Fitting Sample Selection Models for Discrete Response Variables
Description
Tools to fit sample selection models in case of discrete response
variables, through a parametric formulation which represents a natural
extension of the well-known Heckman selection model are provided in the
package. The response variable can be of Bernoulli, Poisson or Negative
Binomial type. The sample selection mechanism allows to choose among a
Normal, Logistic or Gumbel distribution.