In microeconometric models, the variables of interest often feature restricted distributions—for instance with discontinuous support—, which necessitates specific models. Typical examples are discrete-choice models (binary, multinomial, ordered outcomes), sample selection models (censored or truncated outcomes), and count-data models (integer outcomes). The course describes the estimation and interpretation of these models. It also shows how the discrete-choice models can emerge from (structural) random-utility frameworks.

This course has been developed by Jean-Paul Renne. It is illustrated by R codes using various packages that can be obtained from CRAN. This AEC package is available on GitHub. To install it, one need to employ the devtools library:

install.packages("devtools") # in case this library has not been loaded yet

Useful (R) links: