When discussing mergers, both Commission officials and private practitioners often characterise these as a “N to N-1”. We all regularly talk and hear about “6-to-5s”, “5-to-4s”, “4-to-3s”, “3-to-2s”, and even “2-to-1s”. This blog discusses the growing view that within the Commission, a “4-to-3 or fewer” distinction has become a corridor-talk presumption of anti-competitive effects. On this view, mergers leading to three or fewer alternatives in the market are regarded at first sight as being likely to bring about significant price increases.
First, it is important to make clear that the Commission has recently shown readiness to clear 4-to-3 mergers unconditionally. In FedEx/TNT, the Commission cleared the proposed transaction without remedies, and it did so in an industry where three years before a different transaction (UPS/TNT) had been blocked. This example in and by itself shows that merger analysis at the Commission is of course much deeper than just counting the number of alternatives available in the market.
Second, it is equally important to highlight that a “4-to-3 or fewer” presumption (and even if it was just a presumption) would not be based on (nor supported by) economic theory.
The simple counting of alternative suppliers – also known as “fascia count” – was formally introduced in the context of multi-overlap retail mergers. In such cases, and due to the unduly large number of markets to be tackled, (national) competition authorities have come up with “screening” rules based on the number of alternatives present in a given local area. In many of these cases, the authorities have focused their more detailed assessment on those overlapping areas in which post-merger there would be three alternatives or fewer, where they considered anti-competitive effects could not be ruled out and further analysis was required.
The definition and use of simple “screens” in the context of local mergers is understandable – there is a need to pragmatically focus the work of the Case Team and the Parties somehow. But it is also worth keeping in mind that any classifying rule based on a simple “fascia count” is arbitrary. While some theoretical models indeed suggest that a reduction in the number of (symmetric) firms will result in a reduction of competition in the market, there is nothing in the economic literature that suggests that it is precisely in a 4-firm setting where a (or any) merger becomes likely to generate significant anti-competitive effects.
Economics tells us that the scope for mergers to give rise to significant price increases depends on a host of different factors, including not only the level of switching between the merging parties and the margins they generate pre-merger, but also the shape and form of the demand curve they face, capacity constraints, product repositioning, cost efficiencies, customer reactions, ease of entry, and others. In this context, the number of firms present in the market does not really say much about the practical relevance of each of these firms to the competitive process.
Even when ignoring some of the important factors mentioned above and looking at simple static models such as Cournot, Bertrand or the UPP/GUPPI formulae used by the Commission lately, there is nothing that suggests that it is in a 4-to-3 situation where (predicted) price increases get “significant enough”. It really depends on the values that the input parameters take in the market in question (such as margins, diversion ratios or demand characteristics) – in some cases a 4-to-3 or a 3-to-2 could look rather unproblematic, whilst in others a 5-to-4 or a 6-to-5 could give rise to significant (predicted) price increases.
Indeed, the use of a corridor-talk presumption of anti-competitive effects purely based on the number of firms would not only be unfounded, but it would also be at odds with the evolution of the Commission’s economic thinking in the past few years. The Commission’s analysis of unilateral effects, largely based on traditional oligopoly theory, is currently primarily focused on assessing whether the merging parties themselves compete “closely enough” so as for the merger to give rise to significant price increases, and this cannot be inferred either directly or indirectly by simply looking at “how many” alternatives co-exist in the market.
It has been already many years that merger analysis has been – slowly but surely – moving away from old simplistic presumptions purely based on market shares or HHI levels. In this context, it would not seem appropriate to get into making similar judgements based on the mere number of players operating in a given market.