Richardson’s Multilateral Military Expenditure Model Augmentation via Sentiment Analysis of News Data in the Framework of Complexity Eco-nomics

Abstract

Complexity economics incorporates factors that are not accounted for in classical economic thought. With this paper, we intend to demonstrate the potential of using the complexity economics frame of reference by revisiting an existing model of the past, especially one that has been proven to be effective and, most importantly, could utilize the recent developments in advanced analytics as a tool to address a more realistic social milieu. We propose the introduction of more complex dynamics into an already existing model by using Deep-Learning-based NLP analysis of large news article data to fill in the action-reaction matrix for Richardson’s multilateral arms race model. By introducing said complex dynamics we also opened a discussion about the effect that media has on democratic societies’ military spending and concluded at the first level of analysis that media has a significant influence on the electorates’ decision-making of democratic nations in the matter of arms race and defense budgets. Also, we were able to demonstrate the final form of the augmented multilateral arms race model and its predictive capacity. We hope that our findings will encourage the use of advanced analytics in the framework of complexity analysis and improve the existing models’ performance via big data insights.