Diffusion of behavior in network games with threshold dynamics

In this paper we propose a generalized model of network games to incorporate preferences as an endogenous driving force of innovation. Individuals can choose between two actions: either to adopt a new behavior or stay with the default one. A key element is an individual threshold, i.e. the number or proportion of others who must take action before a given actor does so. This threshold represents an individual's inclination to adopt the new behavior. The main novelty of the paper is to assume that the thresholds are endogenously determined. Agents change their inclination by exposition to other inclinations in the social network. This provides a coupled dynamical system of aggregate adoption rate and inclinations orchestrated by the network. With our model we are able to explain a variety of adoption behavior. Of particular interest is the existence of non-monotonic behavior of the aggregate adoption rate which is not possible in the benchmark model without inclination. Our model is therefore able to explain “sudden” outbreaks of collective action. This suggests to reinvent the common static and exogenous concept of a tipping point by defining it endogenously generated by the network.