Complex economic problems and fitness landscapes: Assessment and methodological perspectives

摘要

Recently, a new line of interdisciplinary research in economics, coined complexity economics, has emerged premised on the assumption that the economy should be treated as a complex system. Complexity economists borrow concepts and theories from evolutionary biology such as emergent-phenomena, positive-feedback, self-organization, and diversity. The stated goal of this new subfield is to create more realistic and detailed models of the economy. One popular approach to modeling involves the use of "fitness landscapes" to generate solutions spaces in the search for optima and to model adaptive behavior. Here, in place of production or utility functions, a fitness function provides a simplified, Darwinian evaluation of performance. But recently, there has been a turn toward self-reflection and criticality in the literature; specifically, a debate has surfaced over the limitations of the fitness landscape approach. Par example is the 2014 article titled "Economic Opportunity and Evolution: Beyond Landscapes and Bounded Rationality" appearing in Strategic Entrepreneurship Journal. The authors contest that existing fitness landscapes neglect the emergence of novelty - implicating an ever-changing solution space. However, novelty is not the only limitation of modeling economic phenomena using the fitness landscape. In this paper, I present a methodological assessment of the extant literature in an effort to redraw our attention to the limitations of the fitness landscape in achieving the goal of more realistic and detailed models. Are fitness landscapes serving this goal? If no, what are the limitations? And most importantly, how do we face the challenges head on? (C) 2019 Elsevier B.V. All rights reserved.