How the pre-merger notification regime came about and why we should be careful about discouraging useful mergers.
December 2020 was an extraordinary period in US antitrust law. On December 9, 2020, two cases—one by the Federal Trade Commission and one by state attorneys general from almost every state—were brought against Facebook. One week later, a group of states, led by Texas, brought a new case against Google. And the following day, states, led by Colorado, brought an additional suit against Google. That means Google is now facing three government cases, as the federal government and another group of states had filed an antitrust case against Google on October 20, 2020.
The two cases filed against Facebook focus on the company’s purchases of Instagram and WhatsApp, though both complaints have a tag-along issue as to whether Facebook has manipulated access to the Facebook platform in an anti-competitive manner. And the Texas complaint against Google is noteworthy in that it alleges a conspiracy in violation of Section 1 of the Sherman Act between Facebook and Google regarding advertising markets and yet doesn’t name Facebook as a party to that action. (If you want more on that allegation, The New York Times has dug into the background and a private party has now brought a case based on these allegations.)
I want to focus on the Instagram and WhatsApp mergers. Considering those mergers raises important questions about how we should organize merger inquiries, why we should favor or disfavor mergers and when and what remedies might be sensible. A key focus for my analysis is the timing of when merger review should take place and how that timing matters for investment decisions. (This is a natural issue to think about the Facebook complaints, as my former student Dirk Auer emphasizes in his piece on these issues.)
Facebook bought Instagram for $1 billion on April 9, 2012. Mark Zuckerberg’s original statement on the purchase situated the two firms: “For years, we’ve focused on building the best experience for sharing photos with your friends and family. Now, we’ll be able to work even more closely with the Instagram team to also offer the best experiences for sharing beautiful mobile photos with people based on your interests. We believe these are different experiences that complement each other.”
Both Facebook and Instagram make it possible to share photos, but somehow do so differently—OK, if you say so, Mark. The reality, of course, was that Instagram was born on mobile and Facebook was not, and platform transitions are key points for competition. Lotus 1-2-3 was the leading spreadsheet of its day but it bungled the transition from the text interface of MS-DOS to the graphical user interface world of Windows. Microsoft came to the internet late and saw the Netscape browser as creating a risk to its operating system monopoly.
Notwithstanding all of that, given the chance to object to the Facebook and Instagram deal, antitrust officials passed. The FTC closed its investigation on August 22, 2012 with a brief statement. The UK Office of Fair Trading closed its investigation the same day with a more extended statement of why it didn’t find the purchase problematic, noting that “[i]n summary, the evidence before the OFT does not show that Instagram would be particularly well placed to compete against Facebook in the short run.”
Facebook’s February 19, 2014 purchase of WhatsApp follows a similar pattern. The deal involved more money, $16 billion, but again, the FTC largely passed on it and the European Commission set forth an extensive analysis before greenlighting the merger.
That gets us to the crux of the matter. Prior to 1976, the US mainly reviewed mergers after the fact. When the US broke up Standard Oil in 1911, it did so pursuant to Sections 1 and 2 of the Sherman Act and only after Standard had already completed its acquisitions. The breakup was intended to restore the state of competition that existed prior to Standard’s buying binge. As new laws were passed in 1914 and 1950, the legal framework for evaluating mergers evolved but again the pattern was merge first, challenge and breakup later.
That all changed with the passage of Hart-Scott-Rodino in 1976. That created a pre-merger notification regime. The legislative history to HSR focuses on the difficulty of restoring the ex ante competition conditions that existed prior to the merger. “Unscrambling the eggs” is the phrase often used to capture the challenge. Congress believed that the purpose of antitrust laws was being undercut in having mergers go through that could not be undone effectively later, so it switched from ex post review to ex ante merger review.
How should we think that the timing of merger review matters? Consider a hypothetical. A firm faces an investment decision in how it might enter a new market. It could build the new product purely internally and would do so outside of the risk of a subsequent antitrust review. Alternatively, it could buy a preliminary version of the product and build on that. This is a standard buy vs. build framework, with the added wrinkle of a possible antitrust review if the firm chooses to buy rather than build.
What exactly is at stake here? Take the extreme version of this to see the idea. One firm, call it the incumbent, does a poor job of creating new products but it has great scaling skills. It can take preliminary products and really run with them in a way that creates firm value and social value. A second firm, call it the entrant, is great at first drafts but can’t scale things at all. If we won’t allow the second firm to sell its drafts, those drafts might never get built in the first place as the second firm doesn’t have a path to monetizing those other than selling them to a scaling firm. I don’t think that is an odd pattern: In the world of patents, we don’t necessarily expect inventors to be good at commercialization. Specialization in different parts of the market may be valuable and we would want the invention to move between firms.
Add another risk to the mix. You are a firm with a nice market position. You would like to add a new second product, but there is some risk that that product as built out would be a competitor for your original product. That isn’t a problem if you control the product, but it would be a real problem if you were forced to spin off that product. Again, we might create social value in creating the new product but the firm might not be willing to do that if the possible breakup risk was high. We care about competition, but firms and their shareholders care about private value and don’t want to invest in creating competitors.
This obviously isn’t an argument for no antitrust review of mergers, as the horizontal merger guidelines set out a clear framework to evaluable mergers that could reduce competition. Instead, the point here is about the timing of review and the possible risks of later reviews in discouraging otherwise sensible buy v. build investment choices. Doing the review at the point of purchase rather than later means that the prospective purchaser won’t face the risk of buying a firm, investing in it and thereby creating a much stronger competitor when the firm faces a subsequent merger challenge and a possible breakup.
The legislative history of Hart-Scott-Rodino is again worth reviewing. One of the Senate Report on the draft bill set forth part of the testimony of then-Assistant Attorney General Thomas E Kauper. Kauper noted the problems with subsequent divestitures in practice but then turned to a second point: “Pre-merger notification will also advance the legitimate interests of the business community in planning and predictability. It will enable firms to make post-acquisition changes with much more confidence than they can at present.” This is a system where we block mergers at birth and not years down the road.
There is some chance that those favoring more ex post reviews of mergers would regard the possibility of discouraging mergers as a feature of the system and not a bug. Breaking up Facebook and discouraging other firms from buying young firms with new products? That is a twofer. To reach that conclusion, it seems to me, you have to have strong priors about the disutility of mergers and that has to turn on views about the distribution of product formation and scaling skills.
We should worry about discouraging useful mergers, as it isn’t easy to create social value. We should expect that the willingness of firms to undertake mergers and invest in the assets that they will acquire will be shaped by the risk that they will later be forced to divest those assets, especially if the new and improved firm will immediately start competing with it. It would be nice if firms and their shareholders were interested in competition for its own sake, but they aren’t. The trade-off here is losing beneficial mergers vs. creating more competition ex post. You can change those tradeoffs if the buying firm knows before making the investment that once the investment has been reviewed, it doesn’t face a break-up risk.