selective play
DESCRIPTION
Au, W. T., Law, Y. K., & Lee, Y. H. (2009, August). Effect of information display on cooperation in a public goods dilemma under a selective-play protocol. Paper presented at The 13th International Conference on Social Dilemmas, Kyoto, Japan. Selective Play. - PowerPoint PPT PresentationTRANSCRIPT
Effect of Information Display on
Cooperation in a Public Goods
Dilemma under a Selective-Play Protocol
Department of Psychology
The Chinese University of Hong Kong
Hong Kong, China
Winton Au
Fion Law
Lestin Lee
How to promote cooperation under
endowment heterogeneity?
Au, W. T., Law, Y. K., & Lee, Y. H. (2009, August). Effect of information display on cooperation in a public goods dilemma under a selective-play protocol. Paper presented at The 13th International Conference on Social Dilemmas, Kyoto, Japan.
Selective Play• An option for leaving the
relationship and choosing a new partner (Hayashi & Yamagishi, 1998; Orbell & Dawes, 1991; Tesfatsion, 1995)
– “Choice to exit” A trinary-choice game in a PDG (Orbell, Schwartz-Shea, & Simmons, 1984; Hauk, 2003)
– “Fully-fledged selective play” (Yamagishi, Hayashi, & Jin, 1994; Yamagishi and Hayashi, 1996)
• Switch among groups (Ehrhart & Keser, 1999)
• Bid for partners (Coricelli, Fehr, & Fellner, 2004)
• Rank partners (Page, Putterman, & Unel, 2002)
• Hypothesis: Freedom to choose partners enhances cooperation
Yamagishi & Hayashi (1996)
Display of Endowment Information– Endowment heterogeneity increases cooperation (e.g., Chan, Mestelman, Moir, & Muller, 1996;
Bergstrom, Blume, &Varian,1986)
– … decreases (e.g., Cherry, Kroll, & Shogren, 2005; Ledyard, 1995)
– … does not change (e.g., Warr, 1983; )
• Rich contributes more…– More efficacious, critical or efficient (e.g., Van Dijk & Wilke, 1994; Rapoport, 1988; Diekmann, Przepiorka, &
Wehrli, 2009)
– Altruistic (Becker, 1974)
– Inequality aversion (Bolton & Ockenfels, 2000; Fehr & Schmidt, 1999)
• … less …– Fair share (e.g., Buckley & Cronson, 2006; Kroll, Cherry & Shogren, 2007)
– Anticipatory reciprocity (e.g., Cherry, Kroll & Shogren, 2005)
– Group identification [reduced when inequality justified] (Smith, Jackson, & Sparks, 2003)
• Hypothesis: Display of endowment information affects cooperation
Display of Cooperation Information• Visual anonymity
– Anonymity of a person• Identifiability (Kerr, 1999; Cress & Kimmerle, 2008)
– Anonymity of a person’s behavior or decision
• Hypothesis: Display of [individualized] cooperation information enhances cooperation (e.g., Sell & Wilson, 1991; Cress & Kimmerle, 2008; Vuolevi & Van Lange, 2009; Weisel & Bornstein, 2009)
Experiment• Task
– Students allocating time between group project [cooperation] vs. individual examination [defection]
• Voluntary contribution mechanism• Linear payoff
• Endowment size• Randomized across trials
– 6 hours per week– 9– 12
• Display of cooperation information– Yes– No
• Display of endowment information– Yes– No
[cooperation]
[defection]
D D: 9 hrs. D: (38%) D: 9 hrs. (38%)
No info Endowment info Cooperation info Full info
120 seconds to select members
Experiment• Task
– Students allocating time between group project [cooperation] vs. individual examination [defection]
• Voluntary contribution mechanism• Linear payoff
• Endowment size• Randomized across trials
– 6 hours per week– 9– 12
• Display of cooperation information– Yes– No
• Display of endowment information– Yes– No
• Protocols of play– Self-selected– Forced-play
Self-selected play– 303 participants– 13-24 participants in an one-hour session– 10 games– HK$50 ( ¥ 620) show-up fee +– Bonus HK$40 - $102 (average $73)
• ¥ 500 - ¥ 1500 (average ¥ 900)• Two randomly selected participants
Forced-play– 303 participants– Yoked to the same groupings of self-
selected play
Select group members
Make contribution decision
Receive feedback
10 trials
GroupProjectHours
IndividualExamHours
TotalHours
GroupProjectScore
IndividualExamScores
TotalScores
TotalPayoffs
Results of Trial 2
Coop Info x Endow Info Display
Endow Info No Endow Info0.0
0.5
1.0
1.5
2.0
2.5
3.0
Amount of CooperationCoop Info No Coop Info
Endow Info No Endow Info10%
15%
20%
25%
30%
Proportion of CooperationCoop Info No Coop Info
Coop info enhanced cooperation
Coop Info x Endow Info Display
Endow Info No Endow Info0.0
0.5
1.0
1.5
2.0
2.5
3.0
Amount of CooperationCoop Info No Coop Info
Endow Info No Endow Info10%
15%
20%
25%
30%
Proportion of CooperationCoop Info No Coop Info
Hiding endow info enhanced cooperation
Coop Info x Endow Info Display
Endow Info No Endow Info0.0
0.5
1.0
1.5
2.0
2.5
3.0
Amount of CooperationCoop Info No Coop Info
Endow Info No Endow Info10%
15%
20%
25%
30%
Proportion of CooperationCoop Info No Coop Info
Effect of coop info most prominent in the absence of endow info
Endowment x Coop Info x Endow Info
6 9 1210%
15%
20%
25%
30%
35%
40%
33% 33%
30%
24%26% 26%
13%
16%14%
13% 13% 12%
Coop Info OnlyCoop & Endow InfoNo InfoEndow Info Only
Endowment
Coop
erati
on R
ate
When cooperation information was present, absence of endowment information affected low and medium endowment players the most
Endowment x Coop Info x Protocol
6 9 1210%
15%
20%
25%
30%
35% 34%32%
29%
22%
26% 26%
11%12%
11%
15% 15% 15%
Coop/Selected
Coop/Forced
NoCoop/Selected
NoCoop/Forced
Endowment
Coop
erati
on R
ate
Low and medium endowment players most affected by possibility to choose partners when cooperation information was displayed
When heterogeneity
exists
• Allow selection of players• Hide endowment info and• Display cooperation info
– Most effective for less rich people