The role of time preference in human action can be a difficult subject to grasp correctly. This essay will seek to resolve some common misunderstandings that are essential before one can consider the full implications of the concept in economics. First of all we shall attempt to correct a few particular errors or myths before explaining the true, praxeological foundations of time preference.

Classes of Goods

The first misunderstanding we must address is that the concept of time preference is nearly always expressed with the statement “present goods are more valuable than future goods”. However such a formulation is only shorthand at the very best as it violates some well accepted and understood truths with which “Austrians” are well acquainted and have no difficulty in applying to other concepts. Humans do not have any relation at all to whole categories of goods in their physical embodiment – all of the gold, all of the iron, all of the bread in the world and so on. Rather, humans only act in relation to specific quantities, or units, of goods in order to meet their ends and it is these specific quantities to which value is imputed. Hence the so-called paradox of value – i.e. why a diamond, a seemingly trivial ornate luxury, is more expensive than a bottle of water, which is essential for life – was solved after having confounded the classical economists. The categories “present goods” and “future goods” are precisely this kind of holistic, indiscrete and meaningless concept that has no relevance to action. No human ever acts in relation to all of the present goods in the world, nor to all of the future goods. Rather, we have to examine the precise circumstances in action from which this shorthand derives.

Present Ends and Future Ends

Secondly we must realise that an understanding of time preference cannot come about from any comparison of present ends with future ends, that is, ends that must be met now compared to ends that must be met at some point in the future. Economic laws are only true when they conform to the ceteris paribus rule – that all else is equal. In understanding an economic phenomenon, it is necessary to hold all independent variables constant and to alter only the dependent variable under examination. With time preference, the variable under examination is goods, the means used to extinguish an end, and more specifically the time at which they become available. In testing this variable and making alterations to whether a good takes effect in the present or the future, the end itself, another variable, must remain constant. To talk of present ends and future ends in trying to understand time preference, however, is to make an alteration to a variable other than the one that is under examination. It is to change both the nature of the good and the nature of the end simultaneously, the equivalent of trying to understand the effects of an increase in the quantity demanded while also varying the quantity of supply. If demand was to rise beyond the valuation of the marginal buyer yet supply was to rise beyond the valuation of the marginal seller at an equal rate then price would, all else being equal, remain constant. One would derive from this the conclusion that an increase in demand has no effect upon price, which is clearly incorrect. With time preference, therefore, the examination is to determine the difference between the ability of a present good and a future good to satisfy the same end.

To elaborate on this point, a human has needs that arise at different times, some in the present and some in the future, depending on the length of his period of provision. He may, for example, realise that he needs to satisfy his hunger not only today but also tomorrow, the next day, next week and so on. However, humans themselves exist only in the present and all decisions, choices and actions must be made in the present – not tomorrow, not next week and not next year – and the ends to which they strive must all be ends that exist now. Simply because a need takes effect in the future and may be described as a “future need” does not mean, praxeologically, that it is a future end – end being a category of action that can exist only in the present. Therefore all ends that are sought after must take a place in a human’s rank of values now, and the urgency of their satisfaction will be determined by that rank. For example, I may know that I need to satisfy my hunger today and also that I will have to satisfy my hunger tomorrow. I have two loaves of bread now, one of which I devote to satisfying my hunger now so I eat it now; the other I direct towards the end of satisfying my hunger tomorrow so I store it in a bread bin. Or, in place of the latter, I may arrange to acquire a second loaf of bread tomorrow rather than having one available immediately. However one of these ends is not a present end of satisfying my hunger now and the other a future end of satisfying my hunger tomorrow. I can only make choices and decisions that lead to actions now, in the present, as I do not exist in the future. Therefore all ends must be expressed as present ends. The two ends are, therefore, correctly described as follows: the end of satisfying my hunger now; and the end of providing for the satisfaction of my hunger tomorrow. For the first end, the relevant action is eating the first loaf of bread today. For the second, it is directing the second loaf into the bread bin for storage (or arranging for the acquisition of the second loaf tomorrow). Both ends are therefore present ends met through present actions and if the second end is sufficiently high in my value rankings then it will need to be fulfilled now also and the stored loaf bread, or the expected acquisition of a second loaf of bread, is fulfilling this end now. Crucially, however, the importance that each end may have could be higher or lower than the other. There is no necessity for the second loaf of bread, simply because it will feed me tomorrow, to be less valuable than the first. If I am desperately hungry today then the first end, satisfying my hunger today, may be very high on my rank of values and the second end may be low. Alternatively, if I believe that tomorrow will bring excruciating hardship then the end of providing for tomorrow might be the highest end and the one with which I will be preoccupied. Solely because one end concerns the present and the other the future does not automatically mean that the end concerning the future is a less valuable and provides any explanation of time preference. And there is, consequently, no necessity for the second loaf of bread to be ranked lower in value than the first. Indeed, if providing for tomorrow was the more important end then if one loaf of bread was to vanish this loss would be shifted to the least valuable end – hence I would go hungry today and use the remaining loaf to eat tomorrow.

This analysis explains why, at any present moment in time, a set of fireworks for July 4th may be more valuable than the same set for May 4th; or why ice cream in winter is less valuable than ice cream the following summer; or why someone may engage in plain saving without any expectation of interest. Indeed it is quite conceivable that someone on May 4th would exchange a set of fireworks in return for acquiring the very same set (or even a set with a lower quantity or quality) back on July 4th. The understanding of time preference does not come from situations where the goods are available either now or in the future and where the ends also take effect at varying points of time also. Rather, it comes from those situations where the ends must be met now but where the goods are available at different points in time. In short, we are comparing the ability of a good available today with a good available at a point in the future to satisfy the same end.

Psychology and Physiology

Related to the previous discussion is the fact that psychological and physiological explanations of time preference are not sufficient to establish the necessary truth of the phenomenon. The notion that people may underestimate their future needs, that they may care less about the future than the present, or that their aging bodies will simply be less capable of enjoying satisfaction in the future may all be true but they needn’t necessarily be so. Further, much of this would again be varying the end rather than the type of good. Moreover as we shall see further below, the fact of uncertainty is not sufficient to explain time preference either. Rather, our investigation will concern why time preference arises praxeologically. In other words, what is it about action that causes the law of time preference to arise as a necessary result?

Goods and Serviceability

A step forward towards understanding the difference between a present unit of a good and a future unit of the same good is the difference between their serviceability. All goods derive their value from the ends that they service. Ends are ranked in order of urgency, that is a human will devote goods to fulfilling his most highly valued end first, the second highest next, and so on. As goods to fulfil ends are always scarce, any devotion of a good to one end involves the foregoing of other ends. Where goods can be devoted to either end A or to end B, for example, B will be foregone if the value of attaining A with the goods is ranked higher. Where a particular good is able to accomplish the fulfilment of an end alone (or in combination with very few other goods – there will always, at the very least, be labour) we can derive two things. First, as the good will be sharing its service towards the fulfilment of an end with very few other goods, close to the full value of the end will be imputed to the good. Secondly, because so few other goods have to be used to fulfil the end then there are more goods to be devoted to other ends, hence there are fewer ends that need to be foregone in the pursuit of this, most urgent end. Hence this latter end will be relatively more highly valued. Let’s say, for example, that there are five ends, A, B, C, D, and E, and that there are five goods a’, b’, c’, d’ and e’ to service these ends. If good a’ can service end A without any use of the remaining goods then this leaves all of these goods to service ends B-E. Not only will good a’ be accorded the full value of end A, but the relative value of end A and compared to ends B-E is high. We may say, in this instance, that the good possesses a high degree of serviceability. Where, however, a good requires a higher number of complementary goods to fulfil an end then a lower value will be imputed to that particular good as the full value of the end must now be imputed to a greater number of goods; furthermore, the necessary devotion of more goods towards fulfilling the end will mean that a greater number of other ends will have to be foregone. For example, if good a’ was not able to fulfil end A alone but, rather, needed to act in concert with goods b’-e’, then all of the ends B-E would have to be foregone in the pursuit of end A. While end A may be the highest individually valued end, losing all of these other ends will serve to reduce its relative value and, indeed, the cost may be so great that end A will simply be abandoned.

Let us examine this first of all by exploring an analogy to time, which is distance. Let us say that I strive towards the end of quenching my thirst and that this is my most highly valued end so that I want to act to fulfil it immediately. If I have a bottle of water right next to me that will satisfy this end then, ignoring the cost of labour, the value of the bottle of water will equate to that of the end itself1. The bottle of water has served to fulfil this end with a high degree of serviceability as it has not required the use of any other goods in order to accomplish its task. This means that more goods are left over for the fulfilment of other ends. So let us then say that, as I have easily fulfilled that end, I have a second end of going to pick apples for the day. I then, having had my first end fulfilled, can proceed merrily with the fulfilment of my second with the remaining stock of goods available. And having proceeded with this second end I may have more goods left over for the pursuit of a third end of baking bread. However, what if, in a second scenario, I still desire the same end of quenching my thirst but now the bottle is not right next to me but is ten miles away? This bottle is the same, physically homogenous resource as the bottle that was right next to me but if the distance of ten miles makes, in my mind, an appreciable difference what now is the value of the bottle? The distance means that an appreciable cost must be borne in order to utilise the bottle, costs that are not shared by the utilisation of this bottle in scenario one, rendering the bottle in the second scenario with a lesser degree of serviceability. These costs, clearly, are those that must be borne in order to transport the bottle to me or me to the bottle. Because of this necessity of transportation, complementary goods must now be brought in order to service the end. But these goods were goods that could have been devoted to ends other than quenching my thirst – namely, picking apples and baking bread. The lower serviceability of the bottle means that, in order to utilise it, additional ends to which means could have been devoted now have to be foregone. From this we can derive two conclusions. First, the degree of remoteness caused by distance means that the bottle in scenario two must share its fulfilment of the end with a greater number of goods compared to the bottle in scenario one. The lower capability of the distant bottle in scenario two means that the value of the end of quenching my thirst must be imputed to a greater number of goods2. The value of the bottle in scenario two, therefore, must be discounted accordingly. Secondly, the loss of the other ends – picking apples and baking bread – serves to impose a relatively lower value on the end of quenching my thirst. If this loss becomes too great – i.e that I am not prepared to forego the loss of picking apples and baking bread in order to quench my thirst – then the then the latter end will simply be abandoned and the bottle will cease to have value (or it may be earmarked for a lower valued end to which it may be more suited). In either case in scenario two – whether I proceed to bring the distant bottle to me or I abandon the end of quenching my thirst entirely – the value of the distant bottle in scenario two is lower than that of the bottle right next to me in scenario one.

It is this kind of understanding that is the foundation of an explanation for the phenomenon of time preference – a present unit of a good has greater serviceability in satisfying an end than a future unit of the same good. We will now explore this in detail.

Time and Serviceability

Although analogous, the remoteness of time presents a challenge more difficult than that of distance and there are some important differences. Whereas with distance, the lower value of the distant good could be explained by the option of foregoing lesser valued ends in order to overcome it, an acting human does not necessarily have this luxury with time. Nothing can be done to “speed up” time and its passage must be borne at a constant rate. We therefore have to look to the particulars of action that we touched upon earlier to explain why “remoteness” in time causes an otherwise equally serviceable unit of a good to have lower value.

An action is the result of a choice to satisfy ends with means available. But as we noted above human exists only in the present and must live through the present before the future arrives. A person cannot act in the future; he has to do so in the present. All decisions are therefore present decisions to act towards present means towards present ends. In other words, the very fact that a human acts at all means that he wants an end to be extinguished now or soon, not in the future or later – to act always means to meet an end sooner rather than later. The contrary position – to seek satisfaction in the future – is antithetical to action for if a person desires to meet an end later rather than sooner then he would never act. The present could pass without action but as soon as the later period of time came around it would itself then become the present and the person would be faced with the same conundrum – he would, at that moment, either have to act (in which case he would revert to preferring satisfaction sooner rather than later) or delay action again, in which case he would never act. The logic of action therefore requires sooner satisfaction rather than later. Indeed, even where the action concerned may not bring satisfaction for a long period of time, to begin the action is to demonstrate a preference for the satisfaction of the end to be brought closer in time. It follows also that the end to which action is directed first must be the one that is, in the eyes of the acting human, in the most urgent need of fulfilment, i.e. it is the highest valued end.

What does this mean for the value of a present unit versus the value of a future unit of a good? All goods, as we know, derive their value from the ends that they satisfy. If a human acts now in relation to a good – say a bottle of water – in order to achieve the end of extinguishing his thirst it means that, now, at this moment, this end is his most highly valued end and the good must be accorded (in the absence of other appreciable costs) the same value as the end. To act now means that this end must be fulfilled now, or at least brought closer in time to fulfilment. However, if we take the same moment in time – the present – but remove the good from present availability and move it to a future availability then what does this entail for action? It means that the most highly valued end at that moment cannot be fulfilled by that good. It completely lacks any serviceability towards this end compared to the serviceability of the presently available good. One of several things may happen as a result. If the end is to be satisfied now, substitute present goods must be found. These, however, must be drawn from the satisfaction of other ends and the urgency of these ends must be reweighed against the urgency of satisfying the human’s thirst in light of the fact that the present bottle of water is no longer available. It is quite conceivable that the end would be either abandoned entirely or satisfaction of it would be delayed – in either case it necessarily ceases to be the most valuable end. As other ends now become the object of action so they become more valuable and hence, the future good reduces in value accordingly3. Furthermore, if the end is either abandoned or satisfied by substitutes, the future bottle of water may be earmarked for a lesser valued end such as providing for tomorrow’s thirst – the end being necessarily lesser not because it takes effect in the future but because it is not the most valuable end to be met at the moment when quenching my thirst is most pressing, the very moment when the relevant valuation under scrutiny is occurring.  In all of these cases – substitution, abandonment, delay and direction of the good to a lower valued end – the future bottle of water derives a lower value than the present bottle of water. It is these facts, arising from the logic of action, that is the cause of the phenomenon of time preference, the future bottle being imputed with a discount to reflect its lower utility. We can therefore state the law of time preference as being as follows: a unit of a good that is available to satisfy an end immediately (or sooner) will be more valuable than a unit of a good that can only satisfy the same end in the future (or later).

We can also understand from this why there are gradations of serviceability of future goods – for example, a present unit of a good may be more valuable than a unit available one year from now, a unit one year from now more valuable than a unit two years from now, a unit available in two years more than one in three, and so on. For if the logic of action is to bring ends closer to their satisfaction the nearer in time a good is to that satisfaction the lighter will be the discount applied. If, for instance, a person chooses to delay satisfaction, then the lower that satisfaction will slip down the rankings the longer it must remain unfulfilled, as the cause of that delay is, by necessity, a decision to devote action to other, more highly valued ends in the meantime. The very fact of delay implies a lower value as to act is to place a higher valuation on the object of action now and to seek satisfaction now or sooner where as to not act or delay action is the precise opposite. From this we can also understand the capitalised value of durable goods – why, for instance, uses that are delivered in future slices of time incur a heavier discount the further they stretch into the future. For, at the moment of valuation, each separate use of the durable good must seek out its ability to fulfil an ever diminishing pool of ends that a human holds, each end reducing in value until they are dissipated. Hence the reason why land that is, for all intents and purposes, a permanent good that can yield utility for all eternity, trades for a finite price – to the extent that the remotest future uses can fulfil any end the human holds at all they will be of such infinitely small value as to be negligible.

What if a person deliberately and constantly decides not to act? Do we not here have a definitive example of where a person can persistently prefer future satisfaction? Not at all. To not act is itself an action that must have an end to fulfil. If so, whatever end this may be – peaceful meditation, reflection, or the strength gained through the bearing of hardship – it is more important than the end that some other present good could satisfy. To continue delaying, for example, the quenching of my thirst by not opening a bottle of water doesn’t mean that I prefer a future bottle of water to the present bottle of water. It simply means that not drinking is more valuable than drinking. As soon as, however, drinking becomes my most valuable end it would be the case that the present bottle of water would be more valuable than a future bottle of water in satisfying that end. The situation of choosing not to act therefore has no bearing on the phenomenon of time preference.

Finally, what about the situation where, for example, my most highly valued end is to provide for next week’s hunger and I want to ensure that this is met now, either by storing goods now or by arranging, now, for their acquisition next week? I have an apple available now but it will rot before next week comes and will not fulfil this end. An apple that becomes available next week however, will not be rotten and will fulfil the end. Surely, therefore, we now have a clear case of where a future unit of the same good is able to better satisfy the same end more than a present unit and won’t, in this instance, the future unit be accorded a higher value? Unfortunately not, because the fact that the present apple will rot imposes upon it a qualitative difference from the apple that will not. In other words, an apple that is rotten before the end is fulfilled is not the same good as an apple that is not rotten before the end is fulfilled. We are therefore altering a variable other than the one under examination and hence we can conclude nothing about the latter from such a situation.

Human Appreciation of Time

It must be emphasised that the difference in the elapse of time between the availability of a present unit of a good and a future unit is determined praxeologically. All actions do, of course, take place through time and all goods are remote in time to different degrees. If I decide to drink a bottle of water I first of all have to pick it up, open it and then bring it to my mouth, all of which has to occur through time. But in order to have any relevance in economics the difference has to be appreciated by the human – there has to be a conscious awareness of its passage. With the opening of the bottle all of the actions may happen so quickly that, in my mind, they are praxeologically simultaneous and I therefore impute no lower value to the unopened bottle sitting on the table to the water that I am swallowing and enjoying. On the other hand, the passage of a week before I can drink the water would probably make a lot of difference, especially if I had no other access to water in that time. Further still we can see that £100 received in five minutes will probably not be valued lower than £100 received in this very instant, whereas £100 received in one year’s time would be valued markedly lower. Moreover it should be obvious that it will never occur with units of free goods – a unit of present air is just as valueless as a unit of future air.

Does this fact mean that our analysis of time preference is circular? That we are explaining the fact that humans appreciate time by the fact that humans appreciate time? Not at all, for what we are trying to explain is why a future unit of a good must necessarily be of lower value than a present unit of a good. In other words, using a human’s appreciation of the factor of time as a given, we are concluding from the logic of action that time preference must always be in favour of a present good ahead of a future good. We are not begging the question by reaching this conclusion.

Uncertainty

Time preference has often been explained by the fact that the period of time that elapses between now and the availability of the future unit of the good is fraught with uncertainty – that because the future is always uncertain a person does not know whether the future unit will, in fact, become serviceable and hence this risk possibly serves to discount the utility of the future good. This uncertainty has two sources – a) uncertain future circumstances; b) the uncertainty of the future good actually becoming available. While it is true that uncertainty pervades all human action and that, generally, the longer the period of time that must elapse before an action is complete the greater the uncertainty, it is not in and of itself the cause of time preference. Even if uncertainty was reduced to the point of negligibility, to act now would still mean to prefer satisfaction now rather than later. A good that becomes available in the future must still either be the cause of the delay of satisfaction of the end, or, in the event that the end is satisfied with substitute goods, seek to fulfil a lower valued end or not end at all. In all cases the value of the future good would diminish.

This does not mean that uncertainty is redundant in a complete understanding of time preference; the height of uncertainty could certainly affect the rate of a person’s time preference as it imposes a psychic cost on a human which will affect the valuation of either the delayed end or the new end which a future good could satisfy. In other words, the fact of uncertainty would cause these ends to diminish further in value at the present moment in time, this further reduction being imputed back to the future good. But so too could total certainty serve to increase time preference. If, for example, it was certain that the world would be destroyed tomorrow time preference, far from falling as a result of the certain future, would rise to an astronomical height, with a heavy discount applying to goods that may become available as little as an hour into the future. On the other hand, if there was only a reduced chance of the world being destroyed the discount might be a little lighter. The effects of uncertainty are not therefore uniform upon the phenomenon of time preference and as an explanation of its ultimate cause it is neither necessary nor sufficient.

Exchange between Present and Future Goods

If what we have concluded above is true, that a unit of a future good must be less valuable than a unit of a present good, in which circumstances would a person exchange a present unit for a future unit? After all, we see this every day, mostly clearly in the lending of money at interest and almost certainly engage in the practice ourselves. What is it that could entice us to regard a future good as more valuable?

The key to understanding this is that, compared to our scenarios above, there must be an alteration to the serviceability of the future good that, in the eyes of the acting human, serves to increase its value above that of the present good. It cannot be the case that the same unit of a good available in the future is more valuable than the same unit available right now. What, therefore, is this alteration in serviceability to the future good? The answer should be familiar to us. Nearly always it is an increase the quantity of the future good while the quantity of the present good remains constant. So with the lending of money, for example, the present good may be £100 but the future good for which is exchanged may be £110. £110 has greater serviceability in terms of quantity compared to the £100, however the £100 has greater serviceability in terms of time compared to the £110. A human has to decide which of these two imbalances is of greater value to him. Typically we say that if he prefers a larger unit of a future good to a smaller unit of a present good he possesses “low time preference”. Conversely, if he prefers a smaller unit of a present good to a larger unit of a future good he is said to have “high time preference”. While this is useful shorthand for determining whether a person will have a propensity to save and invest rather than spend and consume (or indeed, when judging the direction of a society’s economic development), it does not tell us the whole picture. For to express a high or low time preference by trading present goods for future goods is an exchange like any other and a high value attached to the good received in exchange must correspond with a low value attached to the good given up in exchange. If, therefore, someone has a low time preference he must, conversely, have what we may term a relatively high “quantity preference” – the increased quantity of the future good being more valuable to him than the end that must be delayed, abandoned or met through substitutes today in order to receive it. On the other hand, if a person has high time preference he has a relatively low quantity preference, preferring to meet an end now with a smaller quantity of a good rather than delay it, abandon it or meet it through substitutes. We might say, therefore, that time preference and quantity preference are negatively correlated.

The concept of time preference is not necessarily limited to a single, homogenous good. It would, for example, be possible to exchange a quantity of present apples for a quantity of future oranges. In this case, while it would not be possible to determine a “rate” between the two quantities exchanged in the way that we can express an interest rate, we can say that a present apple would fetch in exchange a greater number of present oranges than a future apple. Or, conversely, a present orange could be sold for more present apples than a future orange could. There is also the possibility of a qualitative difference as opposed to a quantitative difference. A present apple may, for example, fetch a quantity of the ripest and most luxuriant present oranges whereas a future apple may only fetch the same quantity of lower grade, bog standard present oranges. All of these possibilities are expressions of the law that a present unit of a good is more valuable than a future unit of the same good.

Conclusion

What we have determined, therefore, is that the common expression “present goods are more valuable than future goods” is, at best useful shorthand that can muddy the waters when determining the fundamental truth. Neither also does an understanding of time preference arise from psychological considerations nor from the fact of uncertainty. Rather it is the logic of action itself that means a present unit of a good must always be more valuable than a future unit of a good when comparing their abilities to satisfy the same end. Only an advantageous change in the serviceability of the future good – such as an increase in its quantity – can serve to render the future good more valuable than the present good.

We have not explored the further implications of time preference in economics – particularly its role in interest and the business cycle, which is of great import to “Austrians”. However, a clear understanding of the fundamentals of the phenomenon should serve to enable one to tackle these difficult questions.

View the video version of this post.

1We are, of course, assuming that the bottle cannot be substituted in the event that it is lost in order to avoid the implications upon value that substitution has.

2Exactly the same would be true if, for example, the bottle was, as in scenario one, right next to me, but is now of an appreciably different quality or quantity (i.e. appreciable to the extent that the end cannot be satisfied to the same degree). Once again its serviceability, its power, as judged by my mind, to extinguish an end is diminished and other goods must be brought in to fully satisfy the end.

3It is of course true that in the case of the possibility of substitution the value of the present bottle of water would equate to that of the substitute goods and not from the end of quenching my thirst but this has no bearing upon our analysis of the relatively lower value of the future good as compared with that of the present good.

Advertisements