Central bankers and monetary economists have become intrigued with the idea of reducing the use of paper currency, or even entirely eliminating the circulation of large-denomination notes. Sweden has already begun the process of switching to an all-digital system in which debit cards substitute for cash, while the European Central Bank ceased issuing 500-euro notes at the end of 2018.

The Case Against Cash

Ken Rogoff of Harvard has presented the most comprehensive case for dropping cash, so the details of his approach are worth reviewing. He would phase out (over a decade or more) all large-denomination notes in developed countries. In the United States, $100, $50 and $20 bills would go, and perhaps $10 bills along with them. That would leave smaller-denomination currency in circulation — though Rogoff would consider eventually replacing even these with “equivalent-denomination coins of substantial weight” to make it “burdensome to carry around and conceal large amounts.”

To put this in perspective, $100, $50 and $20 notes make up about 97 percent of the value of U.S. currency. Rogoff acknowledges that it is far too soon to phase out cash entirely in less-developed countries — though here, too, he would still eliminate the largest notes.

Why the switch from currency? First, because cash is widely used in underground economic activity, advocates contend that the elimination of large-denomination notes would significantly diminish criminal activities such as tax evasion, illicit drug trade, illegal immigration, money laundering, human trafficking, bribery of government officials and possibly even terrorism. They also argue that suppressing such activities would have the additional advantage of generating tax revenue.

Second, they maintain that, in a future with low inflation, central banks will need to be able to push nominal interest rates below zero in order to make monetary policy effective in recessions. And this will be difficult or impossible if the public can hoard large amounts of cash as an alternative to bank deposits.

Second Thoughts

These arguments may seem compelling on their face. But a close look suggests they are at best weak — and perhaps entirely mistaken.

To show that the currency phase-out would be good public policy — in econ-speak that it would generate net gains in societal welfare — one must consider the impact on the wellbeing of all affected. For the United States, this includes nonresident users of U.S. currency as well as Americans.

Not counting currency held in bank vaults, as of December 2016, there was roughly $4,200 in cash per U.S. resident in circulation, fully 80 percent of it in $100 bills. Estimates of how much of this currency is held in other countries range between 45 and 60 percent. If (as is almost certainly the case) most of the dollar currency held abroad is in the form of $100 bills, then U.S. residents, on average, must each possess around 12 $100 bills. Yet a 2014 study from the Boston Federal Reserve suggests that only 1 in 20 adult Americans owns any $100 bills. So the bulk of these domestically held high-denomination bills must be buried in the U.S. underground economy.

Or maybe not. As Lawrence H. White of George Mason University notes, those surveyed have incentives to under-report their currency holdings, whether legal or not. A 2017 study from the San Francisco Federal Reserve concluded that “cash was the most, or second most, used payment instrument regardless of household income,” and that it is even used to make 8 percent of all payments of $100 or more. Such ongoing use of cash for these large transactions at least suggests that high-denomination bills still facilitate licit economic transactions.

No one denies that a lot of cash circulates underground. But many of these transactions are merely unreported and otherwise perfectly legal. Thus there could be real welfare losses if forcing the use of less convenient media of exchange impedes or hinders these transactions. Consider too that, while cash supports truly predatory criminal acts, in many cases it is used in illegal activities that are nonetheless welfare-enhancing — among them, evasion of exceptionally high taxes that choke off productivity, and regulation that exists only to protect rent-seeking or to enhance the power of regulators looking for bribes.

But opponents of currency use have made little effort to differentiate the impact of different sorts of cash-fueled underground transactions. To be sure, underground commerce reduces tax revenues. Keep in mind, though, that taxes do more than simply transfer funds from taxpayers to the government. They also discourage people from doing whatever transaction would be taxed. If this inhibits otherwise productive activity, taxes impose economic losses as well as generate revenue. Thus, the downside of any revenue gains from restricting cash is the potential loss from discouraging underground economic production.

Admittedly, tax evasion also distorts economic output by shifting investment into the underground economy and away from other businesses that have higher pre-tax returns. But the quantitative impact of this distortion depends on multiple factors. If phasing out cash increases tax collection, will the added revenue finance genuine public goods whose benefits offset the productivity loss linked to higher tax rates on market activities? Simply assuming that shifting income from the private to the public sector — and in the process, distorting allocation of productive resources in the private sector — generates a net gain to society seems naïve.

In any event, it isn’t clear that raising tax revenue by phasing out cash would in fact generate net gains in government claims on financial resources. Currency is like a bond issued by the government that pays no interest and can never be redeemed. Economists call the resulting benefit to the government “seigniorage.” It can be thought of as an implicit tax on people’s cash balances equal to the interest the government might have been forced to pay if the public held their safe liquid assets as short-term Treasury debt.

Of course, this implicit tax weighs most heavily on cash-intensive underground businesses. Phasing out cash would thus not only change the amount of financial resources extracted from these unreported, productive activities. It could subject them to burdensome regulation that imposes costs without generating net revenue.

Even if the initiative did generate net government revenues, the losses (especially from eliminating smaller-denomination notes) would fall disproportionately on poor people, particularly immigrants, who lack debit cards, checking accounts, and so on. Leaving small-denomination notes or coins in circulation would help somewhat — though not forever, as inflation steadily erodes their real value. In 1950, a $5 bill had purchasing power about equal to that of a $50 bill today. And even as recently as 1980, a $5 bill had the purchasing power of about $15 in today’s prices.

Rogoff recognizes the downside for the poor and advises that any drastic scale-back in cash be compensated with subsidized debit card accounts and perhaps smartphones for all. If the government took the less costly option of merely providing 80 million free, basic electronic currency accounts for low-income individuals, he estimates the cost would be $32 billion per year. That represents another erosion of net benefits and thus needs to be included in the overall cost-benefit analysis.

The relative size of the underground economy is larger in almost all other countries than in the United States. Friedrich Schneider, an economist at Kepler University in Austria, estimates that, among high-income countries, the figures run from 22 percent of GDP in Italy to 15 percent in Sweden to 9 percent in Japan to only 7 percent in the United States. (The figures exclude clearly criminal activities.) These higher figures outside the United States are generally attributed to higher tax rates and more burdensome regulation. Thus the loss from inhibiting the shadow economy in Europe would be considerably larger than in the United States.

Although the most serious levels of tax and regulation evasion occur in less-developed countries such as Brazil and India, even in some relatively advanced economies — think Greece and Italy — the underground economy represents a wide swath of output and employment. Phasing out cash in an economy in which unreported transactions lift the economy’s total output by as much as one-fifth would clearly be disruptive, even if the transition were slow.

Schneider’s estimates are confined to underground activity that is unreported but would in most times and places be considered legitimate. He attempts to exclude criminal activity. Yet governments, hungry for revenue and regulatory authority, frequently make little distinction between the two. This makes the dividing line between productive and predatory underground activity hazy. Schneider’s estimates thus include the output produced by illegal immigrants, while excluding output from the trade in illegal drugs.

Schneider concludes that “a reduction of cash can reduce crime activities as transaction costs rise, but as the profits of crime activities are still very high, the reduction will be modest (10-20 percent at most),” with the bulk of this reduction coming out of the drug trade. Rogoff, for his part, concedes that corruption and bribery are really serious problems only in poorer countries — precisely where a premature elimination of cash would also generate serious economic dislocation. With regard to terrorism, he concludes that eliminating cash would have, at best, minor effects.

Taxes Without Tears

One major cost that opponents of cash take seriously is the lost seigniorage from issuing cash. Between 2006 and 2015, the U.S. government averaged 0.4 percent of GDP annually in seigniorage from printing new notes and spending them. To phase out all existing currency by replacing it with interest-earning Treasury securities would increase the U.S. national debt by nearly 7.5 percent (yes, 7.5 percent!). Assuming a real interest rate of 2 percent on the additional debt, the combined annual cost of eliminating both existing and future U.S. currency would be $98 billion per year, or more than 0.5 percent of GDP.

If the U.S. eliminated only $100 bills and continued to provide the remaining 20 percent of currency, lost seigniorage would fall to $77 billion annually. But no matter how you play with these numbers, phasing out cash does not render significant net gains to the Treasury.

True, developed countries that lack a foreign source of demand for their own currency have much lower rates of seigniorage than the United States. In both Canada and the United Kingdom, seigniorage from future issues of cash (and ignoring the interest cost of eliminating existing cash) potentially amounts to only 0.18 percent of GDP. Therefore, if they phased out cash, government revenue losses would be smaller than in the United States. But some developed countries generate rates of seigniorage that are as high as or higher than the U.S. because of either robust foreign demand for the currency or greater customary use of currency in the domestic economy. If Japan were to eliminate only its 10,000 yen note, which is worth about $90 and represents a remarkable 88 percent of the value of its cash in circulation, the government would lose about 19 trillion yen in seigniorage (more than 3 percent of GDP) annually.

Reducing or eliminating U.S. dollars in circulation would also have a negative effect on individuals and businesses using the approximately 50 percent of dollar notes held abroad. Of course, some of those dollars represent ill-gotten wealth and facilitate malign behavior. Yet advocates of phasing out cash have so far ignored the potential dislocation in economies that openly use the U.S. dollar as currency (Panama, Ecuador, El Salvador, East Timor, the British Virgin Islands, the Caribbean Netherlands, Micronesia and several small island countries in the Pacific), or are partially dollarized (Uruguay, Costa Rica, Honduras, Bermuda, the Bahamas, Iraq, Lebanon, Liberia, Cambodia and Somalia, among others).

The Negative Interest Rabbit Hole

The second main argument for phasing out currency is that doing so would make it easier to stimulate the economy with injections of credit during periods of very low (or negative) inflation. The reasoning is as follows: when an economy sinks into recession, the central bank normally can stimulate lagging demand for goods and services by expanding the money supply and thereby reducing interest rates. But if interest rates are already extremely low, the mechanism becomes very inefficient because banks and other financial intermediaries will choose to accumulate liquid reserves rather than making loans at little profit.

Now, central banks generally have the authority to charge negative interest rates on the reserves that commercial banks and other financial institutions hold as deposits at the central bank. Denmark, Switzerland, Sweden, the Eurozone and Japan have all attempted this. Negative rates on bank reserves in turn put pressure on commercial banks to pay negative rates on the deposits held by their customers.

But note that this gives depositors an incentive to exchange bank balances for currency. Banks can do the same by replacing their excess reserves held as deposits at the central bank with vault cash. So the elimination of cash — or even just the withdrawal of large denominations — would close off this route, making it practical to impose negative interest rates on the private sector.

The term “negative interest rates” somewhat obscures the nature of what is contemplated. If negative rates ripple through to households and businesses as negative interest on bank accounts, they would function as a tax on the public’s monetary balances. Negative rates could generate revenue as the Fed collects from banks’ reserves, and thus would partly offset the seigniorage lost from restricting cash.

Note an important distinction here, though. Inflation functions as an implicit tax on holding (non-interest-bearing) cash balances. Negative interest rates, in contrast, would directly tax money in order to encourage banks to loan out excess reserves and stimulate demand for goods and services.

But negative rates as a tool of monetary policy may be much ado about little. Research on the “zero bound” issue has not yielded a consensus about whether central banks really need to go into negative territory to do their jobs. Moreover, some economists question whether negative interest rates would do any good. Stephen Williamson, formerly at the St. Louis Fed, and John Cochrane of the Hoover Institution argue that, paradoxically, a negative nominal interest rate would actually depress inflation rather than push it back into the territory in which monetary policy becomes more effective at stimulating demand.

The evidence from the experience with negative rates on bank reserves (in Japan and Europe) when changes in the general price level turned negative has not been encouraging. The Fed considered driving rates negative in 2010 during the painfully slow recovery from the Great Recession, but opted for other unconventional approaches to make monetary stimulus work. It thus seems stretching the point to argue that blocking access to currency as a means of escaping negative rates is plainly worth foregoing seigniorage.

Then there are the nuts and bolts problems associated with implementing a negative rate policy even when buttressed by a scarcity of currency. In a world with all but the smallest denominations of currency eliminated, James McAndrews, formerly of the Federal Reserve Bank of New York, foresees complications such as “redesigning debt securities; in some cases, redesigning financial institutions; adopting new social conventions for the timeliness of repayment of debt and payment of taxes; and adapting existing financial institutions for the calculation and payment of interest, the transfer and valuation of debt securities and many other operations.”

The Policeman May Not Be Your Friend

If we grant for a moment that phasing out all but small-denomination notes would accomplish what proponents claim — a marked reduction in crime, particularly crimes that are predatory in nature — would it nonetheless be desirable? Not necessarily. Even when gains appear to be greater than losses in both efficiency and seigniorage, other political-economy considerations may tip the scales into negative territory.

Pierre Lemieux, a senior fellow at the Montreal Economic Institute, likens the protection of the underground economy to constitutional rights, such as protections against self-incrimination and “unreasonable search and seizures,” that may also be invoked by criminals but still protect a free society. “As regulation increases, more people — consumers, entrepreneurs, unfashionable minorities — move to the underground economy. Thus, government cannot regulate past a certain point.” Cash, in other words, enables people not only to escape harmful or misguided government intrusions, but also, in an indirect but effective way, to express their political concerns.

Would alcohol prohibition in the United States have been repealed without routine evasion by countless Americans? Would the United States have moved toward marijuana legalization without the underground economy to introduce it to the middle class? Obviously none of these considerations excuses human trafficking and other forms of violence or brutality that are facilitated by the underground economy. But those in favor of restricting cash offer no more than impressions about the magnitude of these acts compared to harmless or beneficial uses of cash.

One should be very cautious about drastic government impositions that indiscriminately impinge on almost the entire population, no matter how deplorable the outrages they are intended to curb. Perhaps no issue illustrates these public-choice concerns more strikingly than the threat to financial privacy.

Consider the huge quantity of personal financial information that cash’s abolition would make readily available to government. We should be wary of schemes that enhance state intrusion into the remaining spheres of anonymity, despite the advantages these spheres may give criminals.

Several governments have engaged in compulsory currency swaps designed to rein in the underground economy by forcing tax evaders, money launderers and others holding large sums of cash to either face government scrutiny or find their cash hordes stranded. Although not intended to eliminate large-denomination notes permanently, these currency reforms offer instructive lessons about potential pitfalls of eliminating cash.

The most noteworthy case occurred on November 8, 2016, when India’s government gave people two months to exchange the country’s two highest-denomination notes for new ones. But in the wake of the resulting economic disruption, even the government’s annual economic survey euphemistically conceded that the experiment entailed “inconvenience and hardship.” Even enthusiastic supporters of the swap were somewhat taken aback by the consequences. Jagdish Bhagwati of Columbia University and his former students, Vivek Dehejia and Pravin Krishna, acknowledged that it probably “failed in achieving its primary goal” of curtailing tax evasion and corruption.

The United States has already experienced the unfortunate consequences of two crusades to crush behavior deemed illicit: alcohol prohibition and the war on drugs. These crusades have shared some of the same justifications that are made for phasing out cash. And just as the war on drugs has inflicted untold damage outside the borders of the United States, advocates hope the elimination of hand-to-hand currency will become an international campaign as well. What guarantees do we have that this war on cash will not have parallel unintended consequences in granting an enormous range of powers to the state?

* * *

In the final analysis, it seems only prudent to put the burden on advocates of currency restrictions to show that the benefits in terms of crime suppression and the extension of monetary policy in a deflationary environment are worth the costs in terms of the loss of seigniorage. And, of course, hovering in the background is the question of whether free societies can afford to extend the reach of government and financial services corporations any farther in an era in which privacy is threatened on numerous fronts.

In my own view, advocates are far more sanguine than is justified about the benevolence and competence of governments, particularly in developed countries, and unreflectively adopt what the UCLA economist Harold Demsetz has characterized as the “nirvana approach” to public policy. Their analysis gives short shrift to the potential for unintended consequences when the public relies entirely upon the farsightedness of policymakers.