Time Preference
Mises’s chapter on “Action in the Passing of Time” is quite possibly the most confused chapter in the entire book. Why is it so confused? First off, it’s flat out contradictory. For instance, Mises says: “[Acting man’s] only concern is to make the best use of the means available today for the best possible removal of future uneasiness.” But just a little later he asserts something completely different: “The uneasiness which acting man wants to remove as far as possible is always present uneasiness.” Which is it? Does acting man only want to remove future uneasiness or does he always want to remove present uneasiness? Are the present and the future the same thing for Mises? That might explain why his analysis is so static, so confused about the nature of time.
But that’s just the start of the problems. Just try to entangle this bit:
He who consumes a nonperishable good instead of postponing consumption for an indefinite later moment thereby reveals a higher valuation of present satisfaction as compared with later satisfaction. If he were not to prefer satisfaction in a nearer period of the future to that in a remoter period, he would never consume and so satisfy wants. He would always accumulate, he would never consume and enjoy. He would not consume today, but he would not consume tomorrow either, as the morrow would confront him with the same alternative.
What does this even mean? If someone who consumes something reveals a preference for satisfaction in a nearer period of the future, then someone who does not consume something must then reveal a preference for satisfaction in a later period of the future. But someone who prefers satisfaction in a later period will never consume. Does this mean then that since I am not eating breakfast right now that I will never eat breakfast? Where does Mises get this idea that people always want to consume sooner and that later means never?
Then Mises notes an apparent exception to his time preference rule:
There are enjoyments which cannot be had at the same time. A man cannot on the same evening attend performances of Carmen and of Hamlet. In buying a ticket he must choose between the two performances. If tickets to both theaters for the same evening are presented to him as a gift, he must likewise choose. He may think with regard to the ticket which he refuses: “I don’t care for it just now,” or “If only it had been later.” However, this does not mean that he prefers future goods to present goods. He does not have to choose between future goods and present goods. He must choose between two enjoyments both of which he cannot have together. This is the dilemma in every instance of choosing. In the present state of his affairs he may prefer Hamlet to Carmen. The different conditions of a later date may possibly result in another decision [emphasis mine].
It seems to me that this exception would apply at all times (and which the bold sentence seems to admit). One can never do everything at once. We must always choose between various forms of satisfaction. One might think Mises is differentiating between satisfaction in general and dissatisfaction (which would mean all he is saying is that people prefer satisfaction to dissatisfaction). But this, he doesn’t seem to be doing either. For he says that the act of provisioning (and with it I assume production and saving) is itself satisfying:
If action is primarily directed toward the improvement of other people’s conditions and is therefore commonly called altruistic, the uneasiness the actor wants to remove is his own present dissatisfaction with the expected state of other people’s affairs in various periods of the future. In taking care of other people he aims at alleviating his own dissatisfaction.
It is therefore not surprising that acting man often is intent upon prolonging the period of provision beyond the expected duration of his own life.
Therefore, all action is satisfying and one is never choosing between present and future satisfaction. One is always choosing between counterfactuals of the same period of time. The correctness of this choice brings satisfaction.
What then is an example of this supposed preference of nearer satisfaction? Mises provides only one example, one that once again clashes with his other principles:
Those contesting the universal validity of time preference fail to explain why a man does not always invest a sum of 100 dollars available today, although these 100 dollars would increase to 104 dollars within a year’s time. It is obvious that this man in consuming this sum today is determined by a judgment of value which values 100 present dollars higher than 104 dollars available a year later. But even in case he chooses to invest these 100 dollars, the meaning is not that he prefers satisfaction in a later period to that of today. It means that he values 100 dollars today less than 104 dollars a year later. Every penny spent today is, precisely under the conditions of a capitalist economy in which institutions make it possible to invest even the smallest sums, a proof of the higher valuation of present satisfaction as compared with later satisfaction.
First, he seems to forget that he doesn’t consider having money (or even spending it) to be a form of consumption—that is, a direct means of satisfaction. The fact that I don’t invest $100 doesn’t necessarily say anything about my desire for present consumption. Second, there is no reason to presume that investing my money is the best way to maximize my satisfaction in a year’s time. Perhaps spending the money at a bar to develop friendships, or going out on a date with a potential romantic partner will lead to greater satisfaction in a year’s time.
A more concrete example is found in Daniel James Sanchez’s article “Of Time and Marshmallows.” Sanchez mentions an article in the New Yorker about a study conducted with children. The children are given a choice between having one marshmallow now or having two marshmallows after waiting an indefinite time (about 15 minutes it turns out) for the experimenter to return. Most children apparently can’t resist and eat the marshmallow within the first 3 minutes. Sanchez concludes that this is an example of how we prefer present satisfaction over future satisfaction. The problem with this conclusion is that it views time within a vacuum—as if we can simply add time to something without changing what is being chosen. The children are not merely choosing between one marshmallow now and two later. They’re choosing between one marshmallow or two marshmallows plus waiting 15 minutes (and for all they know it could be longer) by themselves in a boring room. They might be choosing between sitting in a room and playing outside. One might think a better way to conduct the experiment would be to measure the preference of spending mundane time before a supposed satisfaction to spending the mundane time after the satisfaction. One could, for example, allow people the choice of either waiting in line for an hour before they go on a roller coaster to waiting in line for an hour after they get off of the roller coaster. But even this can’t measure what it claims to because any period of waiting occurs between supposed periods of satisfaction, and thus the contrary proposition that people always prefer future satisfaction would fit the data just as well.
The fact is, we can only do one thing at a time, and at all times, we must be doing something. Therefore, anytime we choose to do something now, we are also choosing to defer something we could be doing to a later time. This is not to say that time preference can’t take the form of a pathology. Procrastination is certainly a real phenomenon, but no one procrastinates 100% of the time.
Sanchez extends the marshmallow example into a hypothetical scenario involving interpersonal exchange:
Time preference plays just as fundamental a role in interpersonal exchanges as it does in “autistic” exchanges. Let us imagine Craig and Carolyn as parties in a potential exchange. Craig has no marshmallows, but is expecting to get two from his mother in 15 minutes. Carolyn has one marshmallow. They both consider the following potential exchange: Craig “borrows” Carolyn’s marshmallow to eat now in exchange for “paying her back” with the two marshmallows he’ll get in 15 minutes. This potential exchange is precisely analogous to the “autistic” exchange of the Stanford experiment, except for all the additional considerations that are wrapped up in relations with others (wanting to befriend/please/not please/etc., the other person).
This is a good illustration of how Austrian economists err in equivocating between autistic and interpersonal exchange. The error here is that autistic exchange is always about using while interpersonal exchange is always about owning. This leads to two conceptions of consumption: one in which an object is used up and another in which an object is removed from circulation and from producing anything which can be circulated. Owning in fact does not refer to the relationship between a person and an object, but between a person and another person. A person who owns an object has the power to forbid another from using it. When Sanchez moves from the concept of autistic exchange to that of interpersonal exchange, he also moves from the concept of using to the concept of owning but does not acknowledge it. Craig and Carolyn are not in fact negotiating between when each will eat (use) the marshmallows but when each will own them. (Also note how the goods aren’t produced by either of the participants, nor do they give up anything they produced for them. They simply have them coming as a gift in the future—just like in the prison economy.)
To make this point clearer, let’s first up the count of marshmallows. Carolyn still only has 1 marshmallow but Craig has 20 marshmallows coming. However, they aren’t coming until 5 hours. Assuming the same interest rate compounded every 15 minutes, Craig now has to give 20 “future” marshmallows for Carolyn’s present marshmallow. Now let’s reverse the number. Carolyn has 20 present marshmallows and Craig has no marshmallows but 1 coming in 5 hours. In order for Carolyn to exchange the 20 present marshmallows for Craig’s 1 future marshmallow, a negative interest rate must be in play. From the perspective of owning, this is clearly absurd and unrealistic. “All else being equal,” Mises would say, one would prefer to have more now than less later. Thus, Mises concludes that one prefers the satisfaction of wants in the present to those in the future. But wait, when he refers to the satisfaction of wants, doesn’t he mean use and not ownership? What is the use of a marshmallow? Surely it is to eat it. So let’s reconfigure the scenario.
There are 20 marshmallows which Carolyn must eat now unless she negotiates with Craig. There is 1 marshmallow which Craig must eat in 5 hours unless he negotiates with Carolyn. Now the tendency towards a positive interest rate disappears entirely. While most would prefer to own 20 marshmallows as opposed to 1, most would rather eat 1 at a given moment than eat 20. Even if we change it to something more agreeable, like French fries, it is still no more likely that one will prefer 20 French fries now to 1 in 5 hours than 1 French fry now to 20 in 5 hours. Ownership is the power to determine use and not use itself. In the original scenario, Carolyn could eat 1 marshmallow now or eat that same marshmallow after 15 minutes. Craig could only eat his two marshmallows after 15 minutes. Thus, the reason the interest rate is always positive is not that we always prefer to use something now as opposed to later, but because we prefer to have the ability to use something now or later to just the ability to use something later. This becomes even clearer when we talk about money, which is of course what we are actually trying to explain here. Money is one of the only things—perhaps the only thing—that is completely imperishable. One dollar received now and saved “under the mattress” buys exactly the same things in 5 years that a dollar received after those same 5 years does. Though perhaps incomplete itself, the Keynesian concept of liquidity preference captures this reality much better than the Austrian concept of pure time preference. The other thing we have to consider is the fact that in a capitalist economy, money is capable of turning over through the production process. This itself allows for the possibility of interest (more on the importance of turnover in a future appendix).
This insight has a range of profound implications. We can no longer see the lender as making a sacrifice of present use. The fact that I’d prefer not to eat my 20 marshmallows right now does not infringe upon my ability to lend them at interest. I’m in fact being rewarded for something I’d prefer not to do in the first place. What the interest rate really does is compensate me for my loss of social power—my power to control others through the regulation of the use of objects. Second, the borrowers do not necessarily decrease the amount of future use-values by their spending. Housing, transportation, food, healthcare, education—in short, what those with “high time preferences” spend most of their money on—may actually be the best way to maximize one’s future use-values. For the average worker, $5000 spent on a car to get her to a job will result in much more future satisfaction than $5000 put in the bank. Yet the difference in time preference is supposed to be the explanation for the inequality of wealth.
So how did Mises get to this theory of time preference? Let’s return to his faithful disciple Rothbard, who has the merit of revealing Mises’s mistakes in more explicit fashion:
A fundamental and constant truth about human action is that man prefers his end to be achieved in the shortest possible time. Given the specific satisfaction, the sooner it arrives, the better. This results from the fact that time is always scarce, and a means to be economized. (Man, Economy, and State, ch. 1)
And there we have it. Time preference is derived from a concept we have already debunked as nonsense. Rothbard’s view is that we have a certain amount of time that we are “spending.” And if we don’t achieve our end right away, we are wasting scarce time. But this is only true if we accept Rothbard’s conception of wants conceived outside of time. In a way, this is simultaneously a rejection of the truth of time preference, namely that people prefer different things at different times. It frames things as if we want these particular things whenever, and it’s just a matter of economizing our time so that we can get them as quickly as we can. The falsity of such a conception should be clear to anyone who is willing to think about it critically.