Kenneth Rogoff (2014a, b, c) has proposed phasing out cash – at least large-denomination notes in developed economies. The goal would be to:
- Eliminate the zero lower bound on nominal interest rates that may interfere with monetary policy; and
- Eliminate anonymous transaction that may obstruct the fight against crime, money laundering, tax evasion, and the like.
His intervention is set to revive an old debate about the role of cash that lately resurfaced in the blogosphere. Different contributions to this debate may be represented in terms of a two-step approach to monetary regime change:
- First, to grant the general public access to central bank reserves; and
- Second, to phase out cash.
Yes on one, No on two
I argue that the first step – and only the first – should seriously be considered.
Letting the general public hold reserves at the central bank and use them for electronic payments would lower the risk of bank runs and strengthen financial stability. Since deposits held at the central bank are base money – in contrast to deposits at commercial banks, which merely constitute claims on base money – they are run-proof. As a byproduct, public ‘insurance’ of bank deposits could be scaled down and the moral hazard risks it entails could be limited.
Similarly importantly, granting the general public access to reserves would eliminate a disturbing contradiction in public policy. To deter tax evasion, governments increasingly declare the use of currency illegal for specific transactions and require payments to be made electronically using bank deposits. That is, governments outlaw payments using legal tender, insist on the use of bank deposits although they are not legal tender, and in the process expose households and firms to various financial risks. (These risks are partly borne by the public sector, due to deposit ‘insurance’, but this creates other problems.) Broader access to central bank reserves would bring this situation to an end.
But a change of regime may also give rise to problems. Inducing savers to replace part of their deposits at commercial banks with central bank reserves could undermine deposit-financed credit creation – a central activity of the banking sector. Douglas Diamond and Philip Dybvig (1983) showed that deposit contracts can usefully provide insurance against liquidity shocks. How large are those benefits? And how large are the associated social costs, reflected in the distortions caused by deposit ‘insurance’ and losses due to run-induced fire sales and bankruptcies? Macroeconomics is still in the process of providing a reliable answer.
A related question concerns the transition from the regime in place, where non-banks hold deposits at commercial banks, to a new regime where they also hold central bank reserves. Depending on the magnitudes of the desired portfolio changes, this transition could substantially increase the demand for base money and decrease the money multiplier. As a result, liquidity in the banking system might dry up and bank runs could be triggered unless the central bank accommodated sufficiently.
Technical questions that would arise
Questions of a more technical nature would also have to be addressed. One relates to the payment system. Under the regime currently in place, only financial institutions pay reserves to each other through a clearing system. In a regime with more dispersed reserve holdings, wider access to this clearing system could be granted or a parallel network for retail reserve payments could be built. Alternatively, financial institutions could execute reserve payments for households and firms on their behalf, for example by administering off-balance-sheet reserve accounts for their customers and settling net payments between these accounts.
Other technical questions concern the conduct of monetary policy under the new monetary regime: Would central banks lend funds only to financial institutions or also to the broader public, and at the same policy rates? Or would new reserves be brought into circulation by transfers to the private sector, as envisioned in some narrow banking proposals? These questions are far from trivial, and serious reflection might lead to the conclusion that broader access to reserves should be granted only gradually and cautiously. But the regime currently in place with sharply segmented markets for reserves is hardly optimal. ‘Reserves for everyone’ should not be discarded lightly.
Problems with eliminating cash
In contrast, the more ambitious proposal to not only broaden access to reserves but also phase out cash goes too far. Its proponents likely overstate the additional benefits net of costs.
Consider first the benefits. As far as relaxing the zero lower bound is concerned, the fundamental objective – to lower effective real interest rates in order to incentivise earlier consumption and investment – can not only be achieved through monetary policy but also through tax policy.
An increase of the consumption or value added tax rate over time decreases the effective real interest rate. If one wished to relax constraints on monetary policy without relying on tax policy, this could be accomplished without phasing out cash altogether (and without moving to a higher inflation target which is costly for other reasons). As explained by Willem Buiter (2009), all that would be needed is a floating exchange rate between reserves and cash, implying an unbundling of currency from the unit of account. Abolishing currency thus is not necessary unless one fears negative consequences of such a floating exchange rate (see e.g. Goodfriend 2000, fn. 23). In fact, it may not even be sufficient. As pointed out by John Cochrane (2014), prepayment of taxes or bills allows securing a zero nominal rate of interest; even in the absence of cash, some form of zero lower bound thus persists.
As far as the second objective – curbing tax evasion and criminal activity – is concerned, phasing out currency does not guarantee success either. Those seeking anonymity vis-à-vis the government when transacting would still be able to avoid surveillance, for example by hiring an agent and incorporating a shell company. And since the costs for such circumvention are largely of a fixed type, the implicit tax on black market activity due to a phase-out of currency would likely be regressive. Criminals directing large businesses could continue to operate in the dark and the very wealthy would still find it advantageous to evade taxes, but the tax-dodging middle-class contractor would probably come clean.
On the cost side, phasing out currency would severely undermine privacy. (Buiter’s 2009 suggestion of ‘cash-on-a-chip cards’ could limit this concern somewhat.) This is generally acknowledged in the debate, but not always given sufficient weight. The potential consequences of conferring information about nearly all market transactions to a branch of government can barely be overlooked. Even if those consequences were properly understood and a majority of the population were willing to bear them, the protection of minority rights and other considerations should weigh against surrendering information on this scale unless very important reasons render it absolutely necessary. Citizens have a right to privacy and protection not only against rough government but also against rough government employees.
Phasing out cash would also have more mundane negative repercussions, for instance related to reduced financial literacy and the negative consequences thereof. Many households appear unable to manage financial affairs in their own longer-term interest, and the ensuing problems typically get worse when electronic forms of payment and credit are available. In a world without cash but with gift cards and bank checks, children might find it even more difficult to acquire an understanding of budget constraints or the notion of saving.
Finally, there is the important issue of enforceability. Phasing out cash would not only lead to more electronic payments but also likely imply that substitutes for cash are adopted for small transactions – think of cigarettes, baseball cards, or foreign currency. This substitution process would run counter to the aim of the phase-out. Rather than tightening control over payments and the price of money, government and the central bank would relinquish it.
In free societies, the extent of government control over money is limited by money’s usefulness. Proposals to fundamentally change the monetary regime must take this into account. A phase-out of cash is disproportionate and might backfire. But granting access to reserves would raise the use of government-provided money and help resolve important problems. ‘Reserves for everyone’ deserves serious consideration.
Buiter, W (2009), “Negative interest rates: when are they coming to a central bank near you?”, Maverecon blog, 7 May.
Cochrane, J H (2014), “Cancel currency?”, Grumpy Economist blog, 30 December.
Diamond, D W and P H Dybvig (1983), “Bank Runs, Deposit Insurance, and Liquidity”, Journal of Political Economy 91(3).
Goodfriend, M (2000), “Overcoming the Zero Bound on Interest Rate Policy”, Journal of Money, Credit, and Banking 32(4).
Rogoff, K S (2014a), “Costs and Benefits to Phasing Out Paper Currency”, NBER Working Paper 20126, May.
Rogoff, K S (2014b), “Paper money is unfit for a world of high crime and low inflation”, Financial Times, 28 May.
Rogoff, K S (2014c), “Rethinking the Global Currency System”, Munich Lectures in Economics 2014, 18–20 November.