How to Tame the Tech Giants: Reverse the Burden of Proof in Merger Reviews

Most tech acquisitions are approved without a hitch, despite growing evidence that they bring little benefit, because regulators are waging an uphill battle to get information in the face of obfuscation by dominant firms. The solution is to reverse the burden of proof: Instead of making the regulator prove that a merger will cause harm, make the merging parties prove that it won’t.

As you read this, please remember two numbers: 1,000 and zero. The former number represents the 1,000 acquisitions made by GAFAM (Google, Amazon, Facebook, Apple, and Microsoft) in the past 20 years. The latter is the number of those acquisitions actually blocked by regulators worldwide. 

Mergers and acquisitions have directly increased market concentration across Europe and the US, including in brewing, supermarkets, hospitals, car rentals, eyeglasses, crop seeds, industrial chemicals, meat packing, wireless carriers, and many others. Detailed studies document a rise in concentration and market power, a decline in the rate of new firms entering markets, and a fall in labor’s share of economic output.

When I was Chief Competition Economist at the European Commission from 2016-2019, I found myself in an unbalanced tug of war. My team of 30 (really excellent) economists, supervising the economic analysis of every major merger, antitrust, and state aid case in Europe, faced dominant companies that could hire any number of lawyers against us. They paid an army of consultancies to create doubt around an issue, while the burden of proof was on us to dismiss their claims. This burden was made worse by asymmetric information: the companies had all the incentives to hide vast amounts of information from us and use it selectively to create more obfuscating smoke.

As we tried to build our cases, the dominant companies bogged us down in endless, esoteric arguments about definitions of markets and other matters. This has a long history. When Facebook bought Instagram in 2012, for instance, the discussion between the enforcers and the parties ultimately revolved around the “supply of photo apps,” which would be of limited interest to advertisers because “eyeballs are not on the app for a significant period of time” and “limited user data is captured.”  Of 8,000 cases analyzed by the European Commission since 1990, only 30 have been prohibited.

Most acquisitions have been approved without a hitch. In some cases, remedies have been imposed. But adverse experiences have exposed the difficulties of imposing and monitoring effective remedies, while academic studies have documented the frequency with which remedies have failed. This is particularly true for conduct or behavioral remedies, which prescribe specified actions on the part of the merged firm. Examples in the US include the conduct remedies in the merger of Ticketmaster and Live Nation in 2010, the merger of Comcast and NBCU in 2011, and, more recently, the 2020 merger of Sprint and T-Mobile. 

In the case of the Ticketmaster/Live Nation merger, the remedy was that Ticketmaster would not condition the provision of Live Nation’s “entertainment events” on whether the purchaser had also used Ticketmaster’s ticketing services. Ticketmaster nonetheless did precisely that, contending that the language only prevented it from conditioning on provision of all Live Nation events, not any single one or a few. In this manner, the merged company effectively nullified the remedy for nine years, and was subject to no penalty for its strategic interpretation of the clause. 

In one of the few digital mergers analyzed by the EU enforcer, when it acquired DoubleClick in 2007, Google promised to keep its database of web-browsing records separate from the names and other personally identifiable information collected from Gmail and its other login accounts. This was changed a few years later when, as Julia Angwin wrote in 2016, Google “literally cross(ed) out the lines in its privacy policy that promised to keep the two pots of data separate. In its place, Google substituted new language that says browsing habits ‘may be’ combined with what the company learns from the use Gmail and other tools”. Given its dominance in this and many other markets, consumers simply had no choice but to swallow yet another privacy loss. 

The implication that runs through these and similar experiences is the difficulty faced by the competition agency in fashioning a remedy that a determined and incentivized firm cannot evade or avoid.

Problems also arise when it comes to divestitures or structural remedies. Despite their promise of being a “third way” between outright rejection and unconditional approval of a merger, the reality is that divestiture remedies routinely fail. The Federal Trade Commission has conducted two studies of the effects of divestitures in actual merger cases. The first study, in 1999, declared divestitures to be successful if the divested assets remained in the market and were financially viable. Even by that weak standard, a considerable fraction were failures. A second FTC study in 2017 suffered from a number of methodological flaws, but nonetheless again found a similarly large portion of divestitures to be unsuccessful. Incidentally, studies about how well a regulator has done should be more numerous and systematic, and probably not be conducted by the regulators themselves.

Since Robert Bork and his views that most mergers are either competitively neutral or efficiency enhancing prevailed in many antitrust circles, we hear a lot about the “efficiency” promises of big mergers, as scale will allow the resulting larger company to bring down costs to the benefit of consumers. The reality is that this proposition is hardly brought to test with data. An ex post assessment of merger-specific efficiency is very rare. And when it’s done, despite the data limitation, the results show the opposite, from electricity to transportation, to name just a few. If you think about it, this makes sense, as we are talking about mergers involving the largest dominant firms in a certain industry: these companies have most likely already achieved their efficient scale—how much larger should they become? If size drives efficiency, the Soviet Union should have won the Cold War, and we know how that went. 

The solution is to reverse the burden of proof with what we call a “rebuttable structural presumption.” Instead of making the regulator prove that a merger will cause harm, make the merging parties prove that it won’t. Start with a list of firms that are super-big in size, systemic importance, or economic power (that’s the structural bit) and revise the list every five years. Next, presume that we do not want these giants to grow even bigger by gobbling up more companies. But then we pass the ball to them, with the rebuttable part: “Can you prove that this merger will benefit consumers and that it is the only way to bring these benefits?” If not, the merger is blocked.

If we reverse the burden of proof, we would not waste time on endless market definitions. It also efficiently deals with asymmetric information: to succeed, merging parties must produce the information that they currently hide. Many mergers would be blocked (or would not even be attempted, vastly reducing regulators’ workloads). This is just what our societies need: economics tells us that once you are a large company, further concentration creates little benefit to consumers, and the social gains from having strong alternatives to Google, Facebook or Amazon would be immense.

The advocates of Big Tech will, of course, attack this proposal. “It will punish Big Tech firms for their success!” Seriously? These companies have never been so profitable. Why should Google have been allowed to buy Fitbit? Couldn’t it have achieved the same technological result itself, leaving Fitbit to compete with it (either alone or purchased by another firm)? Prove to us, Google, that you really cannot do it without buying Fitbit.

“It will kill the small entrepreneurs!,” they will also cry. Well, small, successful, innovative startups currently get purchased and then help dominant companies entrench their positions. Entrepreneurs have started talking openly of “kill zones,” where venture capitalists are reluctant to fund entrants that might compete with dominant platforms. In any case, my proposal will not stop startups being bought by other competitors of the super-big firms. We could also help the founders of tech startups realize their dreams of going public by making the IPO process easier, thus making it easier for them to grow their companies and reducing incentives to exit by acquisition.

The idea of reversing the burden of proof is starting to gain traction. It is a fundamental pillar of the recent five bills introduced in the House of Representatives by Rep. David Cicilline (D-RI) and advanced by the House Judiciary Committee, in particular the Platform Competition and Opportunity Act, requiring companies, not the government, to prove their case. Similar proposals are now being considered by the UK’s Competition and Markets Authority (CMA). France, Germany and the Netherlands are also pushing for greater intervention powers, putting pressure on the European Commission that has recently been silent on this.

The Covid-19 pandemic shows that we need stronger, more resilient economic ecosystems, not under the thumb of a few giant companies. Economic research shows that we want as much rival innovation effort as possible in concentrated industries, and that buyouts of promising innovators deprive us of that. This proposal would also give American and European startups and local businesses a chance, instead of entrenching US firms’ dominance.

Ultimately, policy should reflect our views as a society. We have proceeded for years on grounds that over-enforcement chills innovation and is worse than under-enforcement, which supposedly gets quickly corrected by the growth of rivals or new entrants. The recent track record in tech put that argument convincingly in the ground.

Remember those numbers: 1,000 and zero. It is time for a bold new approach.