Sound Money in Theory and Practice

Bretton Woods. Gold standard, NGDP targeting, sound money
William McKinley's 1896 campaign poster.

Bretton Woods. Gold standard, NGDP targeting, sound money

In this commentary, I will analyze the concept of sound money and its relevance today.  The concept evolved in the 19th century as many countries adopted the gold standard.  It became associated with commodity money or “hard currency.”  For example, Mises (1966: 782) stated:

The principle of soundness meant that the standard coins — i.e., those to which unlimited legal tender power was assigned by the laws — should be properly assayed and stamped bars of bullion coined in such a way as to make the detection of clipping, abrasion, and counterfeiting easy.  To the government’s stamp no function was attributed other than to certify the weight and fineness of the metal contained.

There was no requirement that “standard coins” be the exclusive or even preponderant means of payment in day-to-day transactions.  So long as banks of issue (private or central banks) maintained convertibility, then the monetary system had the characteristics of sound money.  As Mises suggested, government’s role was minimal.

As a matter of history, sound money is associated with commodity money.  Mises’ characterization assumes commodity money.  Can there be sound fiat currency?

Most 19th century writers on money assumed a system of commodity money with allowance for temporary suspensions during wartime and a return to the standard in peacetime.  The suspension of specie payments by the Bank of England during the Napoleonic Wars prompted banker Henry Thornton to author in 1802 a treatise on managing a paper currency: An Enquiry Into the Nature and Effects of the Paper Credit of Great Britain.  Thornton was a successful banker with an economist’s understanding of banking, finance, and the real economy.  He pioneered analytical distinctions that would not be rediscovered for almost another century: the distinction between real and nominal interest rates, and the concept of an equilibrium or natural rate of interest.

The volume should have become a standard reference work.  But it came to be forgotten because the Bank of England re-established convertibility, and other countries gradually adopted the gold standard over the course of the 19th century.  How to manage a fiat money system (paper credit) ceased to be a practical issue.

When the Federal Reserve System was created at the end of 1913, the United States was on the gold standard along with most of the rest of the world.  The Federal Reserve was not created to manage money in the modern sense, but to provide a national currency as part of a gold standard.  No one was thinking of managing a fiat currency on the eve of World War One.  That was all soon to change with the requirements of wartime finance.

After World War One, the world returned to a global, pseudo gold standard that was chronically short of gold reserves.  The currencies of many countries were overvalued relative to gold.  When the system collapsed in the 1930s, countries were thrust into a fiat currency world without a playbook.  For a time, there were efforts to restore the global gold standard but they came to naught.  World War Two interrupted any effort to craft a new international monetary system.

The post-War, Bretton Woods system constituted the new global monetary order.  Volumes have been written on it.  I do not share the nostalgia of some for it.  It was even less of a gold standard than existed in the interwar period.  In truth, it was a dollar standard.  The dollar was pegged to gold and other currencies pegged to the dollar.  There were numerous exchange-rate adjustments.  The system contained inner contradictions.  Inevitably, the producer of the dominant currency was bound to abuse its “exorbitant privilege” and the United States did so.  The system collapsed and the world was then on a fiat standard.

There was no accepted theory of managing money in a fiat money world.  This was not Henry Thornton’s world in which fiat money was a temporary expedient with an expectation of a return to specie conversion.  It was not the world of the classical quantity theory, which was constructed in a commodity-standard world.  The quantity theory demonstrated the limits of monetary expansion (or changes in the demand for money) before prices would begin to rise sufficiently to threaten convertibility.  In a classical gold standard, the supply of money is endogenous and the price level fixed in the long run.

Milton Friedman and the monetarists offered a restatement of the quantity theory and a model of monetary control for a fiat currency.  Friedman, his students and colleagues believed they had discovered stable empirical relationships among the monetary base, broader measures of money (especially M2), and the demand for money (Friedman 1956 and Friedman and Schwartz 1956).  When monetary targeting was finally implemented by the Volcker Federal Reserve in the 1980s, the posited empirical relationships broke down.  The Fed abandoned monetary targeting.

What followed was a period that John Taylor dubbed the Great Moderation, in which the Fed and other central banks seemed to get it right.  There was enhanced macroeconomic stability (as measured by decreased variance in prices and output).  Taylor discerned that the Fed was following a tacit rule, which others called the Taylor Rule.  But the Fed and then other central banks began to deviate from the rule by lowering interest rates in response to the Dotcom bust.  Taylor (2009) argued the housing bust was the consequence of the boom created by the policy of low interest rates.  “No boom, no bust.”  Central banks have not returned to a monetary rule.  Instead, they have engaged in monetary improvisation.

Money in the 21st century is proving immune to control by central bankers.  The relationship between monetary reserves and various monetary aggregates (the money multiplier) has broken down.  More precisely, central banks appear to have lost the ability to control inflation.  In the United States, Europe and Japan, inflation rates have remained chronically below central bank targets over the course of the economic recovery from the Great Recession.  (The growth of real GDP has also been subpar.)  Economists as diverse as Jerry Jordan (2016) and James Bullard (2016) have questioned whether our textbook models of money creation and inflation control are any longer valid.  That is not to say that future inflation rates will not rise to two percent or beyond.  If they do so, however, it will likely not be the consequence of any central bank policy actions (Jordan 2016).

To reiterate, I question whether we ever had a practical theory of how to manage money in a fiat money world.  The proponents of monetary rules believe they have such a theory. One class of such rules involves NGDP targeting.

The specific question I pose for advocates of NGDP targeting is how today will anything the Federal Reserve does to its balance sheet alter the growth rate of NGDP in a predictable fashion?  The answer to such a question could be that the central bank should do more.  How much more?  And what, then, becomes of the rule?  It sounds like a recipe for discretion.  In any case, central banks have been unable to get either component of NGDP to grow in a normal or predictable manner.

Monetary institutions and policies vary among the major central banks.  For instance, both the European Central Bank and the Bank of Japan have instituted negative interest rates on commercial bank deposits at the central bank.  Meanwhile, the Fed has been paying interest on bank reserves for some time.  The institutions and policies differ, are even opposed to each other, but the policy failures are common.  (The policies have failed on their own terms, regardless of whether one agrees with them.)

Let me return to the classical idea of sound money.  Sound money is a rule, but of a different kind than modern monetary rules such as the Taylor Rule or NGDP targeting.  Sound Money was not a rule based on empirical relationships among economic variables.  It was not invented, but discovered.  It is more analogous to the rule of law.  Mises (1971: 414) made this point clearly. “Ideologically it [sound money] belongs in the same class with political constitutions and bills of rights.  The demand for constitutional guarantees and bills of rights was a reaction against arbitrary rule and non-observance of old customs by kings.”

The argument for sound money is not merely a technical economic argument, but a political economy and even constitutional argument.  When classical economists contended that commodity standards were a bulwark against inflation, they did not suggest that there would be no variability of inflation under a gold standard.  Their own experience told them otherwise.  Rather, they recognized that a gold standard was protection against arbitrary actions by sovereigns to depreciate the currency.  Protection against arbitrary and capricious governmental actions is what constitutions are meant to provide.

Is there a way to avoid arbitrariness in monetary matters in a fiat money system?  Can a monetary rule of some type today provide the protections that existed in the classical, pre-World War One gold standard?  These questions are central to the debate over monetary rules.  They apart from the technical ones I raised above.  Both sets of questions need to be addressed in debates over monetary policy.



Bullard, J. (2016) “Permazero.” Cato Journal. 36 (Spring/Summer): 415-29.

Friedman, M., ed. (1956) Studies in the Quantity Theory of Money. Chicago: The University of Chicago Press.

Friedman, M. and A. J. Schwartz (1963) A Monetary History of the United States, 1867-1960. Princeton: Princeton University Press.

Jordan, J. L. (2016) “The New Monetary Framework.” Cato Journal 36 (Spring/Summer): 367-83.

Mises, L. v. (1971 [1952]) The Theory of Money and Credit. Irvington-on-Hudson: The Foundation for Economic Education.

________ (1966) Human Action, 3d ed. Chicago: Henry Regnery.

Taylor, J. B. (2009) Getting Off Track. Stanford: The Hoover Institution Press.


  1. Isn't having the separation of fiscal and monetary institutions a level of protection from government devaluation? It seems historically that members of the Federal Reserve Board are appointed more on their academic/professional merits rather than political views on how the economy should be managed, but this does seem a weak point in the system where political officials can push for "their people" to occupy leadership positions and influence monetary policy based off political ideology rather monetary theory.

  2. Gerry, thank you for this very helpful overview. It provides an indispensable long-term perspective sorely missing in current discussions of monetary problems. Completely agree with your comments and judgment re: Henry Thornton.

  3. Mr. O'Driscoll as usual hits the nail on the head. And a tip of the hat to Tom Humphrey, also. A rival school of thought involving monetary policy and the banking system is the 100 percent reserves system that has older roots but was brought into prominence by the University of Chicago faculty in the 1930s. See, e.g., Ronnie J. Phillips, The Chicago Plan and New Deal Banking Reform, Armonk, NY: M.E. Sharpe (1995). In recent years, John Cochrane of the University of Chicago also has been writing about this topic. Murray Rothbard also wrote about this topic, but in the context of a 100 percent gold reserve. Short answer: Mr. O'Driscoll is right that central banking without a commodity anchor tends to go off the price level rails. But if government will not allow gold (or silver, or both) into the banking system, then either pure free banking or 100 percent reserve banking might be required because everything in between becomes purely political and negotiable, with consequences we all can observe around us now. — Walker Todd, Chagrin Falls, Ohio

  4. "What followed was a period that John Taylor dubbed the Great Moderation, in which the Fed and other central banks seemed to get it right. There was enhanced macroeconomic stability (as measured by decreased variance in prices and output)."

    I've come to regard the idea that the Federal Reserve got better at stabilizing the economy as a kind of insidious myth, a mistaken interpretation of flawed data.

    In the first place, let's deal with the data as is: There has, apparently, been some general reduction in output volatility in the post WWII US Economy. I would argue, however a significant chunk of this decline in volatility-especially volatility in employment-reflects in part the growing role of sectors relatively less susceptible to the business cycle in our economy-broadly speaking, the rise of services relative to manufacturing.

    To see this effect, one need only analyze the Establishment Survey's data on employment in different sectors. You can get most of the effect just by separating out Goods Producing Industries and Service Providing Industries, but I went ahead and separated out 17 different series:
    Transportation & Warehousing (Trade, Transportation, and Utilities minus Retail, Wholesale, and Utilities), Retail Trade, Wholesale Trade, Utilities (estimated w/ constant growth rate before 1964-sector is acyclical), Federal Government, State & Local Government, Education and Health Services, Financial Activities, Information Services, Leisure & Hospitality, Professional & Business Services, Other Services, Nondurables, Construction, Mining & Logging, Boats, & Non-Boats Durables. I used an HP filter with a lambda of 1.44*10^6 to detrend each series and the total to A) find percent deviations from trend and B) find the current, non cyclical component of of the trend of each sector as a percentage of total employment.

    Doing this, I can create two series: the first is simply the percent deviation from trend of total employment, the second is what those percent deviations would have been if the economy had the same structure as it does today-a weighted average of the percent deviations from trend in each sector. This is what that looks like:

    The red series is the original percent deviations of employment from trend, in blue is the weighted average. Ignoring for the moment the effects of World War II itself, clearly the shift away from manufacturing to services has contributed to the "moderation"-a shift that cannot plausibly be linked to monetary policy improving. A similar picture emerges if I take the weighted average percent deviations, multiply the original trend by them, and then compare year over year growth rates:

    In the second place, especially in a larger historical context, it may not in fact be the case that volatility was exceptionally low during the "Great Moderation"-that may simply be an artifact of the data. Consider this recent paper, for example of an alternative analysis that comes to a different conclusion:

    Ritschl, Albrecht, Samad Sarferaz, and Martin Uebele. "The US Business Cycle, 1867-2006: A Dynamic Factor Approach."

    Which found that, if an output index is constructed allowing time-varying weights on the various series, the post World War II period is not less volatile than the Classical Gold Standard, and the "Great Moderation" is in fact comparable in volatility to an earlier moderation in the 1960's. Given that this analysis does not include the effect of the Great Recession The conclusion could presumably be even stronger. Note that the apparent moderation not being an improvement on the 60's is contrary to the sort of Whiggish view that the Fed first caused the Depression, the the Great Inflation, and then finally learned how to really manage the economy. On the contrary it seems more likely to me what looked like improved performance was really just luck.

Comments are closed.