Agent-Based and Macroscopic Modeling of the Complex Socio-Economic Systems

Abstract

Purpose – The focus of this contribution is the correspondence between collective behavior and inter-individual interactions in the complex socio-economic systems. Currently there is a wide selection of papers proposing various models for the both collective behavior and inter-individual interactions in the complex socio-economic systems. Yet the papers directly relating these two concepts are still quite rare. By studying this correspondence we discuss a cutting edge approach to the modeling of complex socio-economic systems. Design/methodology/approach – The collective behavior is often modeled using stochastic and ordinary calculus, while the inter-individual interactions are modeled using agent-based models. In order to obtain the ideal model, one should start from these frameworks and build a bridge to reach another. This is a formidable task, if we consider the top-down approach, namely starting from the collective behavior and moving towards inter-individual interactions. The bottom-up approach also fails, if complex inter-individual interaction models are considered, yet in this case we can start with simple models and increase the complexity as needed. Findings – The bottom-up approach, considering simple agent-based herding model as a model for the inter-individual interactions, allows us to derive certain macroscopic models of the complex socio-economic systems from the agent-based perspective. This provides interesting insights into the collective behavior patterns observed in the complex socio-economic systems. Research limitations/implications –The simplicity of the agent-based herding model might be considered to be somewhat limiting. Yet this simplicity implies that the model is highly universal. It reproduces universal features of social behavior and also can be further extended to fit different socio-economic scenarios. Practical implications – Insights provided in this contribution might be used to modify existing policy making tools in order to cope with the social transformations in the contemporary society. Originality/Value – The relationship between the inter-individual and the collective behavior is an interesting topic considered to be coming from rather different fields by many scientists. Yet the topic has received due attention only in the recent years. Consequently, the truly systematic approaches directly bridging between these two concepts are somewhat rare. These approaches also differ among themselves – some of the research groups consider questionnaires to understand the individual incentives of the humans, some suggest varying applications of the known physical models and some have roots in the behavioral economics and utility optimization. Our approach in this sense is unique as we start from a simple agentbased herding model and use the ideas from the statistical physics to obtain its macroscopic treatments for the different socio-economic scenarios. In this contribution we present our previous approaches, namely considering new product diffusion in the market and also a financial market model, and also our most recent results, related to the leadership in the social communities and predator-prey type competition in the socio-economic systems. To the best of the authors’ knowledge, the correspondence between the considered simple agent-based herding model and the considered macroscopic models was not previously discussed by the other research groups. Research type: research paper.