Essays on social preferences, competition and cooperation : an experimental approach

• Main Author: Zhang, Ruike
• Other Authors: Yohanes Eko Riyanto
• Format: Theses and Dissertations
• Language: English
• Published: 2018
• Subjects: DRNTU::Social sciences::Economic theory
• Online Access: http://hdl.handle.net/10356/73217
• Institution: Nanyang Technological University

"Man is a social animal and his choices are not rigidly bound to his own preferences only. An act of choice for this social animal is, in a fundamental sense, always a social act." - Sen, Amartya (1973, pp.252-3) Assuming that humans are only motivated by self interests seems at odds with the fact that humans are not solitary beings. They interact with others, and in their social interactions they would often show concerns for others and (or) receive kindness from others. A variety of models that depart from the pursuit of self-interests assumption have been developed over the last few decades to incorporate individuals’ concerns towards others. Examples of those models are social preferences models based on inequality aversion (Fehr and Schmidt 1999, Bolton and Ockenfels 2000), spitefulness and altruism (Levine 1998), and affinity towards others belonging to the same identity group (Akerlof and Kranton 2005, Chen and Li 2009, Bénabou and Tirole 2011). This thesis focuses on the evaluation of the impacts of group identity in competitive and cooperative setting (the first two chapters) and also on the mechanism to promote cooperation in a social context (the third chapter). Specifically, in the following two chapters, we embed social identities into a variety of individual and group decision making environments. We apply the Minimal Group Paradigm (MGP), which is the standard identity inducement method used in a laboratory setting, in our analysis (see Tajfel and Turner (1979) and Chen and Li (2009) for further details on the MGP procedure). In particular, the first chapter establishes a framework with two-dimensional induced identities and focuses on how people allocate resources to others with multiple identity attributes. The second chapter adopts group Tullock rent-seeking contests with two prize sharing rules, namely the equal and the proportional sharing rules. The former refers to the case where the winning prize is divided equally among members of the winning group, while the latter refers to the case where it is divided proportionally to members’ effort level. An equal sharing rule would encourage contestants to cooperate with other group members, whereas a proportional sharing rule would heighten the competitive feeling among members. The second chapter investigates how the interplay between the identity compositions of contest groups and the sharing rules influences cooperation and competition among contestants. The Tullock rent-seeking group contest under the equal sharing rule has a similar incentive structure akin to the one found in the public goods games. Public good games capture a classic social dilemma in which there is a conflict between self and collective interests (see Ledyard (1995) for a literature survey on public goods games). In such a game, theoretically it is better if everyone cooperates by making full contribution, but it is not achievable because individuals would face strong temptation to free-ride on others (Fischbacher, Gächter et al. 2001). Literature has focused on various mechanisms that can be employed to boost individual contribution. They are, for examples a peer-to-peer (second party) punishment (Ostrom, Walker et al. 1992, Fehr and Gächter 2000, Fehr and Gächter 2002, Cason and Gangadharan 2015), third-party punishment (Fehr and Fischbacher 2004, Carpenter and Matthews 2012), and a centralized punishment system (Andreoni and Gee 2012). The centralized punishment system is widely found in the modern society as some delegated or appointed parties to monitor people’s behaviors, such as government officials, boards of directors, and church elders, but there are only few research studies which focus on that (Putterman, Tyran et al. 2011, Kamei, Putterman et al. 2015, DeAngelo and Gee. 2017). The third chapter intends to delve further into the role of centralized punishment in increasing contribution in a public goods setting. The detailed abstracts of each chapter are presented below. The first chapter reports findings from a laboratory experiment eliciting two-dimensional social identities: a horizontal identity determined either randomly or by preferences and a vertical identity defined by income status and determined either by luck or performance. We also vary income gaps between vertical identity groups. Participants make allocation decisions between two others differing in identity attributes. We find robust evidence of in-group favoritism and that both the identity distance between the allocator and the in-group recipient and income gaps influence the degree of in-group favoritism. The second chapter aims to delve further into the role of within- and between-group identity diversity in influencing individual behaviors in a team competition. We model the team competition as a Tullock rent-seeking group-contest. We apply the Minimal Group Paradigm (MGP) to induce group identity in a lab setting. We are interested in evaluating: 1) how a within-team identity diversity influences individuals’ effort decisions, 2) how the identity composition of an opponent team influences individuals’ effort decisions, and 3) how the identity effect interacts with the prize-sharing rule employed in the competition. We find that a homogenous team yields intra-team cooperation, while the within-team identity diversity escalates competitiveness among the team members. Additionally, for those competing teams sharing a common identity, a sense of rivalry between the competing teams is largely eroded. The third chapter evaluates the role of centralized punishment in promoting collective cooperation to provide public goods. To avoid the race to the bottom in the provision of public goods, this centralized punishment mechanism relies on the use of the unilateral and tie punishment imposed on the lowest contributor(s). In this paper, we aim to examine how severe this unilateral and tie punishment should be to achieve the full-contribution equilibrium. Specifically, we are interested in investigating the size of the punishments that should be meted on the lowest contributor(s). We theoretically derive a range of punishment mechanisms which would lead to full contribution and put them into experimental test. Our results generally substantiate the theoretical predictions except for the more lenient punishment parameters. This discrepancy is successfully explained by individual evolutionary learning.