This book focuses on the structural analysis of demand under block rate
pricing, a type of nonlinear pricing used mainly in public utility
services. In this price system, consumers are presented with several
unit prices, which makes a naive analysis biased. However, the response
to the price schedule is often of interest in economics and plays an
important role in policymaking. To address this issue, the book adopts a
structural approach, referred to as the discrete/continuous choice
approach in the literature, to develop corresponding statistical models
for analysis.
The resulting models are extensions of the Tobit model, a well-known
statistical model in econometrics, and their hierarchical structure fits
well in Bayesian methodology. Thus, the book takes the Bayesian approach
and develops the Markov chain Monte Carlo method to conduct statistical
inferences. The methodology derived is then applied to real-world
datasets, microdata collected in Tokyo and the neighboring Chiba
Prefecture, as a useful empirical analysis for prediction as well as
policymaking.