Herbert Simon and Agent-Based Computational Economics

The seventh in a series of excerpts from Minds, Models and Milieux: Commemorating the Centennial of the Birth of Herbert Simon.

Shu-Heng Chen and Ying Fang Kao

Herbert Simon is a quintessential interdisciplinary scholar who has made pioneering contributions concerning the notion of bounded rationality, has built models based on it, and has also made important advances in understanding complex systems. His importance in the field of artificial intelligence, which is in turn the inspiration of agent-based computational economics (ACE), is discussed in detail in Chen (2005). Among all the Nobel Laureates in Economics, there are at least three whose work has been acknowledged by the ACE community. They are Friedrich Hayek (1899-1992), Thomas Schelling, and Elinor Ostrom (1933-2012). The last two directly worked on ACE. Schelling’s celebrated work on the segregation model is considered as one of earliest publications on ACE (Schelling, 1971). Ostrom had contributed to the development of the empirical agent-based models (Janssen and Ostrom, 2006). Hayek did not work on ACE, but the connection of his work to ACE has been pointed out by Vriend (2002).

We believe that there is a strong connection between the development of ACE and Herbert Simon and that his influence on ACE is not less, if not more, profound than the previous three. However, to the best of our knowledge, there seems to be no single document that from a holistic perspective addresses this linkage explicitly.[i] We conjecture that the burgeoning of ACE was too late for the time of Simon, who ended his professional life in 2001. However, even so, it still surprises us that so few attempts have been made to connect Simon and ACE, particularly considering that the latter was founded on artificial intelligence and cognitive psychology, the two pillars to which the former has contributed substantially.

In this chapter, we attempt to explore and identify the connections between Simon’s contributions and the development of ACE. We concentrate on his influence on the conception of an individual within an economic or social system, his philosophy regarding how the social systems are organized and can be understood, and finally about how the underlying rules that govern social interactions can be unearthed by the investigator, in this case a social scientist. We also suggest ways with which the future developments within ACE can be geared to be more Simonian in character and to be closer to his vision.

The rest of the chapter is structured as follows. Section 2 provides an overview of the setting in which we place our arguments. We then divide our arguments into three main departments: individuals, complex systems and the epistemology of ACE. In Section 3, the modeling of software agents in light of Simon’s bounded rationality is discussed. In Section 4, various aspects of complex systems are included here. In Section 5, we elaborate on ACE’s potential as an alternative to neoclassical economics. We conclude the paper in the last section.