By Richard A. EpsteinThe 2010 enactment of the Durbin Amendment as part of the Dodd-Frank Act set into motion an extensive round of administrative rulemaking and litigation that may well have run its course with the recent unanimous opinion of the Circuit Court for the District of Columbia, written by Judge David Tatel for himself and Senior Judges Harry Edwards and Steven Williams in NACS (formerly National Association of Convenience Stores) v. Board of Governors of the Federal Reserve. The outcome of the case was to sustain the decision of the Federal Reserve to allow the banks that issue debit cards to recover $0.21 cent on average in debit card transitions. In so doing, the Court reversed the decision below by Judge Richard Leon, which was openly contemptuous of the arguments of the Fed that carried undue weight in the Court of Appeal. It is a long saga in which no one is covered with glory. To set this in context, it is therefore regrettably necessary to review some of Durbin's tangled history.
The Durbin Amendment The debit card was one of the great commercial innovations in American banking. Starting from a standing start in 1995, it managed by 2009 to become the dominant form of payment in the United States, eclipsing the venerable credit card both in number of transactions and in dollars transferred. One might have thought that this enviable record of success would have won plaudits across the board, for no program can enjoy such success if it does not create net gains to all the parties who contribute to the system.
In this case, those parties numbered five. In the middle of the picture lay the credit card companies, chiefly Visa and MasterCard, which orchestrate transactions between two sides of the market. On the one side lie the credit card holders who received their cards from issuing banks. The key feature of the pre-Durbin arrangement was that the debit card holder paid no monthly or swipe fee for the use of the card. Instead the cost of servicing and recruiting the debit card holders was funded by an interchange fee that was paid to the issuing banks from the retail merchants who accepted the cards. These merchants also paid a fee to the acquiring banks that serviced their accounts, and a smaller fee to the credit card companies that orchestrated the transaction from the middle.
In NACS, Judge Tatel accepted the Durbin fairy tale that this entire arrangement reeked of market failure because of the high level of interchange fee charged for the occasion. But at no point does he explain what the correct fee ought to be, for his only account of market failure is that merchants discovered that they could not do without the card, from which, however, it does not follow that they will pay anything to get it. Rather, what happened was that the credit card companies in discharge of their contractual obligations set the interchange fees at a level that allowed all parties to prosper. The use of that payment in these two-sided markets in effect put the cost of running the system on the parties for whom demand was inelastic (i.e., relatively unresponsive to price changes). The lower prices offered to cardholders thus increased the number of card users, which in turn allowed the fixed costs of running the system to be amortized over a larger customer base. And those interchange dollars funded the special benefit packages that kept debit cardholders coming into the system. In a word, the system was not broken, and the Durbin Amendment did not fix it.
More specifically, the Amendment introduced its own novel inefficiencies by its government command. The relevant text has to be set out in full in order to understand the bizarre nature of the Circuit Court's decision. It reads as follows:
Section 920 (2) Reasonable interchange transaction fees The amount of any interchange transaction fee that an issuer may receive or charge with respect to an electronic debit transaction shall be reasonable and proportional to the cost incurred by the issuer with respect to the transaction.
In prescribing regulations under paragraph (3)(A), the Board shall--
(4) (B) distinguish between--
(i) the incremental cost incurred by an issuer for the role of the issuer in the authorization, clearance, or settlement of a particular electronic debit transaction, which cost shall be considered under paragraph (2); and
(ii) other costs incurred by an issuer which are not specific to a particular electronic debit transaction, which costs shall not be considered under paragraph (2).
The correct reaction to this sorry provision is that it is both clear and misguided. The initial material in subsection (2) is deeply uninformative because setting fees that are "reasonable and proportional to the cost incurred by the issuer with respect to the transaction" gives no hint of the horror to come. That capacious phrase clearly covers all costs, both variable and fixed, associated with the transaction. On this view, the provision does not put any constraint on the fees that could be charged above and beyond those found in a competitive market, which would lead to that result.
The entire sense of the provision takes on a darker meaning in the light of Section (4)(B), which gives a definition that is far more restrictive than the general statement above. It divides the world into two kinds of costs and makes it clear that only the "authorization, clearance, or settlement" costs for a particular electronic transaction should be considered under paragraph (2) while all other costs are removed.
Judicial Obscurantism in the Court of Appeals It does not take a genius to conclude that the listing of these three transaction-specific costs excludes all the overall costs needed to design, operate and maintain the system. By design, those were to be cast back on the issuing banks to recover from their own debit card customers. Try as one might, it is not possible to see any gaps in the statutory structure. The only way in which this could have been made clearer is to have inserted the word "all" before "other costs" in paragraph (ii). But it is hard to resist the conclusion that Senator Durbin, perverse though he be, knew exactly what he was doing with his own Amendment. The Senator was devoted beyond all measure to Walgreen's and other retailers and equally intransigent with respect to the banks, so it is a virtual certainty that he meant what he said--and said what he meant. The retailers had excellent lawyers to help Senator Durbin along his appointed path. Judge Tatel called the Durbin Amendment a badly drafted statute, but that charge is surely wrong. Incompetently conceived, surely, but accurately drafted, regrettably, is a much better account of Durbin's regulatory calamity.
At this point, the contrast between the learned obscurity of Judge Tatel and the blunt clarity of Judge Leon's opinion below is a sight to behold. The key argument of Judge Tatel is that this text could "easily" be regarded as ambiguous so that it is correct for the Board to allow "issuers to recover, equipment, hardware, software and labor costs since [e]ach transaction uses the equipment, hardware, software and associated labor, and no particular transaction can occur without incurring these costs." Judge Leon rightly dismissed that claim in one word: "Please."
Leon's terse view of statutory interpretation makes infinitely more sense than the tendentious reading Judge Tatel, who relied on this identical passage to incorporate the semiotics of Jacques Derrida or the post-structuralism of Michel Foucault into modern administrative law. Finding, or inventing, ambiguity where none existed, he gave the views of the Federal Reserve undeserved prominence under the regrettable Chevron doctrine that has courts defer to agencies when statutes are found ambiguous.
To conjure up that needed ambiguity, Judge Tatel launches into an extended, prolix, and tedious discussion of restrictive and nonrestrictive clauses, which, he claims, allows the Fed to infer this third class of expenses lurking in the shadows that the Fed by rule recover through debit interchange.
We are in an ethereal world. These unspecified objects might be called "fixed, variable costs". But suppose that these costs, like the Loch Ness monster, do exist. It nonetheless remains true that the impatient Judge Leon offers the only tenable reading of the Durbin Amendment: these fixed costs of running the computer network were excluded along with every other business cost needed to keep the program going, without which any particular transactions would not happen.
Economic Redemption, of Sorts As a matter of statutory interpretation, Judge Tatel's opinion is an intellectual train wreck. But functionally, it supplies a most welcome result, because of the hopelessly confiscatory nature of the Durbin Amendment, which on its face would have make made it impossible for the banks to recover their extensive invested costs in their operational system through interchange, without supplying them any alternative. To be sure, there was extensive talk of how banks should charge their own customers monthly or swipe fees. But those were never collected, after they were buried in an avalanche of abuse, starring the ubiquitous Senator Durbin who wrote the heads of Bank of America and Well-Fargo gratuitously nasty letters asking that they rescind the fees that only months before were supposed to be their salvation.
The net result was that the banks could not recover their invested capital sunk in these debit card systems. This whole statutory system borders on the farcical because it overlooks its long-term stability and success. In many places I have urged that the entire statute should be struck down as a confiscatory taking. That decision was resisted in the earlier and misguided 2011 decision of the Eighth Circuit in TCF National Bank v. Bernanke (on which I worked as a consultant to TCF through the trial stage) that suggested that the issuing banks could make up their lost revenue somehow by charging their own customers, which never happened.
As an economic matter, it is clear that the higher the allowable debit interchange fees, the less disruptive the Durbin Amendment is to the operation of the debit card interchange market, and in that sense at least the decision in NACS performs a useful public service that was no part of its intention. Indeed, I suspect--or just hope--that Judge Tatel's misguided bit of statutory interpretation will not be challenged down the road.
Remember that the panel decision was unanimous, and it may prove unlikely that either the entire District of Columbia Court of Appeal or the Supreme Court will have any appetite to untangle the tortured arguments that persuaded so distinguished a panel of the Court of Appeals. And if they did look at this statute, they should start by revisiting the constitutional issues from TCF, which were decided on an assumption that has proved false, namely, that the debit card companies can recover their lost fees from their own customers.
That seems highly unlikely at present, so at present the best achievable resolution for this issue is a large dose intellectual bed rest after all the legal twists and turns of the past four years. But who am I to say? I thought that the chances that Judge Leon's decision would be overturned were close to zero. And never in my most fevered moment could I have imagined the grotesque and improbable way in which the Court of Appeals saved the bacon of the Federal Reserve, and yes, of the issuing banks. Wonders never cease.