PTTP: When is everything else constant?

The pure theory of time preference holds that, if everything else is constant, people prefer present goods to future ones. But, if a teacher were telling that to a student hearing it for the first time, the student might start thinking of some of the obvious counter-examples, like the classic one, that, if someone were able to choose between just two things, ice right now, in the winter, and ice later, in the summer, he might choose the latter over the former, the future good over the present one. But, if he were to then voice that counter-example, the teacher probably would quickly respond that “ice in the winter” and “ice in the summer” are two different goods, which of course violates the important stipulation “if everything else is constant”.

I understand that defense, and I think that I understand the at least the gist of the proof for the pure theory of time preference, but my problem is that it leaves me wondering what the point of it is. Isn’t the stipulation “if everything else is constant” a bit too strong? When would you ever encounter a situation where a present good and a future one aren’t different goods because of the different situations? Basically every case that I can think of, if I were choosing between something now and that the same thing later, things probably would be different, and that probably would affect how much I want it or not. So what is the point of the pure theory of time preference? What use is it? When, if ever, does the stipulation “if everything else is constant” hold up?

Isn’t PTTP one of the subsidiary axioms?

We mentally hold variables constant because controlled laboratory experiments aren’t valid for economics. There’s all sorts of “imaginary constructions” used to identify causal laws. This is just a foundational technique of AE to arrive at ceteris paribus laws.

I have no idea.

Are you saying that the pure theory of time preference is an imaginary construction?

Time-preference is implied by the action axiom, since if people did not prefer present goods to future goods, they would not act. Think of Crusoe deciding whether to catch fish with his bare hands, to build a net, or to construct a trawler. Since greater productivity comes from the longer production processes, the fact that he acts implies there must be some reason why he prefers less goods sooner, rather than more goods later.

Every individual has a personal time-preference rate; the extent to which he prefers consumption now, to consumption in the future. The combined time-preferences of the individuals in society is the ultimate determinant of interest rates.

How about when receiving a gift? That’s pretty much a straight gain with the only variable being time.

As to its value, I think it is useful because it is true. It’s how we work and understanding it is a valuable stepping stone to understanding things like savings, investment and interest. The fact that a pure “everything is equal” situation of time preference doesn’t come up often in the real world doesn’t diminish that. It’s like learning how to solve a certain set of math problems to give greater understanding of a more complicated equation that has better practical application.

I have no idea.

I think it is, along with “always preferring more to less” and some other ones. I thought I just saved some bookmark on this but can’t find it. Mises said everything can be deduced from the fact that we act. Rothbard said that a few subsidiary axioms were needed and I tend to side with Rothbard on this.

Are you saying that the pure theory of time preference is an imaginary construction?

Where did you get this “pure theory” terminology from? It sounds like Menger. I really don’t spend much time on economics so I am not sure I can help.

PTTP sounds like a subsidiary axiom. Otherwise, it would be a law deduced from another axiom.

I assume you are familiar with Mises’ imaginary constructions like the “evenly rotating economy” (I didn’t read that article).

This should help a lot on methodology too.

edit:

That makes sense too. I guess we could then say that ice on the Arctic Circle wouldn’t be sought after, therefore it would not be called a “good”.

Holding “everything else constant” is the only way to isolate the effects of economic phenomena.

Mainstream economics has not realized this, and as a result relies far too much on empirical historical economic data which cannot be interpreted without the aid of a priori reasoning (and even then, only to a limited extent) because they represent an inseparable mix of changing and indefinite conditions.

I never understood the thing “always preferring more to less”. An obvious counter-example: I prefer less sickness to more.

I don’t think that Mises said that. He admitted a lot of “subsidiary axioms” into his work. But he didn’t use that term. Here is an example:

sickness is not usually considered a good.

I don’t know who coined it. But it is a pretty common term at least for the LvMI guys.

The reason why they call it the “pure theory” of time preference actually is precisely because they don’t think that it is a “subsidiary axiom”. They think that it is “a categorial requisite of human action”, that it comes straight from the “action axiom”.

I don’t like the term “subsidiary axioms” very much. There is only axiom: humans act.

But in order to get anywhere with the study of praxeology, we have to think about humans acting within an environment. So we posit the following conditions: 1) humans exist, 2) there is diversity among individuals and the environment, 3) humans prefer leisure to labor, etc. Then we posit further conditions to get further in our study: 4) humans have recognised the benefits of trade and the division of labor, 5) humans have established a general medium of exchange, etc.

We can do praxeology by reversing these assumptions, but it wouldn’t be terribly interesting. These conditions direct our study to the actual world we live in. Praxeology becomes economics (or “catallactics”) when we posit that humans trade.

How did you jump from “people[…] prefer present goods to future goods” to “[e]very individual has a personal time-preference rate”? How did you jump from saying that everyone prefers A to B to saying that some people prefer one A to two B, some people prefer one A to three B, and so on? The problem that I see with that is that it seems that you jumped from talking about “abstract classes”, like water, diamonds, present goods, or future goods, to talking about “concrete things”, like a certain amount of water, a certain amount of diamonds, a certain amount of a certain present good, or a certain amount of a certain future good.

You’re right on “subsidiary”. I have no clue why I even used that word. I think people usually call those “supporting axioms”. “Subsidiary” implies some sort of dependence (different than “being deduced from”, which doesn’t make sense). There is some information on this stuff, I just can’t find it offhand.

But the gift could be something like ice, like in the classic example. So, if the guy could choose when to get the gift, he might choose to get it later, in the summer, instead of now, in the winter.

Mises says the pure-time-preference explains why consumption exists as a phenomenon. for any good (all things being equal), if you do not prefer its consumption now to its consumption later, you will never perform an act of consumption. e.g. for a sandwich-when-you-are-hungry, if you prefer its consumption later, then when later comes around, you still prefer consuming it later…

Humans act implies humans have time-preference. So if an individual did not have time-preference, he would not act. All humans act therefore all humans have time-preference. Part of assuming diversity among humans is assuming diversity between individuals’ time-preference rates. It is clear that, while we all prefer present goods to future goods, some of us are more future-oriented than others.

Where did I say this?

Everyone prefers having 1 fish now to 1 fish in the future, assuming that 1 fish now and 1 fish in the future are considered equally serviceable goods (i.e. the same good), except with respect to time. (This is obviously not true, of course, in the ice cream in winter/summer example). I don’t see the problem.

Well sure, but it could also be, say, a keyboard. I may not need a keyboard now, but when mine fails it would be handy to have another in the house already. The ice example adds its own peculiarities because storing ice takes work, and so the assumption is the other person stores the ice for you between now and whenever. If your ice example happened at a research outpost in Antarctica I doubt they’d care if it’s summer or winter when they get the ice, and therefore their time preference would shift towards the sooner rather than later naturally because there’s no apparent downside to getting it now and that’s the “default” preference.

My point isn’t that an example of everything being constant is easy to find, my point is just that it’s a foundational principle that sheds light on other, more complicated behaviors that are easy to find in the real world and so it is still quite useful.

I never understood the thing “always preferring more to less”. An obvious counter-example: I prefer less sickness to more.

I wrote about this recently but forgot about the thread. See #4 from this part of Mises’ Theory and History.

Sometimes the utterance of a judgment of value is elliptical and makes sense only if appropriately completed by the hearer. “I don’t like measles” means "I prefer the absence of measles to its presence. Such incompleteness is the mark of all references to freedom. Freedom invariably means freedom from (absence of) something referred to expressly or implicitly. The grammatical form of such judgments may be qualified as negative. But it is vain to deduce from this idiomatic attire of a class of judgments of value any statements about their content and to blame them for an alleged negativism. Every judgment of value allows of a formulation in which the more highly valued thing or state is logically expressed in both a positive and a negative way, although sometimes a language may not have developed the appropriate term. Freedom of the press implies the rejection or negation of censorship. But, stated explicitly, it means a state of affairs in which the author alone determines the content of his publication as distinct from a state in which the police has a right to interfere in the matter.

Action necessarily involves the renunciation of something to which a lower value is assigned in order to attain or to preserve something to which a higher value is assigned. Thus, for instance, a definite amount of leisure is renounced in order to reap the product of a definite amount of labor. The renunciation of leisure is the means to attain a more highly valued thing or state.

You prefer more being healthy to less, at the same time as less being sick to more. The “good” is health not sickness, nonetheless.

I don’t think that Mises said that. He admitted a lot of “subsidiary axioms” into his work. But he didn’t use that term.

He didn’t use that term, it was my mistake. Still there is a difference on what axioms apply in the opinion of Mises vs. Rothbard. Read from page 10 here:

For a critique of the time preference origin of pure interest and an alterantive theory based in means ends analysis read Hulsmann’s excellent Theory of Interest.

This is the hook. The item must be the same at both times. Otherwise you are comparing a block of ice now to a puddle later.