The myth of the “cashless society”: How much of America’s currency is overseas?

The rapid growth of substitutes for cash, particularly debit and credit cards, has led economists to predict the advent of the “cashless society”. Yet cash holdings in most developed economies continue to grow and in the U.S., per capita currency holdings now amount to $3000. This paper revisits the long-standing controversy concerning the whereabouts of U.S. cash. Specifically, we employ a previously confidential data source on net shipments of U.S. currency abroad to re-estimate the fraction of U.S. currency held overseas. Contrary to the widely cited figure that 65 percent of U.S. currency is abroad, we now find that direct evidence supports the notion that overseas holdings amount to less than 25 percent. Currently, the official figure for the percent of U.S. currency held abroad as published by the Federal Reserve in their Flow of Funds Accounts and by the Bureau of Economic Analysis in the U.S. Balance of Payments Accounts is 37 percent. This official figure is a proxy variable that is supposed to mimic the previously confidential data series maintained by the New York Federal Reserve. Judson (2012) made this series public enabling us to discover that the official estimates of currency abroad require downward revision to reflect accurately the newly released data on actual cash shipments abroad. We also review the “indirect” approaches to estimating the fraction of currency overseas employed by Porter and Judson (1996) and Judson (2012). We find that these indirect methods to be innovative but deeply flawed due to violations of their restrictive assumptions. Moreover, sensitivity analysis reveals the estimates highly sensitive to alternative specifying assumptions. We conclude that the large estimates of currency abroad obtained by these methods are spurious. The paper also examines the temporal pattern of overseas holdings of U.S. currency and finds that the observed decline in the demand for U.S cash abroad coincides with the growing popularity of the Euro and its growth as a second currency held outside the Euro area between 2003 and 2008. These new findings have significant implications for estimating the domestic money supply and other domestic monetary aggregates; for estimating the net benefits of seigniorage earnings of the Federal Reserve; for forecasting changes in output and prices and for estimating the amount of unreported income and tax evasion in the U.S.