1

Do Auction Bidders “Really” Want to Win the Item, or Do They Simply Want to Win? JAMES R. WOLF HAL R. ARKES WALEED A. MUHANNA*

2

*James R. Wolf Jr., is a graduate student at the Fisher College of Business, Ohio State University, 400 Fisher Hall, 2100 Neil Avenue, Columbus, OH 43210; email: [email protected], and an Assistant Professor of Information Systems, School of Information Technology, Illinois State University, Campus Box 5150, Normal, IL 61790-5150; phone: (614) 893-3140; fax: (309) 438-5113; email: [email protected]. Hal R. Arkes is a Professor of Psychology, Department of Psychology and Senior Scholar in the Center for HOPES, Ohio State University, 240N Lazenby Hall, 1827 Neil Avenue Mall, Columbus, OH 43210-1222; phone: (614) 292-1592; fax: (614) 688-3984; email: [email protected]. Waleed A. Muhanna is an Associate Professor of Management Information Systems, Fisher College of Business, Ohio State University, 400 Fisher Hall, 2100 Neil Avenue, Columbus, OH 43210; phone: (614) 2923808; fax: (614) 292-2118; email: muhanna.1@ osu.edu. This research was partially funded by a grant from Ohio State University’s Jensen-Wallin-Young Fund and by Illinois State University’s Caterpillar Scholars Award.

3

ABSTRACT

This article describes three experiments designed to investigate the effects of competition and attachment on auction bidder behavior. For each study, in order to isolate bidders' desire to win the item from bidders’ desire to simply win the auction, we constructed survey experiments with participants divided into two groups. Both groups bid on a certificate that could be exchanged for an attractive gift, but for one of the groups the certificate number was changed, although the gift for which the certificate could be exchanged remained the same. In each experiment, the group with the changed certificate number bid significantly less.

4

INTRODUCTION

Recent research suggests that the overbidding in online and traditional auctions may be the result of two psychological factors. The first is competitive arousal: bidders get caught up in the competition and bid past their valuations of the item. The second is pseudo-endowment: bidders become attached to the item up for bid, and this attachment causes the item’s valuation to change during the auction. The basic premises of both factors were nicely summed up by winning bidders in Ku, Malhotra, and Murnighan (2005). According to the authors, both winners bid higher than they intended originally. One winner stated that she, “REALLY wanted that pig.” Another overbidder explained that “auction fever took over.” The first winner’s statement suggests that her bidding was affected by her attachment to “that pig,” whereas the second winner’s statement indicates a temporary loss of rationality. The first winner’s overbidding appears motivated by the desire to win the item (i.e., that pig), and the second winner’s overbidding seems motivated simply by a desire to win. This article is an experimental attempt to separate the desire to win the auctioned item from the desire to simply win the auction. Furthermore, this study adds to the growing work investigating endowment-like effects without actual ownership. We conducted three survey experiments in which participants bid on a certificate that could be traded for an iPod. During the auction the certificate number was changed, although the new certificate could still be exchanged for the iPod. To the extent participants developed an attachment to the original certificate number even though they did not yet possess that specific certificate, the change in the certificate number should diminish their willingness to bid. This would be evidence for a “pseudo-endowment effect,” because no attachment had developed to

5

the new certificate number compared to the attachment that had developed to the original one (Ariely & Simonson, 2003). On the other hand, if participants continued to bid unabated despite the change, this would be evidence for participants’ simply wanting to win the auction. We begin with a brief review of endowment and pseudo-endowment effects.

ENDOWMENT AND PSEUDO-ENDOWMENT EFFECTS

Historically, examinations of the endowment effect have focused on situations involving actual ownership (e.g., Thaler 1980). Kahneman and Tversky (1984) used the term “loss aversion” to describe the phenomenon that a loss generally has a greater subjective effect than does an equivalent gain. Thaler (1980) was one of the first to extend the concept of loss aversion to riskless decisions. He noted that people often demanded more to give up an item than they were willing to pay to obtain it and suggested that loss aversion was behind this “endowment effect.” Kahneman, Knetsch, and Thaler (1990) conducted a well-known experimental illustration of the gap between buyers’ willingness to pay (WTP) and sellers’ willingness to accept (WTA). In their study the researchers gave one group coffee mugs (“sellers”) and asked those participants about the lowest sum for which they would agree to exchange the mug. Researchers asked a second group, who did not receive mugs (“buyers”), about the highest sum they would exchange for the mug. In Kahneman et al.’s study, and in many similar investigations, the average selling price demanded was significantly higher than the average buying price. The phenomenon seemed to affect participants’ valuations the moment that they gained procession of an item. For this reason, Kahneman et al. use the term “instant” endowment effect.

6

However, several more recent studies show that that the endowment effect may impact valuation even sooner, before actual ownership even takes place (e.g., Wolf, Arkes, & Muhanna, 2005; Heyman, Orhun, & Ariely, 2004; Simonson & Ariely, 2004; Ariely & Simonson, 2003; Carmon, Wertenbroch, & Zeelenberg, 2003). Ariely and Simonson (2003) suggest that pre-factual feelings of ownership and the accompanying emotional attachment toward the auctioned item may lead bidders to overbid. They posit that psychological ownership may take place during the auction process, i.e., the consumer with the current high bid may develop prefactual feelings of ownership toward the item and begin to feel a valuation-altering attachment to the item. Ariely and Simonson (2003) term this phenomenon the “pseudo-endowment effect” and suggest that once an item has become part of a bidder’s psychological endowment, if the bidder is subsequently outbid, the attachment to the item may increase the bidder’s willingness to bid again to reclaim his or her lost endowment. Supporting this, Wolf et al. (2005) performed a variation of Kahneman et al.’s mug experiment and found evidence that participants’ valuations of the mugs were altered in a manner similar to the endowment effect even though participants were not actually endowed with the mugs. Instead of dividing participants into owners and buyers and comparing participants’ WTP and WTA, the authors divided the participants into long-duration buyers and short-duration buyers and then examined both groups' willingness to pay. Wolf et al. found that participants interacting with the mugs for longer periods of time bid higher and were more likely to retain their mugs than participants interacting with the mugs for shorter durations. These results are similar to those reported by Strahilevitz and Loewenstein (1998), who found that the length of actual ownership affected participant valuations. Strahilevitz and Loewenstein also found that currently owning an item was not a prerequisite for participants to

7

be affected by the endowment effect. They observed that that past ownership of an item and the duration of the past ownership positively affected participant valuations. Although several researchers have recently started to investigate attachment or the pseudo-endowment’s effect on auction bidding behavior, there is a long history of work examining endowment-like effects without actual ownership. For example, Hoch and Loewenstein (1991) discussed the change in reference point that accompanies partial possession of a good. The authors noted that after partially adapting to the possession of a good, consumers may feel deprived if they are subsequently unable to fully acquire the item. As a result, Hoch and Loewenstein suggested that a consumer who partially adapts to the possession of a good may have increased motivation to acquire the item. Carmon et al. (2003) showed that simply thinking about an option can created a sense of prefactual ownership of the option and an increased sense of loss from choosing an alternative option. These findings are consistent with earlier work by Dhar and Simonson (1992), who observed that that making an option the focus of a comparison enhanced the choice’s perceived attractiveness and the probability the item would be selected. Similarly, Casey (1995) discovered a gap between participants’ compensation demanded and their willingness to pay for lottery tickets even in the absence of actual ownership. Likewise, Sen and Johnson (1997) found that merely processing a coupon for a product increased a consumer’s preference for that product in a manner similar to ownership. More recently, Wolf, et al. (2005) examined bidding in eBay Motors’ vehicle auctions and found evidence that bidding in the auctions examined was positively affected by a pseudoendowment effect. Specifically, the authors observed that the longer bidders were in the lead of an auction, the more likely they were to re-bid if they were later outbid. Further the authors

8

found that the effect was reduced, but not eliminated, with market experience. In related work, Simonsohn and Ariely (2005) examined eBay auctions for DVD movies. They found that in auctions with low starting bids (i.e., between $0.00 and $1.00), 17% of winners placed multiple bids. However, in auctions with higher starting bids, Simonsohn and Ariely found that winners placed a lower number of multiple bids. The authors suggested that the winners placed fewer repeat bids in these latter auctions because the higher starting-bids reduced the opportunities for bidders to develop an emotional attachment or pre-factual feelings of ownership toward the auctioned items. Similarly, in two experimental studies of auction bidder behavior, Heyman, Orhun and Ariely (2004) uncovered evidence that “quasi-endowment,” or a sense of ownership, affects auction re-bidding. The first study involved hypothetical auction scenarios and manipulated the duration of perceived ownership by having participants imagine that they had been the leading bidder for varying lengths of time. Heyman and his associates found that participants told to imagine they were in the lead for longer periods of time submitted higher bids. In Heyman et al’s second study, participants participated in real-dollar auctions. The authors found that that the number of rounds in which a bidder was allowed to participate positively affected bid levels.

COMPETITIVE EFFECTS

Heyman et al. (2004) noted two factors that support the notion of a competitive or “opponent effect.” First they explained that bidders commonly discuss the outcome of an auction in terms or winning or losing. They posited that these references to winning or losing in auctions suggest that bidders get caught up in the momentum or otherwise derive some utility from the

9

competitive aspect of the auctions. Heyman and associates noted a second type of evidence supporting the idea of competitiveness and sensitivity of bidders to others: the presence of sniping. Sniping is the practice of bidding at the last second of an auction, when others will not have an opportunity to react before the auction closes. In two distinct sets of auction experiments, Heyman et al. (2004) found a third type of evidence supporting competition effects: an opponent effect. In both sets of experiments, participants in auctions with greater competition (i.e., more bidders) bid higher than bidders in auctions with less competition. In a now-famous example of competition-induced overbidding, Murnighan (2002) conducted a series of auctions for $20 bills with Executive MBA students bidding with their own money. Murnighan reports winning bids in these auctions of $28, $54, and even $2000. The “Cows on Parade” auctions described in Malhotra and Murnighan (2000) provide additional examples of irrational overbidding. The authors reported that, on average, the winning bids in these Internet and live auctions reached 575 percent and 791 percent of expert estimates of the items’ value. Similarly, Ku, Malhotra, and Murnighan (2005) reported overbidding in public art auctions for fiberglass pigs in Cincinnati and Snoopy figures in St. Paul. Several researchers suggest that the competitive nature of auctions may account for this excessive bidding (e.g., Ku, Malhotra, & Murnighan, 2005; Murnighan, 2002; Malhotra & Murnighan, 2000). These researchers suggested that competitive effects cause bidders to get caught up in a kind of auction fever, and this leads to excessive bidding. Ku et al. (2005) defined auction fever as irrational, emotionally charged overbidding. Several studies have suggested that auction sellers may be able to take advantage of these competitive effects by employing low starting bids, which allow bidders to get caught up in the bidding, and this results in higher

10

winning bids (e.g., Ku, Galinsky, & Murnighan, 2006; Ku et al. 2005; and Katkar & LuckingReiley, 2001).

EXPERIMENT ONE

In order to investigate the relative strength of competition and pseudo-endowment on bidder behavior, we constructed a survey experiment with 164 undergraduate students enrolled in a 100-level Information Systems course at a large public university in the Midwest. The goal of the experiment was to isolate participants’ desire to win the auction from their desire to win the item. To reduce participants’ suspicions about the trustworthiness of the seller, researchers told participants to imagine that the auction seller was the university’s School of Information Technology and that all proceeds would go to the school’s scholarship fund. To reduce participants’ suspicions about the physical condition of the item auctioned, researchers told participants to imagine that the item for sale was a certificate redeemable for an iPod Mini (retail value $199.00). The participants were randomly assigned to either the treatment group or the control group, and both groups completed their survey experiments at the same time and place. We gave both the control and the treatment groups the following instructions: “Imagine you are participating in an {school’s name} fundraising auction held on eBay. You are bidding on a certificate redeemable for an Apple iPod Mini. The certificate is good anywhere iPod Minis are sold. {The school’s name} is the seller of the certificate and all proceeds will go to the school’s scholarship fund. To prevent fraud, each certificate is numbered. You are bidding on certificate #189548. The retail value of the certificate is $199.00.

11

Imagine that you are the current high bidder for iPod Mini certificate #189548. Your bid is $111.00.” Next, we instructed both groups: “Now, imagine, you have just received an email telling you that another bidder has outbid you. The high bid is now $120.” At this point, we asked the control group to record the maximum amount that they would be willing to bid to win the iPod Mini certificate #189548. Instead of asking the treatment group to record their bids, we instructed them: “Now imagine that you receive an email from the seller telling you that there was an error in the auction listing an iPod Mini certificate #189548 is not available. The seller informs you that they have replaced the unavailable iPod certificate with an identical certificate #189721. Just like the previous certificate, this new certificate is redeemable for an Apple iPod Mini anywhere iPod Minis are sold.” At this point, we asked the treatment group to record the maximum amount that they would be willing to bid to win the iPod Mini certificate #189721. After completing the survey experiment, each participant was entered into a drawing with prizes of $50, $25 and $25. The drawing was held immediately after all the surveys were collected.

Experiment One Results

The average bid for the control (no certificate number change) group was $143.60, and the average bid for the treatment (number changed) group was $119.69. This difference was significant, t(162) = 3.39, p < .001.

12

EXPERIMENT TWO

In the previous experiment, it is possible that participants in the no switch treatment group were under the impression that there was only one iPod certificate available for auction. As such, the elevated bidding by this group may be explained, not by any attachment to the item but, by this perceived scarcity. To reduce this possible effect, we conducted a second nearly identical survey experiment. However, in this experiment, we clearly stated that the iPod certificate being auctioned was one of many that the school was auctioning for the fundraiser. We conducted this second survey experiment with 106 undergraduate students enrolled in a 100-level Information Systems course at a large public university in the Midwest. The students participated in the study for course credit. The two treatments in this study were identical to the previous study except that on the first page of both treatments, students were told that the certificate was one of several that the seller would be auctioning off as part of its fundraising efforts.

Experiment Two Results

The average bid for the control (no certificate number change) group was $145.80, and the average bid for the treatment (number changed) group was $105.53. This difference was significant, t(104) = 4.59, p < .001.

13

EXPERIMENT THREE

In the previous two experiments, the introduction of seller error in experiments may have generated doubts in the minds of the participants about the character or capability of the seller or caused the participants to suspect that the seller was trying to cheat them or that the substituted item was damaged or of lower quality. To mitigate these doubts, we conducted a third survey experiment and eliminated seller error as the cause of the switched certificate. We conducted this third survey experiment with 127 undergraduate students enrolled in a 100-level Information Systems course at a large public university in the Midwest. Again participants were randomly assigned to one of two treatment groups. The “no switch” group is presented with the identical scenario used in experiment two. However, we changed the scenario presented to the “switched” group altering the rational behind the certificate switch. In experiments one and two, participants were told that the certificate was unavailable due to an error in the auction listing. In this experiment, the certificate becomes unavailable because another bidder has purchased the certificate with the auction’s buy-it-now feature. Just as in experiment two, both groups are initially told to “Imagine you are participating in an {school’s name} fundraising auction held on eBay. You are bidding on a certificate redeemable for an Apple iPod Mini. The certificate is good anywhere iPod Mini’s are sold. {school’s name} is the seller of the certificate and all proceeds will go to the school’s scholarship fund.

14

The certificate is one of several that {school’s name} will be auctioning off as part of its fundraising efforts. To prevent fraud, each certificate is numbered. You are bidding on certificate #189548 . The retail value of the certificate is $199.00” Next, as in the previous two experiments, participants in both groups are told that they are the high bidder for iPod Mini certificate #189548 and that their bid is $111.00. However, participants in the switched group are then presented with the following: “Now, imagine you have just received an email telling you that iPod Mini certificate #189548 is no longer available because another bidder has purchased the certificate using the auction’s Buy-It-Now feature. However, the email also informs you that the seller has another auction currently underway for an identical certificate (#189721). Just like the previous certificate, this new certificate is redeemable for an Apple iPod Mini anywhere iPod Mini’s are sold. The current high bid for this new certificate (#189721) is $120.” At this point, we asked the treatment group to record the maximum amount that they would be willing to bid to win the iPod Mini certificate #189721. After completing the survey experiment, each participant was entered into a drawing for three $20 prizes. The drawing was held immediately after all the surveys were collected.

Experiment Three Results

The average bid for the control (no certificate number change) group was $154.27, and the average bid for the treatment (new certificate) group was $141.67. This difference ($12.60) was significant, t(125) = 2.09, p < .05.

15

DISCUSSION

In each of the three experiments, participants reported that they where willing to pay significantly higher to win a specific certificate number than they are willing to pay to win an identical, but different, certificate. The data from each of these experiments suggest that the desire to win a specific item affects bidders over and above any desire to simply win the auction. That is, pseudo-endowment or attachment affects bidding ceteris paribus. We acknowledge several weaknesses in the reported studies but have taken steps to attempt to mitigate any effects these shortcomings may have had on the study’s results. In the first experiment, participants in the no switch treatment group may have been under the impression that there was only one iPod certificate available and perceived scarcity might have contributed to their elevated bidding. To reduce this possible effect, in the proceeding experiments we clearly stated that the iPod certificate being auctioned was one of many that the school was auctioning for the fundraiser. Next, the introduction of seller error in experiments one and two may have generated doubts in the minds of the participants about the character or capability of the seller or caused the participants to suspect that the seller was trying to cheat them or that the substituted item was damaged or of lower quality. We attempted to mitigate any suspicion of the seller by informing participants that the seller of the auction was the school and that the proceeds were going to the school’s scholarship fund. Second, to reduce suspicion that the substituted item was inferior, we chose to auction coupons redeemable for iPods at local retailers instead of auctioning physical items. In addition, we chose iPod certificates as the auction item because the students are familiar with the product, the items are readily available at several local area stores, they are

16

prevalent on campus, and they constitute a common give-away used in on-campus contests, surveys, and research projects. Finally, in the third experiment, we eliminated seller error as the cause of the switched certificate. Novemsky and Kahneman (2005) find evidence that people can experience loss aversion for a good that they never owned. The current work provides further evidence of this observation and suggests that regardless of any competitive effects, attachment accentuates bidders’ desire to win the item and their sense of loss should they lose the item. As a result, due to their prefactual feelings of ownership, bidders are more concerned with winning a specific item than they are concerned with winning the auction.

17

REFERENCES

Ariely, Dan and Itamar Simonson (2003),"Buying, Bidding, Playing or Competing? Value Assessment and Decision Dynamics in Online Auctions," Journal of Consumer Psychology, 13 (1&2), 113-23. Carmon, Ziv, Klaus Wertenbroch, and Marcel Zeelenberg (2003), "Option Attachment: When Deliberating Makes Choosing Feel Like Losing," Journal of Consumer Research, 30 (1), 15-29. Casey, Jeff T. (1995), "Predicting Buyer-Seller Pricing Disparities," Management Science, 41(6), 979-99. Coursey, Don L., John L.Hovis, and William D. Schulze (1987), "The Disparity between Willingness to Accept and Willingness to Pay Measures of Value," The Quarterly Journal of Economics, 102(3), 679-90. Dhar Ravi, and Itamar Simonson (1992), “The Effects of the Focus of Comparison on Consumer Preferences”, Journal of Marketing Research, 29 (November), 430-440. Heyman, James, Yesim Orhun, and Dan Ariely (2004), “Auction Fever: The Effect of Opponents and Quasi-Endowment on Product Valuations,” Journal of Interactive Marketing, 18 (4 ) 7-21. Hoch, Stephen J. and George F. Loewenstein (1991), “Time-Inconsistent Preferences and Consumer Self-Control,” Journal of Consumer Research, 17 (4), 1-16. Kahneman, Daniel, Jack L. Knetsch, and Richard H.Thaler (1990), "Experimental Tests of the Endowment Effect and the Coase Theorem." Journal of Political Economy, 98(6), 132548.

18

Kahneman, Daniel and Amos Tversky (1984), "Choice Values and Frames," American Psychologist, 39 (4), 341-50. Katkar, Rama and David Lucking-Reiley (2001), “Public Versus Secret Reserve Prices in eBay Auctions: Results from a Pokemon Field Experiment,” NBER Working Paper, 8183. Ku, Gillian, Adam D. Galinsky and J. Keith Murnighan (2006), “Starting Low but Ending High: Congruity, Competitive Arousal, and a Reversal of the Anchoring Effect in Auctions,” Journal of Personality and Social Psychology, Forthcoming. Ku, Gillian, Deepak Malhotra and J. Keith Murnighan (2005), “Towards a Competitive Arousal Model of Decision-Making: A Study of Auction Fever in Live and Internet Auctions,” Organizational Behavior and Human Decision Processes, 96 (2), 89-103. Malhotra, Deepak and J. Keith Murnighan (2000), “Milked for all Their Worth: Competitive Arousal and Escalation in the Chicago Cow Auctions,” Working paper, Kellogg School of Management, Northwestern University. Murnighan, J. Keith. (2002), "A Very Extreme Case of the Dollar Auction." Journal of Management Education, 26(1), 56-69. Sen, Sankar and Eric J. Johnson (1997), “Mere-Possession Effects without Possession in Consumer Choice,” The Journal of Consumer Research, 24, 105-117. Simonsohn, Uri and Dan Ariely (2005), "eBay's Happy Hour: Non-Rational Herding in On-Line Auctions," Working Paper http://ssrn.com/abstract=722484 Strahilevitz A. Michal and George Loewenstein (1998), “The Effect of Ownership History on the Valuation of Objects,” Journal of Consumer Research, 25 (3), 276-289. Thaler, Richard H. 1980 "Toward a Positive Theory of Consumer Choice." Journal of Economic Behavior and Organization, 1 (March), 39-60. Wolf, James R., Hal R. Arkes, and Waleed A. Muhanna (2005), "Is Overbidding in Online

19

Auctions the Result of a Pseudo-Endowment Effect?" Working Paper http://ssrn.com/abstract=735464