Beyond Gaussian averages: redirecting international business and management research toward extreme events and power laws

Abstract

Management research relies predominantly on Gaussian statistics, which assume independent data points and finite variance, effectively marginalizing extreme events as outliers. In contrast, practitioners frequently encounter extreme outcomes driven by positive feedback loops and interdependent interactions. These phenomena typically follow power laws and Paretian distributions characterized by near-infinite variance and scalability. By applying concepts such as self-organized criticality and fractal geometry, researchers can identify the underlying mechanisms of extreme events across diverse natural and social domains. The international business context is particularly susceptible to these dynamics due to the high tensions imposed by globalization, cultural diversity, and rapid technological connectivity. Consequently, traditional statistical models that prioritize averages over extremes produce inaccurate science and offer limited relevance to management practice. Scientific rigor in business studies requires the integration of Pareto-based statistics that account for interdependence and scalability rather than relying on assumption devices that suppress the significance of extreme events. Redirecting research toward the study of power laws ensures a more faithful representation of the complex, interconnected environments in which multinational enterprises operate. – AI-generated abstract.