We present a new platform for large scale networks experiments in continuous time and conduct three experiments on it: groups range from 8 all the way to 100 subjects. These experiments involve pure linking games as well as games with linking and public goods provision.
The major finding is that subjects create sparse networks that are almost always highly efficient. In some cases networks have a very unequal distribution of connections and exhibit small average distances, while in others subjects create equal and dispersed large distance networks. In some cases highly connected nodes earn vastly more while in other cases they earn significantly less than their less connected cohort. Informational overload helps in explaining why highly connected nodes make excessive investments but earn less than the spokes. (Joint work with Syngjoo Choi and Frédéric Moisan)
Contact Person: Michael E. Rose