A few months ago, one of us (Steve, who lives near Boston) needed to send some money to his son (Dan, who lives in Chicago). Hoping to use the most up-to-date technology, Steve checked out the person-to-person (PTP) transfer options on his bank’s website. The first three choices required standard banking information—payee’s account number, bank ABA routing number and the like. They differed in cost and speed: you could implement a transfer in 3 days for $3, next day for $10, and same day for $30. All three used versions of the 20th century payments system: the automated clearing house (ACH) for the first two, and the Fedwire for the third.
But there was also a fourth option: transferring the funds using Zelle. By entering either the payee’s email address or mobile phone number, Zelle will transfer domestically up to $2,500 per day or $20,000 per month immediately at no additional charge. The rise of mechanisms like Zelle hints that financial innovation will finally deliver on the promise of providing cheaper financial services to end users. Keep in mind that, as our friend and colleague Thomas Philippon emphasizes, the unit cost of U.S. financial services has been stuck in the 1½% to 2% range for the past 130 years (see our earlier post)!
When it comes to domestic payments, the U.S. financial system still lags the efficiency in several advanced economies. In 2015, paper checks—a 14th century innovation—still accounted for roughly 15 percent of U.S. noncash payments―both in number and value. (The bulk of transactions in number are by credit or debit card, and in value by ACH.) Europeans typically use very few checks. In Belgium, Germany, the Netherlands, Sweden, and the United Kingdom, for example, they account for less than 1 percent of the total value of all non-bank transactions (see recent BIS data here). Not only that, but many countries have something that the United States does not: a real-time, ubiquitous electronic payments system where people make and receive payments of all sizes almost instantly. Put differently, Zelle is not comprehensive.
Why is the home to Microsoft, Google, Amazon, and Facebook behind in an information-intensive area like payments? The reasons are easy to find. First, other countries have leapfrogged outdated technologies. Much as some parts of the world bypassed land-line telephones, some advanced economies either never had paper checks or they replaced paper checks as they rebuilt their financial infrastructure following WWII. In the United States, checks remained dominant well after their technological sell-by date partly as a result of government support. At its founding in 1914, the Fed created a unified check-based payments system and took responsibility for moving paper checks around the country. Both the legal framework and transportation arrangements supported this system: cancelled paper checks were legal proof of payment, while the Fed leased a fleet of airplanes to fly bags of the paper around the country every night! (See the historical survey here.)
The other key factor delaying a shift to alternative payment mechanisms is the importance of what economists call a network externality. That is, the more people who use one form of payment, the more valuable that method is to the people who are already using it. And, by the same token, the more expensive it is for someone to move away from the prevailing mechanism. Incumbent suppliers of payments technologies—like big banks and credit card firms—have sunk costs for which they seek to be compensated. They may even have an incentive to stifle the development of new technologies that would render their existing systems obsolete.
Against this background, the most practical way to bring about change may be to bring together all the interested parties and encourage a coordinated shift to the new technology. However, as we all know, there can be a conflict between such collaborative efforts, which typically favor incumbents, and the introduction of radical new technologies that would undermine them.
With these considerations in mind, two years ago the Fed convened the Faster Payments Task Force (FPTF), a group of more than 300 experts and interested parties from a wide range of backgrounds with the objective to “identify and evaluate alternative approaches for implementing safe, ubiquitous, faster payments capabilities in the United States.” Earlier this month, the FPTF issued its second and final report, which contains a set of 10 recommendations for making the payments system faster, cheaper and more secure.
Before we get to the outcome of this process, it is worth briefly describing the challenge of reforming the payments system. While it is easy to understand what happens when we pay with cash, most people are unaware of the routine (but complex) procedure involved in completing noncash payments (such as those with a debit card, a credit card, or a check). The following table outlines the eight basic steps.
Anatomy of a payment
At some level, this sequence is trivial―someone has to start the process, the identity of the participants has to be verified, accounts have to be identified and the payer must have the funds, confirmation has to be received, obligations discharged, and the transaction verified. Yet, if even one of these steps fails, the payment will not go through. So, the challenge is to streamline and modernize this eight-step process, getting from the beginning to the end more quickly, cheaply and safely.
The FPTF began by agreeing on a broad set of criteria that define an effective system for completing a payment. They outlined six major categories. The system should be ubiquitous, with universal accessibility and cross-border functionality; efficient, enabling competition, scalability, and the addition of other services; safe and secure, including payer authorization, payment settlement and dispute handling; speedy, with fast approval, clearing and availability of good funds; legally sound, with credible rules that secure privacy, protect consumers and delineate intellectual property; and have a governance structure that is both effective and inclusive.
Using these criteria, the task force then called for reform proposals and engaged McKinsey & Company to assess them. Of 22 initial proposals, 19 were reviewed, and the results for 16 have been published. Interestingly, the posted proposals do not include submissions from VISA, Mastercard, or a number of other large companies that currently facilitate electronic funds transfer worldwide. We have no way of knowing if any of these firms participated at some point in the process. What we do know is that the services offered today are expensive―VISA charges between one and three percent of the value of a credit-card transaction―so a faster, safer, cheaper system could pose a threat to their business. The same goes for existing payments systems for small businesses: for example, on their website Square lists charges of between 2¾% and 3½% + $0.15 per transaction.
It is worth noting that the challenge of reducing costs is not just domestic, but international as well. Consider, for example, cross-border remittances. These are small-value payments―equivalent to a few hundred or a few thousand dollars—usually sent by workers in advanced economies to their relatives or friends in emerging market countries. And yet, they are very expensive. For example, someone sending $1,000 from the United States to Mexico will pay Western Union a charge of roughly 4% (most of which is hidden in the exchange rate used for conversion of dollars into pesos). Neither the sender nor the receiver of such remittances is wealthy.
Returning to the FPTF, from the public disclosures we constructed the following chart to summarize McKinsey’s assessment of the 16 available proposals. For each of the 37 subcategories, we numerically coded the evaluations as follows: Very Effective―4, Effective―3, Somewhat Effective―2, Not Effective―1. We then averaged the scores within each of the six major categories, which we weighted equally in constructing a summary score out of 100.
Qualified Independent Assessment of Payments Proposals
Using this ranking, we took a closer look at the top two: Wingcash and The Clearing House/Fidelity National Information Service (labeled “TCH/FIS” in the chart). (Were there but world enough and time, we’d have looked in detail at all 16.) The Wingcash system starts with the Federal Reserve issuing digital money in which each currency unit―the equivalent of every coin and note ($0.01 to $100)―has its own web address (URL) based on its serial number. Each webpage then has an owner who can transfer ownership to any other registered person or business. Wingcash facilitates these transfers.
The Wingcash model requires that individuals and firms access the system through an agent that authenticates their identity and enforces common standards, such as anti-money laundering (AML) and know-your-customer (KYC) rules. This agent can either be a bank, retaining many features of today’s financial system; or it can be some other organization, including the Federal Reserve itself.
Regular readers of this blog will be unsurprised by our principal reservations about the Wingcash system. First, it lacks anonymity, posing potential risks to civil liberties (for example, see our arguments against the elimination of paper currency.) Second, in the central bank-centric version of the proposal, the banking system could lose its core retail deposit base as people shift from demand deposits into digital Fed notes, leaving the government to run the retail payments system. We address the resulting political economy concerns, including the risk that the Fed will become a state bank that lends to individual households and firms, in an earlier post (see here).
The TCH/FIS proposal generally relies on existing infrastructure—not surprising given that TCH is owned by a set of large banks. As with the bank-centric version of Wingcash, individuals would access the system through their bank or non-bank regulated entity. But, as far as we can tell, everything else remains the same as it is in today’s system, including the use of anonymous paper money.
Are there reasons to be concerned about the TCH/FIS proposal? Possibly. By relying on the interests of incumbents, we may be overlooking a dramatically different technology that, in the long run, could serve us better. Another issue is that in this system there may be very few operators: if they are able to accrue the benefits of improved technology for themselves, rather than passing them on, the new system would sustain the stubbornly high unit costs of finance for end users.
Our point is that every payments mechanism makes choices and has limited flexibility. As a result, we face tradeoffs among various objectives about which is there no clear consensus. Which is more important: anonymity or crime prevention? Lowering unit costs for end users or maintaining a private retail deposit system? Even a glance at the 16 proposals suggests that no single one will make everyone happy.
That brings us back to the role of the Fed. As we mentioned earlier, by convening the FPTF, the Fed has moved to overcome barriers to the introduction of technology into the U.S. payments system in a manner that is similar to what they did for checks 100 years ago. But that reform process has just begun, and will inevitably involve sensitive policy questions that ought to be guided by an informed view of the public interest. The ultimate solution should not be limited by the interests of today’s industry incumbents and start-ups. And, it remains to be seen whether the current batch of new technologies will make payments (both domestic and cross-border) fast, secure and cheap, while securing other values like privacy and freedom.
Acknowledgement: The authors are grateful to their friends and colleagues, Professors Hanna Halaburda and Paul Tucker, for thoughtful discussions regarding payment systems.