John Maynard Keynes once marveled at "how, starting with a mistake, a remorseless logician can end up in Bedlam." (Bedlam was the nickname of a London madhouse.) Keynesian or New Keynesian macroeconomists who start with the mistaken premise that a central bank cannot fight recession except by lowering nominal interest rates have been remorseless logicians in Keynes's sense. In the hope of further empowering central banks to fight recessions, presumably for the benefit of the public, they have ended up like mad social scientists with schemes that would deliberately punish the public for holding currency.
From the premise that nominal interest rates must be cut, together with the fact that nominal interest rates are currently low by historical fiat-currency standards, one readily finds that the "Zero Lower Bound" on nominal interest rates is a looming obstacle to anti-recession policy. At the ZLB the central bank supposedly "runs out of ammunition." Economist Lawrence H. Summers, thinking of the US Federal Reserve's policy-making under the nominal Fed Funds Rate targeting approach that it used in previous recessions, has warned that "typically interest rates come down 500 basis points to contain recessions" but "there isn't going to be 500 basis points of room any time in the foreseeable future." Thus central bankers "don't really have the fuel in the tank to respond" to a new recession.
The new inflationist proposal
Some of the economists who are convinced that the central bank needs the ability to cut rates deeply propose to restore sufficient "room" above the ZLB by pushing short-term nominal interest rates back above 5%, where the nominal Fed Funds Rate could be found between 1968 and 1991. Why was the nominal Fed Funds rate almost continuously above 5% during those years, but not before or after? Because expected inflation was higher during those years. Thus Olivier Blanchard, Giovanni Dell'Ariccia, and Paolo Mauro (2010), and other authors whom I have called "The New Inflationists," logically enough – discovering a new benefit of higher inflation while downplaying its costs – have suggested however tentatively that central banks should raise their inflation-rate targets. Blanchard et al. wrote:
The inflation tax is clearly distortionary, but so are the other, alternative, taxes. Many of the distortions from inflation come from a tax system that is not inflation neutral, for example, from nominal tax brackets or from the deductibility of nominal interest payments. These could be corrected, allowing for a higher optimal inflation rate.
The "optimality" of a higher inflation rate is clearly not a Pareto-optimality in which ordinary currency-holders are no worse off. When currency pays zero interest, a higher inflation rate clearly makes ordinary members of the public worse off by reducing their net benefits from holding currency.
Making cash more costly to store
Other New Keynesians do not want to raise the central bank's inflation-rate target, but instead want to provide more room for nominal rate-cutting by pushing the Lower Bound on the nominal interest rate below zero. They note that the current ZLB exists because deposit-holders and bond-holders can reject yields much below zero (farther below than the percentage cost of storing currency) by switching wealth into currency hoards that pay a zero nominal yield. By doing so they can keep deposit and bond yields above zero. The freedom to flee into currency having thus been identified as an obstacle to negative interest rates, Kenneth Rogoff and other currency prohibitionists have seriously put forth schemes to eliminate the use of paper currency, or at least to make its storage more costly by making the currency notes worth any large nominal sum more bulky. They propose to abolish large (and then medium-sized) notes to make cash storage more costly (think of storing $10 bills instead of $100 bills) and thereby enable deeper negative nominal interest rates in recession before cash storage is triggered.
Alternatively, Rogoff could have proposed to print $100 on cardboard 10 times as thick as a present-day $100 bill. When I offered this idea to him during a debate we had in December, he dryly remarked that he'd heard it before.
Measures to raise the cost of storing currency potentially damage the welfare of ordinary currency users no less than a raising the price of holding currency through higher inflation.
Paying a negative nominal return on currency
Two IMF staff economists, Katrin Assenmacher and Signe Krogstrup, have recently revived a third proposed way to combat the ZLB, this time without raising inflation and without abolishing currency. Namely: Make currency unattractive to hold relative to bank deposits by imposing a negative nominal return on currency. Under the system they consider, currency notes and coins would be progressively devalued against the unit of account. To penalize the holding of a $100 bill by 3.65% per year, for example, the bill would be scheduled to decline in value by one cent per day until it reached only $96.35 in redemption value one year hence. I borrow this example from historical Confederate States $100 bonds that were designed to gain one cent per day, thus 3.65% per annum, to make their present values relatively easy to compute. To make a currency note's redemption value today discoverable at other rates of decline, modern notes might come equipped with magnetic read-only stripes or bar codes, as once suggested by Marvin Goodfriend (2000). Redemption in terms of what? In terms of a unit of bank reserve deposits on the books of the central bank. Commercial banks would only give $97 of deposit credit for a note on a particular day when that is all the central bank would give them.
Assenmacher and Krogstrup's proposal is neither novel nor derived from historical examples. They correctly trace its lineage to Silvio Gesell, who was admired by Keynes but generally considered a crank by other economists, and his 1916 scheme of punishing currency holders by imposing a "demurrage fee." Under Gesell's scheme, "money would need to be stamped at regular intervals to remain valid and that these stamps would have to be purchased." Among modern economists they cite earlier proposals by Goodfriend (2000), by Willem H. Buiter (2009)—who discusses all three devices for punishing currency-holders—and by Ruchir Agarwal and Miles Kimball (2015). The device of paying a negative return on currency is potentially just as harmful to currency-holders' welfare as the previous two devices.
Indirectly, proposals to pay a negative nominal return on currency also build on the earlier monetary economics literature that discussed how to pay a positive real return on currency with the goal of achieving an "optimum quantity of money" in the sense of Milton Friedman's much-discussed concept. Proposals to pay a negative nominal return (and given a 2% inflation target, a negative real) return on money may accordingly be considered policies for achieving an inefficiently small a stock of real money balances.
Still more distantly, paying a negative nominal return on currency in circulation by changing its unit-of-account value recalls the medieval mint practices of "crying up" and "crying down" debased silver coins, that is, increasing or reducing their official unit-of-account value without reminting them to add or subtract silver. The aim of the medieval mints was to enhance the prince's seigniorage revenue, rather than macroeconomic policy. But their method, although discontinuous, was similar.
The mistaken starting point
The mistake that all these proposals start from is the premise that monetary policy can't be expansionary unless it lowers nominal interest rates. But a lower nominal interest rate is at most something temporarily achieved by a surprisingly expansionary monetary policy, thus at most a short-run indicator of the stance of policy. It is a poor and misleading indicator at that, for reasons that Scott Sumner has spelled out. A lower nominal interest rate is neither necessary nor sufficient for a more expansionary monetary policy that would help raise an economy out of recession.
We can judge whether a monetary policy is stabilizing only by gauging its effect on moving the economy toward a stable goal. A monetary policy helps to stabilize, in one well-established and coherent view, if it helps to move the economy toward a smooth target path for the level of nominal GDP. Monetary policy-makers should thus worry about the level of NGDP, not about the nominal interest rate. From the equation of exchange, MV=Py, we know that NGDP (or Py) equals total spending on final goods (MV). The central bank can increase MV by expanding the money stock in the hands of the public M (and taking no offsetting action to reduce the turnover or velocity of money, V).
As Leland Yeager taught, the problem of unsold goods and unemployed labor in a recession has its counterpart in an unsatisfied excess demand to hold money. The problem can be relieved slowly and painfully by waiting for prices and wages to fall and thereby raise the level of real balances, or more promptly by a timely and well-measured expansion of money in the hands of the public. If the problem of too-low MV has come about through a sudden drop in M, the obvious solution is to restore M. If the problem has come about from a fall in velocity V, the central bank can offset it by increasing M to restore NGDP. To expand money held by the public, a drop in the nominal interest rate is neither necessary nor sufficient. A Fed purchase of non-financial assets could do the trick without lowering interest rates.
Negative interest rates on deposits and currency can be seen as a way to try to raise V by penalizing money-holding. But the actual effectiveness of deeply negative interest rates at raising V is empirically unclear, the experiment not having been run. One obvious concern is that a tax on holding money of all kinds may not prompt the public to spend more on final goods, but rather prompt them to attempt to preserve their wealth by shifting it out of negative-yielding deposit and currency balances and into nonfinancial assets like gold and other commodities. In the process, the banking system will shrink in real terms, which will constrain rather than stimulate real investment. If the aim is to restore NGDP to its reference path, the central bank should focus on known and reliable ways to raise NGDP.