Current evidence on the resource costs of irredeemable paper money.

• Author: Watts, Tyler
• Position: Report

I do not know of any attempt to measure the real resource costs of an irredeemable paper currency and to compare such costs with the real resource costs of a commodity currency. That is clearly a much needed research project.

--Milton Friedman

In 1986, Milton Friedman published a brief article in the Journal of Political Economy suggesting the possibility that real resource costs associated with the production and use of money could be greater under the current fiat money regime than under the commodity money regimes that preceded it. His article, and the broader implication about the resource costs of paper money, however, received scant attention. For by 1986, the Fed's disinflationary policies had come to full fruition, inaugurating the Great Moderation that ushered in two decades of low inflation, low unemployment, and strong real growth in the U.S. economy. Private accumulation of gold coins and bullion as an inflation hedge, safe-haven asset (hereafter "investment demand" for gold, following the terminology of the

World Gold Council) had ballooned in the 1970s along with the price of gold, but the Voleker disinflation brought the gold price back down to earth, undermining much of the rationale for ongoing monetary gold accumulation. (1)

More recently, the world has experienced severe financial crises, recession, and massive central bank interventions aimed at preventing a recurrence of the Great Depression. Unconventional monetary policies--in the form of ultra-low interest rates, quantitative easing, and macro-prudential regulation--by the Federal Reserve and other major central banks have contributed to asset price booms in commodities. In particular, in 2012, gold--the traditional safe haven asset still widely perceived as a hedge against inflation and economic uncertainty--saw both its average annual nominal and real price soar to new heights, as shown in Figure 1.

The gold boom is not just a price boom, but a production boom. The high price of gold is largely a result of the extra safe-haven investment funds being diverted into gold holdings during times of fiscal and monetary uncertainty, and this high price in turn induces mining entrepreneurs to direct additional resources into gold production. Although many researchers and commentators argue that the world is near "peak gold" production--that is, world gold output must at some point peak and then decline due to the finite amount of gold extant in the earth's crust--it appears that gold output has indeed responded to real price changes over the past several decades, albeit with a significant lag. Gold, like all fixed mineral resources, exhibits rising marginal costs of production in the short run with fixed production technology. Hence, the lagged output response to higher real prices represents a clear increase in real resource use in the gold mining industry. As Butterman (1980: 377) explained in the U.S. Geological Survey annual Minerals Yearbook for 1978-79, "The increasingly strong gold price provided the incentive for extensive exploration for gold deposits and the development of new mines. Retreatment of old tailings dumps, and the heap leaching of low-grade ores, became economically feasible." Figure 2 confirms that higher real prices of gold--driven largely by bouts of increased investment demand--do indeed elicit an increased supply effort, and hence a greater resource cost of gold production.

In light of these rising resource costs of investment gold production, Friedman's unanswered question acquires renewed relevance. By comparing the net annual real value of gold investment in the fiat money era against the classical gold standard era, we seek to begin the undertaking of Friedman's "much needed research project."

The remainder of this article proceeds as follows. First, we present some estimates of the resource costs of the classical gold standard. While standard estimates assume that the classical-style gold standard requires that 1.5-2.5 percent of annual real output be dedicated toward monetary gold production (Friedman 1953, 1960), recent revisions, based on a more realistic fractional reserve banking practice of the actual gold standard era, revise this number significantly downward, to about 0.05 percent of GDP (White 1999). The estimates of scholars using the historically and dieoretically accurate fractional reserves practice serves as a baseline by which to compare resource costs of net monetary gold production in both the modern pure fiat money regime and the classical gold standard.

Next, we present new evidence on die resource costs of gold across both monetary regimes. We look at trends in gold investment both as a percentage of GDP and as a percentage of total gold demand. We find that in die last several years, not only has gold investment increased significantly as a fraction of total gold demand, from roughly a 10 percent average in the pre-2008 years, to nearly 40 percent in the post-2008 period, but gold coin and bullion holding has approached, in real per-capita terms, levels last observed during die classical gold standard era. These data give weight to Friedman's concern that the "direct resource cost of the gold and silver accumulated in private hoards may have been as great as or greater than it would have been under an effective gold standard" (Friedman 1986: 644). (2)

In die final section, we conclude by reiterating the notion that a real resource opportunity cost is not unique to the gold standard. Fiat currencies are much more susceptible to devaluation and deflation, and rational economic actors anticipate those risks by continuing to accumulate quasi-monetary gold. If the costs associated with gold hoarding and other investment vehicles directed at protecting wealth against fiat money volatility are seen to approach the levels of monetary gold production under historical gold standards, then the case for fiat money cannot rest on reducing resource costs in practice. Our results indicate that, in two major fiat money volatility episodes in the last 40 years, fiat money regimes have drawn gold into quasimonetary or investment uses at levels approaching monetary uses during the classical gold standard.

Resource Costs of Money

A chief complaint against the gold standard is that there are opportunity costs associated with the real resources that must be dedicated to the mining and minting of coins. As Friedman (1962: 40) states,

The fundamental defect of a commodity standard from the point of view of society as a whole is that it requires the use of real resources to add to the stock of money. People must work hard to dig gold out of the ground in South Africa--in order to rebury it in Fort Knox or some similar place. The necessity of using real resources for the operation of a commodity standard establishes a strong incentive for people to find ways to achieve the same result without employing those resources. Friedman has supplied widely cited estimates of this resource cost, in terms of the percentage of total national income required to be directed toward gold mining each year to add sufficient monetary gold stocks so as to maintain price level stability consistent with real economic growth. Assuming a 100 percent reserve gold standard monetary system (i.e., one in which all hand-to-hand currency consists of gold coin or fully backed gold certificates, and all bank deposits are 100 percent...