The European Commission is increasingly concerned that market consolidation will harm innovation and has changed dramatically the way it examines the impact of mergers on innovation. Merging parties should be prepared for it.
How the European Commission looks at the effect of horizontal mergers on innovation will be remembered as one of the important policy changes championed by European Commissioner Margrethe Vestager, who reminded the public that EU merger control rules ‘are there to protect innovation’, and that this objective ‘is important in [our] merger policy’.
The Commissioner’s public interventions have accompanied the rise of a novel theory of harm, which posits that horizontal mergers could ‘lead to a reduction of innovation’ in an industry as a whole. This was its position in relation to a merger between Dow and DuPont, where the Commission found the merger would significantly reduce innovation competition for pesticides. This theory led to the divestiture of DuPont’s global R&D organisation in pesticides as part of the remedies required to clear the merger. In the recent Bayer/Monsanto merger, it seems that the Commission applied a similar approach, and found that the transaction would have ‘significantly reduced competition in a number of markets’ and ‘significantly reduced innovation’. As part of the remedy package, Bayer has committed to divest three important lines of its global R&D organisation.
Although assessing the effects of mergers on innovation is not new—the Commission’s Horizontal Merger Guidelines state that ‘effective competition may be significantly impeded by a merger between two important innovators’—under Vestager’s leadership the Commission’s approach to the impact on innovation has changed dramatically.
Traditionally, the Commission has examined whether a proposed merger creates an overlap between a product actively marketed by one of the parties with a pipeline product developed by the other party (‘market-to-pipeline’) or whether the parties developed separately pipeline products that would eventually compete on the market (‘pipeline-topipeline’). Pipeline products include products at a relatively late stage of their development, with a good chance of launch within two to three years.
R&D expenditures do not automatically translate into new products. Predicting the future isn’t so easy after all
In the recent Pfizer/Hospira merger, the Commission found that the proposed transaction raised competition concerns because Pfizer was developing a competing medicine to Hospira’s. Specifically, Hospira was selling Inflectra, an infliximab biosimilar used to treat several chronic inflammatory diseases, notably the inflammation of Crohn’s disease. At the same time, Pfizer was in an advanced stage of developing its own biosimilar for infliximab. The Commission was concerned that Pfizer would likely discontinue its efforts to bring its new medicine to market, reducing competition. As a remedy, Pfizer divested its development programme.
In Johnson & Johnson/Actelion, the merging parties were separately developing new treatments for insomnia, using the same novel mechanism of action. The Commission was concerned that post-merger, J&J would have the ability and incentive to delay or abandon one of these programmes and thus required that Actelion’s insomnia research programme be divested.
In the US, the Federal Trade Commission (FTC) has followed a similar approach. For example, in Thoratec/ HeartWare, Thoratec was marketing a successful ventricular assist device (a heart pump) while HeartWare’s device achieved promising clinical trials. The FTC challenged the merger, alleging harm to innovation and to future price competition. In Nielsen/Arbitron, Nielsen offered a leading TV audience measurement service, while Arbitron offered a leading radio audience measurement service. At the time of the merger, both were developing a cross-platform audience measurement service. The FTC challenged the merger, alleging harm to innovation.
The Commission’s approach in Dow/DuPont marks a dramatic shift in the way the impact of mergers on innovation is examined. The Commission did not focus on specific product overlap, as it did before, but instead it considered the impact on innovation ‘as a whole’. Putting it simply, the Commission found that Dow and DuPont would likely reduce their R&D budget post-merger, which would inevitably lead to a smaller number of new products brought to the market.
Parties should emphasise the complementarity of their R&D assets, which could lead to an increase in innovation
But is it that simple? First, if the Commission found that a ‘merger between important rival innovators is likely to lead to reduction of innovation’, how is innovation measured? Surely by considering the launch of potential new products. But future new products (or products in an advanced-stage of development) were not the focus in Dow/ DuPont. Can innovation be measured by R&D expenditures? Maybe. But R&D expenditures do not automatically translate into a guaranteed number of new products. R&D activities are a risky venture. Billions can be spent, often without immediate, or any, results. In the pharmaceutical industry, a recent study shows that only 9.6 per cent of drugs in Phase I are approved by the US FDA. The chance of success increases to 15.3 per cent in Phase II and 49.6 per cent in Phase III. Second, as the Commission puts it in its March 2017 decision (at 348), ‘innovation should not be understood as a market in its own right, but as an input activity for both the upstream technology markets and the downstream [product] markets’. If that is the case and if the ultimate goal of merger control is to protect competition between ‘downstream products’, would it not be necessary that innovations worth protecting, although inherently uncertain, be at least to some extent identifiable with existing or pipeline products?
Third, merger control is inherently forward-looking and requires making predictions, which become increasingly imprecise the further into the future one looks. In past cases, that the competitive assessment focused on pipeline products seemed justified, in particular when it was expected that these products would hit the market in two to three years. Equally, merger-specific efficiency arguments are typically accepted within such a timeframe, so that they can be verified with at least some degree of predictability. In sharp contrast, between the moment new molecules are discovered and the point that firms can launch a new product, more than 10 to 12 years can pass. It is extremely difficult to make any sound predictions so far into the future.
Fourth, firms undertake R&D investments when they expect sufficiently high returns. Expected returns on such investments depend in part on the number of firms engaged in the same ‘innovation space’. Obviously, the more firms that are involved, the lower the expected returns. Indeed, when a firm is the only one contemplating a research programme, it can expect greater reward than when others are pursuing the same research agenda. This means that a merger, by eliminating a rival, can increase the chance that the merged firm will undertake the necessary investment to pursue an ambitious R&D programme.
Finally, what can be an adequate remedy in such cases? Prohibit the merger as in Thoratec/HeartWare or Nielsen/Arbitron? In Dow/DuPont, in addition to product overlap divestitures, DuPont’s pipeline in herbicides and insecticides, its discovery pipeline in fungicides and its entire R&D organisation had to be divested, including some 400 to 500 employees. Contrary to past cases, where pipeline products were sold to third parties, forcing the sale of an entire R&D organisation appears far more radical. But will such a type of remedy be successful? Will the R&D organisation be equally successful under its new owner? Will the scientists stay or leave for new ventures? Only time will tell. In the 1995 AMP/Wyeth merger, the parties divested one of the two development programmes for Rotavirus vaccines to the Korean Green Cross. More than ten years later, GSK launched a rival vaccine, while the Korean Green Cross never did. In Ciba-Geigy/Sandoz, the FTC was concerned that the merger could impair the development of gene therapy, a market that the FTC forecasted to be worth US$45 billion within 20 years from then. Although the parties divested one of the gene therapy programmes to Aventis, the market for gene therapy is still very small and Aventis has not launched a product in this space. Predicting the future isn’t so easy after all.
Clearly, the Commission’s theory on innovation is here to stay. It seems to have applied it again in Bayer/Monsanto. Therefore, companies should consider carefully any overlap in innovation activities in a broad sense, even if there is no prospect of developing concrete products in the near future. Focusing on pipeline products is no longer enough to assess the regulatory risks posed by a transaction. Further, since the Commission’s assessment relies heavily on internal documents, the parties should describe their R&D activities with care, highlighting the potential efficiencies of merging them. The parties should emphasise the complementarity of their R&D assets, which could lead to an increase in innovation. At the same time, highlighting the benefit of eliminating duplicative activities could backfire, as the Commission might interpret such a plan as a clear intent to reduce innovation competition. Finally, mergers that may be prone to the Commission’s theory on innovation are likely to face an even longer pre-notification period.