Simulation based methods in microeconometrics

Stata programming

Course description
The course is focused on some simulation-based econometric techniques suitable for analysing micro-level data, with emphasis on the new generation of discrete choice models made available thanks to the development of simulation-based estimation.

Topics

  • Simulation preliminaries: integration by simulation, drawing from densities
  • Method of simulated maximum likelihood
  • Method of simulated moments
  • Discrete choice with simulation: multinomial and multivariate choice models for cross-sectional data, binary choice models for panel data.

Prerequisites
Introductory Econometrics (Linear regression Ordinary Least Squares) is a fundamental requisite. Knowledge of linear panel data models, binary choice models for cross sectional data and Maximum Likelihood methods is strongly advised to take advantage of this course ( see contents of LMEC course Econometrics 3).

Learning outcomes
Student will have understood the potential of the simulation based approach to solve inference problems arising in various contexts, and especially in discrete choice analysis. They will know how to implement some selected estimation procedures by way of software STATA.

Assessment methods
An applied problem set will be assigned to groups of 2-3 students. After having prepared their assignment at home, student will have to discuss it with the instructor.

Syllabus

  • Cameron, C.Trivedi, PK (2005), Microeconometrics: Methods and Applications, Cambridge University Press
  • Cameron, C. Trivedi, PK (2009, 2010), Microeconometrics Using Stata, Stata press
  • Train, K., (2003), Discrete Choice Methods with Simulation, Cambridge University Press

Additional references to relevant papers and STATA software will be provided during the course.

Teaching methods
Throughout the course, the presentation of theoretical methods and models will be complemented by their applications to real data using the software STATA.