Warning: This post is going to be a bit esoteric (perhaps “wonkish,” as Paul Krugman would say). It also may not be very long. It may even be a bit obvious.

Okay, so now that I’ve scared away all possible interested readers, I had a mini-epiphany while watching the ABA Antitrust Section Spring Meeting panel on dominance in novel markets (i.e., market power in the tech sector).

Unfortunately, it’s going to take a bit of set up to get there. Remember way back when we had a multi-part discussion on effeciencies? To just do the briefest of review, in antitrust an “efficiency” is when a merger or other combination will result in synergies or cost savings. Such efficiencies are a defense under the Horizontal Merger Guidelines, which essentially weigh the value of any efficiencies against any projected anticompetitive effect of a transaction. If the efficiencies are large enough, the agencies are supposed to take a pass on attempting to block a transaction so that consumers can enjoy the benefits of the new, more efficient, combined business.

If you’ll stretch your memory just a bit more, you’ll recall that there are three criteria that the Guidelines require for an alleged efficiency to be considered in defense of a transaction. To count, an efficiency must be: (1) merger specific, (2) cognizable, and (3) (usually) in the relevant market.

The qualifier “usually” in the third requirement is the result of a footnote added in the 1997 revisions that says that an out of market efficiency can nonetheless be considered, in the agency’s prosecutorial discretion, if it is “inextricably linked” such that it cannot be preserved via a divestiture that eliminates the anticompetitive effect. In effect, the footnote says that the agencies will weigh a harm in one market against a benefit in another market if there is no other way of getting the benefit.

The problem is, what counts as “inextricably linked?” It’s hard to say. There will almost always be some other theoretical means of enjoying the efficiency without harming competition (e.g., a contractual arrangement that shares what’s needed but doesn’t eliminate competition). Coming up with an example of an inextricably linked efficiency is sufficiently difficult that the Merger Commentary discussion of the topic doesn’t really give an actual example. In fact, I’ve had conversations with two gentlemen who claim to be authors of that footnote in which they’ve said they don’t know for sure either.

And now for the mini-ephiphany (please remember that I said “mini”): two sided markets.

What’s a two-sided market? Well, the classic example is a newspaper. A newspaper publisher has at least two sets of customers: advertisers and subscribers. Traditionally, newspapers have sold their services to both, but what makes a two-sided (or multi-sided) market interesting to analyze in the merger context is that a transaction can have different effects on different sides. If a transaction combined the two hypothetical leading daily newspapers in Newstopia, for example, there may be sufficient remaining competition on the subscriber side from national papers and other news sources, but you could still have significant harm to local advertisers who no longer have two options to keep ad prices low.

Or to move into the 21st century, there are many internet business models that show features of two- (or multi-) sided markets. A future combination among them may provide the first real example where the benefits to users were sufficiently great and inextricably linked to a loss of competition on the advertiser side (or vice versa) to save the transaction.

Okay, so now you know what I mean about obvious. If you’ve given even a moment’s thought to efficiencies in mergers with multi-sided markets, and you know the merger guidelines, you’ve probably thought about how yeah, you’ve got a good argument for inextricably linked. It seems I didn’t make it that far until a few weeks ago.