America needs a modern payments architecture

By Felix Salmon
March 30, 2012
Summers's paper is here.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

I was sad that I had to miss Bruce Summers’s presentation at the Kansas City Fed’s payment conference this morning; I was a couple of miles down the road, at the Kauffman Foundation. But I did manage to grab five minutes to summarize his argument for the assembled econobloggers: it’s an important one, which deserves a lot more attention than it’s likely to get.

Summers’s paper is here. It’s a dense 32 pages long, which is positively laconic in comparison to his upcoming book, Payment Systems: Design, Governance and Oversight, which features contributions from no fewer than 23 famous-in-the-payments-world grandees. But the message of the paper is a very simple one. America desperately needs immediate funds transfer, or IFT. And we’re not going to get it.

Summers’s new paper is a longer and more detailed version of the paper I blogged back in August. That paper was co-written with Kristin Wells, of the Chicago Fed, and was published under the Chicago Fed’s auspices. At the time, I wrote:

What Summers and Wells don’t say, perhaps because they work for the Federal Reserve, is that it’s downright idiotic that the Fed doesn’t step up to the plate and take on its natural role as guardian of the national payment system. Why doesn’t it? I’m not sure, but I suspect it’s something to do with the fact that the Fed doesn’t really exist as a unified body: there’s just a network of regional federal reserve banks, with a board of governors in Washington.

In the new paper, Summers is speaking for himself, and makes explicit what was only implicit before. The regional Reserve Banks have the ability to implement IFT; the Federal Reserve Board has the authority to implement it. But somehow the two seem incapable of joining forces to actually do it.

IFT — the ability for me to pay you, and for you to receive the funds within minutes, rather than having to wait until the following business day — is already a fact of life in many countries around the world, from India to the UK. Where it doesn’t already exist, you can be pretty sure that someone is working hard on a plan to make it happen. Except in the US, where no one seems to have even started the process yet.

Summers explains that as economic connections between individuals, businesses, and government entities are being multiplied at an astonishing pace, and that the payments system is doing an atrocious job of keeping up. The problem is compounded by the fact that no one has introduced a new universal payments mechanism since the check, which clears slowly — and sometimes doesn’t clear at all — but which is extremely versatile and pretty much universally accepted. He writes:

The U.S. payment system does not currently support immediate completion of payments, and there are no plans for doing so despite long-standing evidence of the need for such a capability and development of these capabilities elsewhere around the globe. While there is innovation in immediate payments, it is limited to small closed systems operated by non-banks, or to small closed systems operated by individual banks or consortia of a handful of banks…

Effectiveness is influenced by speed, versatility, and universal coverage. The effectiveness of a particular method of payment depends on how well it meets the convenience and needs of individual and business consumers in the digital economy. Among the payment attributes that consumers look for, speed in completing transactions, versatility in the use of a given method of payment, and universal connectivity to accounts held in banks are of special importance in the digital economy.

Speed is an especially important consideration for payments in the digital economy. Consumers expect virtually immediate completion of their digital transactions. The idea that money in transit is digital information which can be processed immediately has not been readily accepted by the banking industry. Most bank-sponsored payment schemes depend on clearing and settlement systems that are designed around batch processing and delayed settlement, and these clearing and settlement arrangements are being nurtured as opposed to being re-designed around continuous, real-time processing.

The problem, he explains, is that there’s essentially no one in a position to implement a new architecture along these lines. There’s no real national governance of the payments system in the US; while it has historically been overseen by the Federal Reserve Banks, newer developments like the Durbin amendment capping debit interchange fees gave all the regulatory power to the Federal Reserve Board in Washington. If debit transactions had been governed by the regional banks all along, he writes, “arguably, the Reserve Banks would never have allowed non-par clearing and settlement for inter-bank debit card payments.”

But there’s a strong deregulatory impetus within the Federal Reserve system, and most governors have been quite enthusiastic about the idea of getting the Fed out of the business of clearing and settlement and payments regulation. The banks innovated credit and debit cards, which are very popular, so what’s the need? Summers concludes:

The Federal Reserve Board is not interested in leading or guiding the development of clearing and settlement capabilities for payments in the digital economy. Moreover, the Federal Reserve Board is satisfied to give up the Reserve Banks’ operational leverage as providers of inter-bank clearing and settlement services.

And if the Fed won’t do it, there’s no realistic way that the private sector is going to get its act together and implement something as ambitious as IFT on its own — the collective-action problems make such a cooperative endeavor effectively impossible.

Which means that the only possible way that we’re going to get IFT in this country is if Congress acts, and passes an act mandating that the Fed build an IFT system.

Congress has done this kind of thing before: in 1974, it created the National Commission of Electronic Fund Transfers, which in turn guided the development of the US payments system for decades. It needs to do so again — with the Fed playing a central role in drafting the legislation. In the meantime, says Summers, the least that the Fed can do is to just start taking payments innovation seriously, including non-bank players who are building platforms which might revolutionize the way we all send money to each other.

The Federal Reserve Board should develop a special-purpose bank charter for providers of specialized payment services, allowing in particular for the inclusion of non-banks that are payment system innovators and payment method providers in the nation’s money and banking system for payments.

This kind of invisible plumbing is rarely sexy, and it certainly doesn’t get a lot of votes, but it’s crucially important, and could easily create tens if not hundreds of billions of dollars in value for the economy every year. The fact that the market hasn’t done it is a very clear market failure; and where there’s a very clear market failure, the government — in the form of Congress and the Federal Reserve — should step in.

Should, but won’t. More’s the pity.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

What dark age period is the US trying to recreate here? I expect you still use pieces of paper with a signature on to pay bills too? What’s next, resurrecting typewriters?

Posted by FifthDecade | Report as abusive

Ohmigod ohmigod ohmigod *socialism*

Posted by oneagainstmany | Report as abusive

My manual typewriter still works eight-plus hours into a power outage, FifthDecade.

Unlike a FiOS “landline,” internet connection, or television, for instance.

Posted by klhoughton | Report as abusive

“America desperately needs immediate funds transfer, or IFT.” (F.S.)

There are some things America “desperately needs”, but ….

Posted by MrRFox | Report as abusive

Also seems that the lack of IFT and persistance of delayed settlement in the US will hamper the move toward a cashless economy that seems to be occurring in other places (like Sweden). Not sure if there are cost estimates of inefficiencies that arise from delayed settlement, but the costs of processing cash payments is non-trivial. A recent Maija Palmer piece in the FT cited a McKinsey study that puts the cost of processing cash payments at €60bn-€100bn annually in Europe. It also cited a study by the Dutch central bank that found the costs to be about €300 per household.

Posted by fuller.brandon | Report as abusive

Felix: I’m generally a fan and agree with you far more often than not. But, with all due respect — and at the risk of sacrificing whatever very-serious-person policy-wonk street-cred I might otherwise have had by even mentioning this word: Hasn’t the thing that you (and former Fed official Bruce Summers) are bemoaning the non-existence of, already been invented and beta-tested outside of and completely independently of the rent-extraction-racket that is the modern U.S. financial system? I know you’re familiar with Bitcoin, because when I googled “Felix Salmon and bitcoin” I found (inter alia) your snarky tweet from last June deriding “anarcho-libertarian techno-utopians get[ting] their comeuppance” in the “bitcoin fiasco.” And yet somehow, despite the multiple episodes involving hacked exchanges and stolen bitcoins (which have nothing to do with the integrity of the bitcoin system itself — by analogy, those incidents undermine the viability of bitcoin in precisely the same way that the Jesse James gang’s spree of bank robberies doomed paper currency to failure in the late 1800s, i.e., not at all) is still around and functioning just fine as, at a minimium, proof of concept of a viable peer-to-peer “virtual currency,” or, if you prefer, “immediate funds transfer system.”

Posted by TaxWonk | Report as abusive

A cashless economy … is that a double entendre?

I think one thing America does NOT “desperately need” is immediate funds transfer, or IFTs. While speed, versatility and universal coverage might be important when doing overseas transactions, what makes clearing fund transfers slower also makes them safer to use.

That safety is more important to those of us who don’t wish to be saddled with technology that is simply ‘good for business,’ yet have to use it as the old fashioned ways are saddled with higher fees to pressure Universal use of things like IFTs.

A cheque is something you can guarantee, post date, rip up, void, be stopped/cancelled and needs a signature. People who commit cash cheque fraud are often caught and charged whereas electronically you and they are a blip away.

Cash is something you can see and hold and can count your change, and if there is an error you are holding the proof and YOU get instant access to the cash which belongs to you, back, whereas the business now holds your “money” instantly and the individual is now responsible for the money being taken from their ewallet, or ecard, or ebank…

I am mourning the penny here and so cannot give 2 cents for bigger and better ways to to have access to instant electronic ‘money’ as business will benefit with no cash to steal, but not the individual, who will be a another blip waiting to be exploited by electronic thieves chomping at the bit to have easier access to your ebank.

Posted by youniquelikeme | Report as abusive

How many individuals really have this “desperate” need? Most of my personal transactions are either cash for local and debit for online, both of which can be carried out quickly. I get direct deposit at work, and the check is always in the bank when it’s supposed to be. It may be an inconvenience for the wealthy or large businesses that need to move a lot of money through inster-institutional transfers, but it’s hard to really see it as much of a “desperate” problem as some of the other things we have on our plate right now. (And I think someone has nearly IFT-like capacity–when we file our taxes electronically, the payments are processed _very_ quickly indeed!)

Posted by Moopheus | Report as abusive

Well Moopheus, when you file our taxes electronically, the payments are processed quickly unless you are one of the quarter million whose returns were stolen by identity thieves.

In that case it takes a mere year, or more, to get back what is rightfully yours. x-identity-theft-needed-social-security- number/story?id=15834826

Posted by youniquelikeme | Report as abusive

Just what we needed, a faster, more efficient way to go into debt.

Posted by PseudoTurtle | Report as abusive

Felix, the supposed reason why “America desperately needs immediate funds transfer, or IFT.” is so that the banks don’t lose the “float” while that money is in transit, so your article by not pointing that fact out is quite disingenuous. The “float” from trillions of transactions amounts to a considerable amount of money the banks are losing every minute of every day. THAT is what their problem really is.

Posted by PseudoTurtle | Report as abusive

IFT is no more necessary than conversion to the metric system. A good transaction should take its time and be savored. Nothing better than listening to a telex machine confirming payment details with that sound of ink smashing the paper.

Posted by egreen | Report as abusive

“Felix, the supposed reason why “America desperately needs immediate funds transfer, or IFT.” is so that the banks don’t lose the “float” while that money is in transit, so your article by not pointing that fact out is quite disingenuous. The “float” from trillions of transactions amounts to a considerable amount of money the banks are losing every minute of every day. THAT is what their problem really is.”


Posted by RealBadAttitude | Report as abusive

@ RBA… in a 5% world a trillion in float would be worth something… in a 0% world it’s not worth the trouble.

@ Felix… debit transactions memo post in a matter of minutes at my small mutual bank. True the funds don’t move until the next business day but if you are a merchant knowing the money is comming in tomorrow as sure as the sun will rise is pretty much as good as actually having it.

It would be nice if I could text a neice or nephew $10 on their birthday. If I can text $10 to a desaster releif effort why can’t I do it to a relitive on the other coast?

Posted by y2kurtus | Report as abusive

I agree with TaxWonk here. Concerning your comment, “And we’re not going to get it”, you’re clueless.

Posted by zylstra | Report as abusive