by Gavin Robert, Senior Consultant
There has been a lot of discussion recently, both in the competition world and in the wider media, as to whether merger enforcement by the Competition & Markets Authority (CMA) has got tougher over the last few years, e.g. Financial Times on 2 March 2020 ‘UK competition watchdog is most active antitrust enforcer – Competition and Markets Authority frustrated highest share of deals worldwide’. This has been exacerbated by the controversy surrounding its position on Microsoft’s acquisition of Activision Blizzard (especially in circumstances where it was cleared subject to remedies by the European Commission), e.g. the Guardian on 26 April 2023 ‘Activision Blizzard calls UK “closed for business” after Microsoft takeover veto‘.
So, with thanks to help from Maxwell Curtis at Euclid Law, I thought we would take a look at the statistics, comparing the most recent five years with the five years previously. Since this more or less covers the decade since the CMA has been around, this seems an appropriate cut. I am also personally interested in this comparison exercise as I was a Panel Member deciding Phase 2 merger cases, i.e. those more complex cases referred after initial review for a detailed investigation, during the first five-year period.
Of course, any statistical analysis (especially one conducted by lawyers) should come with multiple caveats. Naturally, how you cut the periods will distort the outcome – what if we look at the most recent three years compared to the previous seven years? By necessity, enforcement statistics are a blunt instrument. As anyone involved in one of these cases will testify, mere statistics may conceal the real story of the case, e.g. “clearance subject to remedies” could involve a case where the market subject to a remedy may be trivial, while the key market worth much more to the client was ultimately cleared after a long and hard struggle…We can’t hope to delve into the real stories of every case but, hopefully, taking into account around 585 cases over ten years mitigates that concern. But given the much smaller number of Phase 2 cases (96 over ten years), conclusions on Phase 2 cases should still be treated with some caution.
So what do the statistics tell us?
On the face of it, CMA merger enforcement has got a lot tougher. The % of cases referred to Phase 2 for further investigation has almost doubled (12 to 22%), while the % of Phase 2 cases that result in unconditional clearance has plummeted from 50% to just 20% (and prohibitions have increased from just three to ten).
Let’s look at each of these in turn.
Although the % of cases referred to Phase 2 has more or less doubled, the total number of Phase 1 decisions has dropped substantially: from 330 to 255 (i.e., by around a quarter). This could of course reflect a drop in M&A activity. However, by way of comparison, the total number of notifications to the European Commission actually increased in the last five years compared to the previous five years (by around 15%). One explanation, and one that the CMA appears to support (as endorsed by Sarah Cardell, the CEO of the CMA, at our Euclid Law Breakfast Seminar earlier this year), is that over the last few years the CMA has actively encouraged a greater use of briefing papers rather than preceding directly to notification. The implication is that, where the CMA has indicated that it has no further questions and has not “called in” the case, mergers that would previously have been notified are no longer being notified. Thus, having weeded out more unproblematic mergers through the briefing paper route, it is not surprising that a larger proportion of those that are notified may justify referral to Phase 2 (even if the substantive threshold for referral remains unchanged). Our own anecdotal experience at Euclid Law of dealing with the CMA on briefing papers also tends to support that explanation, at least directionally.
However, the absolute number of referrals has also increased (from 39 to 57). And if one were to use the total decisions of the previous five years rather than the most recent five years as the denominator (thus seeking to ignore the briefing paper effect on total number of decisions), the % of referrals would still increase from around 12 to 17%, i.e. by around half as much. So perhaps the conclusion is that the CMA has got tougher in referring cases to Phase 2 – but by not as much as the raw statistics might suggest.
It is also worth noting that the proportion of cases resolved through remedies at Phase 1 (so-called “UILs”, i.e. Undertakings In Lieu of a reference) has remained relatively constant, with a small increase from 10 to 13% (and the actual number of cases identical in each period).
While the chance of getting referred to Phase 2 appears to have increased, the chance of surviving Phase 2 unscathed has significantly reduced from 50% to 20%. (Together with my colleagues at Euclid Law, I need to try to ignore as purely anecdotal evidence our own much better track record of two out of two Phase 2 unconditional clearances in the last year: LSEG/Quantile and most recently, in the Copart/Hills Motors merger.)
Interestingly, the number of cases that are cancelled because the deal is abandoned after referral has also increased substantially, three-fold from 5 to 15. Abandonment following Phase 2 referral is usually regarded as an enforcement outcome, as it is reasonable to assume that the deal was abandoned because of the referral. But of course, it tells you nothing about the “toughness” of the CMA towards Phase 2 cases. If these cases are excluded, the unconditional clearance % is 57% in the previous five years compared to 28% in the most recent period. Indeed, against what is broadly a similar number of Phase 2 outcomes (i.e., excluding abandoned cases), the number of unconditional clearances has still halved. We might speculate that the reason for the greater number of mergers cancelled following Phase 2 referral is the market perception that the chance of emerging unscathed is so low – a 28% chance will not be attractive to most CEOs. That said, 15 cases over five years is still a small number so we must be careful not to draw general conclusions.
I tend to regard the Phase 2 unconditional clearance % as a guide to how effective an authority has been in eliminating confirmation bias at the second stage. Prior to the CMA’s establishment, when Phase 1 and Phase 2 were conducted by totally separate institutions, the Office of Fair Trading and the Competition Commission respectively, the unconditional clearance % over their last 10 years or so was around 40%. This may therefore be a reasonable guide as to what the unconditional clearance rate should look like without confirmation bias. No doubt the CMA would argue (and Andrea Coscelli, the former CEO of the CMA, has argued), and I would tend to agree, that anything above a 50% unconditional clearance rate in Phase 2 is too high. With a greater than 50/50 chance of clearance, surely something has gone wrong at the referral stage?
It is worth noting that the EC Phase 2 unconditional clearance rate is even lower and has also fallen over the same period, from 15% to under 12%. Again, the smaller number of cases in Phase 2 makes drawing reliable conclusions quite difficult, as so much turns on the facts of individual cases, but the European Commission’s greater flexibility towards remedies may be part of the explanation (e.g., extracting behavioural remedies in vertical cases which might otherwise have been cleared unconditionally).
But what has caused what is still a dramatic fall in the chance of escaping unscathed after Phase 2 in the UK? It is not that case teams are getting better at filtering out unproblematic cases at the end of Phase 1 – as we have seen, Phase 2 referrals have also increased. Phase 2 cases are, of course, decided by independent Panel Members. So even if CMA policy has changed over the last five years to take a more aggressive approach to enforcement, it is not clear how that can be implemented in individual cases. The CMA executive have, however, placed greater emphasis in recent years (the process was just beginning at the end of my term in 2018) on “educating” Panel Members on key areas of competition policy, likely including, for example, reasons for scepticism as to the likelihood of new entry or effectiveness of behavioural remedies. So, while Panel Members no doubt continue to exercise independent judgment in individual cases, their policy approach may be more closely aligned than previously to that of the CMA executive.
So, my conclusion? Perhaps unsurprisingly, yes, the CMA has got tougher: a conclusion that is probably unsurprising to those dealing with the CMA at the coal face day after day.