American Economic Association

Negative Time Preference Author(s): George Loewenstein and Drazen Prelec Source: The American Economic Review, Vol. 81, No. 2, Papers and Proceedings of the Hundred and Third Annual Meeting of the American Economic Association (May, 1991), pp. 347-352 Published by: American Economic Association Stable URL: http://www.jstor.org/stable/2006883 . Accessed: 22/07/2011 12:37

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=aea. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

American Economic Association is collaborating with JSTOR to digitize, preserve and extend access to The American Economic Review.

http://www.jstor.org Negative Time Preference

By GEORGE LOEWENSTEIN AND DRAZEN PRELEC*

The man who lives within his income reached at the end. Thus a declining series is naturally contented with his situa- of consumption levels ought to be prefer- tion, which, by continual, though small able to an increasing series, holding total accumulations, is growing better and consumption constant. better every day. He is enabled gradu- In this paper, we present a short selection ally to relax, both in the rigour of his our 1990 parsimony and in the severity of his of findings (reported more fully in application; and he feels with double paper) that sharply contradict the normative satisfaction this gradual increase of sequencing rule just described. To most per- ease and enjoyment, from having felt sons, a deteriorating series of levels is before the hardship which attended a rather close approximation to the least the want of them. attractive of all possible patterns, regardless Adam Smith of the nature of events that are being or- The Theoryof Moral Sentiments dered. As a secondary violation of the dis- counted utility model, the preferences of Planning for the future invariably re- many people are not additive. Such additiv- quires one to choose among alternative se- ity violations often reflect a concern for quences of outcomes. Even simple short- spreading utility levels evenly over time that term scheduling decisions about work, play, is not attributable to diminishing marginal chores, vacations, etc., involve choosing be- utility within periods. tween sequences, because events that take up time cannot be rescheduled without I. Sequences vs. Simple Outcomes changing the timing of other activities. Most economic analyses of preferences Several recent studies have documented between temporally spaced sequences rely an apparently negative rate of time prefer- on the discounted utility model, along with ence for choices among outcome sequences. the assumption of positive time preference Loewenstein and N. Sicherman (1991) found and diminishing .1 Barring that a majority of museum visitors preferred any preferential interactions across different increasing wage profiles over those that are time periods, the predictions of this model flat or decline over time (holding total value for determining the optimal sequencing of a constant). Pointing out that the flat and given set of events are simple: Place the declining wage profile could produce a best event at the start, then proceed in dominating consumption stream through a descending order until the worst event is suitable savings program, did not have much impact on preference. C. Varey and D. Kahneman (1990) found that subjects strongly preferred brief sequences of de- *Department of Social and Decision Sciences, creasing discomfort, even at the cost of ex- Carnegie Mellon University, Pittsburgh, PA 15215, and periencing overall greater discomfort, while Harvard Graduate School of Business Administration, Boston, MA 02136, respectively. W. T. Ross and I. Simonson (1990) showed 1The applications of this model have been remark- that people prefer sequences that end on a ably diverse, ranging from savings behavior, educa- good note. tional investment, labor supply, bargaining, to crimi- Preference for improvement is an over- nal behavior and addiction. In recent years, however, driven in its descriptive validity has come under challenge (see determined phenomenon, part our forthcoming article, and Loewenstein and Richard by anticipatory savoring and dread (Loew- Thaler, 1990). enstein, 1987), and in part by loss aver-

347 348 AEA PAPERS AND PROCEEDINGS MAY1991

sion (A. Tversky and Kahneman, 1990) and We anticipated that more people would de- habit forming changes in the adaptation lay the fancy French dinner when it was level (J. Duesenberry, 1949). Savoring and combined in a sequence with the Greek dread contribute to the preference for im- dinner than when it was expressed as a provement because, for gains, improving se- single outcome prospect. This is indeed what quences allow decision makers to savor the happened. Of the 86 percent of subjects best until the end of the sequence. With who preferred the fancy French dinner, 80 losses, getting the worst outcomes over with percent preferred a more immediate dinner quickly eliminates dread. Adaptation and (option C) over a more delayed dinner (op- loss aversion induce preference for im- tion D). However, when the French dinner provement because, over time, people tend was composed into a sequence with the to assimilate to ongoing stimuli and to eval- Greek dinner, a slight majority (57 percent) uate new stimuli relative to their assimila- preferred to have the better dinner come tion level. Thus, changes in, rather than later. Even with single-outcome events there levels of, consumption are the carriers of is some motivation to defer the French din- value. Improving sequences afford a contin- ner-witness the 20 percent of subjects who ual series of positive departures (gains) from opted for the longer delay. However, this one's adaptation level; declining sequences tendency is stronger for sequences than for provide a series of relative losses. individual items. Savoring and dread apply to single out- A similar pattern is observed when "Di- comes as well as to sequences, but assimila- nner at home" is substituted for the Greek tion and loss aversion apply only to se- dinner. Since most people eat dinner at quences. The fact that two motives operate home on most nights anyway, embedding for sequences but only one for simple out- the French dinner in an explicit binary se- comes suggests that the tendency to defer quence does not introduce any real modifi- desirable outcomes will be stronger when cation of the problem, but the subject is those outcomes are embedded in sequences. reminded that the choice is "really" be- Such a pattern is illustrated by a survey tween complete sequences. Like other fram- conducted with undergraduates at Harvard ing effects, such reminders cause prefer- University, who were asked the following ences to shift, in this case in favor of the three questions: improving sequence. 1. Which would you prefer if both were free? II. WhatIs a Sequence? A. Dinner at a fancy French restau- rant (86%) If impatience and the desire for i