In response to my recent reply to Henderson and Hummel’s defense of Alan Greenspan’s Fed, I received a number of e-mail messages. They included one asking me for my two-cents concerning a dispute my correspondent was having with a friend of his who insisted that monetarism, and Milton Friedman’s monetarism in particular, had been a complete failure.
Although I realize that Austrian-school economists have themselves been highly critical of monetarism, many of its most fundamental claims are in fact fully consistent with their own understanding of monetary theory. Indeed, back in the 1970s the difference between Austrian and monetarists writings about money seemed trivial compared to the difference between them and the writings of other (broadly “Keynesian”) economists. I recall very well how I myself got “deprogrammed” from mainstream thinking about money and inflation by reading Henry Hazlitt’s wonderful book, The Inflation Problem, and How to Resolve It. Hazlitt was, of course, a thoroughgoing Misesian. Yet no one who reads his book can fail to note the many crucial similarities between his arguments and those of Milton Friedman concerning the same subject.
Consequently I came to the defense of monetarism, and of Milton Friedman. Here is that defense, accompanied by the message that elicited it.
Professor Selgin,
I’m a relatively new reader of mises.org, and read your recent post about Greenspan.
I thought maybe you could help me settle an argument a friend and I are having about monetarism (as defined by Friedman in the 70s). My knowledge of it dates from about then, but my friend says that there were further developments in the 90s and earlier this decade when Friedman renounced it.
This is what my friend wrote:
So first, we can’t ignore the fact that the modern father of monetarism (Milton Friedman) acknowledged in an FT interview in 2003 that it doesn’t work. Importantly, Friedman knew of what he spoke because Israel, the Fed and the Bank of England all tried it in the late ’70s and early ’80s with disastrous results. That Volcker’s three-year flirtation with it caused the worst recession since the Great Depression is not even debatable.
Secondly, Friedman as early as ’84 had to admit how wrong he was (this was when the economy started growing again – the experiment ended in October of ’82, but not soon enough to stave off massive GOP losses in Congress thanks to the collapsing economy) because with the falling gold price suggesting a rising dollar in terms of value, Friedman went to everyone who would listen on Wall Street and said we were inflating. His reasoning was M1 figures, but as anyone who understands M1 knows, it’s when a currency is more useful and worth holding onto that people hold it. Friedman saw skyrocketing M1 figures as evidence that the Fed was creating too much money (despite the collapsing gold price), but in truth the signal was just the opposite. I can send you the quotes where Friedman admitted how incorrect he was when I get back to the office.
Beyond that, even if monetarism were useful, as in if the M signals told us anything worthwhile, there’s still no way the Fed can control the amount of dollars within these fifty states. That is so because dollars are used around the world, and there are markets for dollars around the world. Assuming the Fed sought to reduce the supply of dollars here, money would simply flow here from somewhere else. Indeed, that’s what happened in the US and England when the dunces at the Fed actually believed Friedman’s theory might work.
So unless you can tell me how the Fed can control money supply here despite overwhelming evidence and logic showing it to be an impossibility, there’s not much I can do. You can send me critical e-mails, but the ship on this subject has sailed. If even Friedman finally admitted publicly that it didn’t work (in private, he admitted it long before ’03), isn’t it time for the holdouts to give in?
So, is my friend right? Is monetarism dead?
I’m happy to share some of my reactions to your friend’s statements about Friedman’s monetarism, for whatever they may be worth.
I should begin, though, by noting that “monetarism” has a broader meaning than the one your friend appears to assign to it. To refer to Wikipedia’s definition, which I think is consistent with most economists’ understanding, including my own, monetarism as formulated by Friedman “argues that excessive expansion of the money supply is inherently inflationary, and that monetary authorities should focus solely on maintaining price stability.” That “excessive” begs the question, but the point is that monetary authorities are in principle always capable of avoiding inflation through the exercise of sufficient monetary restraint, that is, by limiting the availability of currency and bank reserves, which they alone have the power to create or destroy.
I should think that, far from being considered “dead,” this perspective appears today perfectly banal.
Certainly it is as least implicitly accepted by all central banks that practice inflation targeting (which includes those committed to Taylor rules and like formulas that include an inflation-target component). In accepting responsibility for targeting inflation, central banks admit their ability to do so in principle by means of the instruments normally at their command, which mainly consist of devices for either expanding the stock of currency and bank reserves or altering the value of broad monetary assets capable of being supported by any given quantity of bank reserves.
But back in the 70s, monetarism understood this way was anything but banal. I don’t know whether your friend was around then, or old enough to have been aware of either the economic circumstances or the state of economic theory at the time. But whether he personally experienced it or not he is presumably aware of the rampant, very destructive, and accelerating inflation that had the entire world in its grip by the end of that decade. The world inflation rate then approached 16%. England’s rate peaked at 25%. The U.S. rate was lower–just shy of 13%–but even at that rate the price level would double every 5.5 years or so. Israel’s inflation rate, finally, was 132%! Up to that time central bankers, encouraged by some really godawful economists, insisted on treating inflation as if it were some sort of plague from without, that they were struggling valiantly to combat; while politicians were blaming it on consumers, who were urged to stop spending so much (remember those WIN buttons?), and on union leaders, who were scolded for demanding higher wages for their members. In a famous publicity ploy Fed Chairman Arthur Burns posed with a baseball bat, as if he were set to knock inflation out of the economy.
Of course, Burns didn’t do any such thing, and for a reason that should now seem obvious to all, namely, that that bat couldn’t possibly do any good unless someone used it to knock some economic sense into Burns and other central bankers like him, who were busy all along churning out vast quantities of money. That, essentially, is what Friedman’s arguments for monetarism did, albeit not by actually educating clueless central bankers like Burns, but by helping to see to it that others with a better grasp of the cause of rising prices took their place.
Your friend, rather perversely I think, refers to the “disastrous results” that followed the monetary tightening of Volcker and other more monetarist-minded central bankers without even hinting at the facts that that the tightening was aimed at bringing down inflation rates, and that it succeeded remarkably well in doing precisely that. In other words, the tightening did precisely what monetarism said it would do, and what monetarists’ critics at the time, wedded to the view that monetary policy was ineffective, and that inflation was entirely caused by OPEC (or by unions, or by anything except monetary policy) insisted it could not do. And yet your friend imagines that the experience proved the monetarists wrong! (In one respect, however, I agree that monetarism was wrong, namely, to the extent that it failed to recognize the need to allow inflation to occur to the extent that it was _solely_ due to falling output or productivity–a case I have argued, along with the symmetrical one for letting prices fall as productivity improves, in my IEA pamphlet _Less Than Zero_.)
It is indeed true that the reduction of inflation came at a big price–that Volcker’s tightening, for instance, succeeded in lowering inflation only in the wake of a severe recession. But this possibility was, first of all, one concerning which monetarists were perfectly aware: they understood the danger that, once inflation, and accelerating inflation especially, came to be anticipated by the public, putting the breaks on money growth would result in prices and wages continuing for some time to rise beyond their new, less-rapidly rising equilibrium values, with a consequent rise in unemployment. That is the tragic situation an economy finds itself in when it has allowed inflation to go on for an extended period of time. No monetarist, certainly not Friedman, welcomed this side-effect of monetary restraint. But then again, had Friedman been listened to earlier (and had the likes of Paul Samuelson been ignored), things would never have come to this tragic stage. By way of analogy, should we blame those who would end the war in Iraq, and who opposed it all along, for the tragic consequences that are likely to follow any rapid U.S withdrawal?
So much for monetarism in the broad sense of the term. It also has a narrower meaning, referring to the practice of implementing monetary policy by targeting a monetary aggregate rather than an interest rate. Volcker’s anti-inflation campaign at first involved a turn to monetarism in this as well as the broader sense of the term; and it was only monetarism in the narrow sense of money stock targeting that Volcker eventually abandoned, and that Friedman himself admitted (around the same time) to be unworkable. The failure of monetary targeting was due to the fact that the demand for and velocity of money, as traditionally defined, which had once appeared to be a stable function of a small number of variables, had ceased to be so during the course of the 70s and early 80s, when high inflation led to numerous financial innovations, including the emergence of new financial assets that competed with and ultimately reduced the demand for traditional forms of money. Here again, it is perfectly possible that money targeting might have continued to work well had it been practiced all along, and in a manner such as would have avoided the outbreak of severe inflation in the first place. But in any event Friedman can hardly be blamed for having failed to anticipate financial innovations that no one else anticipated, either; and it is to his credit that he rejected his own earlier practical prescriptions for implementing (broad) monetarist ideas in favor of new ones arrived at in light of new empirical evidence.
Your friend’s remarks include some statements of a more theoretical nature that seem to me incorrect. Thus he observes that “as anyone who understands M1 knows, it’s when a currency is more useful and worth holding onto that people hold it. Friedman saw skyrocketing M1 figures as evidence that the Fed was creating too much money (despite the collapsing gold price), but in truth the signal was just the opposite.” But while it is correct to say that the real quantity of currency (or any other monetary measure) in existence depends ultimately on the public’s real demand for currency (or whatever), and that the Fed is therefore unable to control such real quantities, the growth rate of the nominal quantity of money in any of its forms is ultimately independent of real demand and dependent on Fed policy. Currency is not an exception, despite the fact that the Fed only issues it in response to special demands from the banks to convert reserve credits into cash, because those demands themselves are conditioned on the overall growth of nominal money, with consumers seeking to retain a certain proportion of their broader money holdings in the form of currency. In short, if the Fed wants to make the “demand” for currency grow, it is perfectly capable of doing so by pursuing an easy-money policy.
By the way, your friend’s opposite claim is precisely the defense lodged by hyper-inflating central bankers throughout modern history, including the administrators of the Reischbank in Weimer Germany and of the Reserve Bank of Zimbabwe lately. “We are only trying to keep up with the demand for our notes!” Yes, any monetary economist worth his or her salt would reply, but the demand for currency is itself rising on account of rising prices!
As for the related suggestion that conventional measures like M1 do not reveal anything worthwhile for the conduct of monetary policy, here is a careful study of the matter that says otherwise:
http://www.richmondfed.org/publications/research/economic_quarterly/2000/fall/pdf/dotsey.pdf
Finally, as for the claim that the Fed can’t possibly control the quantity of dollars in the U.S. (and, by implication, U.S. inflation) because substantial quantities of U.S. dollars are held abroad, it is rather beside the point, for what matters in determining the value of the dollar is in fact the world supply and demand for dollars. (Imagine someone claiming, concerning a hypothetical world oil monopoly–say, OPEC in a world without any rival oil producers–is unable to influence the price of oil by restricting output because doing so wouldn’t result in any substantial change in oil consumption in the OPEC nations themselves.)
Well, I doubt that any of this will persuade your friend to become a monetarist, but hope it will at least convince him not to be quite so inclined to dismiss it.
Sincerely,
George Selgin
{ 57 comments }
← Previous Comments
Anyone:
“Why make a movie about something one understands completely? I make movies about things I do not understand, but wish to.” (Seijun Suzuki.)
Some thoughts on money. The context is a FREE market, entirely free, no central bank, no legal tender laws, anyone can issue promises to pay in any form whatever. There is enforcement against crimes.
Almost anything can serve as a medium of exchange. We could price everything in terms of Hershey candy bars and exchange them. Gold can serve. Silver can serve. Copper can serve. They all can serve at the same time. In the American colonies, there were sometimes 3 or 4 different coin prices for an item, depending on what the exchange medium was. Plus they used wampum, paper, furs, and hides.
Using these items is not essentially different from bartering. They are divisible, portable, and of reasonably known value as they carry their backing in the item itself. Money is really barter of a more convenient form. Attempting to define a money supply and controlling it both seem like fruitless endeavors.
A promise to pay can be money too. The value of it depends on fulfillment. A barber can pay the grocer with a ticket good for a haircut. This is like barter too, but a more abstract form of it. A bank’s note has to be a promise to pay something or else no one will accept it.
I have been told that Wal-Mart gift certificates are being used as money. Being paid a bonus in these avoids taxation. They are accepted because of Wal-Mart’s promise to redeem them. And people believe the promise will be fulfilled. Wal-Mart has the money paid for these gift certificates, just as American Express has the money paid for travelers’ checks. They don’t segregate it. They use the float. They know that not all the certificates or checks will be used or presented at one time. They’re good for the promise, but there is a chance of temporary suspension.
A market I know offered free gifts each week by a coupon. It sometimes ran out of these gifts because it didn’t know how many people might use the coupons. Then, to maintain credibility, it had to pay them the following week. It suspended payment and then restored the next week.
A bank cannot get its notes accepted without people believing it has backing. But there has to be convertibility to act as an assurance that the backing is genuine.
People do not always operate on there being perfect assurances. They use all sorts of methods to measure the validity of a promise or an assurance. They will pay more for an assurance that is more certain, and less for one that is less certain. Most debts can default, and the pricing of them depends on both the chances of default and how much can be recovered when there is a default. Meanwhile the debts circulate and are traded.
People often use credit cards. They buy something with the promise to pay, even though they do not have the money at the time they buy. Companies do the same when they issue debt. The creditors believe, for various reasons, that the debts will be paid. But they also price in the risk that they will not be paid. They accept the risk because they are compensated for it in higher interest.
A bank cannot issue notes convertible into gold and have anyone accept them unless it compensates people for the risk of non-conversion. If it gave them nothing in return, they may as well hold the gold themselves. Gold would dominate holding the note.
The matter depends on the costs of holding and using gold versus the costs of holding and using a bank account. Using gold coin and silver coin has its drawbacks and its advantages. Coins get worn, clipped, stolen, and lost. The user has to verify their goodness. Storing them in a 100% reserve location is such that you pay for storage. You also pay for transfers. Adam Smith describes this in detail in his account of the Bank of Amsterdam.
A bank in a free market perhaps offers clearing and checking services. Sometimes it pays explicit interest. Its paper money is easier to carry. It can be counterfeited, but may be easier to check than gold. The customers weigh the pros and cons of using paper money. The bank promises to redeem on demand, but there is a chance it will suspend. There is a chance of outright loss. All these things get priced in when people decide to use the bank account or not. I am speaking of a FREE market here, not the existing situation. It’s important not to confuse them.
There could be an entirely new kind of bank arise in a free market. This would be a bank that united all one’s accounts: bank, brokerage, mutual funds, pension holdings, money market funds. You would have all your securities in this account. The value would vary daily. When you bought something, the bank would do the clearing. It would liquidate some of your money market fund (or if you had none some other holding per your instruction) and debit your account while crediting someone else’s account. Transfers would be by debits and credits. Defining money in this system would be well-nigh impossible. There would be no authority to create money.
You could have overdraft privileges, and that would be a loan to you. Where or who from? Who are the creditors? Other persons could indicate that they are willing to make loans of certain types. They do this by buying shares in a mutual fund that offers these kinds of loans. You could buy shares in a fund that offers automobile loans. It is more than a fund, as this takes management. It is really a finance company. By buying, you are supplying loans to those who want them. The bank is now a clearing house for these brokerage transactions that are capital market transactions. So far, everything is 100% reserved.
A mutual fund can start up with the following terms. You buy shares in it. The fund then is enabled to lend your money to others. Furthermore, it can issue more shares to entrepreneurs that it thinks will succeed at their businesses. The entrepreneurs agree to repay the shares and loans with interest. This fund is a fractional reserve bank.
frb means that ultimately even prudent bankers must take increasingly riskier loans (in an absolute sense) to avoid being driven out of the credit business in a competitive race to the bottom.
and yes, fully reserved banks could still go broke for all the usual business reasons, ineptitude, fraud, and so forth. but there is no structural flaw that makes them susceptible to a run. as mike sproul has pointed out, the bank of amsterdam only had a brief period as a fully reserved bank, before its directors breached its charter and allowed certain accounts to be overdrawn.
i don’t have any problem with the mutual fund/ hedge fund set-up you’ve described. my problem is essentially semantic. “bank deposit” means just that, a non-interest bearing, fee-accruing account. and a “bank” is the institution that provides such a safekeeping service.
it’s only fair to apprise the public of the nature of the institution (through common language), many would be wary of keeping ready money in a mutual fund or hedge fund, but tolerate bank accounts because the risks are obscured (or socialized, as seems the case today).
i believe bank and deposit have common meanings much the same way that bleach and cornflour don’t need to be explained because they are understood in common parlance.
i like to think of the quantity theory as like a rubber band linking the supply of money and its value. there’s a lot of stretch and no predictable short-term relationship. the decay may be retarded and then a spectacularly rapid catch-up may occur, probably overshooting. but over longer periods the correlation becomes more reliable.
“frb means that ultimately even prudent bankers must take increasingly riskier loans (in an absolute sense) to avoid being driven out of the credit business in a competitive race to the bottom.”
Why does a rational prudent banker do this if it ultimately drives him out of business? Why not let the other guy go out of business and then expand and pick up the pieces? In this respect, banking is no different from any other industry. The good firms do not necessarily imitate what the bad firms are doing. There are insurance companies that underwrite bad risks, but the good ones do not do this simply to get ultimately unprofitable business.
I keep trying in my mind to model how fractional reserve banking works or does not work, and I think at the moment that the notes follow a non-linear valuation pattern as the bank adds loans while holding constant its reserves. If it adds reserves as it adds loans, then it can counteract the declining portion of the curve.
The value of a note first rises as the bank provides the services of clearing and checking and paper money. To pay for these, it lends some of the deposits and maintains the rest as reserves. The lending provides the return that flows through to the depositors. It also provides some profit. The convertability enhances note value and helps keep it at par. If the bank’s loans are collateralized, that helps too. If they are callable, that helps. If the called assets are liquid, that helps. As the bank adds loans without adding reserves, the note value will reach a maximum (where the marginal value of the additional loan income equals the marginal cost in terms of the chance of having to suspend conversion.) At that point, the note value starts to decline. To avoid that, the bank has to build up reserves. They enhance the value of the conversion, or lower the marginal cost of suspension. The bank might pay higher interest on its deposits, and that too might enhance note value.
The ownership and control structure of the bank will make a difference. So do the compensation methods. Many banks have quite a bit of inside stock ownership. This gives the management a longer-run perspective. Otherwise, they may try to maximize their short-run salaries and bonuses, especially if they are compensated by incorrect metrics such as loan growth. If they have too much inside ownership, they will tend to consume more perquisites, but they also may be more conservative. The mutual form of ownership may be better than stock ownership. It worked better for S & Ls. There were more failures in stock-owned S & Ls than in mutuals.
“i like to think of the quantity theory as like a rubber band linking the supply of money and its value.”
In the free banking world that is possible, wherein you have one account for your assets, a wealth account, and you pay people by liquidation, by debits and credits, what’s the meaning of the supply of money?
If anyone can issue promises and these can be used as means of payment, what’s the meaning of the supply of money?
to michael rozeff:
touche! you’re right that in an ideal world, free of state interference, rash bankers would pay for their imprudence by being gobbled up by their more careful competitors.
sadly, history has shown that governments can never resist the easy spoils of money debauchment, and so banking is always awash with moral hazard.
i would hope that at some stage in the not-so-distant-future, that the present monetary disorder becomes so chronic that there emerges a parallel money (gold/silver/platinum?). naturally the state will resist this mortal threat to its last breath, and so this means that it is likely to remain a specie- only money. (ie gold contracts, electronic gold deposits etc. are not going to be welcomed).
this may have the positive effect of re-accustoming people to dealing in physical money. if eventually the state decides to incorporate gold into the official payments structure, then at least people will have had plenty of experience of dealing in tangible money, and perhaps may approach paper money with a good deal more circumspection than is the case today.
money supply in this situation would be merely the total number of oz of easily recoverable precious metals in the country. i’m assuming that deposit accounts are are treated legally as bailments, not as loans.
Correct me if I’m wrong but
Gov’t ask Fed for xDollars
Gov’t gives Fed xTreasuryBonds for xDollars
Gov’t promises to repay xDollars plus Interest
Gov’t owes xDollars+Interest
Interest doesn’t exist
Monetarism is a paradox as only the principal exists in the money supply, so if the principal+interest is owed, where does one find credit to pay the interest?
If every debt in The United States was settled, there would be no money, correct?
How can anyone see this as healthy?
Thanks for the interesting information that ipicked up!
← Previous Comments
Comments on this entry are closed.