William H. Greene

Overview

William H. Greene is Professor of Economics at the Stern School of Business, New York University, one of the most widely read applied econometricians through his graduate textbook Econometric Analysis and the LIMDEP/NLOGIT software. His research centres on microeconometrics: limited-dependent-variable and discrete-choice models, panel data, stochastic frontier / efficiency analysis, and latent class (finite mixture) modelling.

Key Contributions / Features

• Econometric Analysis (Prentice Hall, multiple editions): a standard graduate reference across the discipline.
• LIMDEP / NLOGIT: econometric software implementing the maximum-likelihood limited-dependent-variable and discrete-choice estimators he helped popularise.
• Latent class / mixed models: with Brown and Harris (2014) proposed enforcing the expected-value ordering of latent classes in estimation and an ordered (rather than multinomial-logit) parameterisation of the class-assignment probabilities.
• Stochastic frontier analysis: contributions to true fixed/random-effects frontier models for efficiency measurement.
• Count-data models (Greene 1994): zero-inflated Poisson/NB, a test for zero inflation vs. over-dispersion, and a Heckman-style sample-selection model for counts (Count-Data Regression).

Related

• Latent Class Model
• Discrete Choice Models
• Count-Data Regression
• Selection Model
• Greene (1994) Accounting for Excess Zeros and Sample Selection in Poisson and Negative Binomial Regression Models
• Brown-Greene-Harris (2014)
• Sarah Brown
• Mark N. Harris