lin et al the role of emotions in the endowment effect

Journal of Economic Psychology 27 (2006) 589597

www.elsevier.com/locate/joep

The role of emotions in the endowment eVect

Chien-Huang Lin a, Shih-Chieh Chuang b,,Danny T. Kao a, Chaang-Yung Kung b

a Department of Business Administration, National Central University, Chung Li, Taiwanb Department of Business Administration, Chao-yang University of Technology, 168, Gifeng E. Road,

Wufeng, Taichung County, Taiwan

Received 29 July 2004; received in revised form 7 September 2005; accepted 21 October 2005Available online 27 December 2005

Abstract

This paper argues that the endowment eVect the tendency for minimum selling price to exceedmaximum buying price for a particular object might be minimized when a negative emotion isinduced. The Wndings from two experiments on 400 participants support this hypothesis: the endow-ment eVect only occurs when participants are induced to feel happy, and is absent when people areinduced to feel sadness. Thus, the importance of emotional states is highlighted. 2005 Elsevier B.V. All rights reserved.

JEL classiWcation: M31

PsycINFO classiWcation: 3920

Keywords: Endowment eVect; Emotion; Value; Willingness to accept (WTA); Willingness to pay (WTP)

1. Introduction

Economists are often concerned about the estimated value of various goods or ser-vices, and they attempt to derive estimations regarding said value. The assessment of objec-0167-4870/$ - see front matter 2005 Elsevier B.V. All rights reserved.doi:10.1016/j.joep.2005.10.001

tive values is typically derived from two concepts: WTA and WTP. People are required

* Corresponding author. Tel.: +886 5 5223843; fax: +886 4 2374233.E-mail address: [email protected] (S.-C. Chuang).

590 C.-H. Lin et al. / Journal of Economic Psychology 27 (2006) 589597

to estimate the least price they are willing to accept to sell (WTA), and the ceiling price theyare willing to pay (WTP). Most economists argue that the gap between WTA and WTPshould converge for most commodities. However, experimental research shows that thevalues of WTA and WTP might be remarkably diverse, even for an identical object. Spe-ciWcally, research on the valuation of objects by laboratory experiments indicated that peo-ple set a higher WTA on an owned object than WTP for a not-owned-yet object(Kahneman, Knetsch, & Thaler, 1990; Knetsch & Sinden, 1984; Thaler, 1980).

The gap between the buying price and the selling price of an object is dependant on thestatus of ownership, and is termed the endowment eVect (Kahneman et al., 1990; Thaler,1980). The endowment eVect has been highlighted in studies employing diverse methodolo-gies, including laboratory experiments (Kahneman et al., 1990), economics style marketexperiments (Franciosi, Kujal, Michelitsch, Smith, & Deng, 1996), and Weld studies (e.g.,Johnson, Hershey, Meszaros, & Kunreuther, 1993). Previous research has shown that emo-tional states are inXuenced directly by subjective valuation for an object (Forgas & Ciarro-chi, 2001). This research also asserts that emotional states may moderate the endowmenteVect, and two experiments prove the endowment eVect only occurs with positive emotions not with negative emotions.

2. Literature review and building hypotheses

2.1. The endowment eVect review

Research on endowment eVect (Franciosi et al., 1996; Johnson et al., 1993; Kahnemanet al., 1990; Strahilevitz & Loewenstein, 1998) has shown that people tend to demand ahigher selling price for a commodity that they own than they would be willing to pay ifthey were looking to acquire it as a buyer. The endowment eVect also appears to be moder-ated by regulatory focus: participants were reluctant to exchange an endowed object whenthey focused on preventative measures, but not when they focused on self-promoting goals(Liberman, Idson, Camacho, & Higgins, 1999). Mandel (2002) found that the endowmenteVect was observed when participants imagined another individual wanting to buy from orsell to them, but not when they imagined wanting to by from or sell to another individual.Recent research has focused on emphasizing the importance of aVective response (i.e., pos-itive and negative feeling evaluations).

Loewenstein and IssacharoV (1994) found that the manner in which people obtainedparticular objects aVected their valuation of the objects. SpeciWcally, individuals whoreceived goods based on exemplary performance valued the goods more than if theyreceived the goods by chance or based on poor performance. Loewenstein and IssacharoVsuggested that aVective roles create positive associations with the items in question. SpeciW-cally, much research in decision making has emphasized the importance of aVectiveresponse in judgment and choice (Damasio, 1994; Loewenstein, 1996; Peter & Slovic, 2000;Slovic, Finucane, Peters, & MacGregor, 2002). Given this data, the following will focus onexploring the relationship between emotional states and the endowment eVect.

2.2. Building hypotheses

Reference point dependence and loss aversion are key roles in the endowment eVect

(Kahneman & Tversky, 1979). In this article, reference point dependence and loss aversion

C.-H. Lin et al. / Journal of Economic Psychology 27 (2006) 589597 591

are applied, and the adaptation-level theory (see also Hoch & Loewenstein, 1991) has beenadded into our theoretical framework to examine the control of emotional states on theendowment eVect.

Adaptation, in the context of object ownership, is the tendency for people to becomepsychologically accustomed to changes in their material situation. Adaptation has twomain consequences: (1) a diminution in ones hedonic response to the change and (2)increased sensitivity to departures from the new state of aVairs (Helson, 1947). For exam-ple, one typically appreciates the comfort of a new car, but only for a limited period. If thenew car breaks, forcing one to go back to the original one, the old car may seem intolerablyslow (see in Strahilevitz & Loewenstein, 1998).

Loss aversion refers to the prevalent trend that people tend to place greater weight onlosses than on gains of equal magnitude (Kahneman & Tversky, 1979). Namely, loss aver-sion occurs when losses are evaluated more than gains, which implies that people desire ahigher WTA than WTP.

We adopt Strahilevitz and Loewensteins (1998) model that applied the value functionintegrated in Kahneman and Tverskys prospect theory (1979) to illustrate how adaptationand loss aversion can account for the endowment eVect to build our hypothesis. In Fig. 1,the value function represents the relationship between the cognitive status of ownershipand the value as a function of a discrete outcome level (no ownership and partial adaptedownership) relative to some reference point. The reference point is identiWed when thevalue function is assumed steeper for negative departure (losses) than for positive depar-ture (gains) (Bell, 1985; Kahneman & Tversky, 1979).

Consider an object that would yield a value of V1 if it were owned. The no-adaptationvalue function, represented by the lighter line that goes through the origin, is drawn suchthat owning the good yields V1, and non-owning it yields zero. This no-adaptation func-tion depicts the case of the individual who would derive value from owning an object, butwho has not adapted to the notion of possessing it. With no adaptation, the steeply slopednegative region of the value function is not relevant.

The value functions usefulness lies in showing what happens when the people partiallyadapt to possession of the object. The partial adaptation case is depicted by the bold-line

Fig. 1. The value function for no adaptation, partial adaptation, and full adaptation (see in Strahilevitz & Loe-

PartialNo adaptation Full adaptation

V1Ownership

Value

V2

No adaptation

Partial adaptationwenstein, 1998).


value function. As compared with the no-adaptation situation, the zero point of the valuefunction has moved to the right, signifying that the individual is now in an intermediatestate between no and full adaptation ownership of the object.

Failing to possess the object is no longer aVectively neutral, but instead results in depri-vation (the distance between the origin and where the bold-line value function I intersectthe y-axis). The endowment eVect is explained by the fact that due to loss aversion (V1), thevalue of possessing an object for someone who has not possessed it yet, is smaller thanthose who have possessed it and are adapted to such possession but are unwilling to lose itto someone, V2.

Therefore, Fig. 1 illustrates that an increase in adaptation will move the reference pointand change V. The magnitude of value depends on how strong the motivation is to adaptto possession of an object. As to the ownership with no adaptation, it is homogeneous toneutral with no ownership.

According to the emotion congruency model (ECM), the evaluation of objects andevents is negatively inXuenced by bad emotions and positively inXuenced by positive emo-tions (Iness-Ker & Niedenthal, 2002; Rusting, 1998, 1999). Bower (1981) applied a networktheory of memory encoding to explain this phenomenon, and suggested that each speciWcemotion corresponds to a speciWc memory node comprised of multiple dimensions of emo-tions. The emotions stored in nodes can be activated by various stimuli, which may ignitethe chain reaction to activate other connected memory nodes. Therefore, we assume that ifthe object is accompanied with negative emotions, people will recall various negative expe-riences and project them onto the object; and in turn are willing to accept the loss of thepossible beneWts just as if they had never possessed it (refer to V1). In contrast, if the objectis accompanied with positive emotions, people will recall various pleasant events and pro-ject them onto the object, and they are willing to partially (V2) or fully adapt the owner-ship; they are unwilling to accept giving up the object because they are focused on thebelief that it will cause a loss of beneWts. Our hypothesis is that the endowment eVect issmaller in negative than positive emotional states.

3. Study 1

3.1. Method

One hundred and sixty undergraduate students (91 females and 69 males) participatedin this experiment as part of the course requirement in marketing management class. Theywere randomly and equally assigned to one of four diVerent experimental conditions. Atthe beginning of the experiment, half of the students were individually endowed with amug, and the rest received none. Then, participants were induced to yield either happy(positive) emotions, or sad (negative) emotions. The emotional induction that we intend tomanipulate on emotions and memory is adapted from the processes used by Smith andEllswoth (1985), and the same procedure is followed for both pleasant and unpleasantemotional experiences. Initially in the emotional induction procedure, participants wereasked to recall some past experiences or events, which activated a particular emotion; then,six general questions were asked about that particular experience.

Immediately after the emotion was induced, participants were asked, How happy doyou feel right now? to rate their current emotions. The scales ranged from 1, being

extremely unhappy, to 7, being extremely happy. Then, all of the participants were


asked to look at the mugs. They were told to evaluate how much they would be willing toaccept for selling an owned mug, or how much they would be willing to pay for a mug thatthey did not own. After estimating the price of a mug, the participants current emotionswere subsequently rated with the same scales. This experiment lasted approximately30 min.

4. Results and discussion

4.1. Manipulation checks of emotions

The responses indicated that the two self-rating emotion-scales are highly correlated(r D 0.81). Participants in positive emotional condition (Mpositive D 4.5, SD D 0.89) felt hap-pier than in negative emotional condition (Mnegative D 3.9, SD D 1.3) immediately after anemotion was induced (158) D 2.39, p < 0.05. At the end of the experiment, participants whowere manipulated by the positive emotions (Mpositive D 4.27, SD D 0.91) were happier thanthose who were manipulated by the negative emotions (Mnegative D 3.95, SD D 0.95),t(158) D 2.01, p < 0.05. These results conWrmed the eVectiveness of emotional manipula-tion.

4.2. EVect of emotions on value assessment

Our analysis focused on the examining the eVect of elicited emotions on the perceivedvalue for a mug. The statistics F (1, 156) D 56.89, p < 0.0001 showed the main signiWcanteVects of ownership: people who own a mug perceive a signiWcantly higher price for themug ($2.59) than the non-owners do ($2.07). This Wnding is consistent with prior evidencefor the endowment eVect (Thaler, 1980). More importantly, a signiWcant interaction wasfound between the endowment eVect and the emotions, F(1, 156) D 39.96, p < 0.0001. Addi-tionally, the price proved to be an irrelevant factor between those who owned a mug butwanted to sell it ($1.8, SD D 0.31), and those who did not own a mug but wanted to buyone ($1.72, SD D 0.21) in the negative emotional condition, t(78) D 1.38, p > 0.1. However,a signiWcantly higher price for a mug was perceived by the owners ($3.35, SD D 0.67) thanby the non-owners ($2.54, SD D 0.36) in the positive emotional condition, t(78) D 7.76,p < 0.0001. These statistics strongly support our assumption that the endowment eVect innegative emotions is smaller than in positive emotions.

Experiment 1 supported our hypothesis that emotions can inXuence the level of per-ceived value and the degree to which the endowment eVect exists. However, the Hawthorneexperiment argues that the value is mainly judged by the participants involuntary reac-tions to the stimuli on the spot, instead of by objective evaluations. Similarly, participantsinduced by recalling past events may suVer from issues of self-esteem, and may be forced tomake uncomfortable and reluctant decisions, which in turn aVect the perceived value of anobject. In this context, the manipulated value may be judged by means of the cognitive andmotivational reactions of the recalled events, rather than simply by the induced emotionalstates. Therefore, it is important to use diverse emotional inductions to verify that the emo-tion is a critically independent variable involved in the endowment eVect. That is why emo-tional induction procedures need to be conducted repeatedly. Experiment 2 aimed toincrease the range of emotional inductions by using methods diVerent from those used in

Experiment 1.


5. Study 2

Based upon the success of Experiment 1, Experiment 2 sought to conWrm and extendthese Wndings. In order to prove that the endowment eVect is moderated by emotionsrather than by other variables, we use another emotional induction method in Experiment2. Some audiovisual emotional induction methods were employed successfully to elicitmanipulated emotions in experiments by Adaval (2003) and Forgas and Ciarrochi (2001),and we followed those processes.

5.1. Method

Two hundred and forty participants (131 females and 109 males) were involved inExperiment 2 as part of the course requirement. They were randomly and equally assignedto one of four grids: (ownership: owners vs. non-owners) (emotions: positive vs. negative).Initially, half of the participants were endowed a mug and half were not. Then, theywatched a 25-min Wlm to induce the happy or sad emotional states. After emotions wereinduced as in Experiment 1, participants were asked to assess their current emotional statesby the question how happy do you feel right now? The scales were described with endpoint 1, being extremely unhappy, to 7, being extremely happy. Subsequently,they were asked to express how much they would be willing to sell the mug for that theyowned, and how much money they would be willing to buy a mug for that they did notown. After determining the price of that mug, all participants completed emotions self-rat-ing scales again and left the laboratory room. This experiment took approximately 30 min.

5.2. Emotional manipulation

In Experiment 2, we used audiovisual emotional induction to manipulate emotions,which was diVerent from Experiment 1. Initially, the participants were told that we wereexamining the impacts of various stimuli that might be used for future studies. The experi-menter then showed the participants one of two Wlm clips. In the positive emotional condi-tion, a segment from the feature Wlm Pretty Woman was shown to induce a happy emotion(n D 120). The previous studies had shown that this Wlm reliably induced happy emotions(Adaval, 2003). In the negative condition, a clip from the movie Ordinary People wasshown (n D 120) to induce a sad emotion. Previous studies had shown that it induced sadfeelings (Adaval, 2003). Participants were instructed to watch the videos as if they werewatching videos at home and allow them to be fully involved in the Wlm. Film as an emo-tion manipulator helps exclude potential motivational and cognitive preoccupations inmemory (the so-called autobiographical task) in Experiment 1.

5.3. Manipulation checks of emotions

The emotions of participants who watched the happy Wlm were rated signiWcantly morepositive than those of the people who watched the sad Wlm. This conclusion was the same,whether it was instantly measured after watching the emotional induction Wlms,t(238) D 13.61, p < 0.001, Mpositive D 4.82, Mnegative D 3.41, or at the end of the experiment,t(238) D 5.98, p < 0.001, Mpositive D 4.39, Mnegative D 3.67. These statistics conWrmed the eVec-

tiveness of the emotional manipulation.


5.4. EVect of emotions on value assessment

Our analyses focus on examination the eVect of elicited emotions on the perceived valuefor a mug. Results showed a main signiWcant eVect on the owners, F(1, 236) D 41.52,p < 0.0001, which implied that people perceived a signiWcantly higher price for the ownedmug ($2.5) than the not-owned-yet mug ($2.07). This Wnding is consistent with Experiment1 and prior evidence for the endowment eVect. Moreover, as we expected, an obvious cor-relation was found between the endowment eVect and the emotions, F(1, 236) D 23.75,p < 0.0001. More importantly, the price proved to be an irrelevant factor between thosewho owned a mug but wanted to sell it ($2.03, SD D 0.47) and those who did not own butwanted to buy one ($1.93, SD D 0.38) in the negative emotions, t(118) D 1.93, p > 0.1. How-ever, a signiWcantly higher price was perceived by the mug owners ($2.96, SD D 0.5.98) thanthe non-owners ($2.21, SD D 0.55) in the positive emotions, t(118) D 7.03, p < 0.0001. Theseresults support the hypothesis of this paper.

These statistics provide clear support for our hypothesis that the degree of the endow-ment eVect with positive emotions is greater than with negative emotions. People who ownan object under negative emotions and people who do not own the object project similarestimated values for the object (namely, the endowment eVect does not occur when thenegative emotions are elicited). This suggests that negative emotions may reduce the per-ceived value of an owned object, and pinpoints the potential critical role for emotions inmoderating the endowment eVect.

6. General discussion and conclusion

Typically, people as sellers tend to pursue a higher selling price for an owned object thanthey would as a buyer determining a buying price (e.g., Kahneman et al., 1990). However,as demonstrated in this article with both experiments, a transient emotional state can aVectthe magnitude of the endowment eVect. A higher price is estimated only when people areinduced into positive emotional states rather than negative emotional states. Thus, theemotions can be classiWed as a moderator in the endowment eVect.

Our research successfully identiWed the tendency of the endowment eVect to depend onhow happy people felt at that time. SpeciWcally, while experiencing positive emotions, peo-ple seem to be more willing to possess an object at hand and adapt to full ownership, andin turn will subjectively overestimate the value of an object. Accordingly, sales reps oftenattempt to induce good emotions to potential customers by manipulating music, lighting,ambience and so forth. However, negative emotions could produce a marked undervalua-tion of ones material achievements and could even accentuate experience of depressionand dysphoria (Ottaviani & Beck, 1988). Potential customers with negative experiencestend to underestimate the value of the new products for a newly launched product. There-fore, marketing practitioners should avoid triggering the negative emotions of customers,in order to prevent underestimation of the value of the new product. Business consultantsand managers should deWnitely be interested in both advantages and disadvantages of theemotional eVect.

Previous studies have reported mixed and contradictory Wndings for the behavioral con-sequences of diVerent negative moods (Luomala & Laaksonen, 2000; Raghunathan &Pham, 1999), which suggests that their eVect is complex. For example, Lerner, Small, and

Loewenstein (2004) found that two negative emotions, sadness and disgust, had opposing


endowment eVects: disgust enhance the endowment eVect, whereas sadness decreased it.Our results, in contrast, show the endowment eVect is diVerently aVected by happiness andsadness. Therefore, our results are restricted to the happiness of positive emotions and thesadness of negative emotions.

References

Adaval, R. (2003). How good gets better and bad gets worse: Understanding the impact of aVect on evaluations ofknown brands. Journal of Consumer Research, 30, 352367.

Bell, D. E. (1985). Disappointment in decision making under uncertainty. Operations Research, 33, 127.Bower, G. H. (1981). Mood and memory. American Psychologist, 36, 129148.Damasio, A. R. (1994). Descartes error: Emotion, reason, and the human brain. New York: Avon.Forgas, J. P., & Ciarrochi, J. (2001). On being happy and possessive: The interactive eVects of mood and personal-

ity on consumer judgments. Psychology and Marketing, 18, 239260.Franciosi, R., Kujal, P., Michelitsch, R., Smith, V., & Deng, G. (1996). Experimental test of the endowment eVect.

Journal of Economic Behavior and Organization, 30, 215226.Helson, H. (1947). Adaptation-level as frame of reference for prediction of psychological data. American Journal

of Psychology, 60, 129.Hoch, S., & Loewenstein, G. (1991). Time-inconsistent preference and consumer self-control. Journal of Consumer

Research, 17, 492507.Iness-Ker, A., & Niedenthal, P. M. (2002). Emotion concepts and emotional states in social judgment and catego-

rization. Journal of Personality and Social Psychology, 83, 804816.Johnson, E. J., Hershey, J., Meszaros, J., & Kunreuther, H. (1993). Framing, probability distortions, and insurance

decisions. Journal of Risk and Uncertainty, 7, 3551.Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica, 47, 263

291.Kahneman, D., Knetsch, J. L., & Thaler, R. H. (1990). Experimental tests of the endowment eVect and the coase

theorem. Journal of Political Economy, 98, 13251348.Knetsch, J., & Sinden, J. A. (1984). Willingness to pay and compensation demanded: Experimental evidence of an

unexpected disparity in measures of value. The Quarterly Journal of Economics, 99, 507521.Lerner, J. S., Small, D. A., & Loewenstein, G. (2004). Heart strings and purse strings carryover eVects of emotions

on economic decisions. Psychological Science, 15, 337341.Liberman, N., Idson, L. C., Camacho, C. J., & Higgins, E. T. (1999). Promotion and prevention choices between

stability and change. Journal of Personality and Social Psychology, 77, 11351145.Loewenstein, G., & IssacharoV, S. (1994). Source dependence in the valuation of objects. Journal of Behavioral

Decision Making, 7, 157168.Loewenstein, G. F. (1996). Out of control: Visceral inXuences on behavior. Organizational Behavior and Human

Decision Processes, 65, 272292.Luomala, H. T., & Laaksonen, M. (2000). Contributions from mood research. Psychology and Marketing, 17, 195

233.Mandel, D. R. (2002). Beyond mere ownership: Transaction demand as a moderator of the endowment eVect.

Organizational Behavior and Human Decision Processes, 88, 737747.Ottaviani, R., & Beck, A. T. (1988). Cognitive theory of depression. In K. Fiedler & J. P. Forgas (Eds.), AVect, cog-

nition and social behavior (pp. 209218). Gottingen: Hogrefe.Peter, E., & Slovic, P. (2000). The springs of action: AVective and analytical information processing in choice. Per-

sonality and Social Psychology Bulletin, 26, 14651475.Raghunathan, R., & Pham, M. T. (1999). All negative moods are not equal: Motivational inXuences of anxiety

and sadness on decision making. Organizational Behavior and Human Decision Processes, 79, 5677.Rusting, C. L. (1998). Personality, mood, and cognitive processing of emotional information: Three conceptual

frameworks. Psychological Bulletin, 124, 165196.Rusting, C. L. (1999). Interactive eVects of personality and mood on emotion-congruent memory and judgment.

Journal of Personality and Social Psychology, 77, 10731086.Slovic, P., Finucane, M. L., Peters, E., & MacGregor, D. G. (2002). The aVect heuristic. In T. Gilovich, D. GriYn,

& D. Kahneman (Eds.), Heuristics and biases: The psychology of intuitive judgment (pp. 397420). New York:

Cambridge University Press.


Smith, C. A., & Ellswoth, P. C. (1985). Patterns of cognitive appraisal in emotion. Journal of Personality and SocialPsychology, 48, 813838.

Strahilevitz, M. A., & Loewenstein, G. (1998). The eVect of ownership history on the valuation of objects. Journalof Consumer Research, 25, 276289.

Thaler, R. H. (1980). Toward a positive theory of consumer choice. Journal of Economic Behavior and Organiza-tion, 1, 6074.

The role of emotions in the endowment effectIntroductionLiterature review and building hypothesesThe endowment effect reviewBuilding hypotheses

Study 1Method

Results and discussionManipulation checks of emotionsEffect of emotions on value assessment

Study 2MethodEmotional manipulationManipulation checks of emotionsEffect of emotions on value assessment

General discussion and conclusionReferences

lin et al the role of emotions in the endowment effect

Documents