August 30, 2007

Why are analyses of cash v. debit card so fundamentally flawed?

Several weeks back, Dave Birch commented in depth about Leo van Hove's article. (Unfortunately not free to read.) Dave does the job of bringing the paper to the public, hopefully faithfully.

What follows is a somewhat critical response to Leo's article (as viewed through Dave's spectacles). In summary, I would suggest that the analysis is flawed, because it fails to consider the costs of subsidies. Fundamental questions in the analysis, which for sake of polemic thought I'll claim as being answered in the negative, are raised below.

The cash v. everything else debate hinges on the costs to banks of the various instruments. Occasionally, when convenient, the costs to the merchants are brought in as well. If the analysis is stretching out to be fair, it also includes the costs to the users. This is rarer, but Leo talks about this.

What is never brought in is the opportunity costs to those without these tools. The central banks hint at this issue when they talk about societal costs (and who's to know how they calculate this) but what they fundamentally fail to recognise is the cost of a credit card transaction for someone without a credit card. Or, likewise, without a bank account.

By way of example, how much does it cost to hire a car without a credit card? In many countries, you simply cannot do it. The cost of something you can't do is so high it breaks the model; ignoring that cost, while popular, is not the appropriate response for policy or science. Dave and Leo hint at this:

One of the most obvious implications stemming from this observation is that there is a low price elasticity: the consumer demand for goods and services does not depend greatly on the cost of the payment instruments.

The large group of society who can't pay the cost are sometimes called the unbanked, in the banking world. That's part honest recognition, and part marketing: Banks and sometimes central banks say, we must "bank" everyone, and the unbanked are are future growth area.

Sounds simple, and perhaps it is. Why then haven't they done this? Why do most countries exhibit large (double-digit) portions of the population outside the bank net? Part of the answer, if not all of it, is the risk, and banks are, if nothing else, very careful in the risk business.

And now we get closer to the nub of the efficient payments problem. Payment systems are not banking. Payment systems are more technical systems, almost turn-key devices, that can be built for standard levels of risk (and in this context, I mean, the risk that all businesses except banks take on).

That is, by normal principles, payment systems should be outside banking. If payment systems were outside banking, then the "unbanked problem" wouldn't exist, because the risk would be properly allocated to those people doing transactions.

However, payment systems are not outside banking, for one very good reason, and that is this: Banks need a way to borrow consumer's money. They need the hard cash to build reserves which allows them to make loans ... back to the same public. (That's what banking is, BTW.) And, it turns out that if you offer a payment system alongside a deposit account, this makes for a ready source of those demand deposits. The synergies are very high, as the MBAs would say. The need to control competition is very high, central bankers would say, because a strong, non-leaky bank sector is in society's interests.

We must keep firmly in mind, however, that payment systems is not the same business as banking (remember, always, payment systems are not risky, unless you make them so), and therefore, the tying of payment systems to banking is a cross-subsidisation. So that means that the entire analysis of the costs of cash are wrong unless they start with the assumption that we deliberately run a cross-subsidisation system by definition, and we have to eat that cost as a society.

This is seems an appropriate time to bring in Leo / Daves version of the story:

We've agreed that the market for payment instruments is very different from other markets. So different, in fact, that "market" is probably an inappropriate description. This issue is a basic structural problem: central banks are charged with improving the efficiency of the payment system while being responsible for the most inefficient mechanism. Inefficient here means, just to be clear, "has highest social cost".

So we are all agreed that payment systems are not open, even if we disagree on how to discuss the foundation. I'm pretty sure that the entire unbanked world would vote for my thesis, and a large proportion of the banked would do so too, once properly appraised of the costs of the subsidy (Paypal charges what percentage per transaction?).

And, that's the punchline: without including the costs of that cross-subsidisation system, all other analyses are not only flawed, but meaningless. It matters not whether cash, debit cards and credit cards compete, because we have chosen by policy to run cross-subsidies. Only if we are to drop all the subsidies are we likely to come close to a meaningful competition, and only if we measure the costs of subsidies are we close to comparison or policy.

And this finally gets us to the real core of the argument: Nobody's talking about opening up the payments market. Not in Europe, at least. And, since 9/11, the US has taken great strides to close it up, aligning the majority of the rich world.

Nobody's talking about competition as a driver for efficiency in payment systems. Therefore, we can claim, the assumption that the central banks are accepting the responsibility for the efficiency of the payment systems is, on the face of it, wrong. What then is Leo's article about?

Leo focuses on the apparent tension between a central bank's duty to ensure efficient payments systems and its operational activities in providing the least efficient payment system of all.

Why is this? Why so much contradiction? Once we've accepted the above logic, we can predict what this debate is really about.

Central banks are caught between their customers (banks) and their regulators (parliaments, the public, those nasty bloggers, etc). Banks understand one thing: control of market. Europeans (by this I mean the peculiar policy ones in or near Brussels) understand one other thing: everything should be the same price across Europe. And one price that hasn't fallen to Brussels is the cross-border payment.

The central banks then find themselves at the poker table with the banks, trading chips with labels like SEPA, CASH and FRANCHISE. SEPA is "Single European Payments Area" and means one price across Europe for all retail payments, more or less. The game works like this:

"Who'll take SEPA off my hands in exchange for CASH?"

Which, if you follow the highly unspoken cross-subsidisation of payment systems above, is why the debate seems so surreal.

Whoever plays with SEPA and CASH is playing for high stakes poker. The problem is, we have already seen the cards in the hands of other players, and they've got TRAINWRECK written all over them.

Posted by iang at August 30, 2007 09:34 AM | TrackBack
Post a comment

Remember personal info?

Hit preview to see your comment as it would be displayed.