Limits and Criticalities of Predictions and Forecasting in Complex Social and Economic Scenarios: A Cybernetics Key

Abstract

Predictions play a key role in assuring the status of “rationality” in decisions. Nevertheless, in the field of social sciences and economics, predictions fail to correctly depict the oncoming scenarios. Why is it so difficult to achieve quantitative prediction of social and economic systems? Can science provide reliable predictions of social and economic paths that can be used to implement effective interventions? As in the notorious “El Farol bar problem” depicted by Brian Arthur (Am Econ Rev 84:406–411, 1994), the validity of predictive models is more a social issue than a matter of good mathematics. Predictability in social systems is due to limited knowledge of society and human behavior. We do not yet have worldwide, quantitative knowledge of human social behavior; for instance, the perception of certain issues or the predisposition to adopt certain behaviors. Though tremendous progress has been made in recent years in data gathering thanks to the development of new technologies and the consequent increase in computational power, social and economic models still rely on assumptions of rationality that undermine their predictive effectiveness. Through some theoretical and epistemological reflections, we propose a way in which the cybernetic paradigm of complexity management can be used for better decision-making in complex scenarios with a comprising, dynamic, and evolving approach. We will show how a cybernetic approach can help to overcome the fear of uncertainty and serve as an effective tool for improving decisions and actions.