What Drives Failure to Maximize Payoffs in the Lab? A Test of the Inequality Aversion Hypothesis

Abstract

Experiments based on the Beard and Beil (1994) two-player coordination game robustly show that coordination failures arise as a result of two puzzling behaviors: (i) subjects are not willing to rely on others' self-interested maximization, and (ii) self-interested maximization is not ubiquitous. Such behavior is often considered to challenge the relevance of subgame perfectness as an equilibrium selection criterion, since weakly dominated strategies are actually used. We report on new experiments investigating whether inequality in payoffs between players, maintained in most lab implementations of this game, drives such behavior. Our data clearly show that the failure to maximize personal payoffs, as well as the fear that others might act this way, do not stem from inequality aversion. This result is robust to varying the saliency of decisions, repetition-based learning and cultural differences between France and Poland.

Cognitive ability and the effect of strategic uncertainty

“…Human treatments come from Jacquemet and Zylbersztejn (2014), while all the other sessions were carried out in October 2012 and February and March 2014.…”

“…Baseline Game 1 (BG1, shown on the top left panel of Table 2) was used as the baseline treatment in Beard and Beil (1994);Beard, Beil, and Mataga (2001); Zylbersztejn (2013, 2014). Egalitarian Game 2 (EG2, shown on the bottom right panel of Table 2) was one of the additional matrices introduced by Jacquemet and Zylbersztejn (2014) in an attempt to assess the eect of the relative payo structure on subjects'…”

“…Moreover, from the standard theory perspective, R (L ; m) (H ; h) (R, r) is a natural candidate for a focal point, since it is also risk-dominant. 1 Notwithstanding these predictions, various studies have found a frequent failure to achieve the ecient equilibrium (see, e.g., Beard and Beil, 1994;Beard, Beil, and Mataga, 2001;Goeree and Holt, 2001;Cooper and Van Huyck, 2003;Jacquemet and Zylbersztejn, 2014) both in sequential and simultaneous implementations of this game. Depending on the exact experimental setup, between 20% and 84% of observed outcomes are not Pareto ecient.…”

Learning, Words and Actions: Experimental Evidence on Coordination-Improving Information

“…We have tested this hypothesis through companion experiments, reported in Jacquemet and Zylbersztejn (2011), in which the baseline treatment is compared with a treatment that equalizes payoffs between players in the Pareto-Nash equilibrium. We unambiguously reject the hypothesis that aversion to inequality is enough to account for player Bs' striking behavior.…”

Cognitive Ability and the Effect of Strategic Uncertainty

“…Moreover, from the standard theory perspective, (R, r) is a natural candidate for a focal point, since it is also risk-dominant. 1 Notwithstanding these predictions, various studies have found a frequent failure to achieve the efficient equilibrium (see, e.g., Beard and Beil, 1994;Beard, Beil, and Mataga, 2001;Goeree and Holt, 2001;Cooper and Van Huyck, 2003;Jacquemet and Zylbersztejn, 2014) both in sequential and simultaneous implementations of this game. Depending on the exact experimental setup, between 20% and 84% of observed outcomes are not Pareto efficient.…”