Should we maintain a cash system at any cost?

by Fod Barnes, Senior Adviser to the PSR

“Money makes the world go round”

In Cabaret, for Liza Minnelli, this is clearly physical cash – but does it have to be? And if it doesn’t, do we still need or want cash?

Should we maintain a cash system at any cost? How expensive would it have to be to maintain an operating cash system for it not to be worthwhile?

At one level, the economy clearly doesn’t need cash – many businesses don’t take cash anymore, and many customers and citizens no longer use cash. The impact of the Covid-19 outbreak has also accelerated the trend away from cash use, and now there is even talk, and action, that central banks might switch over digital money. The possibility of an economy fully functioning without cash is clearly with us. And yet …

Quite a few people find it very hard to cope without cash and the government has indicated it will legislate if necessary to ensure that cash is still available for those who need to use it. Sometimes the alternatives suddenly stop working – either in limited ways or in more widespread ways when, for example, one (or even all) of the main electronic payment systems were to go down. Using cash as a means of exchange can be carried out completely independently of the electronic infrastructure – the ultimate “off grid” safety net. Which could come in handy if the next war is a cyber war or even when a next big geomagnetic storm forces our electronics to grind to a halt. These seem to occur about every 150 years. The last one, the Carrington Event, took place in 1859.

So maybe the issue is not so clear cut. It is likely that the vast majority of the UK economy could - at least in normal times - function quite happily without cash, but would this actually be the right outcome? And how would one decide? Looks like a case for economic analysis! A well carried out cost-benefit analysis might be really useful here.

Some of the basics that a good cost-benefit analysis would need include answering some quite simple questions:

How expensive is cash to use?
How expensive are the alternatives?
If cash is more expensive to use, who should pay? (And if it’s cheaper, who should benefit?)
Does a decline in the use of cash mean that the per-use cost of using it goes up dramatically?
And if it does, when does it become “too expensive”? (Even if some people would still like to use cash – or even need to use cash.)

The answers are surprisingly complex. Just take the first two:

Comparing a cash transaction with a debit card transaction throws up all sorts of anomalies and peculiarities, that makes the analysis difficult. For example, in a cash transaction the customer’s bank is likely to have paid out an interchange fee to the ATM operator (say 25p on a cash withdrawal of £100 – or 0.25%). The merchant’s bank may well then charge the merchant to deposit the cash. (say £1.00 – but for any amount? Or 0.5%?) With debit card payments the customer’s bank is likely to receive 0.2%. And the merchant will have paid out more to their card acquirer (out of which will be taken the 0.2%) – for a small merchant around 1-2%, for large merchants considerably less. And the merchant’s bank may also charge a fee for processing the incoming credit.

And then there are the internal costs of the merchant handling cash or card, the cost (if any) of the customer going to the cash machine (or even going into a branch, where they still exist). The change (if any) of the insurance premium for the merchant if they never have cash, etc. Using cards or cash clearly involves the use of real economic resources – small, but real – and it’s not directly obvious which is cheaper. (Nothing, not even in-credit banking, is really free.)

And then how do you value the ability to switch to cash in the (hopefully very small probability) of the non-cash systems failing for any material length of time? If that did happen, and there was no, or very limited, cash infrastructure, the costs to the economy could be very large. Or, alternatively, would it be just like a Sunday in the 1950s when nothing was open? Tedious, yes, but damaging to the economy?

Meanwhile, what about those who can’t use non-cash? Those without a bank account – especially those who can’t open a bank account. It would seem inhumane to allow cash to disappear while also having rules that make it more or less impossible for some people to use non-cash. But do you keep cash, or reform the rules? A similar dilemma would exist where the telecommunications infrastructure doesn’t support non-cash. Do you keep cash or upgrade the infrastructure?

Then there’s also those who have a particular need and for whom cash is easier, but it’s not (completely) impossible to get by without it. Again, do we keep cash or devise better, non-cash, alternatives?

The $64,000 question (1950’s TV show – the most exciting thing on a Sunday – and the prize was, or at least should have been, paid in cash.)

It clearly can’t be right that cash should be kept whatever the cost. On the other hand, if the additional costs of using cash are small, or even negative (i.e. it is actually cheaper than the alternatives) it would be unfortunate if the ability to use cash disappeared either by default or because of the idiosyncratic, and largely a result of historical accident, pattern of who pays the transaction costs of different methods of payment.

There are some potentially very large benefits (e.g. using cash as a back-up when a potential solar storm disrupts the use of electronic non-cash systems), and currently at least large numbers of small benefits (e.g. for those who prefer or need to use cash) to keeping the cash system functional – at least in the short term. Or at least until innovation has progressed us to a point where we can be sure that payment systems are safe, secure, and ensure that all consumers and businesses have adequate access to the systems that enable them to transact in ways that meet their reasonable needs.

So, after all of that, let us apply the usual two-handed economist approach, to answer the exam question set at the beginning of this article: