Utility of income as a random function: Behavioral characterization and empirical evidence

Abstract

The paper proposes a particular approach to model the utility of income. We develop a theoretical framework that restricts the class of admissible functional forms and distributions of the random components of the model. The theoretical approach is based on theories of probabilistic choice and ideas that are used in modern psychophysical research. From our theoretical framework, we obtain the empirical model and the corresponding likelihood function. The empirical analysis is based on a bStated PreferenceQ survey. The model fits the data quite well. Finally, we discuss the concept of cardinality and the implications for consumer demand relations.