SN- Lecture 5
2 likes543 views
Experiments on social preferences show that individuals exhibit other-regarding behavior and do not always act rationally in a selfish manner. In dictator, ultimatum, gift exchange, trust, and public goods games: (1) People share resources and reject unfair offers, even when selfish behavior would maximize their outcomes; (2) Reciprocation and trust exist, as seen by positive responses to generous acts and trust in others; and (3) Conditional cooperation leads to initial cooperation that declines over time without enforcement, even between partners. The ability to punish free-riders can stabilize cooperation.
SN- Lecture 5
- 2. To cover the findings on social preferences & individual behavior To understand the link between microlevel preferences & macrolevel outcomes for these findings Aim Lecture 5
- 3. Experiments run according to a fixed protocol Participants receive written rules of the game Decisions are usually anonymous Subjects get paid depending on their decisions (no flat payment) No deception EXPERIMENTS WE KNOW THAT
- 4. Dictator game Ultimatum game Gift exchange game Trust game Public goods game EXPERIMENTS Measure various aspects of people’s social preferences
- 6. Dictator game Forsythe et al. 1994 What is the rational behavior of the sender? ALTRUISM
- 7. Dictator game Forsythe et al. 1994 What is the rational behavior of the sender? A rational and selfish dictator will keep all the money ALTRUISM
- 8. Dictator game No support for the selfish prediction A rational and selfish dictator will keep all the money What is the rational behavior of the sender? Positive Sharing Altruism Other-regarding Preferences
- 9. Dictator game No support for the selfish prediction 22% transferred a positive amount Dictators share about 28% of their endowment However, it matters: Who is the recipient? Higher transfers to a charity than to another participant Sociodemographical Background Older dictators transfer more money than younger ones
- 10. Ultimatum game Guth et al. 1982 Sender FAIR SHARING Offers around 40% to 50% Receiver Rejects offers below 30% Offers in the ultimatum game are substantially higher than in the dictator game. Why?
- 11. Ultimatum game Proposer FAIR SHARING If the proposer expects that a low offer might be rejected, it is rational to offer an amount more likely to be accepted The fact that people reject offers poses a bigger challenge People might reject for emotional reasons They feel unfairly treated and want to punish greedy intentions
- 12. Methods Low offers Neuro-scientific Activate areas of the brain associated with anger and disgust People reject low offers because they consciously want to reject them Sanfey at al. 2003 Strength of activation of these areas Predicts the probability of rejection quite well Social Preferences:
- 14. Effort is not contractible Fehr et al. 1993 Gift Exchange Effort Profit employer Cost employee Payoff employee Wage Profit employer Payoff employee What is the rational behavior?
- 15. Effort is not contractible Rational players Gift Exchange Irrespectively of the wage, will choose the minimum effort (costly) Employees: Have incentives to pay the lowest wage Employers:
- 16. Positive Reciprocity Findings Gift Exchange tend to reward generous wage offers by high efforts Employees:
- 18. The Trust Game Berg et al. 1995 Trust & Trustworthiness What is the rational behavior of the receiver? What is the rational behavior of the sender?
- 19. The Trust Game What is the rational behavior of the investor and the receiver? Selfish recipients will not return anything, irrespective of the amount received Recipient: Investor: Rational and selfish investors would foresee this and invest nothing
- 20. Measuring Trust Why does the game measures trust? Any amount transferred is tripled Transferring the whole endowment will maximize the joint income of both players Transferring “X” pays off only if receiveing at least “X” back Sending a positive amount signals trust Back transfers measure trustworthiness
- 21. Trust Game Findings Trustors send on average $5.16 (between $0 and $10) Trustees return on average $4.66 Almost 50% of the trustees returned a positive amount (25% exceeded the initial transfer) Older people trust more & are more trustworthy
- 23. Cooperation & Free-riding PD & PG Games They are the most important vehicle for studying cooperation problems in controlled laboratory experiments
- 24. Prisoner’s Dilemma It’s been extensively studied Main aspects of interest: The extent of cooperation in one-shot games The importance of strategic incentives
- 25. Two studies Cooper et al. 1996 & Andreoni and Miller 1993 Design: Ten rounds Two treatments: Stranger & Partner What is the rational behavior?
- 26. Rational PD What is the rational behavior? Why? Stranger: Each play is against a new opponent (one-shot) Partner: Backward Induction- In the last period both will defect, and so on... Under assumptions of selfishness and rationality, all players in both conditions are predicted to defect
- 27. Findings PD Stranger: Cooperation at least 20% of the times Partner: Cooperation at least 50% of the times People are prepared to cooperate even in one-shot games The possibility to behave strategically strongly increases cooperation
- 28. Repeated Cooperation What happens if groups become larger? The strategic gains from cooperation that comes from repeated interactions are a powerful force in explaining real-world cooperation in small and stable groups
- 29. Large Groups? What happens if groups become larger? Bilateral PD: A player can punish a defector by defecting as well Large Groups: Not possible. Defection punishes defectors and other cooperators Punished cooperators might then defect
- 31. Public Goods Game Experiments study cooperation for the provision of public goods What is the rational behavior?
- 32. Public Goods Game What is the rational Behavior? A rational and selfish individual has incentive to keep all the points The group: Better off, as a whole, if everybody contributes all 10 points Tension between individual incentives & collective benefits
- 33. PG Game Common Results Design: 10 rounds 20 tokens Partner & Stranger
- 34. PG Game Main Findings Positive contributions (one-shot) Sociodemographic Variables: Older people cooperate more Rural residents contribute more than urban ones Strategic incentives matters Partners contribute more than strangers Contributions decline over time to very low levels Why?
- 35. Altruistic Punishment The only way a cheated cooperator can avoid being a sucker is by reducing cooperation Punishing everyone, even other cooperators
- 36. If group members can identify a defector and punish her or him Will this Altruistic Punishment Solve the free rider problem Prevent the breakdown of cooperation
- 37. Design: Punishment in Repeated PG Games Second Stage: after subjects made their contribution Information: Contribution of each group member Punishment: Max. 10 points to each group member (costly) For every point, the punisher’s income decreases 10% Treatments: Stranger & Partner Fehr and Gachter 2000
- 38. Results Strong increase in contributions, with punishment Partners contribute more than strangers 100% > 60% of the endowment
- 39. Micro-Macro Link Relation between individual motives & behavior of the collective Cooperation declines over time Results in PG games without punishment Cooperation stabilizes or increases Results in PG games with punishment
- 40. These differences in microlevel motivations produce a macrolevel outcome in which everyone eventually free-rides Why the decline? Some are free riders, independently of others’ contributions Some are conditional cooperators Perhaps people are heterogeneous in their cooperative inclinations
- 41. Micromotives Macrobehavior Fischbacher and Gachter 2010 Design: Indicate contributions in the strategic form Free-riders: No contribution - 30% Conditional cooperators: Increase contribution. -50% Other 20% complicated patters
- 42. Micro-Macro Conditional cooperators cooperate if others cooperate If there are free-riders, conditional cooperators reduce their contribution Social preferences can explain the decay of cooperation Even if not everyone is motivated selfishly, the aggregate outcome is one in which everyone behaves selfishly
- 43. Checklist Individuals cooperate in different strategic conditions - even when the interaction only occurs once Cooperation is affected by strategic behavior - interacting repeatedly with others increases it Cooperation declines over time even between partners Potential punishment increases cooperation and stabilizes it along time
- 44. Questions?