Hulsmann's 'A Theory of Interest'

Economics Economics Questions
1

I have re-read and been reflecting on Jorg Guido Hulsmann’s paper “A Theory of Interest” and the implications for the Misesean / Rothbardian framework.

We previously touched on this paper in these threads:

https://forum.freecapitalists.org/t/pttp-when-is-everything-else-constant/14210

https://forum.freecapitalists.org/t/the-disutility-of-labor-as-a-category-of-action/14050

I think it deserves its own thread.

I have come to the conclusion that the basis of economics (much of the first chapter of MES, and some parts of HA) needs to be significantly reformulated in light of Hulsmann’s paper. In particular:

  • The categories of leisure and labor need to be redefined; the idea of leisure being preferred to labor being an “empirical”, or “subsidiary”, assumption needs to be re-examined; the generality of concepts such as uncertainty and speculation needs to be reconsidered.
  • Time-preference theory needs to be corrected and de-emphasised (if not expunged completely), or at least reformulated within the context of the means-ends value spread concept.

I could say much more about this. But before I do… Does anyone else here feel that the paper is as profound as I do? Does anyone here have or know of any criticisms of his theory they would like to discuss?

2

I read that paper a few months ago and I remember thinking that it was entirely incoherent. Also, any theory of interest which entirely rejects time preference cannot, in any way, be considered Misesian.

3

A tentative critique:

As Guido said, for the actor an end has a higher value than its means. If he could choose between the means or the end, clearly he would pick the end. He would always prefer the end to its means. If he could go straight to the end, why the hell would he even consider taking some sort of useless roundabout approach? If he had the choice between having the means now to an end later, or simply the end immediately, he would pick the former. Of course.

But I’m not so sure that it’s such a quick knockdown in that we couldn’t even conceive of it as being different without running into a bunch of inescapable contradictions. I mean, let’s say that I’m thirsty. If I could pick between drinking some water, and quenching my thirst, which one would I choose? Well, isn’t that perfectly incoherent? If I drink the water, I quench my thirst. And, if I quench my thirst, well, I quench my damn thirst. In both cases, I get what I’m looking for. How exactly could I choose between quenching my thirst and quenching my thirst? If you encounter a choice between two perfectly identical things, it’s certain that you’re just playing with words.

Because the means just is the ends, it’s perfectly meaningless to say whether I prefer one to the other. If I have the means, I have the end. And, if I have the end, I have the end. What would it even mean to choose one or the other? But not so fast. I keep omitting the “will”. It’s not if I drink the water, I quench my thirst: It’s if I drink it, I will quench my thirst. If I have the means, I will have the end. So maybe the time factor has something to do with it? Maybe I would prefer the end to its means simply because I would end up getting what I want more quickly? But then wouldn’t that destroy his point because it would plunge us back into simply talking about preferring present goods to future ones? And then wouldn’t we simply need to abandon his train of thought, and instead start talking about why exactly we have that preference (the time preference)?

Well, I think that the whole answer is to point to uncertainty. If it really were perfectly certain that I would definitely get the end out of what I think is its means, it would be perfectly incoherent to question whether I would prefer one to the other. Because one just would be the other, it would be completely meaningless to ask which I would choose. But it’s not like that at all. There’s always a particular amount of uncertainty. I’m never perfectly confident in any of my opinions. If I could get the end without what I think is its means, of course I would do that. The end is the end, but what I think is the means in only what I think is the means. If I have a 100% chance for getting the satisfaction vs. a <100% chance of getting it, of course I would pick the former. Anything else would plunge us into a sea of contradictions.

4

I think your mistake here is in the description of your end. If I take what you say literally, then you are describing an act of leisure, not labor, so yes it is incoherent to try and distinguish ends from means. It is no different than dancing. In that case, my end would be “to dance” (or more precisely, “to be dancing”), and my means would be “dancing”. So that the means and ends are indistinguishable. Equivalently, your end as you have described it here is “to quench/drink” (or more precisely, “to be quenching/drinking”) and your means is “quenching/drinking”. Your end would be satisfied during the action itself (i.e. it is a present-oriented action) and the means and ends are inextricable. So its leisure. All leisure actions are present-oriented, because the end, being inextricable from the means, is acheived during the action itself.

If I restate your end as “to not be thirsty”, then we can see that the means can be distinguished from the end. You can coherently choose between having your end fulfilled (without any action being done) and having the means to fulfil that end. Then it is an act of labor. Furthermore, drinking (in this sense of it being a labor action) is future-oriented, because your end will be satisfied only after you have acted; it will not be satisfied during the action itself.

So you might think that all leisure actions are present-oriented and all labor actions are future oriented. If this were the case, means-ends value spread theory would be equivalent to time-preference theory. But this is not the case. Here is Hulsmann’s example of present-oriented labor:

Hulsmann’s argument is that originary interest emerges from all labor actions, even those (like this example) which are not future-oriented, in the sense that the means and ends coincide in time. So originary interest exists independently of any considerations of time. Time-preference theory is really just a special case of means-ends value spread theory. Originary interest emerges from the more general category labor actions rather than the special case of future-oriented labor actions.

5

I thought it was a good paper, and I agree that time preference is shaky. Bob Murphy had some criticisms of time preference as an explanation for money interest, too.

6

Yes, Murphy’s criticisms are good. Here is his paper and here is a lecture he gives about it. While they have similar criticisms, their solutions are quite different. In short, while Hulsmann has generalised time-preference theory to means-ends-value-spread theory, Murphy has specialised time-preference theory to a money-time-preference theory. I think Hulsmann is on the right track, not Murphy.

7

Actually, when I said “quenching my thirst”, I meant removing the uncomfortable bodily sensation (for me, a stabbing pain in my throat), and that happens after I make the decision to send a stream of water down my throat.

8

Could you give a quick summary of how he did that?

9

My view of the undeniable element of time preference is that, all other things being equal, people would like satisfaction now and in the future. But it’s not clear this explains money-interest.

10

OK, then you were thinking of it as a labor action. Now you put it like this, surely you can see that you can easily distinguish the means and the ends? You can imagine having the stabbing pain gone without having to pour the water down your throat?

11

Sorry, for some reason I didn’t notice that you already addressed that point.

But what would drive you to choose one over the other if they both perfectly have you where you’re trying to go?

If it were absolutely certain if I send the stream of water down my throat, I would destroy the uncomfortable bodily sensation, I don’t see how I could choose between them. In both cases, I don’t have the uncomfortable bodily sensation. I mean, I know that they are in fact distinguishable in that I would have to choose between one or the other, so I guess that it’s more that I’m saying without anything else, we don’t have anything to break the tie, but if we have uncertainty, there is a reason to pick one over the other.

But if they’re distinguishable, that means that we are making some sort of distinction between the utility of going straight for the goal, or simply taking a roundabout approach. And I contend that’s the uncertainty. In other words, I accept that there is the exact same “value spread between means and ends” that Guido was talking about, but that he acted as if it was totally obvious and unrelated with time. Well, there is that value spread, but it’s because of the uncertainty or whatever. If there were no uncertainty that sending the stream of water down my throat would remove the uncomfortable bodily sensation, I would be perfectly indifferent between the one and the other. And then what? How could I act on a perfect indifference?

EDIT: In other words, I guess when I said “the means just are the ends”, I meant in terms of their utility.

12

I would have to rewatch the lecture. I may get a chance to do that tonight.

13

Sure, but if I were perfectly certain that the one causes the other, I wouldn’t believe it even for a second.

EDIT: Nevermind, this post is useless.

14

trulib, I also liked Hulsman’s paper. I hinted about my ‘theory of interest’ in the first thread linked in your OP (link to my post) but got no feedback on it. Hopefully, will get some feedback on it now.

Here’s the text again:

Hulsmann’s piece was excellent, especially the review/intro part. Do you have a link (pdf?) for Bob Murphy’s angle?

A common thread in all theories of interest seems to be the sooner/later, cause/effect, means/end paradigms, which are merely different aspects of one and same thing: before/after location on the time axis. My (layman) intuition on the source of interest comes from experience and also involves time but from an uncertainty, opportunity angle. Having control over $100 now is always preferable to having such control a year from now because my freedom to use (control, allocate, invest, etc.) the $100 in combination with the opportunities that may come my way over that year is always worth a non-zero amount to me. There’s a saying that: “Success is when opportunity meets preparation.” My control over the $100 (vs lacking it) is an essential part of the “preparation”. If the opportunity came, and I didn’t have the $100, the opportunity is lost.

In this light, to me, interest is inextricably related to opportunity cost. When you ask me to lend you $100 for a year, I will price the loan (interest) to match (or exceed) my perception of potential opportunity lost (in addition to pricing in your credit risk, of course). The more volatile the environment (markets), the larger the number and intensity of potential opportunities over the year (distressed sellers of goods/assets, IBM shares drop 50%, etc.) and/or the higher the probability I’ll need them to cover my own potential losses/risks, hence the costlier the loan. The longer the period I’m denied my control over the $100, the larger the number of opportunities potentially missed (or potential losses/risks), the larger the uncertainty span of all possible outcomes that may affect me (positively or negatively) hence the larger amount of total interest I would demand in proportion to the length of the loan. These uncertainty forces are considered when pricing bonds and are reflected in the shapes of the bond yield curves (typically, higher annual yields for longer maturities, all other things being equal.)

Z.

15

Here, Z: http://mises.org/journals/scholar/murphy2.pdf

16

Very interesting, but you sort of lost me when you started talking about uncertainty.

Do you have any thoughts on how your potential opportunity cost idea relates to the uncertainty thing that I was talking about?

17

I admit I have not read Hulsmann’s paper (although I’m pretty sure I would disagree since I firmly believe time preference determines the interest rate). However, I have read Hulsman’s “Time Preference and Investment Expenditure” which is a much more striking critique of the Austrian structure of Production and the Interest rate, since its something Austrians can not really argue against.. In the paper he says that there is no one to one relationship between the rate of savings and the interest rate. A high interest rate could result from a decrease in saving (a shift in the supply of savings) or an increase in demand (coming from an increase in the labor supply for example). He mentions that the main defect in modern Austrian capital theory (something that extends to Rothbard,and Mises) is that they only considered changes in the supply side of the interest rate graph-never the demand side. As far as I can tell he is right, on page 418 of MES Rothbard lays out that interest rate determining graph. It does not take more knowledge of basic price theory to see that high time preference= low savings and low time preference = high savings does not have to be completely true.

The conclusions of this mean that one needs to distinguish between capital widening and capital deepening, and that the proportions between consumption and investment (both individual and aggregate) do not have a certain relationship with the rate of interest.

His paper: http://www.guidohulsmann.com/pdf/Time_Preference_Investment_Expenditure.pdf

18

Okay, let me give my theory of interest too.

In a pure market economy, the market follows the people’s preferences. If everybody stops liking one thing, and starts liking another, the first will disappear from the market no matter what, and the second will appear if it’s possible. If the average preferences move in one direction, the market tries to move in the same one. It follows their preferences.

So in the 1st thread that trulib linked to, I asked in the the PTTP, where is everything equal? If everything else is constant, people prefer present goods to future ones. But where is everything else constant? Well, I think that the answer is that it’s constant with money. As the commonly accepted medium of exchange, it can get you anything that the market has. But what if you change your preferences from one time to another? In the ice in the winter vs. ice in the summer example, you change your preference for the ice. In the winter, it’s not as useful as in the summer. In short, your preferences shift from one time to another. But in the pure market economy, the market follows your preferences! If you’re preferences change from one time to another, what your money can buy changes in proportion. The fact that the market moves cancels out the fact that your preferences move, and all that’s left is time preference. And as I said in this post, I suspect that the time preference is born out of the fact that you’re never perfectly certain that what you think are the means to your ends really are such.

But certainly the pure market economy doesn’t really follow your preferences? Doesn’t it have everybody else to take into account? Well yeah. I guess that I was just trying to make the explanation simpler. For one loan to the next, the rate of interest might be a lot different. I mean, different people have different time preferences, right? So I’m just talking about the average rate of interest or whatever. In the pure market economy, the market follows the average preferences, so the average rate of interest is where everything else is constant. So I guess that my theory of interest is a “monetary” one too? Or maybe I have no idea what that’s supposed to mean.

Anyway, in the 1st thread, I mentioned that the PTTP starts out with the statement that to act, the person must prefer present goods to future ones (because otherwise they would never do anything), but how that doesn’t lead to the idea that there are different rates of time preference. If I prefer present goods to future ones, how does that have anything to do with any sort of rate of time preference (that I might prefer 1 good sooner to 2 later, 1 sooner to 3 later, or whatever)? Well, that was a rough restatement, so make sure to refer back to the original thread, but either way the point is that I think that my idea that it’s the uncertainty that accounts for the time preference gives us the different rates. I can be more or less uncertain about a production procedure. And there are 2 variables: Uncertainty that each step will lead to each other step, and the amount of steps. If I’m 90% certain that step 1 will lead to step 2, 90% for step 2 to 3, the same for 3 to 4, and so on until 10 (the end), we have to take into account that the overall certainty decreases with each step. Somebody who knew some probability theory could probably tell you by how much. I guess.

Anyway, that’s my tentative theory of time preference and interest.

19

I can’t comment substantively since I have not read the paper for a while however I do remember it being rather incisive. I wanted just to comment that his theory implies value imputation theory is incorrect since there is always an inequailty of value rather than the consumer valuations filtering directly into the means of production.

20

Doesn’t his theory mean that the ends do impute their value to the means, but just that they don’t impute all of their value (just some of it)?