Text Mining in Behavioral Economics with a Focus on the Solution of Social Dilemma Situations

2021-2026, funded by the VolkswagenFoundation

The aim of this project is to establish text mining as a method in the field of behavioral economics. The method of text mining, i.e. the computer-assisted analysis of large quantities of text, has great potential to improve our understanding of economic decision-making. While it is already a firmly established method in other disciplines, there are rather few economists using this method yet. Text mining offers a wide range of possibilities to understand the economic behavior of humans and the deliberation processes behind their decisions. On the one hand, it makes it easier to evaluate very large data sets (e.g. conversations in social media); on the other hand, computer-assisted methods can assess and quantify the differences between different texts (e.g. negotiation protocols) more objectively than human readers can.

A particular focus of this project is on the application of text mining methods to find out more about how people succeed in finding cooperative solutions in social dilemma situations. The basic economic problem in such situations is that cooperative behavior has a positive externality (the private benefit is smaller than social benefit), which has the consequence that individually optimal behavior results in too little cooperation relative to the efficient level. There is plenty of research showing that communication helps solving the dilemma, but apart from the observation that communication helps we know little about why this is the case. This is where this project sets in.