‘The Framing of Decisions and the Psychology of Choice Amos Tversky; Daniel Kahneman Science, New Series, Vol. 211, No. 4481 (Jan. 30, 1981), 453-458. Stable URL: http flinksjstor.orgsici?sici=0096-8075°% 28 198 10130%293%3A211%9A44819%3CAS3SIATRODATHIE2 0.CO%3B23 ‘Your use of the ISTOR archive indicates your acceptance of ISTOR’s Terms and Conditions of Use, available at fp (fw. jstor orglaboutitersihtml. ISTOR's Terms and Conditions of Use provides, in part, that unless You. have obtained prior permission, you ray not download an entire issue of &joumal or multiple copies of aricies, and You may use content in the ISTOR archive only for your personal, non-commercial use. Each copy of any part of a JSTOR twansmission must contain the same copyright notice that appears on the sercen or lnted page of such transmission. Science is published by American Association for the Advancement of Science. Please contact the publisher for further permissions regaiding the use of this work. Publisher contact information may be obtained at up:forwer stor ong/jourmals/aaas, hia Science ©1981 American Association for the Advancement of Science ISTOR and the ISTOR logo are trademarks of ISTOR, and are Registered in the US. Patent and Trademase Office. For mote information on ISTOR contact jstor-info@umich edu, ©2003 IsTOR up:thrwwjstor orgy ‘Mon Sep 1 11:20:04 2003The Framing of Decisions and the Psychology of Choice Amos Tversky and Daniel Kahneman Explanations and predictions of people's choices, in everyday life as well as in the social sciences, are often found ‘ed on the assumption of human rational: ity. The definition of rationality has been. rmuich debated, nut there is general agree ‘ment that rational choices should satisfy some elementary requirements of con- sisteney and coherence. In this anicle tional choice tequites thatthe preference between options should not reverse with, changes of frame. Because of imperfee- tions of human perception and decision, however, changes of perspective offen reverse the relative apparent size of ob. jects and the relative desirability of op- tions. We have obtained systematic rever: ‘Summary. The psychological principles that govern the perception of decision prab- lems and the evaluation of probablities and outconies produce predictable shits of preference when the same problem is framed mn different ways. Reversals of prefer lence are demonstrated in choices regarding monetary cutcomes, both hypotheticat and real, and in questions pertaining to the loss of human lives. The effects of frames ‘on preferences are compares ta the effects af perspectives an perceptual appear- ‘ance. The dependence of preferences on the formulation of decision problems is a ‘sighifcant concern for the theory of rational choice. we describe decision problems in which people systematically violate the te- quirements of consistency and cober- fence, and we trace these violations to the psychological principles that govern the perception of decision problems and the evaluation of options. A decision problem is defined by the acts or options among which one must cchoose, the possible outcomes or con- sequences of these acts, and the contin~ sgeacies or conditional probabilities that relate outcomes to acts. We use the term “decision frame” to refer to the deci siot-maker's conception ofthe acts, out- comes, and contingencies associated with a particular choice. The frame that a desision-maker adop\s is controlled part- ly by the formulation of the prablem and partly by the norms, habits, and personal characteristics of the decision-maker. Itis often possible to frame a given de> cision problem in more than one way. Alternative frames for a decision grab lem may be compared to alternative per- spectives on a visual scene. Veridical perception requires that the perceived relative height af two neighboring moun- tains, say, should not reverse with ‘changes of vantage point, Similarly, ca- SCIENCE, VOL. 21, 30 JANUARY 196 saly of preference by vatiations in the framing of acts, contingencies, or out comes. These effects have been ob- served in a variety of problems and in the choices of differeat groups af respon eats. Here we present selected ilstra- tions of preference reversals, with data ‘obtained from students at Stanford Uni versity and at the University of British Columbia who answered brief question- ‘naires in a classroom setting. The total ‘number of respondents for each problem fs denoted by N, and the percentage who chose each option is indicated in boeackets. The effect of variations in framing is iusteated in problems 1 and 2. robles [ (N= 152 Imagine thatthe U.S. is preparing for the outbreak of 2n unusual ‘Asian disease, which i= expected to kil 600 people. Two alternative programs to commbat the discase have been propased. Assume that the ekaet sciete estimate of the con seauenees of the programs are a8 allows If Program is adopted, 200 people wil be ‘ved. (72 percent) | Program Bis adopted, there i 19 probabil. “ay that 600 people will be saved, and 213 probability that mo people. will be sved. (8 percent, Which ofthe two programs would you fevae? ‘The majority choice ia this problem is isk averse: the prospect of certainly saving 200 lives is more attractive than a risky prospect of equal expected value, that is, a one-in-three chance of saving 600 lives, ‘A second gtoup of respondents was siven the cover story of problem 1 with a different formulation of the alternative programs, as follows: Problem 2{V = 153) If Programs C is adopted 400 people wil die [22 percent] Te Progeam D is adopted there is U3 peobabili- ity that nobody wil die, and 23 probabil (that 600 people will dc. [98 percent) Which of the two programs would you favor? The majority choice in problem risk taking: the certain death of 400 people is less acceptable than the two-in three chance that 600 will die. The pref- ereaces in problems 1 and 2 illustrate a common pattern: choices involving gains are often risk averse and choices in- volving losses are often risk taking However, i is easy to sce that the two problems are effectively identical. The only difference between them is that the outcomes are described in problem | by the number of lives saved and in problem 2 by the number of lives lost. The change is accompanied by a pronounced shift from risk aversion to risk taking. We hhave observed this reversal ia several groups of respondents, including univer- sity faculty and physicians. Inconsistent responses to problems I and ? arise from the conjunction of a framing effect with contradictory attitudes toward risks in- volving gains and losses. We turn now to an analysis of these attitudes, ‘The Evaluation af Prospects ‘The major theory of decision-making under risk is the expected utllity model. “This model is based on a set of axioms, for example, transitivity of preferences, Which provide etitetia for the rationality of choices. The choices of an individual who conforms to the axioms can be de- scribed in terms of che utilities of various outcomes for that individual. The wility of a risky prospect is equal to the ex- pected utility of its outcomes, obtained bby weighting the utility of each possible ‘outcome by its probability. When faced with « choice, a cational decision-maker will preter the prospect thst offers the highest expected utility (, 2). Dr. Teeny if potest of esvchoony wt Sas {grater Suporte S15, and Be ee ee em ie Eee Aeneas Sk a6 SO7UALALTOOASIEOL.WO CopyiBU O 19H AAAS “ssAs will be illustrated below, people ex ribit patterns of preference which appear ‘incompatible with expected utility theo- ry. We have presented elsewhere @) a descriptive model, called prospect theo- +y, which modifies expected uuity theo- 1 s0 as to accammodate these observa tions, We distinguish two phases in the choice process: an intial phase in. which acts, outcomes, and contingencies are framed, and a subsequent phase of eval- uation (). Far simplicity, we restrict the formal Weaiment ofthe teary to choices involving stated numerical probabilities and quantitative outcames, such as mon- ey, time, oF mumber of lives, Consider a prospest that yields out come x with prabability p, outcome y with probability 4, and the status quo with probability 1 p — 4. According (o prospect theory, there are values v() associated with autcomes, and decision weights #1.) associated with proabill- fies, such that the averall value of the prospect equals rp} v(x) + ma) ¥). A slightly different equation should be ap: plied if all outcomes of a prospect are on the same side of the zero point (3) In prospect theory, outcomes are ex- pressed as positive or negative devia tions (gains oF losses) from a neutral tet ‘erence outcome, which is assigned a val- tue of zero, Although sulnective values differ among individuals and attributes, we propose that the value function is ‘commonly S-shaped, concave above the reference point and convex below it, a8 ‘tlystrated in Fig, 1. For example, the if ference in subjective value between agains of $10 and $20 is grester than the subjective difference between gains of S110 and $120, The same relation be- ween value differences holds for the corresponding losses. Another property ‘of the value funetion is that the response to losses is more extreme than the fe sponse to gains, The displeasure associ ated with losing a suc. of money is gener ally greater than the pleasure associated ‘with winning the same amount, as is re ected in people's reluctance ta accept fir bets on a toss of a coin, Several stud ies of decision «2, 6) and judgment (73 have confirmed these properties of the valve function (8). ‘The second major departure of pros: pect theory from the expected tility ‘adel involves the treatment af prabe. bilities. In expected utility theory the uuiity af an uncertain outcome is weight: ed by its probability, in prospect theory the value of an uncertain outcome is maul- tiplied by a decision weight (2), which is a monotonic function of p but is not a probability. The weighting function = 46 vale Losses Gales Fig. 1. A typochetieal value funetion. has the following properties. First, im- possible events are discarded, that is, (0) = 0, and the seale is normalized so that 7(1) = 1, but the function is not well behaved near the endpoints, Second, for low probabilities m(p) > p, but mip) + ml —p) = 1. Thus low groba- bilities are overweighted, moderate and high probabilities are underweighted and the latter effect is more pranaunced than the former. Thivd, mipg¥a(p) < ‘miparitaipr) for all 0 < p.q.r = 1. That is, for any fixed probability ratio q, the ratio of decision weights is closer to tunity when the probabilities are low than when they ate high, for example, se(.Ia.2) > a(,8)¢0(.8). A hypothetical ‘weighting function which satisfies these properties is shown in Fig. 2. The major ‘qualitative properties of decision weights can be extended 0 eases in which the probabililies of outcomes are subjective ly assessed rather than explicitly given. Tn these situations, however, decision ‘weights may also be affected by other ctaracteristies of an event, such as am- Diguity or vagueness (9) Prospect theory, and the scales illus (rated in Figs. 1 and 2, should be viewed san approximate, incomplete, and sim. plified description of the evaluation of risky prospects, Although the properties ‘of and 7 summarize a common pattern, of choice, they are not universal: the preferences of some individuals are not well deseribed by an S-shaped value function and a consistent set of decision Weights. The simultaneous measurement ‘of values and decision weights involves serious experimental and statistical diffi- culties 10). If and » were linear throughout, the preference order hetween aptions would be independent of the framing of acts, ‘outcomes, of contingencies. Because of the characteristic nonlinearities of and , hawever, different frames can lead to diftecent choices. The following three sections describe reversals of preference caused by variations in the framing of acts, contingencies, and outeomes ‘The Framing of Acts Problem 3 [NM ~ 150): Imagine eat yu face tne. following pair of concurrent decisions st examine both dee'sions, then indicate ‘pions you preiee Decision (), Choose, betwees: ‘A. a sure gui of S240 [8 percect} B28 chance to gain $1000, and T3#hchance ta gui nothing 6 percent] Decision (i). Choose between: (Ca sure loss of $780 [19 percent) . 78% ehance co lose $1000, and 125% chance to ase nothing [8 percent) ‘The majority choice in decision (i) is, tisk averse: a riskless prospect is pre+ ferred to a risky prospect of equal or areater expected value, In contrast, the majority choice in decisin (iis risk take ing: 2 risky prospect is preferred to a riskless prospect of equal expected val- ue. This pattern of risk aversion in choices involving gains and risk secking in choices involving losses is attributable to the properties of v and =. Because the value function is S-shaped, the value as. sociated with a gain of $240 is greater than. 24 percent of the value associated with a gain of $1000, and the (negative) value associated with a loss of $750 is smaller than 75 percent of the value ass0- ciated with a loss of $1000. Thus the shape of the value funetion contributes to tisk aversion in decision () and to risk seeking in decision (if). Moreover, the underweighting of moderate and ‘high probabilities contributes to the relative attractiveness of the sure gain ia (i) and tw the relative aversiveness of the suce loss in (i). The same analysis applies to problems 1 and 2. Because (i) and (li) were presented to- gether, the respondents had in effect to choose one prospect from the set: A and C, Band C, Aand D, Band D, The most comman pattern (4 and D) was chosen by 73 percent of respandents, wile the least popular pattern (B and C) was ‘chosen fy only 3 percent of respondents. However, the combination of B and C is definitely superior to the combina. tion A and D, as is readily seen in probe lem 4 Problem 4 [N = 88), Choose between: A &D. 25% chance to win $240, and 75% charee to lose $760. (0 per cent] 25% chance to win $250, and 7856 chance Co ose $750. (100 per cent BEC. When the prospects were combined ‘and the damfnance of the second option all respondents cliose ‘Fae popularity of the inferior option in problem 3 implies that this prablem was framed as a pair of SCIENCE, VOL. 211