Time Preference

BIBLIOGRAPHY

A person with a time preference favors having a good sooner rather than later. As a result, the person also prefers having a good immediately to having a somewhat greater good later. Having a time preference amounts to discounting the value of future goods. It may be a source of improvident behavior later regretted: Because of a time preference, someone may spend next month’s rent money on a party and yet later, when the rent is due, would rather pay the rent than have had the party.

A leading member of the Austrian School of economics, Eugen von Böhm-Bawerk ([1884, 1889, 1921] 1959) appeals to time preference to justify payment of interest. Because people would rather have goods now than have identical goods later, he concludes that people who borrow capital should pay for its use. Refining this view, Irving Fisher (1930) attributes a community’s interest rate, an indication of the community’s preference for a dollar of present income over a dollar of future income, to time preference together with the value of investment opportunity.

Böhm-Bawerk also uses time preference to explain profits from business ventures. An employer pays for labor and other factors of production. Payments occur before products are sold. The payments plus profits equal the products’ price. The profits are the employer’s compensation for postponing consumption of the amount of payments until the products are sold. The prevailing rate of profit depends on the prevailing degree of time preference and is equal to the prevailing rate of interest. Although short-term profits may not equal the rate of interest, longterm profits do. Economists elaborating this view argue that time preference is just a partial explanation of profits. Other relevant factors are the uncertainty of a product’s sale price and the influence of the supply of money on sale price.

Studies in psychology measure time preference. Harrell Chesson and W. Kip Viscusi (2000) infer the temporal discount rate of subjects from their choices. Subjects stated their choices between an amount of money in a fixed number of years and a gamble that pays either that amount of money in a smaller number of years or in a larger number of years. Chesson and Viscusi discovered that the subjects’ choices do not maximize time-discounted expected consumption. More surprisingly, they discovered that some groups have discount rates different from the rates commonly attributed to them. For example, smokers have a lower discount rate than nonsmokers do, and therefore they do not disregard future consequences more than others do. Chesson and Viscusi conjectured that smokers simply have tastes and perceptions of risk different from those of nonsmokers. Also, contrary to the common view that youth is impatient, Chesson and Viscusi found that older subjects have higher discount rates than younger subjects have. Shorter life expectancies may explain older subjects’ greater impatience to receive benefits.

Philip Trostel and Grant Taylor (2001) similarly argue that time preference increases as a person ages. Because of declining abilities, an aging person increasingly prefers consumption now to consumption later. Trostel and Taylor find that this effect is independent of a decline in the probability of survival.

Viscusi and Joel Huber (2006) examined revealed rates of time preference for public goods, such as water quality. They measured a person’s impatience to receive a public good and divided that impatience into a component due to time preference and a component due to the person’s perceived probability of not living long enough to benefit from the public good. According to their data, people display hyperbolic discounting rather than the exponential discounting most theorists assume. According to exponential discounting, if a good arrives in T years and the interest rate is r, then the discount rate for the good is δ^T with δ = 1/(1 + r ). According to quasi-hyperbolic discounting, the form of hyperbolic discounting that Viscusi and Huber investigated, the discount rate is λδ ^T with a parameter λ such that 0 < λ < 1. Their data show a high rate of time preference for immediate improvements and substantially lower rates of time preference for later improvements. This sharp decline in the rate of time preference is inconsistent with exponential discounting, but is consistent with hyperbolic discounting. The rate of time preference also varies among groups of subjects. Subjects who are old especially disliked delays in provision of public goods.

Economists characterize time preference as discounting future goods. Does time preference apply to goods that are dated, for example, dated commodity bundles? Because the value of a commodity bundle with a future date is already discounted for time, discounting the dated bundle for time results in an unwarranted double discount. For generality, philosophers characterize time preference as discounting future desires. A person with a time preference gives less weight now to satisfaction of a future desire than the person will give to satisfaction of the future desire when the desire occurs. The person’s discount applies to reasons for acts rather than to consequences of acts.

Philosophers and economists debate the rationality of time preference. Discounting future goods because of uncertainty is plainly reasonable, so the debate considers primarily the rationality of pure time preference, that is, discounting future goods just because they are future. Robert Strotz (1956) finds that clear-thinking people display pure time preference, and so concludes that pure time preference is rational.

Paul Weirich (1981) takes doing what one knows one will regret as a sign of time preference. Acting that way discounts future desires that generate regret. Foreknowledge of regret eliminates discounting because of uncertainty. Treating only future desires that are rational eliminates another reason for discounting. Suppose that in some cases it is rational to do what one knows one will rationally regret; then it is rational to have a pure time preference. One rationally prefers satisfaction of present desires to satisfaction of future desires just because the present desires are present. Basic goals change over time. Suppose that basic goals now and in the future are similarly rational although they differ. Then one may rationally not care now about satisfying a future rational desire. Suppose that one makes decisions using rational, all-things-considered desires that take account of all relevant reasons, including promotion of future desires’ satisfaction. A decision to perform an act is then rational, even if it leads to foreseen regret because of foreseen changes in basic goals. Doing what one knows one will regret, a manifestation of pure time preference, may be rational.

Derek Parfit reviews time preference’s treatment by Plato, Jeremy Bentham, and David Hume (1984, pp. 158–163). He characterizes time preference as caring less about the further future, and calls it a “bias toward the near.” If two future goods that are identical in features an agent cares about are separated by a period of time, a bias toward the near may lead an agent, relative to the first good, to discount the second good. Suppose that the discount increases as the first good comes nearer, and is highest when the first good arrives. Then the agent will have trouble keeping resolutions. The agent may for receipt of a sum of money agree to forego the first good and wait for the second good when both goods are far off. When the first good is imminent, the agent may pay a larger sum to have the first good rather than wait for the second. The agent’s temporal discounting causes a preference reversal.

Whether an agent is subject to such preference reversals depends on the agent’s type of discounting. If an agent has an exponential discount rate, and discounts the future at the same n percent per unit of time, then relative to the first good, the second good always receives the same discount no matter how near the two goods are. Given an exponential discount rate, no preference reversals arise.

Many philosophers hold that a bias toward the near, which is independent of uncertainty concerning the future, arises from lack of imagination. Because of shortsightedness, an imminent pleasure is more vivid than a remote pleasure. The difference in vividness is a poor reason for a preference, however. A person with a bias for the near, for example, postpones pains at the cost of making them worse. It is irrational to care more about the nearer future just because it is nearer. Although Parfit acknowledges that because of a foreseen change in desires one may do what one knows one will regret (p. 189), he concludes that one should care about one’s self-interest in a temporally neutral way. This neutrality is incompatible with a pure time preference.

SEE ALSO Farsightedness; Interest Rates

BIBLIOGRAPHY

Böhm-Bawerk, Eugen von. [1884, 1889, 1921] 1959. Capital and Interest . Trans. George Hunke and Hans Sennholz. South Holland, IL: Libertarian Press.

Chesson, Harrell, and W. Kip Viscusi. 2000. The Heterogeneity of Time-Risk Tradeoffs. Journal of Behavioral Decision Making 13: 251–258.

Fisher, Irving. 1930. The Theory of Interest . New York: Macmillan.

Parfit, Derek. 1984. Reasons and Persons . Oxford: Clarendon Press.

Strotz, Robert. 1956. Myopia and Inconsistency in Dynamic Utility Maximization. Review of Economic Studies 23: 165–180.

Trostel, Philip, and Grant Taylor. 2001. A Theory of Time Preference. Economic Inquiry 39: 379–395.

Viscusi, W. Kip, and Joel Huber. 2006. Hyperbolic Discounting of Public Goods. Harvard Law and Economics Discussion Paper No. 543. http://ssrn.com/abstract=921425.

Weirich, Paul. 1981. A Bias of Rationality. Australasian Journal of Philosophy 59: 31–37.

Paul Weirich