Identifying the Effects of New Information in Dynamic Models of Individual Choices
I derive a model of dynamic discrete choice-making that provides a framework for studying the effects of changing information on the behavior of individuals. The framework uses the same building blocks as standard cross-sectional models of choice, and shows how these can be generalized to dynamic choice. There are three key features of this model. First, the model has a special case a standard cross-sectional model of discrete choice, such as multinomial logit, as the time between choices becomes small. Second, the model provides a framework that allows for choices to be observed at arbitrary intervals of time, allowing for the analysis of complex panel survey designs that randomize the times of interviews. And third, the model alleviates a technical yet substantively important restriction imposed by other models on how the systematic component of utilities may change over time.