An interesting (if not insidious) argument in Todd Zywicki’s much-discussed* paper [PDF] on credit-card interchange fees is the claim that two-sidedness of the market — in a nutshell, it’s necessary have both buyers and sellers get on a payment platform as a prerequisite to carrying out transactions — upends the traditional efficiency appeal of pricing at cost:
The claim that interchange fees are too high is also problematic as a matter of economic theory, as it fails to appreciate the “two-sided market” of credit card networks in which overall network efficiency arises from the interdependencies of the network’s various actors and the ability to reallocate costs among them in order the maximize the value of the system as a whole on behalf of the beneficiaries of that system as a group.
There’s already one sleight-of-hand here, in that the theory suggests that cross-subsidies from one side of the market to the other may be efficient but not that platform owners’ economic profits are innocent. Moreover, two-sided markets theory doesn’t specify which side(s) of the market get the subsidy as a general matter, so you can’t simply say “two-sided markets hence this subsidy.” I’m not suggesting that readers not believe in potential network externalities, just warning that there’s considerable risk of just-so storytelling.
To that end, I suggest a thought experiment. Imagine a world where the price on the supposedly subsidizing side were lower than it is now, and the subsidized side were higher, and the state of the market that implies. There’s a two-sided markets argument to make for additional subsidy if you would look at the resulting market conditions and say, “Wow, we really should impose a tax to bring up the subsidizing price to today’s level and provide today’s subsidies to overcome obvious network externalities!”
In the case of credit-card interchange fees, this state of the world isn’t terribly hard to imagine. Set the way-back machine to the early- to mid-1990s, and there are lower interchange fees, few rewards cards, and fairly high rates of adoption of credit cards. (The dominant forms of rewards cards at the time are oil company loyalty cards, which as Adam Levitin observes are a conceptually different form of “subsidy,” and airline frequent-flyer affinity cards charging cardholders above-average annual fees.) It’s 1991. Should we impose a tax of a quarter or a half-percent on (at least) a portion of merchants’ receipts to give out cash or frequent-flyer miles to subsidize further credit-card spending?
My guess is that if I walked into a GMU law and economics seminar in 1995 and proposed a sales tax to give perks to the credit-card constrained** to overcome network externalities in payment markets that I’d have been laughed back to College Park. But then I’m just cynical.
** Not necessarily the actual market outcome.