Trust and Performance: Exploring Socio-Economic Mechanisms in the “Deep” Network Structure with Agent-Based Modeling

This paper extends the concept of interaction platforms and explores the evolution of interaction and cooperation supported by individuals’ changing trust and trustworthiness on directed weighted regular ring network from the angle of micro scope by using agent-based modeling. This agent-based model integrates several considerations below via a relatively delicate experimental design: 1) a characteristic of trust is that trust is destroyed easily and built harder (Slovic, 1993); 2) trustworthiness may be reflected on both strategy decision and payoff structure decision; 3) individuals can decide whether or not to be involved in an interaction; 4) interaction density exists, not only between neighbors and strangers (Macy and Skvoretz, 1998), but also within neighbors; 5) information diffusion. In this agent-based model, marginal rate of exploitation of original payoff matrix and relative exploitation degree between two payoff matrices are stressed in their influence of trust-destroying; influence of observing is introduced via imagined strategy; relationship is maintained through relationship maintenance strength, and so on. This paper treats number of immediate neighbors, degree of embeddedness in social network, mutation probability of payoff matrix, mutated payoff matrix, proportion of high trust agents and probabilities of information diffusion within neighborhood and among non-neighbors as important aspects happening on interaction platforms, and the influences of these factors are probed respectively on the base of a base-line simulation.