Skip to content

Lecture 58 Ultimatum & Dictator Games

Dictator Games

  • Responder cannot reject the offer.
  • To prove whether its altruism or fear.
    • Answer: Mostly fear of rejection (strategic)
  • NE = 0
  • In one-shot situation, no strategic thinking involved as no need to consider the responder's reaction
    • Any positive offer is altruistic
  • Another aspect: wealth of the dictators were not earned
  • baseline (unearned wealth) \(\to\) Assumed to have certain amount of money
    • 17% gave zero offer
  • earned wealth: do some task to get some money and then split it
    • GMAT questions correctly got them this.
    • 80% gave zero offer
  • double-blind: the three players don't know each other… put this in a sealed envelope in a box
    • 96% gave zero offer
  • When there was complete anonymity 96% dictators offered zero amount.

Basic Results

  • Baseline design variants:
    1. Repeat the game (new player each tome)
    2. Ask responder to state a minimum acceptable offer (MAO), instead of deciding whether to accept a specific offer, so that he doesn't change his decision based on the offer (like 7 is my lucky number so I will select 7% that he has offered)

Minimum Acceptable Offer

  • MAO advantage: measure likely reactions to all possible offers
  • At MAO, pleasure of getting money = satisfaction from refusing the offer and getting no money, but punishing the proposer for violating the 50-50 social norm.

The number, \(R\) indicates the strength of the Responder's private reciprocity motive.
- Large \(R\) \(\implies\) she cares a lot about fairness/generosity.
- \(R = 0\) \(\implies\) she doesn't care about the Proposer's motives at all.
- Satisfaction at reject a low offer = \(R(50-y)\). The gain from accepting the offer = \(y\).

So, 'reject the offer if \(y \lt R(50-y)\)'

Thus, MAO can be calculated as

$$
y \lt \dfrac{50R}{(1+R)}
$$
- \(R=1\), equal importance on reciprocity and the social norm. Reject any offer less than $25.
- Cutoff at which two motivations, monetary gain and punishing for justice balance out.
- \(R=0.5\), cutoff = $16.67.
- \(R=2\), cutoff = $33.33.
- Actual offers even control for risk-aversion (more generous)

Variant of Dictator & Ultimatum Games

  1. Methodical variables change experiment
    • stakes
    • repetetion doesn't make difference
    • Anonymity lowers allocations
  2. Demographic variables = different groups of people behave.
    • men and econ majors are often more self-interested
    • high-T males reject more… but more generous
    • Young children are more self-interested
  3. Culture
    • stakes
    • language
    • experimenter effects
  4. Descriptive variables change description of the game
    • labelling and context but not structure
    • call it seller-buyer exchange \(\implies\) self-interest is encouraged
    • claim from a shared resource pool \(\implies\) encourage generosity.
  5. Structural variables change game by adding variables (identity, communication, entitlement)
    • Create entitlement by letting a contest winner be the Proposer \(\to\) lower offer
    • Knowing the players makes the allocations higher
    • When they don't know share of pie, responders are reluctant to reject low offers
    • Multi-person games \(\to\) social preferences not based on another's overall generosity… but about another player's fairness
    • In competitions, there is no way for fair minded people to compete.