Endowment eect Marie-Pierre Dargnies March 6, 2013
Marie-Pierre Dargnies
Endowment eect
Kahneman, Knetsch, and Thaler (JPE 1990)
Half of the subjects are given a mug and asked for WTA Half of the subjects are shown a mug and asked for WTP Finding: WTA(5.75$) ∼ 2 ∗ WTP (2.25$) (chocolate bars): WTA(0.5$) ∼ 2 ∗ WTP (0.3$)
Marie-Pierre Dargnies
Endowment eect
Endowment eect is predicted by reference-dependent utility with loss aversion λ>1 x −r r v (x ) = λ(x − r ) ifif xx ≥ 80% keep mug alternative explanation? P & Z use same arguments as in their 2005 paper => asymmetry generated by experimental procedures
Marie-Pierre Dargnies
Endowment eect
Methodology: Plott & Zeiler (AER 2007) Method and Language
"This mug is yours. You own the mug. I am giving this mug to you" Language used to endow subjects may lead subject to perceive it as a "gift" Subjects might be reluctant to trade a gift, resulting in asymmetry
Suggestion of relative value: Giving subjects good X (and not Y) may convey a signal about the relative value of X Public revelation of choice: Most experiment did not implement anonymous decisions. May result in a cascade: if other subjects' decisions signal good X's relative value if subjects care about the view of others
Marie-Pierre Dargnies
Endowment eect
List (QJE 2003) Experience & endowment eect Experiment with sport cards dealers => experience in trading. Protocol: Get people to ll survey Hand them memorabilia card A (B) as thank-you gift After survey, show them memorabilia card B (A) Card A ∼ Card B "Do you want to switch?" "Are you going to keep the object?" 2 experiments with dierent object (sport cards and Mickey Mouse pins) Standard theory prediction: 50% trade (given goods are comparable) Endowment eect: too little trade Marie-Pierre Dargnies
Endowment eect
List (QJE 2003) Experience & endowment eect: ndings Finding 1: Strong endowment eect for inexperienced dealers How to reconcile with Plott & Zeiler (2005)? No training? No, nothing di cult about switching cards No practice? No, people used to exchanging cards No incentive compatibility? No Is it anonymity? Unlikely Gift? Possible Finding 2: Substantial experience lowers the endowment eect to zero Signaling of relative value of cards (Plott & Zeiler) Getting rid of loss aversion? Expecting to trade cards again? Marie-Pierre Dargnies
Endowment eect
List (ECMA 2004): Sport card dealers II Field experiment similar to QJE 2003 but with mugs and chocolate Trading in four treatments: Mug: "Switch to Chocolate?" Chocolate: "Switch to Mug?" Neither: "Choose Mug or Chocolate?" Both: "Switch to Mug or Chocolate?"
Large endowment eect for inexperienced card dealers No endowment eect for experienced card dealers! Learning (or reference point formation) generalizes beyond original domain
Marie-Pierre Dargnies
Endowment eect
Engelmann and Hollard (ECMA 2010): Forced trade
Forced trade succeeds in eliminating EE with very little trade experience, in contrast to List's results (years in eld, minutes in the lab using forced trade?) Consistent with conjecture that trade uncertainty is a major source of EE "Free" market is a poor teacher
Marie-Pierre Dargnies
Endowment eect
Koszegi-Rabin (AER 2006)
experienced traders expect to trade with p=0.5 reference point is (0.5, {mug}; 0.5, {∅}) This implies no endowment eect
Marie-Pierre Dargnies
Endowment eect
Implications
Endowment sets reference point
Status quo based reference point If individual is loss averse there is no trade
Expectations set reference point
If individual has rational expectations about r, she exhibit no endowment eect, even if loss averse E.g., if you expect to trade your item, no endowment eect KKT (1990): "...no endowment eect would be expected, such as when goods are purchased for resale rather than for utilization."
Marie-Pierre Dargnies
Endowment eect
Ericson & Fuster (QJE 2011)
Prospect theory is not explicit about reference point
Status quo Based on recent expectations (Koszegi & Rabin) => may result in dierent implications
Idea: manipulate subjects' expectations about possibility to trade a good => see whether expectations aect exchange behavior
Marie-Pierre Dargnies
Endowment eect
Experiment I: Design
Subjects have a mug and a pen on their table Individual coin ip (labeled 1 and 9) => get an index card with 1 or 9 Instructions say that:
subjects own the mug in front of them They will leave the experiment with either the mug or the pen At the end of the session they may have the opportunity to exchange the mug for the pen according to a speci c procedure
Marie-Pierre Dargnies
Endowment eect
Experiment I: Design (2)
Procedure for exchange: the option for exchange will occur i 10-sided die comes up lower than or equal to the number on their index card Low Treatment: subjects whose coin ip was 1, have a 10% chance of having the option to exchange the mug for the pen High Treatment: subjects whose coin ip was 9, have a 90% chance of having the option to exchange the mug for the pen Each subject made a decision conditional on the die roll => gives an observation for each subject
Marie-Pierre Dargnies
Endowment eect
Experiment I- Prediction
Status quo: If reference point is shaped by status quo the exchange probability should not aect trade, i.e., the proportion of exchanges is the same in Low Treatment and High Treatment Expectations: If reference point are determined by expectations we observe more exchanges in High Treatment than in Low Treatment
Marie-Pierre Dargnies
Endowment eect
Experiment I- Results
Marie-Pierre Dargnies
Endowment eect
Experiment 2 - Design
Same as Experiment 1 except: Get index card (1 or 8) dependent on coin ip If 10-sided die comes up lower than index, they get the mug in front of them for free If die comes up 9 they have option to keep the mug or exchange it for randomly determined amount of money ($0 -$9.57)
Marie-Pierre Dargnies
Endowment eect
Experiment 2 - Design (2) Treatment Low: index card = 1: 10% chance of receiving the mug for free Treatment High: index card = 8: 80% chance of receiving the mug for free In both treatments 10% chance of having choice between mug and money After lling out questionnaire and making decisions new instructions for case in which die comes up with 8:
If the die comes up 8 they will get a choice between a pen and a randomly determined amount of money ($0 â $9.57) Notice that they previously expected to get neither mug or money
Marie-Pierre Dargnies
Endowment eect
Experiment 2 - Prediction
Status quo: If reference point is shaped by status quo the WTA is not aected, i.e., it is the same for a low and high probability of getting the mug for free Expectations: If reference point are determined by expectations we observe a higher WTA when the probability of getting the mug for free is high than it is low
Marie-Pierre Dargnies
Endowment eect
Experiment II- Results
Controlling for WTAPen yields marginally signi cant eect:
WTAMug − WTAPen = $0.92 in High Treatment WTAMug WTAPen = $0.02 in Low Treatment WTAMug increases by 20-30% in High Treatment Marie-Pierre Dargnies
Endowment eect
Discussion Results suggests that the expectation of continued ownership is responsible for low fraction of trades (and not status-quo eects, current formal ownership) Loss averse people may trade to same extent as non-loss averse people, when they expect transactions Plott & Zeiler is compatible with utility function kinked at expected outcomes Subjects in P & Z may have expected to trade or get alternative item If subjects have perception that they will be able to trade, perceptions matter under expectation-based reference points (but not under status-quo) List's experienced traders are also in line with results: If expect to trade, there is no endowment eect Other factors like social norms, salience, etc may in uence reference point (study does not rule out these factors) Marie-Pierre Dargnies
Endowment eect
