Understanding Drivers’ Behaviour at Roundabouts: Do Motorists Approach a Bargaining Environment ‘Playing’ a Natural Ultimatum Game?

In terms of urban and extra-urban driving environments, one of the most common phenomena of the last two decades has been the proliferation of roundabouts, which in some cases have replaced the traffic light systems and, in other cases, the horizontal signs typical of road junctions. In the management of traffic flows, roundabouts are often regarded as satisficing solutions, that is, they increase average traffic speed while reducing the number of road accidents satisfactorily. Besides from the role played by physical, structural features, deeply investigated by designers, engineers, and other scientists alike, the social outcome of roundabouts largely depends on the behaviour they are capable of inducing in motorists—and drivers in general—as well as on social cognition. It is very likely, though, that drivers’ behaviour differs across individual agents because of exogenous factors, such as cultural and local traditions, education, family upbringing, and other social backgrounds and cognitive relationships. Here, we study the behaviour of motorists found in metadata, that is, data collected in other studies, with a particular focus on Italy and the United States. The basic idea is to compare the decisions made by motorists at urban and extra-urban roundabouts to those made by economic agents interacting in an Ultimatum Game with discrete offers. Nonetheless, the novelty and relevance of this work lie primarily in the emergence of roundabouts theoretically understood as bargaining environments examined through the Ultimatum Game. Analysing some metadata, we find that motorists might behave as rational agents, although fundamental heuristic factors still need to be thoroughly investigated in future work. This approach allows us both to add very preliminary findings to the vast economic literature dealing with the experimental evidence about Subgame Perfect Nash Equilibria in Ultimatum Games and, above all, to develop a research project based on large-scale natural experimentation and data mining, from classical to modern behavioural economics, there and back again.