If I Scratch Your Back, Someone Else Will Scratch Mine？Trust Is Not as Easy as You Have Imagined：An Agent-Based Model of Indirect Reciprocity Trust Game

• Main Authors: Li-Fen Huang, 黃莉芬
• Other Authors: Bin-Tzong Chie
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/2rvq98

碩士 === 淡江大學 === 產業經濟學系碩士班 === 106 === This study adds an indirect reciprocity mechanism to the network-based trust model. Literatures have pointed out that the importance of reputation in indirect reciprocity. In order to secure a good reputation, agent must not only establish a good reputation but also continuously cooperate with other trustees to accumulate wealth. In this model, we observe how agents interact under the indirect reciprocity and whether agents can be paired successfully therefore to form a larger network topology affecting wealth creation and distribution. This model makes the connection mechanism closer to real life situation by introducing the act of collecting third-party information into the interlinking process. The agent''s information access is not random but built through a dynamic social network. In the process of trust game, trust plays an important role. Following the definition of existing literature, the trust mentioned in this paper is called myopic trust. It influences the formation of social networks in indirect reciprocity. According to the increasing or decreasing of the capital in the previous round, the agent (trustor) will decide whether to build or eliminate the link in the next round. The decision of building a link is partially based on the rating of that potential link from the agent’s friends (trusted). We then combine friends and agent’s rating to decide whether to build a network with the potential link. The message from agent’s friends is given randomly and then the agent needs to decide the amount of portfolios and feedbacks. According to the return of the previous period, the agent will effectively allocate his investment portfolio and use the logit model to allocate how much feedback to their trustors. The results show, in the society of myopic trust, the agents are less likely to establish trust with others and this has a negative impact on economic growth and wealth distribution. As for the building of connection, the more the friends an agent inquires, the lower the chances of building of connection and the lower the rate of accumulation of wealth, especially at a low technology level. In the society of myopic trust, the social capital multipliers are negatively related to the Gini index. On the contrary, regardless of the size of the capital multiplier, wealth-inequality is present in the society.