#	Pagina
attuale pagina	/open-h2020/projects/209252/results.html

Report

Teaser, summary, work performed and final results

Periodic Reporting for period 1 - BEING (Behavioural Economics in Network Games)

Teaser

Summary

What is the problem/issue being addressed?
Over the last 50 years behavioural economics has provided new ways to analyse how people make decisions without imposing strong assumption on individual rationality. One area in which behavioural aspects are particularly relevant is the interpretation and aggregation of information prior to any decision. For instance, the underlying assumption that justifies democracy is that decisions are made by well-informed voters. Understanding how voters can be equipped with correct information and use it properly, in particular while interacting and exchanging information in unregulated social networks, is the key issue that was addressed in this project.

Why is it important for society?
Events during the run-up to the last presidential election in the United States, or to the Brexit referendum have cast doubts on the overall beneficial effects that social media have on society. Understanding how and why some information can easily spread, and what underlying mechanisms on the individual level foster or dampen the spread of information is basis for urgently needed regulation that prevents attacks on European democracies.

What are the overall objectives?
This project combined insights from behavioural economics and the studies on social and economic networks. The first objective was to set a benchmark of information transmission in networks among fully rational Bayesian agents. The second objective was to understand how people update their beliefs when they do not know the reliability of their source of information. The final objective was to understand how the findings of the first two parts interact, that is, how people use and share information that they find via a social network and whose source is not identifiable.

Work performed

The first working package that came out of this project closes a gap in the literature on farsightedness of coalitions. It provides a way to analyse how society makes decisions with long term consequences that have to be adjusted in the future. In this category fall, for instance, the use of new technologies that have not yet been properly regulated in detail. Failing to provide regulation in the future might have disastrous consequences, and peopleâ€™s belief about their own ability to regulate such technology in the future has a huge impact on their willingness to allow an introduction in the first place.

The second working package sets a benchmark for rational transmission of information in social networks and is the first that does so. Agents have to decide between becoming active in a protest or remaining inactive. They become active only if the number of active agents in the overall society is sufficiently large. However, they cannot observe everybody but only their neighbours in a social network, that is family, friends, colleagues, etc. From the behaviour of those they observe they can deduce what is happening outside their vicinity and use that information for their own decision. This model of social learning provides a theoretical underpinning for stylized facts on revolutions, namely that they are quick and unanticipated. The theoretical findings are illustrated using data on protests, revolutions, and political violence around the globe in a time frame from 1976 to 2014.

Final results

Several parts of this project have extended our knowledge of how people make decisions in a social context. First, the concept of â€œfull farsightednessâ€ in social situation has been introduced and solved a problem in the theoretical literature that was raised already 25 years ago. Second, a social learning model in the context of activism has been introduced that allows for any kind of belief updating in which agents understand that events of probability 0 cannot happen. The experimental work is still ongoing, but we expect insights in peopleâ€™s heuristics when dealing with information whose reliability is unknown. This is particularly relevant as information is not anymore obtained exclusively from neutral news outlets with high journalistic standards, but rather from social media which are unregulated.