Is Action an A Priori Category?

And was Mises a Kantian?

Sometime in the remote future – maybe 10 or 20 years from now, or even 50, if I live that long – I intend to publish an essay or treatise called Praxeology refuted. (If I cannot refute it, the title will be Praxeology validated.) Here is a head start.

The father of praxeology, the science of human action, Ludwig von Mises, did call action an “a priori category”. And he takes this terminology from the father of the a priori, Immanuel Kant. So let me begin with what Kant meant by this term.

“A priori” means “before experience” or “independent of experience”. Kant attempted to prove that such a priori knowledge lies at the base of all our knowledge. The senses provide us with the raw material of knowledge, but this raw material has to be ordered or structured by  a set of categories, which are known to us independent of all experience and, in fact, are required to even make experience possible. Those categories are twelve in number, no more and no less, and they all have to do with the logic of our thinking. “Action” is not one of those categories, so it should be noted that Mises already deviates from Kant by adding “action” to the categories.

Nevertheless, there is much in Mises’ writings on praxeology that reveals a heavy Kantian influence. For one thing, he maintains that truth and falsehood are a matter, not of reality outside of us, but of the logical structure of the human mind. (I will devote a chapter in my upcoming treatise to this.)

But let me turn to this “category of action”. What is meant by it? The simplest formulation I can come up with is that man acts purposively and relates means to ends. When man acts, he does so with a purpose in mind. The purpose may be long-range (such as pursuing a certain career – or writing a treatise on praxeology), or it may short range (such as going down to the grocery to buy some food – or completing this blog post). But there is no such thing as purposeless action: even if a man just takes a stroll, there is the purpose of relaxation.

It might be objected that some actions (such as sneezing, or removing one’s hand from a hot plate) are involuntary. Mises’ answer to such objections is that those are not actions, but reactions. And human action also has to be distinguished from animal behavior: animals act in ways that promote their survival – lions hunt for food, and the zebras and antelopes run away from the lion to find a safer place to graze. But animals do not consciously set themselves goals and purposes; this is the distinctive mark of man.

So far, so good – there is certainly nothing here that an Objectivist should object to (or any thinking person, for that matter). But is it a priori? Is it independent of experience?

The point that man acts purposively may seem so self-evident and so all-encompassing that it eludes analysis. But I would say that we first know it by introspection: every one of us knows introspectively that we act with a purpose in mind and that we relate means to ends in order to achieve this purpose. That other men do the same is evident from their behavior; the assumption that other men act totally mechanically is too preposterous to seriously consider.

But – Immanuel Kant to the contrary notwithstanding – introspection is no more a priori than is extrospection.

Is Mises, then, a Kantian? Let me quote one of the foremost experts on Mises. Jörg Guido Hülsmann (from his Introduction to the third edition of Epistemological Problems of Economics):

The least one can say is that Mises’ theoretical analyses do not fit very well the caricature of the “Kantian” approach – studying the workings of the human mind, and nothing but this, in order to derive a priori insights about the rest of the world. If we want to do justice to what Mises actually said and did, rather than to squeeze his views into some preconceived epistemological scheme, then it seems we cannot avoid the conclusion that the affinities of Mises’s ideas with Kant’s philosophy are mainly rhetorical affinities. Mises is not closer to Kant than he is to any other rationalist philosopher.

Mises always stressed that the propositions of praxeology and economics were not derived from metaphysical (in the pejorative sense of “groundless”) speculation, but from facts of experience – though not experience of the kind that comes from the human senses. (P. liii.)

Well, fine – I like this. (I would like it even more, if it were true.)

It is true that introspective experience does not come through the senses – at least not the kind of introspective experience I wrote about earlier. But it is still experience – not an “a priori” that precedes experience and makes experience possible, the way Kant’s categories are supposed to do.

Let me follow Hülsmann’s advice and see what Mises himself actually said, to find out whether his affinities with Kant are “mainly rhetorical” or not. First from Human Action:

There is no means to establish an a posteriori theory of human conduct and social events. (P. 31.)

And so, such a theory has to be a priori.

In the field of human history a limitation similar to that which the experimentally tested theories enjoin upon the attempts to interpret and elucidate individual physical, chemical, and physiological events is provided by praxeology. Praxeology is a theoretical and systematic, not a historical, subject. Its scope is human action as such, irrespective of all environmental, accidental, and individual circumstances of the concrete acts. Its cognition is purely formal and general without reference to the material content and the particular features of the actual case. It aims at knowledge valid for all instances in which the conditions exactly correspond to those implied in its assumptions and inferences. Its statements and propositions are not derived from experience. They are, like those of logic and mathematics, a priori. They are not subject to verification or falsification on the ground of experience and facts. They are both logically and temporally antecedent to any comprehension of historical facts. They are a necessary requirement of any intellectual grasp of historic events. Without them we should not be able to see in the course of events anything else than kaleidoscopic change and chaotic muddle. (P. 52; italics mine.)

I would also like to quote from Epistemological Problems of Economics (translated from the German original by no less a person than George Reisman):

The science of human action that strives for universally valid knowledge is the theoretical system whose hitherto best elaborated branch is economics. In all of its branches this science is a priori, not empirical. Like logic and mathematics, it is not derived from experience; it is prior to experience. It is, as it were, the logic of action and deed. (P 13; italics mine.)

However, what we know about our action under given conditions is derived not from experience, but from reason. What we know about the fundamental categories of action – action, economizing, preferring, the relationship of means and ends, and everything else that, together with these, constitute the system of human action – is not derived from experience. We conceive all this from within, just as we conceive logical and mathematical truths, a priori, without reference to any experience. Nor would experience ever lead anyone to the knowledge of these things if he did not comprehend them from within himself. (P. 14.)

Those quotes show more than a “rhetorical affinity” between Mises and Kant. Mises’ epistemological framework is clearly Kantian. And saying this is not to “squeeze his views into some preconceived epistemological scheme”; he squeezes himself into that scheme.

Now, the question in my mind is: Is this Kantian framework really necessary to validate all of Mises’ numerous insights. I think not. But this is a large subject, so I will return to it later. Maybe soon, maybe in some remote future.

Fractional Reserve Banking Yesterday and Today

Based on a Facebook note I recently wrote.

Fractional reserve banking in today’s world is the main engine of inflation. But before I go into that, some definitions and preliminary remarks:

Fractional reserve banking means issuing fiduciary media (those expressions are interchangeable). So what is fiduciary media? George Reisman gives the following definition in Capitalism: A Treatise on Economics:

Fiduciary media are transferable claims to standard money, payable by the issuer on demand, and accepted in commerce as the equivalent of standard money, but for which no standard money actually exists. (P. 512.)

And Ludwig von Mises explains the term in Human Action:

If the money reserve kept by the debtor against the money-substitutes issued are less than the total amount of such substitutes, we call that amount of substitutes which exceeds the reserve fiduciary media. (P. 433.)

What is meant by “standard money” and “money substitutes” here?

Under a metallic standard, “standard money” would be physical gold and/or silver. (In today´s world, “standard money” is irredeemable paper money, backed ultimately by government force.) “Money substitutes” are media that serve as money without actually being money, such as banknotes, checks and token coins. If a money substitute is fully backed by standard money (e.g. if a banknote can at any time be exchanged at the bank for the amount of gold/silver that is actually stated on the note), then Mises calls it a “money certificate”. If the banknote is not fully backed, then it is a “fiduciary medium”.

But how does one know whether a given banknote is a money certificate or a fiduciary medium? In other words, how does one know whether it is fully backed or fractional? The crux of the matter is: one cannot know. As Mises explains in the same paragraph I quoted above:

As a rule it is not possible to ascertain whether a concrete specimen of money-substitutes is a money-certificate or a fiduciary medium. A part of the total amount of money-substitutes issued is usually covered by a money reserve held. Thus a part of the total amount of money-substitutes issued is money-certificates, the rest fiduciary media. But his fact can only be recognized by those familiar with the bank’s balance sheets. The individual banknote, deposit or token coin does not indicate its catallactic character [i.e. its nature in exchange].

In simple words: The banknote itself doesn’t tell you whether it is a money certificate or a fiduciary medium; it doesn’t tell you whether it is fully backed by gold/silver or only partly (“fractionally”) backed. A $10 note looks exactly the same, regardless of whether it is fully backed or only partially backed.

This point deserves stressing; in my experience people miss this. The distinguishing mark of a fiduciary medium is precisely that it cannot be distinguished from standard money. If one banknote states: “money certificate, backed by gold (or silver), and another states: “fiduciary medium, not backed by gold (or silver)”, then nobody would touch the second banknote, because it is stated on the note itself that it is worthless. Or say that a banknote states that it is only partially backed; it says: “this note has 90% backing (or 50% or 10% or whatever the actual backing would be)”, then people would accept it for that value, not for the value stated on the note. If a $100 note were backed by 50% of actual money, it would exchange in transactions as $50, etc. If a bank were to do this, it would be a sign of honesty, but it would also wipe out the whole point of issuing fiduciary media.

I have argued (and so have Murray Rothbard and George Reisman) that this practice is fraudulent by its very nature. (Mises, as far as I know, never said it is fraudulent, but this is due to the fact that Mises always shied away from moral judgment.) I won’t repeat my argument here, because those who have not gotten the gist of it before will hardly get the gist this time, if I repeat it. I will simply insult them and call them knuckle-headed and stupid.

Back to today. What has to be understood is that the money supply today is almost totally made up of fiduciary media. It is not even made up of irredeemable paper money (“fiat money”). The fiduciary media today are built upon a basis of fiat money; but that fiat money is only a small fraction of the actual money supply.

Those fiduciary media are created in the very act of lending out money. Here in Sweden the reserve requirement when a loan is made is 5% or thereabouts (I believe it is slightly higher in the US, but not much; the figure I have seen is 7%). This means that if I take out a loan of $100000 (e.g. to buy a new apartment), it is only backed by $5000 of fiat money; the remaining $95000 are created out of thin air. And, since I am not the only person borrowing from banks, you can just imagine what it means in today’s world, where trillions of dollars, euros, etc. are being lent out with virtually no metallic backing.

(It’s even worse. As Thomas Kenworthy pointed out on Facebook:

That’s correct. And it’s even less because the loans granted by the central bank to the banks are not delivered as physical cash; rather they are credited to special accounts which the banks may use as if they were cash when creating new loans.)

In the “good old days” the fiduciary media at least existed as physical objects – banknotes, checks, token coins. Today, only the irredeemable fiat money exists in physical form. The fiduciary media are digital; they exist only on computers!

(For those of you who understand Swedish I recommend this article by Torgny Persson, the head of Liberty Silver.)

What happens next? As I am sure you are all aware of, an inflationary boom cannot last forever. It will either end in hyperinflation (where the currency simply loses its value and is destroyed) or in a deflationary depression. And what happens in a deflation is that those fiduciary media, which were originally created out of thin air, now disappear into the same thin air. Those fractional loans simply lose their value; banks go bankrupt; all the malinvestments made during the boom turn out to be malinvestments; there will be mass unemployment. If the government then lets the depression run its course, there will eventually be recovery; there might even be a return to some kind of gold standard; but what will most likely happen is that governments and central banks start to re-inflate, and then we will have to live through the whole cycle again.

Fractional reserve banking may appear to be practical in the (very) short run; but it is certainly not practical in the long run (or even in the medium run). And – if one holds that “the moral is the practical” – it is hardly moral either.

$ $ $

This note engendered quite a lively discussion on Facebook; unfortunately mostly in the form of rambling from a couple of persons in favor of FRB.

One of those persons (by a tortuous reasoning) arrived at the conclusion that fractional reserve banking does not mean lending out more than one has in the vaults, but lending out less than one has in the vaults. (And I have seen this argument before.) Consider the implications of this:

In today’s world it would mean that no matter how many trillions are lent out, there are always even more trillions in the bank vaults to cover this. In today’s world of fiat money, it is of course conceivable that this money is continually being printed. This is not how it works, but it is at least conceivable.

But what about a world where the fiduciary media are at least built upon a base of gold and silver? Issuing fiduciary media does not magically create more gold and silver. Is a fractional bank one that lends out only part of its gold and silver? But why would a bank do that? This is just an attempt to defend fractional banking by changing the meaning of the term.

The other person brought up an argument that I have also heard before. It goes like this:

A fractional reserve bank operates like an insurance company. An insurance company does not need reserves to cover every insurance policy, because it will never have to pay out all those policies at the same time. If, for example, it is insurance against fire, it will never happen that every policy holder’s house burns down at the same time. Similarly, a fractional reserve bank will never, it is argued, have to pay back every claim at one and the same time, so the fear of a bank run is as unsubstantiated as the fear that all houses will burn down at the same time. And therefore, there is no difference between taking out a fractional bank loan and taking out an insurance policy.

But this argument is sheer sophistry. Insurance companies do not create the various misfortunes or disasters they insure against. It is, for example, not the insurance companies that put houses on fire; it is pyromaniacs. But the disasters inherent in fractional reserve banking – in particular, the possibility of a bank run – are created by the very practice of fractional reserve banking.

So what is it one is insuring oneself against by taking out a fractional loan? The disastrous effects of fractional reserve banking, or what? But you wouldn’t buy insurance against fire from a pyromaniac. Or would you?

$ $ $

This is a comment I got from Stuart Hayashi:

If I, as a vendor, accept notes from a full-reserve bank, then I have to understand that the full-reserve bank is not going to pay me interest on my account. Rather, as long as I hold the full-reserve bank’s notes, I am contractually obligated to pay some periodic fee to the full-reserve bank for the service of its keeping my money in storage.

No, this is a misunderstanding.

Certainly, someone could deposit his gold in a bank just for safe-keeping. Then, in effect, the bank would be a warehouse, and you would have to pay a fee for the safe-keeping.

But if, instead, you allow the bank to lend out your deposited gold, then, in effect, you are lending out the gold, using the bank as an intermediary. You would do this on the premise that the bank knows better than you what investments will become profitable. (If you know this better than the bank, you might as well start your own bank and lend out your gold directly. But most people don’t have the expertise necessary for this.)

In this case, you don’t pay a fee but will receive interest on your money.

The idea that one couldn’t receive interest in a full-reserve system would make sense, if the money supply (the amount of gold/silver in this full-reserve system) were to be totally rigid. But this is never the case. Some new gold/silver is always being mined and minted.

But even if we suppose that the money supply is absolutely rigid, so that no interest could be paid, this problem isn’t solved by the issuance of fiduciary media. In fact, this would only give you counterfeit interest on counterfeit money.

$ $ $

Update September 2016: One of my Facebook contacts, Jim Nelson, has pointed out to me that my reasoning here clashes with George Reisman’s net consumption theory of profit/interest. The theory says that the average profit/interest in the economic system equals the capitalists’ own consumption. If profit/interest were to go to zero, this would mean that the capitalists consume nothing, which would mean that they would starve – an implication that is obviously absurd. And this would hold true also on the assumption of a totally rigid money supply.

There is one more component in the average profit/interest in the economic system, and that is net investment – i.e. the surplus of productive expenditure over business costs. But a main source of net investment is new and additional money, and therefore net investment will tend to disappear under invariable money. It will be modest under a precious metal standard and quite large under our present inflationary paper standard.

Absent increases in the money supply, net investment would occur when capitalists curtail their consumption and invest more. But since capitalists, too – apart from their yachts, their caviar, their Mouton-Rotschild, their twenty-year matured single malt whisky, their Habana cigars and other luxury items – need such consumption items as food, clothes and shelter, they could never stop consuming totally, since then they would starve and/or freeze to death, and then we would finally get rid of them. (This last is meant as sarcasm.)

For the derivation and implications of the net consumption/net investment theory, I refer you to Reisman’s Capitalism: A Treatise on Economics, chapters 16 and 17 (for net investment also the end of chapter 15).

My answer to Stuart above would make sense, if there were only net investment and no net consumption. But in fact, net consumption is the more important component.

$ $ $

There is one thing I glossed over in my note, because I have covered it before and I didn’t think it necessary to repeat it. But it seems I have to.

Changes in the money supply affect the price structure in the economy, but it doesn’t do so immediately, and it doesn’t affect every price in the same way. Newly printed money, or new fiduciary media created out of thin air, gets into the hands of some people first and others only later. The person who gets the fractional loan is in a position to use the new money before prices have risen and stands to gain; others get it only after prices have risen and stand to lose.

But those people who get the new money first have done absolutely nothing to deserve this gain (they merely have been friends of the bank and, in today’s world, the government). The people who get the money last have done absolutely nothing to deserve their loss (having to pay higher prices for virtually everything). This is grossly unfair.

Mises himself wrote about this so many times that anyone who has read him should know it. He didn’t go so far as to call it “fraud” or “counterfeiting”, but that is because he thought that moral judgment has no place in economics. (This is one of my few big disagreements with Mises. I think that moral judgment has a place in economics.)

And this is also the reason I don’t accept the argument that FRB is merely a “calculated risk”. This could be said about the bank and the borrower; but it certainly doesn’t apply to those people who have to bear the brunt in the form of rising prices. Besides, it could be said of any counterfeiter that he takes a calculated risk when he injects counterfeit money into the economy. The risk, of course, is being found out and having to go to jail; but no such risk faces a government-sponsored counterfeiter.

To quote Mises himself:

The clients of the expanding bank [i.e. the bank issuing fiduciary media] […] receive additional credits, they expand their business activities, they appear on the market with an additional demand for goods and services, they bid up prices. Those people who are not clients of the expanding bank are not in a position to afford those higher prices; they are forced to restrict their purchases. Thus there prevails on the market a shifting of goods from the nonclients to the clients of the expanding bank. (Human Action, p. 437.)

Mises should have been more moralistic and say that the clients of the expanding bank skin the nonclients.

$ $ $

I should mention that Mises’ own solution to the problems of fiduciary media was free banking with absolutely no government interference. Under such a system, the market itself would put strict limits on the practice. The reason is that if the customers of an FRB bank lose confidence in its ability to repay, there will inevitably be a bank run, and the bank would go bankrupt. That will make the bank very cautious about issuing fiduciary media.

If there is government interference, on the other hand, the government will bail out the defaulting bank. And this is something that often happened in the “good old days”, when the fiduciary media were built on a basis of gold and/or silver. (That it happens today if fairly obvious…)

$ $ $

Mises also has an interesting observation on the origin of fiduciary media that I’d like to share:

Issuing money-certificates is an expensive venture. The banknotes must be printed; the token coins minted; a complicated accounting system for the deposits must be organized; the reserves must be kept in safety; then there is the risk of being cheated by counterfeit banknotes and checks. Against all these expenses stands only the slight chance that some of the banknotes issued may be destroyed and the still slighter chance that some depositors may forget their deposits. Issuing money-certificates is a ruinous business if not connected with issuing fiduciary media. In the early history of banking there were banks whose only operation consisted in issuing money-certificates. But these banks were indemnified by their clients for the costs incurred. (Human Action, p. 435.)

This is actually the only serious argument I have ever seen against the idea of totally outlawing fiduciary media. As you might know, both Rothbard and Reisman favor such outlawing. Mises himself, I think, is ambivalent on this issue: he thought that in the absence of any government interference, the market itself would set strict limits to the practice (see above). On the other hand, he writes in The Theory of Money and Credit:

Now it is obvious that the only way of eliminating human influence on the credit system is to suppress all further issue of fiduciary media. (P. 447.)

$ $ $

My collected blog posts on fractional reserve banking.

And, in Swedish, Varför “fractional reserve banking” bör förbjudas