Choice modeling and estimation
Michael Fleermann, UT-EWI-SOR
Many economic decisions involve a choice between finitely many mutually exclusive, qualitative alternatives, also called discrete alternatives. Examples include modes of transportation (bike, car, bus), levels of labor force, residential locations (Area 1, Area 2,.., Area n), or generally choosing one product out of a finite set of different products.
To an observer, these choices are often random due to unobservable characteristics of the decision-maker and the choice alternatives. The purpose of the theory of discrete choice is to provide a structure of the choice probabilities that can be justified from behavioral arguments, such as random utility maximization (RUM).
In the first part of the presentation we will point out several paradigmatic frameworks for decision making and will finally arrive at the so called Mixed Multinomial Logit (MMNL) Model. In the second part we will focus on the challenges of conducting estimation in those models. In the third part of the presentation, we will present directions for future research endeavors.