Seeking Monetary Stability: The Elusive Gold Standard

I am not so much opposed to the gold standard itself as I am underwhelmed by arguments asserting the uniqueness of its benefits. My “meh” derives from two propositions: 1) Any policy benefit from a gold standard can be replicated by a set of rules governing the supply of fiat money; 2) the political will necessary to implement a set of rules that mimic the effects of a gold standard would never be greater, and almost surely would be less, than the political will necessary to implement a gold standard. Therefore, it would be better for advocates of sound money to support a set of rules by which the supply of money increases at a set rate—ideally, a rate that would, on average, result in zero inflation.

Advocates of the gold standard—entirely correctly—argue it is a commitment mechanism by which the government’s ability to manipulate the money supply, and so the value of money, would be constrained. The thing is, as we all know by experience, the strength of a policy commitment to a gold standard is no greater than any other statutory (or constitutional) commitment.

The question is, what’s the best way back to stable money?

What’s important about the gold standard is the discipline it provides to support price stability. A rule-based monetary system can provide as great a commitment to price stability—perhaps even more—but would be much easier to implement.

Most economists, including Milton Friedman, argue as long as the rate of inflation or deflation is stable—that is, is fully anticipated—then costs to inflation or deflation are minimal. I would go a bit further, arguing there is no justification for non-zero inflation.

Friedman writes in the opening chapter in his book The Optimum Quantity of Money,

Anticipated inflations or deflations produce no transfers from debtors to creditors which raise questions of equity; the interest rate on claims valued in nominal terms adjusts to allow for the anticipated rate of inflation. Anticipated inflations or deflations need involve no frictions in adjusting to changing prices. Every individual can take the anticipated change in the price level into account in setting prices for future trades. Finally, anticipated inflations or deflations involve no tradeoffs between inflation and employment.

The only reason to set the inflation rate consistently greater than zero is the belief that decreasing the interest rate in response to a recession can reduce the length and severity of the recession. Of course, to the extent that response is anticipated, then Friedman’s analysis above—anticipating the rational expectations revolution in macroeconomics of the 1970s—suggests an interest rate change will have no real effect. (New Keynesian economists have produced models showing sticky prices could result in real economic effects even for anticipated changes in the money supply. But that is a matter for another day.)

Friedman develops a case for a policy of modest, sustained deflation in his book. Friedman argues (modest) deflation—set at the same percentage as the nominal interest rate—would create real returns for holding money. Individuals and firms would therefore minimize transaction costs they otherwise incur when they take monetary transactions to secure positive returns for the funds they would otherwise hold as money.

Overall savings from reducing transaction costs would be modest. And Friedman concedes the “practical consideration . . . that the optimum rate of price decline will change from time to time” means that it would be “unwise to recommend as a policy objective a policy of deflation of final-product prices sufficient to yield a full optimum in the sense of this paper.”

With no compelling reason for a policy of planned, sustained inflation, and no practical reason for a policy of planned, sustained deflation, a policy of stable money—zero planned inflation—would seem at least weakly optimal.

Whether money is denominated as gold or as fiat money, zero inflation requires the supply of money to increase at the same rate as production increases. If the economy produces three percent more goods and services, then the inflation would be zero only if the supply of money also increases by three percent.

But discretion is the problem, and economic measures lag what actually happens in the economy. So the goal would be a non-discretionary increase in the money supply equal to the anticipated average increase in production over a given time span. This would take no more political will, and probably much less, than a return to the gold standard.

Of interest, given variance in the price of gold as supply and demand for gold change, it is entirely possible a rule-based system of fiat money would generate more price stability than a gold standard. To wit, a dollar of fiat currency would be backed by the basket of commodities (including gold, by the way) represented by the entire U.S. economy rather than backed by just one commodity. Hence, the variation of the value of a dollar backed by the basket of commodities reflected in the entire U.S. economy would almost always be less than the variation in the value of a dollar backed solely by gold.

The critical factor, whether for gold or for a rule-based monetary policy, is the commitment to maintain the regime, whether gold or the rule. The political will it would take to get a rule-based change in Federal Reserve policies through Congress would certainly be less than persuading Congress to re-adopt a gold standard.

On the other hand, even if Congress were to re-adopt the gold standard, it would be no more difficult to repeal a gold standard than it would be to repeal a rule-based policy for the Fed. A rule-based policy shares all of the virtues of a gold standard, yet, as a practical matter, would be politically easier to enact or adopt than a gold standard.