Meeting on The Cashless Society

+ Paul Romer

In November of 2012, the Urbanization Project hosted a workshop on "The Cashless Society" that focused both on the implications of cashlessness for monetary policy and crime and on related developments in biometric technologies, with particular emphasis on the status of India’s UID project. This post provides a summary of the general consensus that emerged from these discussions.

Experience has show that any discussion about changes in payment systems will activate strong passions about the balance between tyranny and independence, particularly in the United States. Rational discourse about payment systems will be hindered by these passions and by the technical difficulty of monetary theory.

Creating a strong but accountable state is, of course, the essence of good governance. We need states that are strong enough to enforce laws, which are required to achieve efficient outcomes. We must also be sure that the individuals who are part of the state must themselves follow the law. Moreover, in holding the people who administer the state accountable, we must guard against a political process that creates an opening for powerful private actors to use the state for narrow private purposes, e.g. when incumbents prevent entry by new competitors.

These perennial concerns are closely related to questions about how best to protect privacy and individual freedom and to prevent firms from using strategies that use obscurity and complexity to undermine the power of competition. Given how powerful private actors have become in the online world, it is unclear whether the greater risk now springs from a state that is too strong to be held accountable or too weak to maintain the essentials necessary for consumers to benefit from safety and competitive markets.

It is unclear whether the type of privacy that an anonymous payment system based on hand-to-hand payments of currency is in fact an effective mechanism for holding a government accountable. Maintaining a payments system based on fiat currency seems, on its face, to be an implausible way to hold a state that stops obeying the law in check — the monetary authority could simply use inflation to destroy this payments system.  

Based on extensive experience in India and in smaller experiments in other developing countries, there are feasible, low cost biometric identification systems that could support electronic payment systems, systems that could replace the use of hand-to-hand currency. In many developing countries, where it is the weakness of the state that poses more of a threat to daily life, these systems may be welcomed as more effective mechanisms for citizens to get benefits from the market system and from the state.

Within any jurisdiction, a prohibition on the use of paper currency could displace certain types of criminal activity, e.g. money laundering that brings cash from drug gangs into the banking system. It could raise the cost and/or reduce the incidence of some local criminal activities, but only if some substitute form of payment did not develop that provides an equal level of anonymity. Hand-to-hand exchange of a commodity like gold is one possibility. Electronic systems like bitcoin, coupled with pervasive internet connected mobile devices, might be another. But all such systems leave extensive electronic records of transactions, so they might expose criminal enterprises to more risk of detection and prosecution than systems based on hand-to-hand currency.

Prohibiting the possession of currency would remove the "zero nominal bound" as a constraint on countercyclical monetary policy. Nevertheless, one surprising conclusion from the discussion was that this should not be counted as an advantage of a prohibition on the possession of currency. One of the participants in the conference, Miles Kimball, pointed out that monetary authorities already have feasible mechanisms that they could use to avoid this constraint even when hand-to-hand currency continues to circulate. In the US, we have already adopted one of the key required measures: non-zero interest payments on banks' reserves at the Fed, at a rate that can be set by the monetary authority.

This rate could easily be set to a negative value. The next step under the Kimball proposal is to establish an exchange rate between currency and the reserves held at the Fed and to identify bank reserves as the unit of account. This exchange rate could deviate from 1 unit of reserves per unit of currency. If, for example, one dollar of paper currency purchased only 0.98 dollars in reserves, someone who makes a purchase with paper currency could face a 2% surcharge. (Many travelers who purchase foreign exchange overseas have already encountered a value for a paper bill that is slightly less than a claim on a bank in the form of a travelers check.) In practice, it is likely that many merchants would not impose this charge on small purchases. It would matter only for organizations like banks that make big exchanges of reserves for currency and vice versa. In effect, this would amount to having the monetary authority issue two types of currency and manage the exchange rate between them.

Under the Kimball proposal, central banks would probably have enough room to stimulate the economy if this exchange rate deviated from strict 1 to 1 equality only temporarily and only by a relatively small amount. For example, by letting the value of currency fall relative to reserves by a few percent per year for a year or so while nominal interest rates are negative, then recovering back to par soon thereafter.

The other possibility that emerged from the discussion was that a central bank that wanted to "tax" the use of currency without prohibiting it outright could let the value of the paper currency depreciate forever relative to reserves, e.g. by 10% per year. When the exchange rate between currency and reserves was large enough, merchants would presumably charge more for payment in paper currency, just as merchants near the US-Canada border charged more for payment in Canadian dollars than for payment in US dollars when the exchange rate between the two types of dollars fell to 0.8 Canadian dollars per US dollar.

The surprising conclusion was that by managing the exchange rate between currency and bank reserves, the central bank could remove the zero lower bound and could tax the use of currency, which would tax criminal enterprises that rely on currency. So a full prohibition on the use of currency could be viewed as a limiting version of less extreme policies that tax currency by letting its value depreciate relative to bank reserves. Because prices and inflation would be denominated in units of bank reserves, this type of policy can tax currency without causing inflation and without inducing any of the distortions associated with inflation.

In fact, as Kimball has noted, this policy is the only policy that has been proposed that has any credible claim to offering a monetary environment with zero inflation.*

* Note added 6-13: Conventional monetary policy can of course yield zero inflation (or negative inflation for that matter, as we had during the Great Depression.) Policies that do this without solving the problem of the zero nominal bound will not be credible in the sense that they will eventually be abandoned once people have experience with the high cost that comes from giving up the ability to implement a counter-cyclical monetary policy.

Back to top
see comments ()