Zywicki on Interchange

Todd Zywicki has a new paper out on interchange regulation, just in time to support the banks' push against the Durbin interchange amendment in conference committee. The paper doesn't present any new arguments or evidence.  Instead, it presents a highly polemical form of antiregulatory claims. 

There's an awful lot to criticize about this paper, starting with its complete unwillingness to engage with pro-regulatory arguments and evidence on anything beyond a strawman basis.  The omission of the findings of the Reserve Bank of Australia (and reliance on a MasterCard funded study instead) on the impact of Australian regulation is remarkable.  

But don't take my word for it.  Zywicki gets spanked around pretty soundly by the Australian economist Joshua Gans, who objects to the way his work is used by Zywicki in a "very selective and misconstrued way" in a paper whose "broad conclusions" are "flawed." 

Let me add my own broad objection (I'll probably blog on more of the details later).  Zywicki's general assumption about bank regulation is that if fee type A is regulated, then fee types B and C will increase to offset the regulation.  That might be the result; indeed, it is a variation on the whak-a-mole bank fee thesis (also here), that if fee A is banned, new fees B and C will sprout up. 

But there is another possible regulatory outcome that Zywicki never considers:  banks might simply have to endure lower profit margins.  If the consumer side of credit card pricing markets is competitive as Zywicki believes (I've got my doubts, which is the point of the whak-a-mole thesis), then the result should be smaller profit margins, instead of shifted fees.  Zywicki seems to take it as a given that banks must maintain profitability levels.  But they don't.  That's the nature of capitalism:  bank have a right to make a profit, but only through fair and legal competition. If a bank can't operate profitably under those conditions, should it really be in business? 

Comments

[Todd Zywicki]

Adam:
What is your evidence (1) that these are pure economic rents here and that (2) they are not dissipated, such as in competition for consumers? If there are rents, who is earning them? And if they (whoever it is) can earn economic rents, why don't the networks abuse consumers too?

I thought your whole argument was that card issuers compete excessively for consumers, leading to inefficiently low (zero or negative) prices for consumers. And that the purpose of interchange price controls is to increase costs for consumers and force them to "internalize" the costs of their card use. Are you now saying that price that consumers pay for cards is just right and is not too low? If you still believe they are too low, but that there are pure economic rents here that are not dissipated, then how in the world did the consumer prices get to be too low if they are not being offset by merchants? Otherwise, it seems like your argument assumes that if there are rents, they are dissipated in excessive competition for consumers in the form of a zero or negative price (such as rewards).

Even if there are rents to someone here, the only way that prices would not go up for consumers would be if none of those rents are currently being dissipated. In which case there is also no argument that consumer card prices are too low or that cards are overused as a result.

All of this assumes, of course, that there are actually rents present in the first place, which I haven't seen any evidence that is the case.

As for Gans, his research simply does not show what he claims it shows. Period.

As for whack-a-mole, there is evidence going back until at least the 1930s that regulation of some terms of credit contracts invariably lead to changes in prices of other terms. There are dozens and dozens of empirical studies that support the hypothesis, all of which were done in credit markets less competitive than those today (which is also well-verified by economists).

[Adam Levitin]

Todd, thanks for engaging. Here's some prima facie evidence of economic rents earned by both the networks and the issuers. This spring, Visa's Interlink debit network raised fees by 18-40% (depending on merchant category) and made it more difficult to get into the lower fee categories. No loss of market share. And no increase in rewards programs as a result. That sure looks like economic rents earned by the issuers to me.

There’s also a recent example from the credit card market. While credit card interchange has remained stable (on a weighted average basis) for the last couple of years, the total cost for merchants has risen because there’s another pig at the trough since MasterCard and Visa’s IPOs. The networks are now being run on a for-profit, rather than mutual basis, and the shareholders want to be fed. So network assessments have increased on MasterCard and Visa, but not on Discover and Amex. Again, a small, but significant increase in price without any loss in market share for MasterCard and Visa. Raising fees to feed another mouth via dividends to shareholders is pure extraction of economic rents by the networks.

You don’t have my argument correctly. My argument is not that card issuers compete excessively for consumers resulting in inefficiently low or zero prices. Instead, it is that card network rules effectively fix the cost of all payments system as being equal to each other. That means consumers decide what payment system to use based solely on benefits, not benefits net of cost. Cards generally have more perceived consumer benefits than other systems; this has very little to do with issuer competition. This means that total card usage is inefficiently high.

Here's where I think you make a wrong turn. You say that card networks compete for issuers and that card users benefit from higher interchange fees in the form of lower prices and/or greater rewards. In other words, issuers are taking money (indirectly) from consumers via interchange and then giving it back to them in lower rates or greater rewards. In other words, even if there are rents earned by the issuers, they are dissipated. (Notice that still leaves the problem of the network rents, even if correct). In the Coasean world with perfect competition, full dissipation would mean that 100% of interchange fees above the de minimis cost of actually processing the transaction go back to card users via lower interest rates or rewards.

But we know that’s not the real world. Issuers do not pass thru anywhere close to 100% interchange above processing costs because the card issuance market is not price competitive. Card issuers compete on things like image, rewards, and salient price points, not total price. While networks compete for issuers, that simply does not translate into consumer benefit. Because of imperfect competition in the issuing market, issuers are able to retain most of the interchange fees they levy. Those are economic rents.

Your concern is that if interchange fees are regulated, merchants will effectively become the rent receivers, and these rents won’t be dissipated via lower prices to consumers. No doubt, merchants will not always pass thru all the savings to consumers; the pass-thru depends on how competitive their particular industries are. But merchants are, in general, much more price competitive than card issuers, so there is likely to be much greater rent dissipation via merchants competition for consumers than via card issuer competition for consumers.

Consider Wal-Mart. Its business model is to squeeze suppliers in order to pass thru savings to consumers. The goal is to be able to offer a store that has everything and at low prices in order to bring in lots of business. Low margin, very high volume. Do you really think that Wal-Mart isn’t going to pass thru the bulk of the interchange savings, just the way it passes thru savings from every other supplier?

Finally on Whak-a-Mole™. I think we agree that is a problem for any attempt at credit regulation. I think we differ on the response. The solution, it seems to me, is not to throw up our hands, but instead to adopt a regulatory in which only approved fees classes can be charged. Figuring out what those fees types are is a tricky issue, but the net result is to prevent additional fees types from popping up in response to regulation.

There is still the problem of fees being shifted from one type to another. But because certain types of fees (e.g. annual fees) are more salient to consumers than others (e.g., interchange fees), it isn’t possible for card issuers to do a dollar-for-dollar shift in fees. Forcing fees toward transparent, salient points enhances competition and pushes down total fee levels. Our goal should be a card market with as close as possible to perfect competition and maximum consumer choice. Unfortunately, this isn't a market in which a pure laissez-faire approach will achieve that.