Fairness Perception and Decisions Under Uncertainty

1
Fairness Perception and Decisions Under
Uncertainty

• Min Gong
• Jonathan Baron
• Howard Kunreuther

2
Previous Research on Fairness

• Equity, Envy, and Efficiency (Varian,1974)
• Timing-effect problem (Myerson, 1981 etc.)
• Ex ante and Ex post equity (Keller and Sarin,
  1988)
• Various theoretical models (Gajdoes and Tallon
  2002, etc.)
• But how do people really perceive fairness under
  uncertainty? Compared with no uncertainty, do
  people make and request more or less fair offers
  under uncertainty?
• No research found so far.

3
Questions to be Answered in Study 1

• Is 30 chance of getting the whole pie more or
  less fair than 30 of the pie?
• Does the fairness perception depend on peoples
  roles, e.g. whether they make or accept the
  offer?
• Do people perceive the same offer differently in
  terms of fairness before or after the uncertainty
  is resolved?

4
Experimental Design

• The Deterministic Ultimatum Game (the usual
  ultimatum game, Dgame)
• Splitting 100 beans (worth 5)
• The Stochastic Ultimatum Game (Sgame)
• Two players determine their chances of winning
  100 beans (5)
• The proposer makes an offer on how large a chance
  he is willing to give the responder for winning
  100 beans. The responder decides to accept or to
  reject the offer.
• If the offer is accepted, a random number will be
  generated to decide whether the proposer or the
  responder gets 100 beans. The other person will
  get nothing. If the offer is rejected, then the
  game is over, and nobody gets any beans.

5
Stochastic Ultimatum Game
6
Experimental Design --Continued

• Design
• 112 subjects (56 pairs)
• One-Shot game
• Fairness Rating after Responders make a
  decision, both Proposers and Responders rate how
  fair the offer is on a scale of 0-100, where 0
  represents not fair at all and 100 represents
  very fair.
• Three conditions (Between Subject)
• Ex ante fairness rating in the Sgame by Proposers
  and Responders (before the uncertainty is
  resolved )
• Ex post fairness rating in the Sgame by Proposers
  and Responders (after the uncertainty is resolved
  )
• Fairness rating in the Dgame by Proposers and
  Responders

7
Fairness Perception

• Average Offers in the Dgame (37 of 100 beans)
  and Sgame (36 chance of winning 100 beans) are
  not different, but the fairness ratings are.

8
Major Findings in Study 1

• Proposers believe that x chance is fairer (87
  vs. 66) than x of the pie before the uncertainty
  is resolved(t(36)2.58).
• Responders, however, do not agree with them.
  Instead, they think (ex anti) x chance of
  winning the pie is as fair as x of the pie
  (t(24)-0.79).
• What is more, after the uncertainty is resolved,
  Respondents think x chance they had is less fair
  than x of the pie and they have been unfairly
  treated (66 vs. 46, t(49)-2.01).

9
Questions to be Answered in Study 2

• Do responders request more or less under
  uncertainty than under certainty? (Minimum
  acceptance offer-MAO)
• Do proposers make different offers under
  uncertainty than under certainty if they are
  asked to state an MAO as well as propose an
  offer?
• Do people make different decisions when the
  uncertainty is presented as exclusive chance (as
  in Sgame in Study 1) or independent chances?

10
Experimental Design

• Each subject makes a pre-committed offer and a
  minimum acceptable offer (MAO), with counter
  balanced order
• 152 subjects in betwesen-subject design
• 3 Games
• Dgame
• Sgame (as in Study 1) with exclusive chance in
  which only one player gets 100 beans
• Sgame with independent chances
• Similar to Sgame in Study 1 that two players
  chances add up to 100
• But two players have independent chances
• the outcome can be both get 100, nobody gets
  anything, or one gets 100.

11
Major Findings in Study 2

• Proposed offers are significantly different in
  three games, being the highest in Dgame, and the
  lowest in Sgame exclusive.
• This is consistent with the fairness perception
  in Study 1, indicated Proposers may try to
  maintain a certain level of fairness.
• MAO are not significantly different cross three
  games.

12

Discussion Three factors

• Players Role
• Proposers view x of chance as being fairer than
  x of the pie Responders view them the same
• Proposed offers are different in the three games
  MAO are not
• Timing-effect
• Ex ante and ex post (before and after the
  uncertainty is resolved) fairness perception are
  different for both Proposers and Responders
• Presentation Mode of Uncertainty
• People make more generous offers in Sgame with
  independent chances than with exclusive chances

13
Future Research

• Questions remains to be answered
• Why does the uncertainty matter to the Proposer
  only, but not to the Responder?
• Why do people treat exclusive chance differently
  from independent chances?
• Uncertain pie
• Instead of having 100 beans to split, players
  have to win 100 beans with some probability
• Imitate real-life scenario, such as oil drilling
  companies splitting the pie before the bill
  passes

14

• Gajdos., T. Tallon J.M. 2002. "Fairness under
  Uncertainty," Economics Bulletin, 4(18), 1-7.
• Keller. L.R. Sarin R.K. 1988. "Equity in Social
  Risk Some Empirical Observations," Risk
  Analysis, 8(1), 135-146.
• Myerson, R. B., 1981. Utilitarianism,
  egalitarianism, and the timing aspect in social
  choice problems. Econometrica 49 (4), 883-897.
• Varian, H., 1974. Equity, Envy and Efficiency.
  Journal of Economic Theory 9, 6391