Oslo Summer School in Comparative Social Science Studies 2010


Topics in Applied Microeconometrics

Lecturer: Assistant Professor Matthew J. Wiswall,
Department of Economics,
University of New York, USA

Main discipline: Economics

Dates: 2 - 6 August 2010
Course Credits: 10 pts (ECTS)
Limitation: 30 participants


Objectives
This is a course in applied microeconometrics suitable for PhD graduate students who have completed the first year sequence in econometrics.  The course is intended as a “bridge” between the first year coursework in econometrics and the second year coursework in the applied microeconomics fields (labor, IO, development, trade, etc.).   Other advanced students who have an empirical component to their dissertation research might also find parts of the course useful.  The emphasis of the course is the practical and computational aspects of commonly used econometric methods.

In each lecture, I plan to discuss a specific applications of the methods we have covered.  I will also provide students with some examples of econometric techniques using the software Matlab and STATA.  Prior familiarity with STATA and Matlab is not required but is highly recommended.  I will not provide a general introduction to the software during class time.

 

Essential readings for preparation to the course

For further background, I also recommend:

 

LECTURE OUTLINE

Class Hours: 20 total hours of lecture over the course of 5 days, divided into 4 topics.

Topic 1: Numerical Methods (3 hours)
Topics covered: numerical derivatives, numerical quadrature, Monte Carlo integration, drawing univariate, multivariate, and truncated random numbers, optimization methods including Newton-Raphson, Nelder-Mead simplex, and simulated annealing, constrained optimization, parallel computing.

Required Readings:

Recommended Background Reading:

 

Topic 2: Regression and Selection (6 hours)
Topics covered: conditional expectations and linear projection, short and long regression, sampling weights, measurement error, omitted variables, causal inference/treatment effects, matching, kernel density estimation, local regression, control functions, Heckman two-step method, distribution theory for generated regressors, semi-parametric selection models.

Required Readings:

Recommended Background Reading:

 

Topic 3: Instrumental Variables (6 hours)
Topics covered: 2SLS, GMM interpretation, over-identification tests, optimal instruments, weak instruments, missing instruments, split-sample IV, heterogeneous treatment effects, local average treatment effects, non-linear models, variable treatment intensity, average causal response, natural experiments, regression discontinuity.

Required Readings:

Recommended Background Reading:

 

Topic 4: Simulation Based Estimation (5 hours)
Topics covered: method of simulated moments, simulated maximum likelihood estimation, indirect inference, control variates, antithetic refinements, quasi-Monte Carlo methods, importance sampling; smooth simulators, GHK simulator.

Required Readings:

Recommended Background Reading:

 

The lecturer
Matthew Wiswall is an assistant professor of economics at New York University with affiliations at NYU's Steinhardt School of Education, the Institute for Education and Social Policy, Institute of Human Development and Social Change, and Statistics Norway.  Dr. Wiswall earned his Ph.D. in economics in 2005 from the University of California-Los Angeles and specializes in applied microeconomics and applied econometrics. Dr. Wiswall has conducted research on teacher quality, job training, gender differences in education and labor market outcomes, and family influences on child development. His research has been funded by the US National Science Foundation, the Spencer Foundation, and the US Dept. of Education.

Main page - Current Oslo Summer School Program