Firstly, learning (including both subject domain learning, and what is sometimes called social learning) plays an important role in situations which can be characterized by social dilemmas. However, this role is not well understood, and represents an alternative approach to understanding social dilemmas vis a vis rational utility maximization.
Secondly, our paradigm for simulating learning in SimEd can be seen as a kind of commons problem: learning in a school depends on both students and teachers contributing to the "commons" (the interactive environment of the classroom), but the immediate cost of not contributing is generally less than the cost of contributing. Hence some form of social control is usually required to maintain the commons.
My publications include the following papers on SimEd: