Abstract
After having played a prisoner’s dilemma, players can approve or reject the other’s choice of cooperation or defection. If both players approve the other’s choice, the outcome is the result of the chosen strategies in the prisoner’s dilemma; however, if either rejects the other’s choice, the outcome is the same as if they had mutually defected from the prisoner’s dilemma. In theory, such an approval mechanism implements cooperation in backward elimination of weakly dominated strategies, although this is not the case in the subgame perfect Nash equilibrium. By contrast, the compensation mechanism proposed by Varian (Am Econ Rev 84(5):1278–1293, 1994) implements cooperation in the latter but not in the former. This result motivates the present experimental study of the two mechanisms. The approval mechanism sessions yield a cooperation rate of 90% in the first period and 93.2% across periods, while the compensation mechanism sessions yield a cooperation rate of 63.3% in the first period and 75.2% across periods. In addition, the backward elimination of weakly dominated strategies better predicts subjects’ behavior than does the subgame perfect Nash equilibrium in both mechanism sessions.
Similar content being viewed by others
Notes
This holds provided the payoff asymmetry in the underlying prisoner’s dilemma is not too large (Qin 2005).
In this vein, Saijo et al. (2016), who developed a working paper version of the present paper, characterized the AM using certain no-punishment and no-coercion conditions. Kimbrough and Sheremeta (2013) presented situations in which side payments assumed in the CM—collusion in a market, patent races, and R&D competition—might not be legal. Guala (2012) pointed out that there is little anthropological evidence that humankind has used private punishment.
The AM differs from the money-back guarantee mechanism, as follows. If either player, but not both, chooses C, then the $10 contribution is returned to the cooperator. The money-back guarantee mechanism cannot generate (7,17), while the AM achieves (7,17) if players choose (C, D) and then both choose y. The same argument applies to Brams and Kilgour’s (2009) concept of voting between the outcomes of all-C and all-D in order to overcome PD. The AM has an advantage in terms of welfare compared to the money-back guarantee mechanism and Brams and Kilgour’s (2009) voting. To observe this, consider a case in which only player 1 is sufficiently altruistic that player 1’s utility is given by (player 1’s material payoff) \(+{\uprho } \) (player 2’s material payoff), \({\uprho } >4/3\). Assume that players know each other’s payoff function. Then, the outcome of CD maximizes the sum of utilities of two players among four possible outcomes of the PD in Fig. 1. The AM implements CD, while neither the money-back guarantee mechanism nor Brams and Kilgour’s (2009) voting mechanism does.
A path is a list of what each player does in each stage: player 1’s choice between C and D, player 2’s choice between C and D, player 1’s choice between y and n, and player 2’s choice between y and n. For simplicity, hereafter, we write a path as CDny instead of (C, D, n, y). In the AM, we have 96 SPNEs. See Saijo et al. (2016) for more details.
The risk dominance criterion, even with subgame perfection, cannot select the strategy uniquely, unlike in the case of BEWDS. The off-path prediction by risk dominance is redundant. To observe this, consider subgame CD in Fig. 2. According to risk dominance, an NE with greater product deviation loss is more likely to occur (e.g., Blonski and Spagnolo 2015). Since the product of loss deviating from NE (n, y) and that of (n, n) are zero, both are predicted in light of risk dominance. However, seven out of the eight observed choices in subgame CD are (n, y), as reported in Table 2.
These results are due to the discreteness of strategies.
The pairings were anonymous and determined in advance so that no two subjects were paired more than once.
We used Andreoni and Miller’s (1993) method for statistical testing. We first calculated the average cooperation rate for each subject across periods, followed by the test statistic using the averages to eliminate cross-period correlation.
Additional experiments on the AM and PD, in which subjects play the game in only one period, provide similar results, suggesting that the AM works without repetition. The data and results are available upon request.
Although many people advocated the Paris Accord, the national pledges by countries to cut emissions are voluntary and do not involve penalty. “At best, scientists who have analyzed it say it will cut global greenhouse gas emissions by about half the required amount to avert a potential increase in atmospheric temperatures of 2 degrees Celsius or 3.6 degrees Fahrenheit” (NYTimes, 2015).
References
Andreoni J, Varian H (1999) Preplay contracting in the prisoners’ dilemma. Proc Natl Acad Sci 96(19):10933–10938
Andreoni J, Miller JH (1993) Rational cooperation in the finitely repeated prisoner’s dilemma: experimental evidence. Econ J 103(418):570–585
Banks JS, Plott CR, Porter DP (1988) An experimental analysis of unanimity in public goods provision mechanisms. Rev Econ Stud 55(2):301–322
Beatty T, Crawford I (2011) How demanding is the revealed preference approach to demand? Am Econ Rev 101(6):2782–2795
Blonski M, Spagnolo G (2015) Prisoners’ other dilemma. Intl J Game Theory 44(1):61–81
Bracht J, Figuieres C, Ratto M (2008) Relative performance of two simple incentive mechanisms in a public goods experiment. J Pub Econ 92(1):54–90
Brams SJ, Kilgour DM (2009) How democracy resolves conflict in difficult games. In: Levin S (ed) Games, groups, and the global good. Springer, Berlin, pp 229–241
Charness G, Fréchette GR, Qin C-Z (2007) Endogenous transfers in the prisoner’s dilemma game: an experimental test of cooperation and coordination. GEB 60(2):287–306
Chen Y (2008) Incentive-compatible mechanisms for pure public goods: a survey of experimental literature. In: Plott CR, Smith VL (eds) The handbook of experimental economics results. Elsevier, Amsterdam, pp 625–643
Cooper R, DeJong DV, Forsythe R, Ross TW (1996) Cooperation without reputation: experimental evidence from prisoner’s dilemma games. GEB 12(2):187–218
Davenport C (2015) Nations approve landmark climate accord in Paris. New York Times. Retrieved November 7, 2017 http://www.nytimes.com/2015/12/13/world/europe/climate-change-accord-paris.html
Fehr E, Gächter S (2000) Cooperation and punishment in public goods experiments. Am Econ Rev 90(4):980–994
Fehr E, Powell M, Wilkening T (2014) Handing out guns at a knife fight: behavioral limitations of subgame-perfect implementation, ECON – Working Paper 171, University of Zurich
Fischbacher U (2007) z-Tree: Zurich toolbox for ready-made economic experiments. Exp Econ 10(2):171–178
Forsythe R, Rietz T, Myerson R, Weber R (1996) An experimental study of voting rules and polls in three-candidate elections. Intl J Game Theory 25(3):355–383
Guala F (2012) Reciprocity: weak or strong? What punishment experiments do (and do not) demonstrate. Behav Brain Sci 35(01):1–15
Hamaguchi Y, Mitani S, Saijo T (2003) Does the Varian mechanism work? Emissions trading as an example. Intl J Bus Econ 2(2):85–96
Huang X, Masuda T, Okano Y, Saijo T (2017) Cooperation among behaviorally heterogeneous players in social dilemma with stay or leave decisions. University of Arizona Department of Economics Working Paper series 2017-16
Kalai E (1981) Preplay negotiations and the prisoner’s dilemma. Math Soc Sci 1(4):375–379
Kimbrough EO, Sheremeta RM (2013) Side-payments and the costs of conflict. Intl J Ind Organ 31:278–286
Masuda T, Okano Y, Saijo T (2014) The minimum approval mechanism implements the efficient public good allocation theoretically and experimentally. Games Econ Behav 83:73–85
Midler E, Figuières C, Willinger M (2015) Choice overload, coordination and inequality: three hurdles to the effectiveness of the compensation mechanism? Soc Choice Welf 45(3):513–535
Plott CR, Smith VL (2008) Handbooks in economics 28: handbook of experimental economics results. Elsevier, Amsterdam
Qin CZ (2005) Penalties and rewards as inducements to cooperate. Department of Economics, University of California-Santa Barbara
Saijo T, Okano Y, Yamakawa T (2016) The approval mechanism experiment: a solution to prisoner’s dilemma. KUT-SDES Working Paper no. 2015-12 (Revised), Kochi University of Technology
Saijo T, Shen J (2018) Mate choice mechanism for solving a quasi-dilemma. J Behav Exp Econ 72:1–8
Selten R (1975) Reexamination of the perfectness concept for equilibrium points in extensive games. Intl J Game Theory 4(1):25–55
Selten R (1991) Properties of a measure of predictive success. Math Soc Sci 21(2):153–167
Varian HR (1994) A solution to the problem of externalities when agents are well-informed. Am Econ Rev 84(5):1278–1293
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Additional information
We thank the associate editor and two anonymous referees for their useful comments and suggestions. Masuda is also grateful for helpful comments from Ryan Tierney. This research was supported by the Suntory Foundation; the Joint Usage/Research Center at the Institute of Social and Economic Research, Osaka University; Scientific Research A (24243028) and Challenging Exploratory Research (16K13354) of the Japan Society for the Promotion of Science; the Research Institute for Humanity and Nature (RIHN Project Number 14200122); and “Experimental Social Sciences: Toward Experimentally-based New Social Sciences for the 21st Century,” a project funded by a Grant-in-Aid for Scientific Research on Priority Areas from the Ministry of Education, Science and Culture of Japan. We also thank the Economics Department, Osaka University, for allowing us to use their computer laboratory.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Rights and permissions
About this article
Cite this article
Saijo, T., Masuda, T. & Yamakawa, T. Approval mechanism to solve prisoner’s dilemma: comparison with Varian’s compensation mechanism. Soc Choice Welf 51, 65–77 (2018). https://doi.org/10.1007/s00355-017-1107-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00355-017-1107-z