Agent Based Strategizing PDF Download

Despite its importance, the relationship between an organization and its environment is difficult to study through real-world, empirical analysis because of the ambiguities and complexities involved. This work advances our understanding of how organizations interact with their environments by modeling expected environmental complexities in an multiple agent-based computer simulation. Integrating principles derived from organization theory and computational organization theory, an agent-based simulation rule set is developed, and then extended to the study issues of diversification and CEO compensation, two unresolved areas in strategic management. With regard to diversification, results obtained through various ANOVA tests, difference tests, and multiple regressions found generally that a related diversification strategy was positively associated with profitability, while unrelated diversification was consistently found to be unprofitable, but was related to greater longevity. Furthermore, diversification was found to be a more prevalent strategy in harsher environments, and there appeared to be a tendency for the most profitable agents to choose a very conservative, related strategy under high-harshness conditions, and a moderate, related strategy under low-harshness conditions. CEO pay results using regression indicated that pay through agency interacts positively with pay through stakeholder performance, while a series of histograms for agents in different environmental extremes indicated that profit-seeking agents moved away from a high-stakeholder proportion and toward a more moderate proportion. Conversely, it was apparently profitable to incorporate a higher proportion of fit-based pay when conditions faced by the agent are more difficult. The implications of the results are discussed in light of the extant literature on diversification and corporate governance.