This archived content originally appeared at, the predecessor site to, and does not carry the sponsorship of the Cato Institute.

The middle ground

(Prompted by a recent post by David Glasner.)

A number of advocates of the gold standard claim they want to return to a "real," pre-World War I style gold standard rather than the interwar or Bretton Woods versions, which proved not to be durable. Unlike the pre-World War I gold standard, though, where countries representing only about half of the world's economic output had central banks–and the biggest economy, the United States, was not among them–they propose to leave central banks in place.

Critics of the gold standard point to the disastrous experience of the interwar gold standard as an argument against returning to gold. What only a few acknowledge is that the interwar gold standard worked much differently from the prewar gold standard. In particular, research into the behavior of the Federal Reserve and the Bank of France has shown that, because they were central banks with discretionary monopoly powers rather than competitive commercial banks, they made decisions to accumulate and sit on large stocks of gold that competitive banks would not have done. Dick Timberlake returns to this point regarding the Federal Reserve in his new book Constitutional Money. He points out that when Franklin Roosevelt took the United States off the gold standard in 1933, the Federal Reserve held about substantially more gold than the minimum required by law; moreover, it had the power under the law to go below the minimum if it declared that emergency circumstances required doing so. The Fed could have followed a much more expansionary policy in the years after the stock market crash of 1929. Dick reiterates the theme of an important article he wrote several years ago, that the reason it did not do so was that it was guiding its policy by a version of the "real bills doctrine."  While benign for competitive commercial banks, the real bills doctrine has what Dick terms a "dark side" if implemented by a central bank as the Fed did.

Here's what I propose as the middle ground for the advocates and critics of the gold standard: central banking and a pre-World War I-style gold standard are incompatible. If you want the gold standard in durable form, you can't have central banking. If you want central banking, you should not mix it with the gold standard.

As I have written before, it will not do to use the experience of the interwar period as an argument against the gold standard simply. It is an argument against the gold standard under central banking. The Great Depression was Great precisely because it was uncommon. During all the previous centuries of experience with gold and silver standards there was nothing to compare to it. It is sobering in that context to observe that at present, most of  Europe (Germany, a few of its neighbors, and Russia being exceptions) and Japan are still below their pre-recession peaks of output, and that in some cases they have been below their pre-recession peaks longer than they were during the Great Depression. Prolonged economic sluggishness is just as possible without a gold standard as with it.


  1. The real bills doctrine says that any bank, central or private, should issue money only in exchange for short term real bills of adequate value. A bank that follows this rule will automatically receive a dollar's worth of backing assets for every dollar that it issues, and so the bank's money will hold its value. Whether a bank holds 100 oz of assets as backing for $100, or 300 oz of assets as backing for $300, each dollar remains worth 1 oz.

    The practice of issuing money for real bills has the effect of making the money supply grow and shrink with the needs of business, but it is important to note that the advocates of the real bills doctrine did not usually insist that money should be issued only for real bills. They were usually content that money should be issued only for "solid" assets, including government bonds. In practice, the real bills doctrine assured that when money was tight, people would bring solid assets to their banks to exchange for money, and when money was abundant, the excess currency would reflux to the issuing bank. The currency would be elastic.

    But consider what Timberlake has to say about the Fed's application of the real bills doctrine:

    "Monetary historian Clark Warburton, writing some years later,
    recounted the viciousness of “direct pressure.” In the early 1930s, he
    noted, the Fed Banks

    virtually stopped rediscounting or otherwise acquiring “eligi-
    ble” paper. This [hiatus] was not due to any lack of eligible
    paper. . . . It was due to “direct pressure” so strong as to
    amount to virtual prohibition of rediscounting for banks
    which were making loans for security speculation. . . . Federal
    Reserve authorities had discouraged discounting almost to
    the point of prohibition [Warburton 1966: 339–40]."

    In other words, the Fed was clearly NOT following the real bills rule of issuing money in exchange for solid assets. The Fed was practically refusing to issue ANY money at all. And this is supposed to be a failure of the real bills doctrine? No. This is a failure to understand the real bills doctrine, and Timberlake misunderstands it just as badly as the Fed governors did. Had the Fed been following the real bills doctrine, it would have freely issued cash to anyone who offered good security in exchange.

    By the way, the same error is at the heart of the popular idea that blames the real bills doctrine for the German hyperinflation of the 1920's. When the market interest rate is 200% per year, a central bank that lends at 5% is issuing money for assets of INADEQUATE value, and is therefore failing to follow the real bills doctrine.

    1. When you write "a bank holds 100 oz of assets", do you mean that the bank literally holds 100 ounces of gold (or whatever the standard of value is), or do you mean that the bank holds any collateral with the same market price as 100 ounces of gold at the beginning of the 90 days (or whatever the short term is)? Since a government bond qualifies as a "solid asset", you presumably mean the latter.

      What prevents decentralized banks from continually rolling over an ever increasing volume of government bonds, as a central bank does, and thus creating an inflationary spigot of money for government to spend?

      Don't decentralized banks have an incentive to treat government bonds this way if a government will not raise taxes sufficient to pay the principle and interest on its bonds? won't the "decentralized" banks conspire with government to impose an inflation tax rather than suffer a government default?

      A free bank must incorporate default risk into its expectation of the future value of its holdings when issuing notes in exchange for collateral, including the risk that a borrower's collateral depreciates before a default. If banks generally assume that government (or any particular creditor) cannot default, isn't that the same as having a central bank, and isn't it a self-fulfilling prophecy given the possibility of inflation?

      There is no single market interest rate for every borrower, right? If I assume one borrower is more likely to default and than another, don't I charge the former a higher interest rate?

  2. 1. Yes, "100 oz worth of assets" refers to anything of value that is currently worth 100 oz.

    2. If a bank holds 100 oz worth of bonds as backing for $100, and if those bonds are rolled over while the bank prints another $50 and uses it to buy another 50 oz worth of bonds, then the bank now has 150 oz backing $150, and $1 still equals 1 oz. If that extra $50 was not wanted in the circulation, then several things could happen: (1) $50 will reflux to the issuing bank (2) $50 could reflux to some other bank. (3) Banks would increase the interest they pay on cash holdings, until the public became willing to hold that extra $50. (4) If the channels through which that unwanted $50 were closed, so that the $50 cannot reflux, then that is an effective default by the issuing bank, and the money will lose value because it has lost backing.

    3. If the government won't support its bonds, then 100 oz worth of bonds will fall in value to 90 oz, and the $100 that those bonds were backing will drop in price to 0.9 oz/$. No amount of rolling over will change this.

    4. There is always some risk of default, so imagining that there is no such risk is a mistake. It makes no difference if the bank holding those (defaultable) bonds is a central bank or a decentralized bank. A loss of backing causes inflation in either case.

    5. Yes, high risk=high interest

    1. 3 is a devaluation of the currency. The government bonds still pay the promised "dollars", but these dollars buy fewer ounces and presumably less of everything else valued relative to ounces. In particular, the collateral held by banks against dollar loans is worth more dollars thus improving the banks' dollar balance sheet. The real bills doctrine doesn't prevent this inflation when a state commands nominally "riskless" bonds. The "riskless" bonds create the inflation, right?

      4 We can distinguish an inflationary default by currency devaluation from an outright default that pays fewer dollars then a bond promises. The former is possible because everyone must collect dollars to pay taxes, even if the tax revenue doesn't support the state's spending without continual devaluation of the currency in which everyone owes taxes. If everyone owes taxes in ounces rather than dollars and if the state spends ounces and its bonds promise ounces, then state spending without commensurate taxes requires an outright default, and that's why we now have "dollars" divorced from ounces of silver, right?

      5 So banks can pay ever lower interest rates on nominally "riskless" government bonds as long as other lending is unprofitable even at zero interest rates, and that's why we have "stagflation", right?

      1. 3. (a) If the bank's assets drop from 100 oz to 90 oz, and as a result, each of the $100 it has issued fall in value to .9 oz., then the net worth of the issuing bank is unchanged.
        (b) The RBD says that adequately backed dollars will hold their value. By extension, a loss of backing will cause inflation. So if 'riskless' bonds fall in value, there will be inflation.

        4. Correct. If a state has issued $100, but its assets (mainly taxes receivable) total only 90 oz., then either the dollar must fall to .9 oz/$, or the state must default on 10% of its dollar obligations.

        5. (a) Not sure what you mean. Banks don't pay interest on government bonds. They receive it.
        (b) A likely explanation for stagflation:
        i. The money issuer is losing assets. This causes inflation.
        ii. The money issuer is failing to issue enough money for people to conveniently conduct their business. This causes recession.

        1. 5. I should have written, "So banks can accept ever lower interest rates on nominally 'riskless' government bonds …"

          In the scenario I imagine, the money issuer (a central bank) issues notes backed by a government's bonds, and the government spends the money. The government doesn't raise sufficient revenue to pay promised principal and interest, but the banking system nonetheless rolls over the government's debt by issuing more money for more bonds, because it assumes that the government cannot default in nominal terms ultimately, and it's right, because it always issues as much money as necessary to roll over the government's debts.

          The recession occurs for whatever reason, and during the recession, banks cannot find profitable lending opportunities or even collect all of the principal and interest owed on existing loans, but they never cease rolling over the government's debt, because the government cannot default.

          A bank's other assets can fall in value as its debtors become more likely to default during a recession, but its government bonds never fall in value for this reason, so banks are even more inclined to buy the government's debt during a recession. The government spending that banks fuel this way is inflationary, but the inflation supports the price of collateral securing the bank's other loans, so banks accept it.

          1. That 'rolling over' process would be cut off before it gets started. People who find themselves earning negative real interest rates on government bonds would prefer to buy real assets like land, so the nominal yield on government bonds would rise to the point where a government bond is just as attractive to investors (in real terms) as any other investment of comparable risk and liquidity.

  3. "… most of Europe (Germany, a few of its neighbors, and Russia being exceptions) and Japan are still below their pre-recession peaks of output, …"

    Japan's population is shrinking, its labor force is shrinking even faster and its population of retirees (who consume capital otherwise available for growth) is rising. Since these trends seem unlikely to change, when does it make no sense to expect Japan's total output to increase? Productivity may increase without total output increasing.

Comments are closed.