Contrary to conventional wisdom, central banks cannot control inflation. They should therefore give up inflation targeting and embrace monetary reform. What we need is “active” (or “positive”) money, issued by public and private issuers (e.g., central banks issuing 100% money and private issuers producing bitcoins by validating transactions in the blockchain).
Our understanding of the relationship between money and inflation seems to go around in circles. In the 1960s and 1970s, economists in general and central bankers in particular paid little attention to money. This changed in the 1980s, when money was seen as the key driver for inflation. Today, we seem to have returned to the view that money does not matter (or at least not very much). In my view, the ambiguity on money is related to conventional economics’ lack of understanding of the creation of money. Conventional economics treats money like a good: its exchange value declines when there is more of it. But in reality, money is a liability created by the extension of credit, and its relationship to goods prices is highly complex.
Going around in circles
Arthur Burns was a preeminent American economist and chairman of the Federal Reserve during the 1970s. According to the minutes of a meeting of the Federal Reserve Board on June 11, 1970, Burns “did not believe that the Federal Reserve should be expected to cope with inflation single-handedly. The only effective answer, in his opinion, lay in some form of incomes policy. ” Thus, Burns supported the introduction of wage and price controls in August 1971 by President Richard Nixon to combat inflation that had risen to six percent per year. As we now know, this “form of incomes policy” was an utter failure. The 1970s became the decade with the highest inflation since World-War-II.
In 1979, when inflation was running at 11 percent, Paul Volcker took the chair at the Federal Reserve. Looking back at his term in an interview in 2000, he said: “It always seemed to me that there is a kind of common sense view that inflation is too much money chasing too few goods. You could oversimplify it and say that inflation is just a monetary phenomenon. There are decades, hundreds of years, of economic thinking relating the money supply to inflation, and people to some extent have that in their bones.”  He raised the Fed’s policy rate, the Fed Funds Target Rate, to 20 percent in 1981 in order to contain money supply growth, and he got inflation down to 3 percent by 1983. Volcker seemed to have once and for all proved Milton Friedman’s thesis right that “inflation is always and everywhere a monetary phenomenon”.
But as soon as inflation had been conquered, doubts about the usefulness of money as an intermediate target for the control of inflation began to emerge. The Bank of Canada led the sceptics, with its governor Gerald Bouey famously saying: “"We did not abandon [the monetary aggregate] M1, M1 abandoned us". His and other central bankers’ problem at the time was that financial deregulation had blurred the concept of money and hence obfuscated the relationship between money and inflation.
Three decades later, central bankers seem to have progressed in a dialectic way. They are again closer to Burns than to Volcker, though “on a higher level”. Money apparently no longer matters to them. The Fed has abandoned publication of the broad money aggregate M3 and the ECB has quietly buried its reference value for this measure. Now, they follow again “some form of incomes policy” as they aim to control inflation through wage growth, which they see as driven by changes in unemployment. Wage and price “controls” are now not as bluntly implemented as they were by Nixon and Burns. The central bankers of our time want to steer wage and price inflation by fine-tuning economic growth. But they do not seem to be more successful than their predecessors. Nixon and Burns could not get inflation down; our central bankers can’t get it up. Any form of incomes policy fails.
Why have we gone around in circles on the approach to monetary policy? In my view, the answer is that conventional economics misunderstands the relationship between money and inflation, because it has fallen victim to three fallacies about money.
First fallacy: Central banks control money supply
Economic textbooks claim that banks create book money by holding only a fraction of central bank money deposits in reserve and lending on the rest. Consequently, the central bank can control money supply by steering the volume of central bank money and adjusting bank reserve holding requirements. This is of course a fairy tale from the distant past. In our present, so-called “fiat-money” system, banks create money through credit extension: with the stroke of a pen the banker credits the account of the debtor with the amount lent. No deposits are required before the credit is extended. It is the credit that creates the deposit. Should banks need central bank money to fulfill cash demands or reserve requirements, they can borrow it from the central bank. The credit they have extended serves as collateral.
The influence of the central bank on money creation by the banks is highly indirect. Banks demand central bank reserves to make payments among each other and to the central bank in order to fulfill reserve requirements. The central bank creates central bank reserves against credit the banks have extended, which they now submit as collateral for their borrowing of central bank money. By setting the interest rate on central bank reserves that it lends to the banks, the central bank aims to influence banks’ own lending rates to the general public. These rates in turn affect credit demand. Thus, money is produced in what I call a private-public partnership (PPP): private banks produce money, with the central bank trying to influence the production activity. As we can learn from monetary history, the PPP of money production works well when banks are eager to lend and the central bank can restrain them. This was the case when Volcker tamed inflation. It fails to work, when the banks do not want to lend. This has been the case after the financial crisis, when higher risk aversion and tighter regulation has weakened banks’ appetite for lending.
Second fallacy: More money raises the price level
Economic textbooks also say that an increase of money supply leads to an increase of the price level, though the lags can be long and variable. This is the key statement of the Quantity Theory of Money, which was refined by Milton Friedman’s monetarist community. It is also another fairy tale from the past. Contrary to the economic textbooks, money created through credit extension leads first and foremost to a change in relative prices that may or may not raise a consumer price index designed by statisticians.
Here is why: The borrower, whose account is credited with the amount borrowed by the bank when he signs the credit contract, exchanges the money for whatever he planned to acquire when he decided to borrow. Hence, new money is created to acquire a specific object, and the price of this object increases when the new monetary demand meets the available supply of the object. If the object is a consumer good, consumer goods prices increase; if it is an investment good, capital goods prices increase; and if it is an existing asset of some type, prices of this type of asset increase.
In the recent past, new money has been primarily spent to buy existing assets. The chart (see Pour en savoir plus) shows the developments of consumer, producer and asset prices (as measure by the Flossbach von Storch Wealth Price Index) in Germany.  Clearly, monetary policy easing in and after the financial crisis of 2008/9 has failed to lift either consumer or producer prices materially, but it has boosted asset prices substantially. At the end of 2015, German asset price inflation was running at 7.8% on the year, the highest rate since the beginning of the series in 2003. Consumer price inflation amounted to 0.3%, close to its historical lows, and producer price inflation was -2.4%, the lowest level since 2010.
There are, of course, ripple effects through the economy from the purchase made with new money. More demand for consumer goods will increase demand for capital goods to produce consumer goods, and this may increase the demand for workers to produce capital goods. But the size and timing of the ripple effects are impossible to predict. They may be extremely indirect and weak, if new money is used only for the exchange of existing assets. In this case, the price of the asset facing more demand rises, but little else. Whether new money raises a price index for consumer goods within a certain time period is therefore highly uncertain. It may only raise asset prices, or it may drive asset and capital goods prices for a long time before it affects consumer goods prices. Only one thing is certain: it changes relative prices and therefore distorts resource allocation.
Third fallacy: Central banks can control inflation
Today, central banks pursue inflation targets over “the medium-term”, which often means on a three year horizon. But if they can’t control money supply, if there is no direct link between money and a consumer price index, and “some form of incomes policy” was revealed to be a flop, they are unlikely to achieve their ambition. To be true, consumer price inflation was benign through the decade dubbed the “Great Moderation” that lasted from the mid-1990s to the mid-2000s, but asset prices got out of hand. The collateral damage of inflation targeting were three asset market crashes, in 1998, 2000 and finally in 2007, when the fall-out from the credit crash caused the Great Recession.
Now, central banks seem unable to increase inflation as banks are reluctant to extend credit. Central banks boost book money by buying assets from non-banks (e.g., an insurance company) via commercial banks (they pay central bank money to banks, which create book money for the purchase of assets for the central bank). But the activity of non-banks seems little affected by this operation as they passively exchange one asset they hold against money, which they now hold as another asset.
For central banks inflation is good, when it is the result of economic growth and is moderate. Recent experience has shown that central banks cannot create “good inflation” at will. But they can always create “bad inflation”. Inflation becomes bad, when the general public loses trust in the purchasing power of money and rushes to exchange it against goods, real assets or anything else. Nobody can lose only a little trust in a controlled way. Hence, central banks can create bad inflation by debasing money, but they cannot control it.
What is to be done?
The advice is as simple as it is radical and without a chance of being heeded: End the private-public partnership of money production and stop targeting meaningless consumer price indices. Money should be issued as an asset (not as a liability as in the credit money system), and be backed by trust in the issuer. I call this “active money”. The issuer could be private (like the Bitcoin algorithm) or public (like a central bank issuing its money directly), and the backing immaterial (e.g., trust in the purchasing power of a crypto-currency) or material (e.g., collateralized with gold).
In Switzerland, a referendum on the abolition of the private-public partnership of money production will soon be held. The proponents of monetary reform want to give the Swiss National Bank the responsibility for the issuance of both cash and book money. The task of banks would be reduced to process non-cash payments, take deposits and lend these deposits on to investors. This would be a first step towards an active money regime.
Whatever the outcome of the Swiss referendum will be, policy makers elsewhere will not reform the existing monetary system voluntarily. They are key beneficiaries of a system that supplies them with cheap credit to fund public programs for their supporters. But reform could be forced on them. Most likely, central banks will continue with their ineffective incomes policy aimed at increasing inflation until the next recession hits us. When the economy begins to go down with an inflation rate of close to zero, commercial banks will be unable to keep lending. Instead, they will destroy money by calling in credit or going bankrupt. In response, central banks will issue “helicopter money” to replace defunct credit money. To this end, they may transfer central bank money to commercial banks with the instruction to create book money against it and to pay it into the accounts of their customers. This would mark the switch from credit money to “active” money.
Will people trust the new type of money? This will depend on how issuance is organized. The issuance of money with a view to fine-tuning economic growth and pursuing some target for a general price level is bound to fail, because there is no mechanical relationship between money issuance and these variables. Money issuers pursuing such objectives presume knowledge that they cannot have. What is needed is an inelastic supply of money based on a fixed algorithm (like Bitcoin, Milton Friedman’s k-percent rule of money expansion, or the production rate of gold mines). Ideally, users should be given a choice among different algorithms by having money issuers compete among each other for customers of their products.