Thread by @CarmeloCennamo

Platform envelopment & killer acquisitions in the spotlight as @Meta (i.e, @Facebook) appeals @CMAgovUK decision to sell @GIPHY. Here’s why I think this could be good news….also for regulators! But some background first. A thread 🧵

#platform #regulation #digital #competition

First, Giphy - what’s that? Check it out, push the gif button 👇

Well…. in this case the gif animation is provided by Tenor, @GIPHY direct “competitor”, owned by …. 🥁 🥁 🥁 @Google !!! Oh yes…

Technically, it’s an online GIF search engine and database. Main product: gifs and animated stickers. It’s an integrated functionality in @Facebook & other @Meta #platforms. Is @GIPHY a platform? A potential competitor to @Meta? @CMAgovUK thinks so!

@CMAgovUK argues that @Meta purchase of @GIPHY lessened competition in the UK in the display advertising and social media markets. Giphy would have become *otherwise* a contender of Meta in the display ads mkt. Post-transaction Meta can foreclose its rivals from access to Giphy

This will make rivals’ services less attractive to users, and increase Meta’s dominance in social media and display ads, on the ground that GIFs are a critical part of a customer social experience (sort of an essential facility to digital socialization?) …🤔

So is this #platformenvelopment with the sole purpose of killing potential competitor? Eisenmann, @g2parker and @InfoEcon hold that platform envelopment occurs through bundling of platforms’ functionalities so as to leverage shared user relationships and components.

This kind of bundling can thus create economies of scope and free up increased value for users and undermine small platforms with standalone functionality. They show conditions under which this is more likely - in some cases the target is fully displaced, in others they coexist

You can check the paper here: papers.ssrn.com/sol3/papers.cf…

The CMA doesn’t make any considerations of complementarities for users (and among the two services) or the broader impact on cross-platforms competition. It only argues on the ground of restraints of *dynamic competition* along the vertical relationship

So what’s dynamic competition? Competition but in the future…prospective. So, what looks like a little frog today is expected to become a prince tomorrow and dethrone the king.

@CompetitionProf & David Teece give more complex account of the concept, articulating the distinctive nature and implications….which is not necessarily what the CMA has in mind here academic.oup.com/icc/advance-ar…

The CMA analysis is still sort of static comparative analysis between state A and B, holding everything else equal, except now state B is held in the future, as an IF-Then scenario. But on which ground we fix that possible equilibrium (?) point B in the future?

And here’s where I see the problem with current antitrust tools grounded on narrow relevant market definition. First: how do we assess the potential restraint of dynamic competition without fully accounting for the relevant actors in the competitive landscape beyond the target?

Second: what is “dynamic”? When do we have a restraint? If it suffices to show the mere possibility of future competitive scenario, by its logic, any acquisition by firms in dominant position should be blocked on this basis. So, this cannot be a proper principle.

So the relevant pending question is how do we establish the specific threshold and likelihood in a given case? That’s where the counter factual analysis should come to help. A good starting point could be looking at Giphy rival services. Are they eating at Meta’s ads revenues?

Are they steering users away from Facebook products thanks to their GIFs services? Are they attracting new advertisers, constraining Facebook’s mkt share? That would have been useful analysis to establish the EXENT Giphy could have turned from complementor into competitor

Third: What about the potential cross-platform competition to @Google services? After all, Giphy is a specialized search online service, competing directly with Tenor, owned by Google…is this competition across platforms not relevant?

In a working paper with @ivankavisnjic we show how platform envelopment by “GAFA” platforms over the years has led to increased overlap of services and thus increased coopetitive relationships among them: papers.ssrn.com/sol3/papers.cf…

The @CMAgovUK has considered the market for ads on social platforms as the relevant market, but doesn’t consider online search, which is the primary functionality of the “platform”. Why?

If @Meta is forced to undo the merger, shouldn’t also @Google be forced to sell its @gifkeyboard (Tenor) service? Wouldn’t this create market competitive distortions in the broader platform market competitive landscape, giving @Google even more power over online search (and ads)?

Which brings me to the core of the issue in such kind of antitrust analysis: trade-offs and counter factual! We need better tools to anchor the analysis to critical structural dimensions of the competitive landscape, which in platform markets must consider the within and across

So, why is the appeal good news for CMA? ‘Cause it pushes us to develop more rigorous tools which are able to tease out the bad (purely anticompetitive) from good (pro innovative) platform envelopment [pls stop calling them killer acquisitions - it’s platforms not pharma!]

Also, because it might save one important exit option for VC investors, and with it the tech startup ecosystem that gravitates around tech platforms…not just big tech but also deep tech emerging around industrial leading firms!

Recent work by Prado & Bauer show how acquisitions by big tech firms over the last 10 years have generated positive spillover effects on innovation through further investments by VCs
papers.ssrn.com/sol3/papers.cf…

There are also economic models showing when such acquisitions result in efficient upgrades of target’s technology (Motta & Peitz)
papers.ssrn.com/sol3/papers.cf…

or when making such acquisitions more difficult may result in less VC investment, and thus lower innovation (Cabral). sciencedirect.com/science/articl…

Shall we not consider these broader implications and trade offs? I would submit that, for imposing structural separation, we need stricter criteria, ones that look at the impact of the resulting structure, within the (narrowly) defined market, but also beyond it.

@threadreaderapp unroll