ICES Seminar in Experimental Economics and Game Theory

Endogenous Game Choice and Giving Behavior in Distribution Games

Friday, November 6, 2020 12:00 PM to 1:00 PM
Online Location, Zoom Meeting

The ICES Seminar in Experimental Economics and Game Theory of the Fall 2020 semester will feature:

Emin Karagozoglu

Bilkent University

Endogenous Game Choice and Giving Behavior in Distribution Games 

 

Please contact ICES Office Manager (sbahabib@gmu.edu) for Zoom link.

 

Abstract

List (2007) and Bardsley (2008) introduced the “taking game”, where the strategy space of the dictator allows him to take money from the passive player. Their findings cast doubt on the “pure altruism” interpretation for dictator game giving since subjects who assumed the role of the dictator in the taking game gave less than their counter-parts in the dictator game.  In these experiments, as in most experiments, the games subjects play are exogenously assigned by the experimenter. However, in many real life interactions, agents choose (or at least have a say regarding) the type of contract, protocol, or procedure to be employed. We first conducted an experiment (N = 268) where the game to be played (dictator game or taking game) is not exogenously assigned but it is rather the choice of the passive player, in the experimental treatment. Our experiment has between-subjects design, and has three treatments, two of which serve as control treatments: (i) exogenously assigned dictator game (EX-D), (ii) exogenously assigned taking game (EX-T), and (iii) passive subjects choose which game to be played (EN). We denote the endogenously chosen dictator (taking) game with EN-D (EN-T). We hypothesize that (i) the dictators in the taking game give less than the dictators in the dictator game do, (ii) passive players choose the dictator game more frequently than they choose the taking game, (iii) the mere fact that the game played in EN-D is the passive player’s choice makes passive players responsible for it, and as a consequence dictator game giving is lower in EN-D than in EX-D, and finally (iv) giving in EX-T and EN-T are identical. Our experimental results are in line with all of these hypotheses, and thus highlight the importance of endogenous selection into distributive games. Second, we also conducted a survey (N = 296), and asked respondents to predict the behavior in the experiment. A significant majority of them correctly guessed the results. Moreover, the text analysis of their answers to our open-ended questions show that the potential reasons they gave for their guesses are in line with the supporting arguments we provided for our hypotheses (e.g., responsibility, agency, accountability). The main message of the paper is in line with the recent work of Smith and Wilson (2018), who report very different behavior in voluntary ultimatum games compared to exogenously given ultimatum games.

 

For more information about the Seminar Series, please visit the Seminar Schedule homepage.