The UK’s CMA has often been cast as the tough cop on the mergers beat – quick to intervene, sceptical of behavioural remedies, and frequently at odds with its European peers. But was its hardline reputation ever truly deserved? A closer look at the numbers reveals a more nuanced story.

The CMA in the crosshairs

The Competition and Markets Authority (CMA) is undertaking sweeping changes to its approach to merger control. Following a recent overhaul of its procedures for in-depth investigations, it is now proposing several reforms to early-stage reviews – and, perhaps most consequentially, reconsidering its long-standing preference for structural remedies over behavioural ones, as signalled in its call for evidence in March 2025 on reforming its approach to merger remedies.

This shift did not happen overnight. Amid heightened scrutiny of mergers – particularly in the digital sector – the CMA has found itself at the heart of several high-profile investigations ending in prohibition or abandonment. In some cases, it stood alone in opposing the deal. A pivotal moment came in 2023, when the CMA blocked Microsoft’s acquisition of Activision Blizzard – a move that stood in stark contrast to the European Commission clearing it with behavioural remedies. The merging parties swiftly hit back, declaring the UK “closed for business” – a pointed challenge to the government’s pro-growth narrative.

Facing mounting pressure – including an appeal from Microsoft – the CMA ultimately approved a restructured version of the deal. This was followed by a landmark decision to clear Vodafone’s acquisition of Three, subject to a behavioural commitment to invest in a joint 5G network roll out across the UK. But the political fallout was already underway. In January 2025, CMA chair Marcus Bokkerink stepped down, with the government signalling a desire for a “different strategic approach” more closely aligned with its economic growth agenda. He was succeeded by Doug Gurr, a former Amazon executive.

But was the CMA’s tough reputation deserved? Has it truly been more heavy-handed than its European counterparts in triggering Phase 2 investigations or blocking deals?

Aggressive or selective?

At first glance, the statistics suggest a tougher stance. As Figure 1 below shows, between 2020 and 2024, 16% of CMA merger investigations across Phase 1 and Phase 2 resulted in the deal being prohibited or abandoned, compared to just 2% for the EC.

Figure 1: Comparison of CMA and EC merger investigation outcomes across Phase 1 and Phase 2, 2020-2024 inclusive

Source: Frontier Economics analysis and interpretation. EC data compiled by Frontier from the EC website. CMA data compiled using the Linklaters Platypus platform and the CMA website.

However, this comparison is misleading. As Figure 2 shows, the EC conducted nearly 1,900 Phase 1 investigations during that period – over nine times more than the CMA. This disparity reflects the UK’s voluntary notification regime, which allows the CMA to choose which deals to review. Firms can submit briefing papers for transactions that meet UK jurisdictional thresholds but appear uncontroversial, to “test the waters” without making a formal notification. The EC, by contrast, must assess every notifiable deal, however routine, even though it operates a simplified procedure for deals unlikely to raise competition concerns. Viewed in this context, the CMA’s higher proportion of prohibited deals reflects a more selective, but not necessarily more aggressive, approach.

Figure 2: Volume of Phase 1 investigations by the EC and the CMA, 2020-2024 inclusive

Sources: Frontier Economics analysis and interpretation; EC data compiled by Frontier from the EC website; CMA data compiled using the Linklaters Platypus platform and the CMA website; briefing paper statistics compiled by Ronan Scanlan, Partner at Steptoe International.

Second thoughts on Phase 2

What about Phase 2 outcomes? Again, the data tell a subtler story. As Figure 3 shows, the CMA’s scepticism toward behavioural remedies is clear: nearly one in five EC Phase 2 cases were cleared with such remedies, whereas they remain exceptionally rare in CMA decisions. However, this hasn’t translated into a higher rate of prohibitions in London than Brussels. Roughly 50% of CMA Phase 2 cases ended in prohibition or abandonment – broadly in line with the EC's 45%.

In fact, the CMA has been more willing to clear deals outright. Between 2020 and 2024, it approved 26% of Phase 2 cases without remedies, versus just 9% for the EC.  The CMA has unconditionally cleared several deals where behavioural remedies were tabled, including:

·       Facebook/Kustomer (2021): cleared unconditionally by the CMA at Phase 1; the EC required behavioural remedies.

·       LSEG/Quantile (2022): behavioural remedies offered at Phase 1 were rejected by the CMA, which later cleared the deal unconditionally at Phase 2.

·       Broadcom/VMware (2023): cleared unconditionally by the CMA at Phase 2; the EC required behavioural remedies.

In short: the CMA’s wariness towards behavioural remedies has not translated into a higher rate of prohibitions. If anything, it may have contributed to more unconditional clearances – especially in marginal cases where structural remedies were not viable and competition concerns fell short of justifying a block. In contrast, the EC’s greater openness to behavioural commitments may have led it to impose such remedies as a precaution in similar cases.

Figure 3: Comparison of CMA and EC Phase 2 merger investigation outcomes, 2020-2024 inclusive

Source: Frontier Economics analysis and interpretation; EC data compiled by Frontier from the EC website; CMA data compiled using the Linklaters Platypus platform and the CMA website.

Beyond the numbers

Overall, the data suggest that the CMA has taken a lighter-touch approach to uncontroversial deals than the EC in recent years, and has been more willing to clear complex Phase 2 mergers unconditionally, despite – or perhaps because of – its scepticism toward behavioural remedies. So is its ‘bad cop’ image unfair? Painful as it may be for economists to admit, statistics don’t tell the full story.

First, the dry numbers miss how the process feels. The CMA has recently faced criticism for creating protracted uncertainty through its discretion over calling in deals – illustrated by its assessment of Microsoft’s partnership with OpenAI, which took more than a year simply to conclude that it did not qualify for a formal review. It has also come under fire for its conduct during Phase 2 reviews – particularly for its slow engagement and reluctance to share provisional thinking until late in the process. The CMA’s proposed changes to pre-notification and Phase 1, as well as recent reforms in Phase 2 – including earlier dialogue and more frequent engagement with decision-makers – implicitly acknowledge these concerns.

Second, concerns arise not just from divergence between the CMA and EC, but from their shared caution. Both the CMA and EC have prohibited horizontal and non-horizontal mergers at Phase 2 at similarly high rates – close to 50% between 2020 and 2024 in each case – notwithstanding the EC’s acknowledgment in its guidelines on the assessment of non-horizontal mergers that such transactions are typically less problematic and more efficiency-enhancing than mergers between direct rivals.

Third, a deeper dive into the head-to-head record between the CMA and the EC on cases investigated by both authorities yields insights that the headline statistics gloss over. The CMA has diverged from the EC and other European authorities in several high-profile cases, with the CMA’s more sceptical stance on remedies often being the dividing factor. Examples include:

• Google / Fitbit (2020), cleared by the EC with behavioural commitments that prompted criticism from the CMA’s then-CEO, despite the CMA lacking jurisdiction.
• Cargotec/Konecranes (2022): the CMA rejected remedies that the EC accepted; the deal was ultimately abandoned.
• Facebook / GIPHY (2022) investigated and blocked by the CMA (with behavioural remedies rejected), despite falling below the EC’s thresholds and being cleared in Austria.
• Microsoft/Activision (2023), the climactic CMA intervention – cleared by the EC with a behavioural remedies package that the CMA rejected.

These cases are not the norm. In several other investigations, the CMA and EC reached consistent decisions – and in some instances, the CMA cleared deals that were either blocked by the EC (Booking.com/eTraveli) or abandoned in the face of EC concerns (Amazon/iRobot). Nonetheless, the cases above have attracted significant attention, as the CMA – apparently driven by a rigid position on remedies – blocked global transactions with no clear UK-specific angle that were approved by larger international regulators. As a result, these cases have had an outsized influence on the perception of the CMA as out of step with its peers.

Rethinking remedies

Several of the CMA’s most controversial decisions involved vertical mergers with behavioural remedies that the CMA rejected. If it wants to rebuild its reputation for proportionality, greater openness to such remedies would be a logical place to start.

Currently, the CMA applies a sequential test when assessing such remedies: first, will the remedy be effective in addressing the CMA’s competition concerns? And second – if more than one effective option is available – which option imposes fewer costs? This sequencing risks treating proportionality as a second-order consideration. The CMA’s recent reaffirmation of its ‘4Ps’ – proportionality, process, predictability, and pace – suggests it now sees proportionality as a first-order consideration.

Caution around behavioural remedies is understandable. They are often harder to monitor and more vulnerable to circumvention than structural remedies. But those risks must be assessed against the scale and seriousness of the competition concern. In Phase 2, the test is whether a merger is likely to result in a substantial lessening of competition (SLC) on the balance of probabilities. In that context, an “effective” remedy need only reduce the risk of an SLC below 50% – not to zero. Where the risk is clear-cut and severe, a cast-iron solution may be needed; where the SLC concern is marginal, insisting on such certainty may be disproportionate.

Some have questioned whether a more central role for proportionality in this assessment would require changes to the CMA’s statutory framework, which requires it to “achieve as comprehensive a solution as is reasonable and practicable” in remedying an SLC. There ought to be scope – within the existing legal test of what is “reasonable and practicable” – for the CMA to adopt a more calibrated approach in marginal cases, consistent with its broader commitment to the ‘4Ps’. If not, the government would be well advised to take note.

Rethinking Brussels too

Proportionality cuts both ways. The EC’s tendency to require remedies – structural or behavioural – in nearly every Phase 2 case risks making remedies the default price of clearance. Such outcomes are not always efficient or proportionate, especially given the burden that remedies can impose on firms and regulators. The EC’s ongoing merger guidelines consultation offers a timely opportunity to reconsider this practice.

Both the CMA and EC would benefit from embedding proportionality more deeply into their assessments. But amid the criticism, the UK regime has much to defend: a flexible notification system; a willingness to clear deals unconditionally when the evidence supports it; and meaningful recent reforms to improve the investigation process. There’s work to do – but also progress worth protecting.