Lip service

Merger guidelines often recognise the concept of efficiencies arising from a merger. The recently revised CMA Merger Assessment Guidelines (the “CMA MAGs”) set out that efficiencies could in principle arise in the following ways: “cost savings; the elimination of double marginalisation through vertical integration; greater innovation or quality arising from the combination of unique assets; or better meeting customers’ needs by enabling the integration or interoperability of complementary products” (paragraph 8.2).

In practice, though, merger efficiencies have hardly ever been identified at EC level or in the UK. I am aware of only a handful of instances over the last 20 or so years where some level of merger efficiency has been taken into account, always in the context of a merger that raises concern and typically through reductions in marginal costs (e.g. Asda/Netto, Deutsche Boerse/NYSE, UPS/TNT, Sainsbury’s/Asda). 

The CMA MAGs go on (in paragraph 8.8) to set out that merger efficiencies must:

(a) enhance rivalry in the supply of those products where an SLC may otherwise arise;

(b) be timely, likely and sufficient to prevent an SLC from arising;

(c) be merger-specific; and

(d) benefit customers in the UK.

These criteria are somewhat confused. It is sensible to split them into two categories. First, criteria relating to whether a claimed efficiency should be taken into account. Second, whether the level of such efficiencies is sufficient to offset any SLC finding. In the first camp are the well-known triumvirate of whether a claimed efficiency is timely, likely and merger-specific. I comment on these below. There is only really one issue in the second camp, which is whether efficiencies are sufficiently large to offset any SLC finding; the CMA’s approach essentially repeats this criterion three times (in a), b) relating to sufficiency, and d)).

Turning to the evidential standards for an efficiency, it is fairly uncontroversial that that merger efficiencies should be timely (i.e. they should arise relatively quickly). I observe only that the relevant standard for timeliness should be the same as the standard for the assessment of an SLC. In other parts of the CMA MAGs, the CMA outlines that the adverse impacts of mergers in the digital space may not be felt for several years (e.g. in its section on potential competition concerns, paragraph 5.4). If so, it would make sense to treat efficiencies symmetrically (for more details, please see Frontier’s submission to the CMA on the Draft Merger Guidelines available here).

It is also uncontroversial that merger efficiencies should be likely, in the sense that they should be plausible and verifiable. However, the approach in the CMA MAGs expresses considerable scepticism as to whether efficiencies in fact arise as a result of merger, citing a paper by Professor John Kwoka which comments on the treatment of efficiencies in the US Merger Guidelines and recommends a treatment basically along EC/UK lines. This upfront scepticism appears unwarranted. The evidence highlighted in Professor Kwoka’s report covers only a narrow selection of industries (mostly manufacturing plants and hospital mergers in the US) and, where it does consider a broader cross-section of firms, finds that many mergers do lead to significant synergies, and often to all the synergies anticipated by management or at least 75% of those synergies (more details are in Frontier’s submission referred to above). 

Moreover, in a report by LEAR for the CMA, reviewing past merger clearance decisions in the digital sector, LEAR found that in 40% of the cases it reviewed (admittedly a small sample) it appeared that the merger had led to efficiencies – and specifically in Facebook/Instagram and Google/Waze. LEAR’s findings are not reported in the CMA MAGs.

Perhaps the most troubling area of current practice is the issue of “merger-specificity”. At a high level this concept makes sense. If the efficiencies can be achieved without a merger, then the existence of efficiencies is not a justification for the merger.

(An example might be a merger in the mobile telecoms space, where efficiencies could arise from improved network sharing, but where these might be achievable through network sharing agreements short of full merger, so leaving the firms free to compete at the retail level. Although see this article by my colleagues Goran Serdarevic, Peter Davies and Martin Duckworth for more on whether network sharing agreements can in fact provide the same efficiency benefits as a full merger.)

But at some point the issue of merger-specificity seems to have been reinterpreted as relating to whether the efficiencies only arise from that specific merger (as opposed to some other merger involving the target). This is not a sensible approach from a consumer welfare perspective. Suppose there are four potential acquirers bidding for the same target. Each of the mergers would give rise to a 5% cost saving or equivalent quality improvement, of which half (2.5%) is expected to be passed on to customers, while the GUPPI for each merger implies a 2% price rise. So any of the mergers would give rise to a net reduction in prices of 0.5% and so is beneficial from a consumer welfare perspective. 

However, none of these mergers has a merger-specific efficiency on the revised interpretation, because in each case the efficiencies could be achieved by another merger. Paradoxically, the net result is that no merger would be allowed, which is detrimental to consumer welfare, and hence this revised interpretation of what a merger-specific efficiency is makes no economic sense.

(This is similar to a joke by Yoram Bauman – the “stand up economist” – see here at around 1:38 – where he observes that if he gives you the choice between receiving either of two Snickers bars for free, you are in fact no better off than if you hadn't been offered any Snickers bars, because now if you don’t take one Snickers bar, you get an identical Snickers bar instead…)

In conclusion, therefore, the treatment of merger efficiencies is already unsatisfactory. The CMA MAGs appear to seek to institutionalise an even greater bias against efficiency arguments. This does not appear a good way to maximise consumer welfare.