temporal discounting

n. Temporal discounting (also called delay discounting) is the tendency for behavioral decisions to be relatively uninfluenced by consequences that are deferred in time. All potential courses of action have consequences that are beneficial (rewards) or detrimental (costs), and their subjective value fuels choices among competing courses of action. Delay reduces the subjective value of a consequence, resulting in reduced impact on current action.

For example, a person with cash in hand may choose to save it or spend it now. Saving may generate sizable deferred benefits (e.g., retirement income), while an extravagant purchase today generates immediate but lesser pleasures. Temporal discounting renders this choice between large and small rewards difficult, because the psychological impact of the former is diluted under delay. This is true even for persons who can describe the value of the delayed consequences (e.g., how important it is to be well funded in old age) because the factors that affect choice are not necessarily the ones that drive awareness.

Two features are central to problems associated with temporal discounting. First, many situations require choice between one course of action with smaller-sooner consequences and another with larger-later consequences. Second, unit for unit, immediate consequences, which are relatively unaffected by discounting, have greater impact on choice than delayed ones.

When action prevails in service of shortterm consequences, we say that an individual has behaved impulsively or shown a failure of the will. When action prevails in service of deferred consequences, we say that the individual has exhibited self-control. Yet all individuals act impulsively at times, in part because larger rewards are discounted proportionally more than smaller ones. For reasons that are beyond the scope of the present discussion, this effect has two implications. First, it exacerbates the subjective superiority of smaller-sooner consequences. Second, it assures that the temptation of small, immediate rewards is most pronounced when they are imminent. In the latter case, a decision to break one's diet is more likely when the dessert cart is rolled past the restaurant table than during the half-hour drive to the restaurant prior to dinner, even though the longterm benefits of healthy eating remain the same in the two cases.

Individuals differ in the extent to which they discount delayed outcomes. Children discount more than do adults, and some cultural differences have been detected (e.g., Americans were found to discount more than natives of China or Japan). Overall, research suggests a population distribution that is skewed away from extreme impulsiveness, with those at the impulsive tail of the distribution at risk for a variety of clinical and social problems. Especially strong discounting of delayed rewards has been found in academic procrastinators, pathological gamblers, substance abusers, and persons with attention deficit–hyperactivity disorder – all classes of individuals who are regarded as excessively impulsive. Additionally, consistently with the notion that good social interactions require some tolerance for temporary imbalances in social exchange, individuals who deeply discount delayed rewards do poorly in some social situations.

Temporal discounting is described by a hyperbolic function that approximates the form of \(SV = A/(1 + kD)\). The subjective value (\(SV\)) of a consequence is its objective amount (\(A\)) or size discounted as a function of the length of the delay (\(D\)) until the consequences occur. This function is negatively decelerating, meaning that subjective value decreases quickly under brief delays and then changes relatively little as delays become more extended. That is, even brief delays have pronounced discounting effects. The fitted parameter \(k\) is a measure of how quickly subjective value decreases under delay, with larger values implying greater impulsiveness. Thus, \(k\) differs across groups of individuals (e.g., substance abusers vs. normal controls) and also may vary across circumstances.

Two common situational influences on temporal discounting have been observed. First, rewards are discounted more deeply than equal-sized costs – that is, their effects are weaker in the face of delay – providing one possible basis for the well-known tendency of unpleasant events to be more psychologically potent than pleasant ones. Second, the same person may show different degrees of temporal discounting for consequences of different types. One individual may, for instance, deeply discount money rewards but not outcomes related to good health; another may show the opposite pattern. Thus, to speak of an individual as impulsive or not oversimplifies matters. The cause of domain effects is not known, but there is some evidence that experience can affect discounting tendencies. For example, discounting of money rewards appears to change with rates of inflation in the local economy, which is a measure of how quickly money loses its value over time. Perhaps analogously, substance abusers discount outcomes related to obtaining drugs especially steeply. In this latter case, it is possible that experience with drug taking (e.g., exposure to withdrawal symptoms) creates this tendency, although it is also possible that individuals already equipped with the tendency are simply at special risk of substance abuse problems.

A temporal discounting analysis predicts several hedges against impulsive choice. Here is one illustration. Precommitment consists of locking in a course of action before the temptation of smaller-sooner outcomes becomes imminent. Compare two dieters shopping for food. Shopper A purchases food in person at the market, while Shopper B orders groceries through an online service that delivers the next day. Both must choose between items that taste good now (smaller-sooner rewards) but adversely affect weight and items that taste worse but promote healthy weight (larger-later rewards). Both shoppers can eat items as soon as they are purchased. For Shopper A, this means that buying occurs when the power of smaller-sooner rewards to tempt is near its undelayed, undiscounted maximum. Shopper B, however, benefits from the fact that all consequences are discounted under delay. The delay to delivery may reduce the subjective value of fattening foods to the point where better items are chosen for purchase, and without fattening foods in the pantry, Shopper B is likely to eat better. – TSC