue guidelines on exclusionary abuses.^[24]

First, the firm in question needs to hold a dominant position. For this reason, as with excessive pricing discussed above, market definition and an assessment of market power and dominance is the first step in assessing whether there is risk that a price-based exclusionary abuse could occur or has occurred.

Second, to distinguish whether vigorous price competition (which is generally beneficial) is capable of foreclosing competitors, the EC explains that it will generally intervene if the pricing conduct is capable of ‘hampering competition from competitors that are considered to be as efficient as the dominant undertaking’^[25] – in other words, if an ‘as efficient competitor’ (AEC) would be foreclosed. The EC has since clarified that it ‘recognises that in certain circumstances a less efficient competitor may also exert a constraint which should be taken into account when considering whether particular price-based conduct leads to anticompetitive foreclosure’.^[26]

The status of the AEC test has been much debated recently, notably in the EC’s Intel decision in 2009,^[27] and subsequent appeals.^[28] The European Court of Justice (ECJ) made clear that if an undertaking provided evidence that its conduct was not capable of restricting competition then the EC must examine:

• the extent of the undertaking’s dominant position;
• the share of the market covered by the challenged practice (i.e., the contestable share of the market);
• the conditions and arrangements for granting the rebates in question;
• their duration;
• their amount; and
• the possible existence of a strategy to exclude from the market competitors that are at least as efficient as the dominant undertaking.^[29]

The economics behind the ECJ’s Intel judgment is consistent with the EC’s Article 82 Guidance paper, and also in line with a CMA decision published in June 2015 when it closed a case relating to a suspected loyalty-inducing discount scheme in the pharmaceutical industry. That CMA decision noted that concerns can arise relating to foreclosure in certain circumstances, as given below.^[30]

• Discount or rebate schemes have a loyalty-inducing effect. Rebates that arise in return for exclusive purchasing are an example of this.
• Retroactive rebates may be of particular concern here, particularly if there are units that the purchaser needs to buy from the dominant company while other units are contestable, and the discount or rebate is conditional on the customer purchasing its contestable sales from the dominant company. This implies the rival has to compensate the customer for the loss of the discount across all volumes if it wishes to win business on the contestable portion.
• If the price that the rival would have to charge falls below the costs of the dominant firm, then a firm that is equally as efficient as the dominant firm will be foreclosed from competition for some or all of the contestable market. In some pharmaceutical markets, the contestable portion may relate to new patients while existing patients may continue on their existing course of treatment.

Below we discuss some recent cases in the UK and the Netherlands, which all relate to tumour necrosis factor (TNF)-alpha inhibitors, to illustrate the economic issues connected with rebates.

CMA Remicade case^[31]

The CMA began a formal investigation in December 2015 into Merck Sharpe and Dohme Limited’s (MSD) pricing of Remicade, the brand name for its infliximab product, a type of biological immunosuppressant, used to treat autoimmune inflammatory disorders such as Crohn’s disease and rheumatoid arthritis.^[32] The CMA suspected that MSD had abused a dominant position by offering loyalty-inducing discounts for the sale of Remicade in the UK.^[33]

The Remicade patent expired in the UK in February 2015^[34] and two biosimilar products (Inflectra, sold by Hospira, and Remsima, sold by Napp) were introduced in the UK from March 2015.^[35] In contrast with generic medicines which are generally considered to be bioequivalent to small molecule drugs, biosimilar drugs are ‘similar to’ the originator biological medicine, because no two batches of a biological product are identical.

Prior to the Remicade patent expiring, the Commercial Medicines Unit (which advises trusts in England on procurement of medicines) held a tender process. MSD designed a discount scheme in its response. Simplifying the scheme for ease of exposition: prices varied according to volume thresholds reached; prices were the same across regions in England with volume thresholds tailored to expected demand within each region. If a region purchased less than 85 per cent of total infliximab demand through Remicade, all discounts would be withdrawn and list prices would be charged. Prices after discount would fall as MSD’s share of total expected infliximab demand increased from 85 per cent to 100 per cent.^[36] If a region purchased more than 85 per cent but less than 94 per cent of expected infliximab demand through Remicade, it would face higher prices than previously.^[37] The discounts were not retroactive after MSD introduced a quarterly review process between the original launch of the scheme and its implementation.

Market definition was limited to infliximab, including Remicade and biosimilars, in England.^[38] This was not widened to include other drugs with a degree of therapeutic substitutability due to the difference in mode of administration which meant that Remicade was generally prescribed when other biological immunosuppressant medicines were not suitable.^[39] MSD was found to have held a dominant position in the relevant market during the relevant period.^[40]

The CMA inferred that MSD had an anticompetitive strategy in designing its discount scheme:

MSD’s internal analysis demonstrates that MSD designed its Discount Scheme to result in:

1. Biosimilar suppliers having to charge very low prices in order to match the effective price charged by MSD over the contestable share of demand; and
2. the NHS [National Health Service] having to pay more in total for infliximab products if it chose to switch from purchasing Remicade to purchasing Biosimilars.^[41]

The CMA did not apply an AEC test in its statement of objections and did not consider it necessary to do so.^[42] The CMA assessed factors that it viewed as relevant to assessing the possible existence of a strategy aimed at excluding competitors that are at least as efficient as the dominant undertaking. To do this, the CMA assessed dominance, the share of the market covered by the challenged practice, the conditions for granting discounts as well as their duration and amount.^[43]

The net price of Remicade was ‘very low’ over the portion of the market that MSD expected biosimilars to compete for (mainly new patients).^[44] The CMA found that ‘Biosimilar suppliers would have had to make a loss to win sales from Remicade . . . in which case it would not have been possible for a Biosimilar to compete with MSD unless the NHS was willing to switch a larger proportion of sales from Remicade to Biosimilars’.^[45]

The design of the discount scheme thus in theory appeared to satisfy economic tests for a scheme that had the potential to exclude competition by equally efficient competitors.

The CMA, however, ultimately concluded that an exclusionary effect was not likely. This is because Remicade’s discount scheme made a number of assumptions about the market at the time which turned out to be incorrect, which in turn meant that exclusion did not occur, such as:

• MSD had overestimated the degree of clinical caution within the NHS, which would have acted to limit expansion by biosimilars.^[46] Trusts took a longer-term view of costs and benefits of biosimilars, and risking a short-term increase in infliximab expenditure to reduce overall expenditure as confidence in biosimilar usage increased.^[47] This meant that the financial incentives that MSD had designed were quickly overcome; and
• the strength of the financial incentives turned out to be less than MSD had planned – in particular, the retroactive portion of the discount scheme’s original design was removed when MSD introduced a quarterly review process prior to implementation.

These factors in combination meant that the contestable share for biosimilars turned out to be higher than MSD had anticipated.^[48] Based on these particular factual circumstances – in effect, because MSD’s planning assumptions had proven to be incorrect – the CMA concluded that the conditions for prohibition were not met^[49] and issued a ‘no grounds for action’ decision in March 2019.

Recent cases in the Netherlands

Following the Netherlands Authority for Consumers and Markets (ACM) sector enquiry into TNF-alpha inhibitors, the competition authority investigated discounts on other branded infliximab products: Humira and Enbrel. The ACM was concerned that discount structures could be used to exclude biosimilar competitors and reduce incentives for manufacturers to invest in biosimilars. The potential concerns are similar to those in the UK Remicade case: that discount schemes could be used to limit purchases of biosimilars, and restrict their expansion with the result that payors pay more for drugs instead of benefiting from competition after patents on originator drugs have expired.

AbbVie and Humira

The patent on adalimumab, the active ingredient in Humira, expired in the Netherlands in late 2018. According to the ACM, AbbVie offered hospitals discounts that mandated exclusive purchasing of Humira for existing patients and the ACM launched an investigation in 2019. Hospitals would receive a significant discount only if existing patients continued to use Humira; these discounts would be withdrawn if hospitals allowed some patients to switch to a biosimilar, with the result that they would then have to pay significantly more for Humira. The discount scheme as described is one in which a firm tries to foreclose competition for the contestable portion of the market (new patients) by withdrawing discounts on the potentially non-contestable portion (existing patients). AbbVie did not endorse the ACM’s findings but nevertheless dropped the conditions associated with its discounts in 2020.^[50]

Pfizer and Enbrel

Pfizer’s patent on etanercept, the active ingredient in Enbrel, expired in 2015, following which two biosimilars entered the Netherlands.^[51] In autumn 2021, the ACM opened an investigation into Pfizer’s pricing of Enbrel because Pfizer’s contracts allowed it to reduce the discount it offered if the volumes purchased by the hospital decreased by more than a specified percentage. This aspect of the contracts created a financial barrier to hospitals switching from Enbrel to biosimilars. The hospitals had to continue purchasing Enbrel for existing patients that could not switch or did not wish to switch to biosimilars, and hospitals could have ended up paying almost four times as much for Enbrel for those patients. While Pfizer did not agree with the ACM’s assessment, it amended its contracts to remove the relevant discounting clauses.

Summary of European pharmaceutical cases involving rebates and discounts

Killer acquisitions

A ‘killer acquisition’ is a merger with the intention or effect of eliminating a future competitor. Often, the term relates not to a loss of existing competition but to the acquisition of a target that could compete with the acquirer in future. The term is controversial, often intended to imply a need for change in merger procedures and assessment. For example, Italy amended its merger control law in 2022 to allow its competition authority to review concentrations that fall below turnover thresholds and where the concentration could have a ‘detrimental effect on the development of small companies characterised by innovative strategies’.^[52] Nonetheless, there is some evidence of killer acquisitions having occurred in the life sciences sector.

This is one of two sectors to have attracted recent commentary on possible ‘killer acquisitions’. The other is the tech sector, where the concerns are rather different. The large digital platform providers (GAFAM) often buy innovative start-ups – Google has made 240 acquisitions and 161 investments since its foundation, for example.^[53] These could be ‘killer acquisitions’ if, without the deal, the targets would have grown into a serious competitor to those businesses. However, this is a low-probability (but high impact) outcome that is near-impossible to predict: most such deals are (presumably) not anticompetitive and may serve a valuable function in rewarding innovative entrepreneurs.

In pharmaceuticals, the situation is different. Here, the concern is about acquiring control of an existing R&D process, usually leading to a well-defined future competitor. For example, a small company that is innovating around a highly profitable patented drug might be bought out by the owner of that drug. It is conceptually straightforward to determine which existing products may face competition from a drug in development once a drug has proceeded to clinical trials. The EC has often identified overlaps in product development, in merger assessments, as given below.

The Commission intervened to protect innovation competition in a number of cases which, for example, threatened to thwart advanced R&D projects for life-saving cancer drugs (Novartis/GlaxoSmithKline Oncology) or for pipeline insomnia medicines at an early stage of development (Johnson & Johnson/Actelion). In the Pfizer/Hospira case, the Commission was concerned that the merger would do away with one of the two parallel projects to develop competing biosimilars. The Commission cleared all these transactions but only after the companies offered remedies to ensure that pipeline projects were not dropped and found a new operator to drive them forward.^[54]

However, an academic paper by Cunningham et al. (2021)^[55] and a study conducted by Informa Pharma Consulting and Szücs (2020) for the EC^[56] both present evidence that transactions between pharmaceutical firms resulted in a higher likelihood of drugs in development not proceeding to be introduced, when those drugs overlapped with the existing portfolio of the acquiring firm.

Cunningham et al. use a data set of drug projects initiated between 1989 and 2010, and track their development activities and any acquisition until 2016. About 22 per cent of these projects were identified as affected by a merger and about 25 per cent of those involved an overlap between the acquired development project and products of the acquiring firm, where an overlap is defined as a drug of the same therapeutic class and mechanism of action. The authors find that acquisitions of an overlapping project were about 20 per cent more likely to result in cancelled projects than others. From this, they conclude that 5 per cent to 7 per cent of all acquisitions were ‘killer acquisitions’ in effect and further note that they are particularly concentrated just below the threshold for notification to the US authorities, suggesting this may be a deliberate strategy. They find the effect to be equivalent to around 13 new drug projects per year being ‘killed’.

There may be other, more benign explanations for these results, some of which Cunningham et al. address. Not all overlapping drug projects produce clinical or economic benefits. Some might represent mere additions to an already reasonably competitive market, in which case it might make sense for an existing supplier to discontinue them from an acquired portfolio (but not if this were the target firm’s only product – and the dataset includes such examples).

It is not obvious how this analysis should influence policy. The study found a higher probability of project cancellation, but against a background in which many projects are cancelled for benign reasons – so which deals can be prospectively assessed as ‘killers’? We understand the EC has commissioned an ex post review to identify specific past killer acquisitions in the pharmaceutical sector and advise on possible changes to merger control in response.^[57] At the time of writing, the results of the review have not been published. The EC has also called for Member States to refer to it transactions below the national notification thresholds, using Article 22 of the EU Merger Regulation.^[58] At the time of writing, it remains to be seen whether the EC or another competition authority will identify specific ‘killer transactions’ as part of merger control in the life sciences sector but they clearly remain on competition authorities’ radar.^[59]

Notes