Facebook/GIPHY and the (Un)innovative Theory of Harm
Towards the end of the last decade, the success of a handful of U.S. technology companies caused significant consternation amongst European policymakers. The initial response was a series of reports extrapolating new theories of harm in competition law for digital players. Some of these went even further, resulting in the creation of new regulatory regimes handing Government enforcers’ discretionary powers to intervene in the economy, with limited safeguards. In parallel, competition authorities stepped up enforcement using existing tools, resulting in a number of new enforcement actions at European and Member State levels, including former EU Member States.
As these competition cases are brought under judicial scrutiny, there are lessons to be learned for future policymakers, both in terms of the new theories of harm being tested, and the importance of checks and balances on administrative discretion. The UK Competition and Markets Authority (CMA) decision regarding Meta’s acquisition of GIPHY is one of the first such examples. The CMA’s decision has been heralded as opening up a “new era” of antitrust enforcement against tech, and is based on an innovative theory of harm to dynamic competition. But is this new era actually going to be good for consumers and the economy more broadly? At a hearing at the Competition Appeals Tribunal (CAT) at the end of April 2022, the parties and a handful of third-party interveners (including the Computer & Communications Industry Association) made their views known.
A Dynamic Competition Theory of Harm …
The CMA ordered the ex-post divestment of GIPHY, a GIF database service, on the grounds that it competitively constrained Facebook’s display advertising offerings. Specifically, GIPHY’s supply of brand advertising in GIF search results reflected their innovative capabilities which, as an independent company, compelled Facebook to continue innovating its own advertising services. Post-acquisition, in the absence of this competitive constraint, Facebook would become a lazy monopolist and competition in the UK would be harmed.
Facebook (now Meta), supported by CCIA, argued that the CMA had abused its discretion. In reality, GIPHY was not the competitive threat that the CMA painted it as, and its “paid alignment” advertising offering (which was already active in the U.S., to little success), was not the “next big thing”. Rather, GIPHY was at the end of its innovation runway, it was burning cash (projected to run out of money just a few months after receiving Meta’s offer) and at risk of having to continue funding its operations through a down-round of investment. This could cause it to lose its best employees, damaging its capability to innovate and compete dynamically, and its future competitive significance. It’s up to the CAT now to decide who was right.
Well, not exactly. The CAT doesn’t review the facts. In order to win in the UK, appellants must meet the incredibly high bar of judicial review, i.e. that the decision is tainted by illegality, irrationality or procedural unfairness. This means the CMA can make a lot of factual mistakes, and has a wide range of discretion in its interpretation of the evidence. But there are still some limits. The law still requires a finding of “substantial lessening of competition”, and the CAT’s actual job is to decide the exact contours of that legal test when the CMA asserts a reduction in dynamic competition.
… Or a Return to Structural, Per-Se Prohibitions?
In essence, the CMA’s argument was that it was within the discretion of its economic analysis to determine that GIPHY was a dynamic competitor, and that in the display advertising market the loss of any rival would mean a reduction in competition that would inevitably harm consumers in the UK.
The CMA argued that it was within its discretion to find this substantial lessening of competition even though GIPHY’s brand advertising service
- was not a close competitor to Facebook’s direct-response display advertising (in fact, per para 7.51 of the CMA’s Final Report, GIPHY’s offering was far less effective, and serves a very different purpose)
- was experimentally deployed only in the U.S. and failed to find traction there
- was not substantially valued by Facebook (or even identified as a competitive threat despite a massive trawl of internal documents), and given GIPHY’s demonstrated inability to find external funding, and
- was unlikely to become significant in the future (the CMA also did not address the meteoric rise of TikTok as the dominant social media platform of the near future).
On its face, nearly every piece of evidence would lead an objective and impartial observer to conclude that, in fact, the transaction is unlikely to harm consumers of display advertising in the UK. While there may be vertical concerns, these would typically be handled by behavioral remedies, not outright prohibition and divestment. Furthermore, being able to vertically integrate a GIF library into its social media ecosystem and direct future GIF innovations could substantially improve the quality of Meta’s services, in competition with other platform rivals. There were therefore also important pro-competitive benefits to be considered, particularly for the millions of social media users who would benefit from better integrated services. Nevertheless the CMA blocked the merger, in part on the grounds that it was “protecting millions of social media users and promoting competition and innovation in digital advertising.”
If the CAT upholds this reasoning, it would mean the CMA would be obliged to block the acquisition of any company that has ever introduced an innovative product or service (even where it fails to find product-market fit). This approach amounts to a structural presumption that the acquisition of virtually any company is anticompetitive. Rather than an innovative theory of harm, or even one focused on an analysis of innovative effects, it looks more like a return to out-dated notions of rivalry and structural presumptions. These have been shown to be inappropriate for the industrial-era economy, let alone today’s dynamic digital sector. It’s an approach that could be particularly harmful to investment, entrepreneurship, and innovation.
Improving The Enforcement Toolkit
After so much work has been done to better understand the competitive dynamics of the digital economy, no less by the CMA itself (which notably published new Merger Assessment Guidelines in March of 2021, with an entire section on potential and dynamic competition), it’s a bit disappointing to see an analysis that basically boils down to “more competitors good, less competitors bad”.
This isn’t a new era of understanding, it’s more a return to the same old short-hand presumptions that caused the Chicago School’s rise to begin with. The effects of these simple market concentration doctrines in the industrial era are all the more dangerous in dynamic digital markets. It may lead to quicker enforcement, but more false positives and worse economic outcomes. Clearly the economic and theoretical underpinnings of this new antitrust revolution require quite a bit more work if policymakers want to avoid repeating the mistakes of the past.
But this isn’t about a new law or new approach, this is about the statutory interpretation of existing merger control law in the UK. The CAT’s judgment could touch on a number of important questions, not least of which is: How much can the CMA rely on structural presumptions around market outcomes, and avoid an in-depth analysis of evidence that contradicts these assumptions? All eyes now await the judgment, which is expected in a few months.
Editor’s Note: Kay Jebelli is counsel to the Computer & Communications Industry Association (CCIA). CCIA was admitted as intervener in support of Meta’s appeal.