L6: PGG with punishment Flashcards
What did Fehr and Gachter (2000) examine?
The role of punishment opportunities in sustaining co-operation in social dilemmas
Explain the basic structure of Fehr and Gachter (2000)?
2 stage PGG:
1) Standard PGG
2) Players informed anonymously of contributions of others in their group and can assign ‘punishment points’ to them
n=4, E=20, k=1.6
Explain the design of Fehr and Gachter (2000)?
2x2x2
1) w. and wo. punishment
2) Partners and strangers matching protocol on both cases
3) Order manipulation: all sessions have w. and wo. punishment phases and order changes
10 period game!!!
Explain how the punishment worked in Fehr and Gachter (2000)?
Each player can assign up to 10 punishment points to each of the other group members BUT it is costly for both the punisher and the punished subject
- Each punishment point someone receives reduces their previous stage payoff by 10%
- Cost of punishment increases with each point someone assigns
What were the findings of Fehr and Gachter (2000)? for no punishment
No punishment - normal decaying pattern of contributions
What were the findings of Fehr and Gachter (2000)? for punishment (4)
a) higher avg. contribution
b) sharp jump in co-operations
c) no decay pattern observed
the less a player contributes the more they are punished!
How do the findings of Fehr and Gachter (2000) vary between partners and strangers?
They don’t - results similar for both treatments
What did Fehr and Gachter (2002) want to find out?
Whether co-operation among genetically unrelated is frequent, in non-repeat interactions, when gains from reputation are absent
ie. looks at whether humans engage in altruistic punishment (see slide don’t really get this)
Define altruistic punishment? What does this mean for cost of punishment?
Punishment that is costly and yields no material benefits for the punishers
Makes punishment a second-order public good (who should/will pay for it?
Explain the structure of Fehr and Gachter (2002)’s game?
PGG:
- 2 conditions: treatment and non-treatment
- Group composition changed from period to period so that no subject ever met another more than once (perfect strangers)
3 main findings of Fehr and Gachter (2002)?
1) Punishment acts mostly imposed on defectors BY co-operators
2) Punishment treatment: co-op. increases over time
3) Non-punishment treatment: co-op. sharply decreases
Explain (developed) the welfare implications of Fehr and Gachter (2002) findings?
Monitoring other people’s behaviour makes them more pro-socially oriented tf suggests implications of tax policies etc.
BUT implementation of these systems is COSTLY
Therefore Q) can costly punishment improve group welfare as measured by individuals NET EARNINGS
What did Gaechter et al. (2008) investigate? What were their treatments?
They examined whether the duration of interaction affects the efficacy of the punishment
4 treatments:
- Punishment and non-punishment
- 10 period game and 50 period game
What did Gaechter et al. (2008) find?
10P: no punishment led to higher avg. net earnings
50P: punishment let to higher avg. net earnings
TF in LT is more efficient to have costly punishment
What did Masclet et al. (2003) study? What was the idea behind the study?
Studied if non-monetary punishment affected co-operation (ie. INFORMAL SANCTIONS: peer pressure, gossip, social ostracism etc.)
Idea: if players incur a disutility (not necessarily monetary) from disapproval they may make higher contributions
Explain the structure of Masclet et al. (2003)?
- 30 periods divided into 3x10 period segments
- segments 1-10 and 21-30: no punishment available - subjects endowed with 20 tokens and they choose simultaneously the contribution to the group account
- segments 11-20: either monetary punishment (MP) or non-monetary punishment (NMP) can be used
Explain how monetary punishment worked in Masclet et al. (2003)?
In P11-20, each period consisted of a 2-stage game. First stage as usual. At the beginning of 2nd stage, subjects could assign 0-10 punishment points to each of the other 3 group members after they were informed of the contribution levels of other group members. Each point players is given reduces their payoff by 10%
Explain how non-monetary punishment worked in Masclet et al. (2003)?
Same as for monetary punishment, except the points represented levels of disapproval and have no monetary influence
What were the 4 main findings of Masclet et al. (2003)?
1) Initially MP and NMP sanctions increase co-op. by same amount, but over time MP led to higher over co-op.
2) Once sanctions are lifted contributions fall back to level in P1-10 again
3) NMP more effective in partners than strangers treatment
4) Both sanction types increase avg. net earnings
3 conclusions for Masclet et al (2003)?
1) Informal sanctions can enhance co-op.
2) Avg. earnings relatively similar NMP and MP
3) More punishment points one receives -> higher increase in contributions