Adam Levitin also has a letter in today’s Washington Times today, responding to the op-ed column that Josh Wright and I published on Friday on the antitrust economics of interchange regulation. I’ve followed Adam’s work on this for some time, and I confess that I literally do not understand his theory of the market here. I am not being tendentious or snarky–I mean that I don’t follow the general argument he is making. I’m not sure that he has spelled it out systematically anywhere, so I’ve tried to infer it from the specific points he’s made.
I understand the theory of the Reserve Bank of Australia: they think that the dynamics of the payment card market leads to prices that are too high for merchants and–crucially–too low for consumers. So their view is that precisely because consumer demand is so elastic and merchant demand inelastic, the result is that consumers pay too little for cards. As a result, consumers have an incentive to use cards too often relative to other payment mechanisms such as cash and checks which the RBA believes is more economically efficient for certain types of transactions.
The entire purpose of the intervention in Australia, therefore, was to make pricing more “transparent” and to shift costs from merchants to consumers and thereby to reduce the volume of card use. If you read its reports, hte RBA essentially saw itself as solving a prisoner’s dilemma caused by the fact that merchants “have” to accept cards because of consumer demand. As far as I can tell, the RBA nowhere assumes that there were undissipated rents flowing to issuers or the networks. The argument was one about overall economic efficiency among payment systems.
I disagree with almost all of the RBA’s argument. While it is theoretically possible, I don’t think it is empirically valid. When you say that merchants “have” to accept cards, functionally that is no different from saying that merchants “have” to provide consumers with friendly and helpful sales clerks or clean, well-lit stores. Sure, each merchant could save money if it could pay less and hire surly clerks. But it doesn’t seem to me that merchant stores are too clean or their sales clerks too friendly relative to the efficient level of cleanliness and friendliness. Moreover, in my view, there are many external social benefits from card usage, such as reduced tax evasion, crime, and time savings going to ATM’s, etc.
More specifically, while the RBA’s argument is theoretically possible, I just don’t think the empirical evidence supports the argument that there is a market failure in the payment card network. Given that cross-subsidies are ubiquitous in two-sided markets, the fact that prices don’t match marginal costs is largely irrelevant to the inquiry. Similarly, I don’t think that one can say that just because advertisers subsidize newspaper readers that there is “too much” news being produced and distributed. Furthermore, as I have argued, to the extent that reducing interchange fees leads to term re-pricing in the form of higher annual fees on consumers, that actually has extremely negative effects from a dynamic competition perspective. Annual fees essentially amount to a tax on cardholding, which raises switching costs for consumers, which actually could lead to exactly the sort of economic rents that don’t appear to be present now.
Finally, I disagree with the RBA’s assumption that merchants “must” accept Visa and Mastercard. Costco, for example, only takes American Express and there’s no reason why large retailers couldn’t refuse Visa and MC and pass along the cost savings to consumers (or course they don’t but that’s a different issue). In fact, Adam provides a good example in his letter–apparently George Mason University has stopped accepting Visa for student payments. If Visa were truly a monopolist and sellers had no choice but to accept it, then this couldn’t happen. My argument rests on the idea that the market process at root seems to work here and the pricing anomalies that we see are easily understandable with the theory of two-sided markets. If interchange prices get too high relative to the benefits, merchants will opt out of the network (as apparently GMU has done) and do exclusive deals (like Costco) or accept only cash, check, or something else (like PayPal). Otherwise merchants will continue to accept the card. Looks like competition to me. The network, in turn, seeks to maximize the value of the entire platform or network.
Adam, however, seems to have a wholly novel theory, at least as implied by his letter today. As I understand it, his theory is that the card networks and issuers extract rents from both merchants and consumers simultaneously. In his model, it appears, networks/issuers (I’m not sure who the bad guy is) sets prices that are too high for both sides of the market. Somewhere in here is apparently sustainable economic rents reaped by the issuers and none of the rents are dissipated in competition for consumers. That’s the only way I can interpret his theory that interchange could result in both increasing fees to merchants and higher costs and reduced benefits to consumers.
If that is the model, then I frankly don’t get it. First, I’ve seen no evidence of economic rents. Second, while I at least understand the theoretical argument for why merchants have to accept cards even if they have to subsidize consumers inefficiently (although I disagree with it), I don’t understand as a matter of theory (much less empirics) how issuers could get both consumers and merchants to continue using payment cards when neither side benefits. Why wouldn’t consumer voluntarily use cash if issuers are exploiting them? If consumers “compel” merchants to accept cards, what compels consumers to use cards that neither side wants to use? Especially debit cards, where consumers don’t even have the option to revolve balances.
Third, I understand the argument of how the price of cards might be inefficiently low for consumers because of this rent dissipation, but I don’t understand how the price of cards could be too high or just right. But Adam suggested in a post last week that reductions in interchange fees might just come out of the issuers’ “profits,” with no increased cost to consumers. If that is the case, then it seems like he must be arguing that prices for consumers for cards is not too low, meaning prices are just right, or as I read him today, higher than they should be.
This seems to be the logic of “market failure” that lies behind the Durbin Amendment as well–that there really is a free lunch here that can result in a reduction of costs to merchants but no corresponding increase in costs to consumers. Even though I disagree with the policy of the RBA, at least they understood that reducing costs to merchants was tantamount to increasing prices for consumers. Advocates of the Durbin Amendment, however, seem to believe that they can reduce revenues generated from merchants without finding $20 billion of revenues somewhere else.