Let’s begin with a comment anarcholibertarian made about my article Four Valuable Lessons From Actually reading the Regression Theorem.
Our buddy David Kramer made a comment about this article which I am also wondering about. He says:
“’4. Fiat money has a yesterday, and thus does not violate the regression theorem.’
If by ‘fiat’ he means the paper FRNs [= Federal reserve Notes] which forcefully/fraudulently replaced U.S. Gold and Silver certificates (which, in and of themselves, REPRESENTED something of value and were not the valued commodity itself), he’s wrong.”
Could it be that government force is the only reason why the dollar is in general use? What would happen if the government told everyone to use a random token or they will be shot? I’m thinking that it will be volatile in price, since no one knows how much 1 token is worth compared to 100 tokens, but it would still be in general use out of fear. I’m still reading through your bitcoin articles, so if you have already answered this, just point me in the right direction if you please.
Good stuff. Here’s my reply:
It’s only lately that I’ve come to understand some deep stuff about fiat money. So here is what I know.
A. Kramer’s argument is not with me, but with Mises, whom I quote extensively in the beginning of the article. A slow careful reading shows that Mises is talking exactly about FRN’s when he writes:
The other possible case is that in which coins that once circulated as commodity money are transformed into fiat money by cessation of free coinage (either because there was no further minting at all or because minting was continued only on behalf of the Treasury), no obligation of conversion being de jure or de facto assumed by anybody, and nobody having any grounds for hoping that such an obligation ever would be assumed by anybody. Here the starting point for the valuation lies in the objective exchange value of the coins at the time of the cessation of free coinage.
The situation he describes is that of a country whose money is coins. One day, the govt says it will no longer make any coins, only fiat [=intrinsically worthless] money, not redeemable for anything. And he goes on to say that the reg. thm is satisfied.
That’s my appeal to authority. And the logic behind it is impeccable, as well. What’s the basic question the reg thm asks? In a word, it’s “How does a person know the purchasing power of this new money?” How does he know how many apples the grocer will give him, how many books amazon.com will give him, etc etc etc?
Mises has a simple answer. For commodity money, he knows he will get at least its value as a commodity, the commodity price. For a newly introduced fiat money, he assumes it will get him the same stuff that the old commodity money of the same name used to buy. A silver dollar bought me a gallon of gas? Then this new paper dollar will probably get me the same thing. A dollar is a dollar. I imagine that historically, this has been the case as well.
In short, the reg thm asks, “What number?” and for fiat money, even FRNs, we have an answer. Same number as the old money.
B. This is what I knew at first. Then I found Mises writing in Money and Credit that when a govt in the old days tried to debase the coinage by making the coins only partially gold, their purchasing power instantly dropped to be worth only the amount of gold they actually still contained. Which led me to ask myself, “Isn’t going from a silver dollar to a paper dollar the ultimate debasing of the coin? I mean, instead of making only 90% silver, the way the kings in the old days debased the coinage, they are making a new “coin” with zero percent silver.”
How does that not contradict what Mises wrote in Money and Credit, the very same book, about fiat money being just fine and satisfying the reg. thm?
My answer is that yes, fiat money satisfies the theoretical problem posed by the regression theorem, for the reason above, but it has another, different, more serious problem. People will drop it like a hot potato as soon as they get a chance. “Yes, there is a number that can be assigned to it, but who wants this garbage, if I have a choice?” I wrote about this at length, quoting Mises and explaining his arguments, in my humble article, The Kickstart Fallacy.
Maybe Kramer had this other problem in mind when he said I was wrong about FRNs.
C. So why is fiat money accepted at all? I think the only possible answer is because of some kind of violence, or threat of violence. Why did that not work for the old kings, but did work for FDR and for all modern countries today?
I’m not sure. Here are some guesses.
Peter Schiff says that modern fiat currency has a twisted kind of “intrinsic value” [= value besides trading it for goods and services], in that you need some of it on hand to pay your taxes. Maybe in the old days you could still pay in geese and chickens if you had no cash, and that’s why the old kings failed.
Maybe in the old days there were no legal tender laws. In the Kickstart Fallacy article I appealed to Thier’s Law at one point, which distinguishes between a country with legal tender laws and one without.
Both those theories will need a lot of historical research to see if they are consistent with the facts, research I am totally unfit to do.
A final guess is that maybe it takes some time, maybe depending on the size of the country or some other factor, for the debasement to take place. In other words, we know the dollar and all fiat currencies are constantly losing value. Usually this is attributed to inflation of the money supply, and of course that is correct. But maybe there is another force at work, that exists even with a constant money supply, mainly the Law of Debasement, we may call it, that says money will eventually sink to the purchasing power it has as a commodity.
D. For the repercussions of all this to bitcoin, I refer you to the Kickstart Fallacy article.
For most of my stuff on bitcoin, see https://smilingdavesblog.wordpress.com/2012/08/03/bitcoin-all-in-one-place/
Smiling Dave on Austrian Economics. 
Thanks Smiling Dave. I’m flattered that you responded with a full post. It’s very thought-provoking.
Here are David Kramer’s thoughts on it:
Interesting, but I think Mises is wrong too. To claim that a worthless piece of paper (U.S. gold certificate) that represented a real commodity worth $20 is now de-coupled from that and fraudulently replaced with the worthless piece of paper Federal Reserve Note is now “worth” $20 not because of the market value of the paper commodity but because someone scammed you into believing that the paper itself was “worth” that because someone with a gun printed $20″ on it, is NOT the Regression Theorem.
If it “is,” then why can’t we all just print up FRNs and do that? We can all certainly go out and mine gold and silver (but Bernard von NotHaus will tell you that we can’t legally mint coins with them). Because the FRNs are a forced medium of exchange that are worthless as a commodity (i.e., the paper they are printed on) in and of themselves and the only way the government and the banksters get away with it is through 1) guns 2) the sheeple are too ignorant to understand the scam.
LikeLike
We have to separate the economist within us from the libertarian within. The economist tells us what is, the libertarian what should be.
The regression theorem does not ask what is right or wrong, but rather a totally different question, to wit, people only accept money as payment because they know what they can buy with that money. But how do they know? The answer is because of existing prices out there. But how did those prices know what to charge? Because of earlier pre-existing prices. And how did they know? From even earlier prices. We keep going back in time, but human history does not extend back to infinity. Thus all money had to have an initial value of some kind. Where did that initial valuation come from?
Clearly, this question will be answered if we explain where the initial value came from, whether it was forces of good or evil or somewhere in the middle that made it have that initial value. In other words, the subject of the regression theorem is a technical question that requires a technical answer. And if the answer says that guns and sheeple were involved, that does not invalidate the answer.
Mr Kramer’s observation, that FRNs are a forced medium of exchange, is certainly correct. And indeed I discussed the economic implications of that fact in the second part of the article.
Now we may ask, does the fact that FRNs are forced make them economically different from bitcoin? True, they both lack intrinsic value, but maybe FRNs are doomed to die if and when the force goes away, but bitcoins, which are voluntary, can survive. My thesis is that there is no difference in that respect. The fact that FRNs were imposed by force does not taint them forever when the force goes away. All that happens is that they revert to their free market value. So I see no diff. between bitcoins and FRNs. And the voluntary acceptance of bitcoin has not happened, as I explained at length in my many articles [search for “wide demand’].
This may be the place to state that I try to keep my humble blog limited to the economic side of things, letting the reader draw his own conclusion about right and wrong.
LikeLike
“I mean, instead of making only 90% silver, the way the kings in the old days debased the coinage, they are making a new “coin” with zero percent silver.”
“How does that not contradict what Mises wrote in Money and Credit, the very same book, about fiat money being just fine and satisfying the reg. thm?”
It does, which is why, as people try to use paper money as if it represents the same value as the commodity on which it was once based, distortions occur in terms of real goods.
It’s functionally equivalent to Mises’ response to the Socialists, according to Joseph Salerno, with regard to the idea that Socialism would work if you kept the same prices, perpetually, as the day it replaced Capitalism.
He said that because preferences changed all the time, the old prices would become increasingly misrepresentative of underlying preferences.
In Salerno’s quote of Mises, he’s referring to the price staying the same while preferences changed.
But a distortion can occur if preferences stayed the same and prices changed due to the price inflation caused by monetary inflation.
Either scenario would result in the wrong price, because real [albeit subjective] value is set by consumer preferences.
[Time stamped]
Calculation and Socialism | Joseph T. Salerno
“C. So why is fiat money accepted at all? I think the only possible answer is because of some kind of violence, or threat of violence. Why did that not work for the old kings, but did work for FDR and for all modern countries today?
“I’m not sure.”
The only reason it seems to work for the USD is because we’re currently able to export our inflation around the world because it’s being used by other countries as their reserve currency.
In effect, we’re spreading the losses that would normally accrue to Americans over the citizens of other countries.
But that, too, will end, for the same reason it ended for the old kings.
LikeLike
Thank you for commenting, Guest.
You seem to assume that introduction of fiat money creates problems because the money supply is inflated, which is of course true. However, it is possible, in theory, to introduce fiat money with no inflation of the money supply. Simply do as FDR did and scoop up all the gold coins, and replace them by paper money, and then not print more. Thus inflation is non existent. The article, and the portions of Mises’s book, Money and Credit, quoted in it, is [implicitly] discussing such a scenario. In pther words, the discussion is about the replacement of coins by paper, without regard to inflation.
LikeLike
“However, it is possible, in theory, to introduce fiat money with no inflation of the money supply. Simply do as FDR did and scoop up all the gold coins, and replace them by paper money, and then not print more. Thus inflation is non existent.”
That would presume that “money” is just a measure of value, such that any token at all could be money; In which case, you wouldn’t need a regression theorem to explain the value of money, since you could just arbitrarily assign a value.
(That would also be an implicit concession to the Keynesians that what is actually desired in money is stable prices – that wildly fluctuating prices are necessarily a bad thing. But, while wildly fluctuating prices are disruptive for individual planning, as long as the prices reflect consumer preferences, you actually do want those prices to fluctuate wildly because those would be the correct prices.)
But either Mises or Rothbard (forgot which, but probably both, anyway) rejected the idea that money is a measure of value.
The reason is because money’s trade value derives from its use-value – the same as all economic value.
Further, if you say that you can get rid of all the gold and just replace it with paper, and there’d be no inflation, then you’d have to say that you can do the same thing with bitcoins.
I disagree with that because of the reasons above, but that’s what you’d have to concede.
LikeLike
“That would presume that “money” is just a measure of value, such that any token at all could be money;”
I am going with the definition of money being the generally accepted medium of exchange, following both Mises and Rothbard. By this definition “money” is that which almost everyone, over 99%, will accept in payment for their wares.
Sometimes they are forced to accept it. If FDR pulled that switch and threatened death to anyone who refused to accept the paper money, the article talks about when his plan would work and when it would not.
I don’t see why the definition of money as measure of value has to follow from the above. Perhaps you could elaborate, preferably with a syllogism.
“In which case, you wouldn’t need a regression theorem to explain the value of money, since you could just arbitrarily assign a value.”
You could arbitrarily assign a value, true, meaning you could write on the paper that the paper is equal to one gold coin, or two gold coins, or whatever you wish. But the problem the regression theorem discusses would still apply, because the question would be how does a person know what people will be willing to sell for such a piece of paper. Mises wrote that they would, at first at least, sell in exchange for a paper that says “One Gold Coin” whatever they would have sold yesterday in exchange for a true actual coin. My article entitled The Kickstart Fallacy talks about whether there are other problems besides the regression theorem that FDR would have to face.
“The reason is because money’s trade value derives from its use-value – the same as all economic value.”
The reason for what?
“Further, if you say that you can get rid of all the gold and just replace it with paper, and there’d be no inflation, then you’d have to say that you can do the same thing with bitcoins.”
Yes, if FDR took all the gold and handed out bitcoins instead, there would indeed be no inflation. And if he made a law declaring that one bitcoin equals one dollar, the regression theorem would be satisfied. But there is no FDR telling us how much a bitcoin is worth. Which is exactly why there is a problem of the regression theorem with bitcoin.
I think one thing has to be emphasized. The regression theorem does not say that bitcoins will cause inflation, or will be killed by inflation. It just says that the bitcoins themselves will never become the generally accepted medium of exchange, because of a certain technical problem. [See my article Bitcoin Takes a Beating, which explains what the problem is]. The problem with bitcoins has nothing to due with inflation.
Thank you for taking an interest in my humble articles
LikeLike
Sorry if taking these out of order is confusing, but the following order seems more efficient for my purposes.
“Thank you for taking an interest in my humble articles”
Your articles are awesome! I appreciate you and them.
“And if he made a law declaring that one bitcoin equals one dollar, the regression theorem would be satisfied.”
But Mises said, in Human Action:
“It [The Regression Theorem] deduces a more special case from the rules of a more universal theory. It shows how the special phenomenon necessarily emerges out of the operation of the rules generally valid for all phenomena. It does not say: This happened at that time and at that place. It says: This always happens when the conditions appear; whenever a good which has not been demanded previously for the employment as a medium of exchange, begins to be demanded for this employment, the same effects must appear again; no good can be employed for the function of a medium of exchange which at the very beginning of its use for this purpose did not have exchange value on account of other employments. And all these statements implied in the regression theorem are enounced apodictically as implied in the apriorism of praxeology. It must happen this way. Nobody can ever succeed in constructing a hypothetical case in which things were to occur in a different way.”
Not that Mises can’t ever be wrong, but he’s right about this, because all value is subjective and comes from the individual consumer and is imputed back up the structure of production (Menger’s Theory of Imputation).
To say that FDR could impose a value onto bitcoins is to deny Subjective Value theory, and is to grant the Marxist’s premise that value can be added (again, if all value comes from the individual consumer, then none of it can be imposed; it must already exist on the individuals’ preference scale).
“I don’t see why the definition of money as measure of value has to follow from the above. Perhaps you could elaborate, preferably with a syllogism.”
Because it presumes that its value in trade can derive from some source other than the consumers’ own preference scale, which would be a denial of Subjective Value theory.
That’s why all money is a commodity, and to the extent people try to use non-commodities as money, someone has to be ripped off in terms of real goods and services.
“The reason for what?”
The reason that money can’t be a measure of value. Nothing has “intrinsic” value, in itself. The value of all things derive from its use in fulfilling someone’s ends.
So, money never guarantees a certain purchasing power in terms of real goods. Rather, someone values it enough as a commodity that they are willing to pay someone enough of something else that it becomes cost efficient to save in that commodity.
That’s when it becomes money.
So, goods already have value, and they *become* money. At best, FRNs and bitcoins could be money substitutes, but if you can’t get the stuff they supposedly represent (say gold), then it’s just a worthless paper or set of digits.
(Not to mention the flaw in thinking of digits as “things”.)
The value of gold cannot, therefore, simply be poured into paper by fiat.
The only reason we still use paper, today, is because others have to be ripped off (through price inflation), in order for the rest of us to get anything for them.
“Yes, if FDR took all the gold and handed out bitcoins instead, there would indeed be no inflation.”
But that would be the ultimate inflation – pretending that something is worth another thing that doesn’t even exist.
The real value of gold, in that scenario would be zero, since there isn’t any.
LikeLike
1. Mises wrote in Money and Credit that fiat money, such as FRNs, satisfy the regression theorem. I quote him at length and explain in simple language here: https://smilingdavesblog.wordpress.com/2013/04/16/four-valuable-lessons-from-actually-reading-the-regression-theorem
2. By the usual definitions of inflation [increase in money supply or increase in prices], what FDR does in the hypothetical we are discussing is not inflation. Reprehensible, yes. Theft, yes. But not inflation.
LikeLike
“Mises wrote in Money and Credit that fiat money, such as FRNs, satisfy the regression theorem.”
From your article:
“But what about fiat money. It is also intrinsically useless, no? And Mises gives us the answer. At least it has a yesterday! Bitcoin does not.”
All that’s required for some good to have a yesterday’s price is to have one made up.
As soon as someone buys it at that made-up price, for whatever reason, it now has a yesterday’s price.
(I’m not denying that people *do* exchange FRNs based on their perception that it has a value that’s based on silver; Rather, I’m saying that because the perception is false, their valuation is false, and therefore FRNs aren’t money.
(It is not sufficient that people trade something *as if* it were money, for something to be money; It must also serve the function of money, which is to enable valuation of the goods it will buy relative to an individual’s subjective preference rankings.
(Which is why money must always have a link to use-value.)
You have, in the past, correctly noted that there’s no reason to suppose that a made-up price would last, and is therefore not a reliable basis for economic calculation.
But the same is true of FRNs, since the reason the silver, upon which FRNs used to be based, had monetary value was because of its use-value.
Take away the supply of silver, and silver can’t be used as much, anymore – and therefore the paper substitutes for silver also lose their basis for monetary value.
Also from Human Action:
“A money-substitute can be embodied either in a banknote or in a demand deposit with a bank subject to check (“checkbook money” or deposit currency), provided the bank is prepared to exchange the note or the deposit daily free of charge against money proper. Token coins are also money-substitutes, provided the owner is in a position to exchange them at need against money free of expense and without delay. To achieve this it is not required that the government be bound by law to redeem them. What counts is the fact that these tokens can be really converted free of expense and without delay. If the total amount of token coins issued is kept within reasonable limits, no special provisions on the part of the government are necessary to keep their exchange value at par with their face value. The demand of the public for small change gives everybody the opportunity to exchange them easily against pieces of money. The main thing is that every owner of a money-substitute is perfectly certain that it can, at every instant and free of expense, be exchanged against money. …”
“… The issuing of money-certificates does not increase the quantity of things suitable to satisfy the demand for money for cash holding. Changes in the quantity of money-certificates therefore do not alter the supply of money and the money relation. They do not play any role in the determination of the purchasing power of money.
“If the money reserve kept by the debtor against the money-substitutes issued is less than the total amount of such substitutes, we call that amount of substitutes which exceeds the reserve fiduciary media. …”
“… The issue of money-certificates does not increase the funds which the bank can employ in the conduct of its lending business. A bank which does not issue fiduciary media can only grant commodity credit, i.e., it can only lend its own funds and the amount of money which its customers have entrusted to it. The issue of fiduciary media enlarges the bank’s funds available for lending beyond these limits. It can now not only grant commodity credit, but also circulation credit, i.e., credit granted out of the issue of fiduciary media.”
“While the quantity of money-certificates is indifferent, the quantity of fiduciary media is not. The fiduciary media affect the market phenomena in the same way as money does. Changes in their quantity influence the determination of money’s purchasing power and of prices and—temporarily—also of the rate of interest.
“Earlier economists applied a different terminology. Many were prepared to call the money-substitutes simply money, as they are fit to render the services money renders. However, this terminology is not expedient. The first purpose of a scientific terminology is to facilitate the analysis of the problems involved. The task of the catallactic theory of money—as differentiated from the legal theory and from the technical disciplines of bank management and accountancy —is the study of the problems of the determination of prices and interest rates. This task requires a sharp distinction between moneycertificates and fiduciary media.
“The term credit expansion has often been misinterpreted. It is important to realize that commodity credit cannot be expanded. The only vehicle of credit expansion is circulation credit.”
What to take away from that unfortunately long quote:
– “Money-substitutes” are only substitutes if they can be redeemed in “money proper” – meaning that money-substitutes are *not* money.
That would include FRNs or bitcoins (if somehow bitcoins could be a substitute for a single good).
– Since commodity credit cannot be expanded, neither can commodity money.
It’s not the mere increase – or even *sudden* and/or great increase – of the money supply that causes the boom-bust cycle.
Rather, it’s the creation of what Mises calls “fiduciary media” – notes in excess of specie – that causes it.
(And taking away the gold and replacing them with FRNs would make all FRNs fiduciary media and, at any rate, not at all money.)
Why? Because a sudden increase of commodity money would rightly result in a lower marginal utility for the commodity. A lower marginal utility means a lower value in trade.
And a higher price would correctly reflect the lower marginal utility of the commodity.
LikeLike
I’m not sure we disagree.
My thesis is that there are two distinct problems a new money has to solve in order to succeed. One is that it has to satisfy the regression theorem, which FRNs do [as Mises wrote in Money and Credit] and bitcoins don’t [as my bitcoin articles show].
The other is competition from other kinds of money. My article the Kickstart Fallacy talks about that a bit. The bitcoin enthusiasts enthuse about the advantages bitcoin has over the competition. But those advantages help only if bitcoin passes the very first hurdle, satisfying the regression thm., which it doesn’t. And in addition it has a huge deciding disadvantage compared to all fiat monies, which the govt passes laws that force one to use them, and also to coins, which have intrinsic value.
We do disagree about one thing, though, for sure. You wrote, “As soon as someone buys it at that made-up price, for whatever reason, it now has a yesterday’s price.” A lot of people agree with you, but I have presented my case that the yesterday’s price had to be very widespread. “Someone” is not enough. You need huge masses of people, almost everyone in fact, not one person or a handful of people. In addition, that price has to be assumed by everyone to be unvolatile, an additional requirement for a yesterday’s price. See my article Bitcoin All in One Place for links to where I prove it all.
LikeLike
Sorry for not getting back to you in a while. I thought you had [very kindly] decided not to continue responding, from your response on another article.
“… but I have presented my case that the yesterday’s price had to be very widespread.”
But there’s never a widespead price for money – it buys different kinds and quantities of goods.
A good becomes money because it can enable a double-coincidence of wants – that is, you can buy many kinds of goods with it – not because it has a widespread price.
“One is that it has to satisfy the regression theorem, which FRNs do …”
As soon as FRNs stopped being redeemable in specie, with people still believing it is still worth what it used to be backed by, they stopped satisfying the Regression Theorem.
The reason is because, at that moment, people were not basing their valuations of the FRNs on a reality of identical worth (part of the nature of money-certificates), or even a reality of mere related worth (“in the ballpark”, but still “based on” the value of silver), but rather on the perception that such was the case.
And the nature of that false perception is functionally arbitrary.
They *think* FRNs are worth such and such, but because they aren’t, they’re essentially making up their own values for them.
And once you say that such a state of perception can exist, there is no logical reason for you to deny that such a state can exist at any point in time, even before something is valued for its use as a commodity.
That is, consistency demands that you concede that money does *not* have to have a use-value before it can become money.
Because if you can make up a price after a money has satisfied the Regression Theorem, you can do so before. Because in either case, the value is made up, and you don’t need a Regression Theorem to justify made up prices.
LikeLike
1. Widespread means almost everyone will accept it as payment for their wares.
If only a handful of people are willing to accept payment in bitcoins, say, but almost everyone else will not take them as payment for their wares, then the bitcoins are not in widespread use.
2. “As soon as FRNs stopped being redeemable in specie, with people still believing it is still worth what it used to be backed by, they stopped satisfying the Regression Theorem.”
No, because they knew the prices of things in the stores, like milk and eggs. Those prices were the same as they were before the FRN stopped being redeemable.
“They *think* FRNs are worth such and such…”
They know it, because the supermarket stickers say that is the price for milk, same as yesterday.
Honored Guest, I have to refer you to my article Bitcoin All in One Place, because it’s all explained there.
LikeLike
I understand, Smiling Dave. You are an interesting writer and a fine Austrian Economist.
I had more to say, but you have been very generous with your time, and I appreciate it.
I’ll keep checking back. Keep up the great work.
LikeLike
Thx.
LikeLike