Moments-based methods for structural estimation: theory and applications
with Alexandros Theloudis, Luxembourg Institute for Socio-Economic Research and University College London.
Luxembourg Institute of Socio-Economic Research (LISER)
Maison des Sciences Humaines
11, Porte des Sciences
L-4366 Esch-sur-Alzette / Belval
MSH LISER Seminar room 1- 1st floor
09:30 am
05:00 pm
For inquiries:

This one-day course will present four interrelated estimators often used in the estimation of structural models: Generalized Method of Moments, Minimum Distance, Method of Simulated Moments, and Indirect Inference. We will examine the theoretical background of these methods and their relevance for various modern applications, such as the estimation of earnings dynamics, consumer demand equations, and dynamic discrete labor supply models. The overall focus of the course will be towards providing practitioners with instructions needed to understand and use these methods; the course will abstract from proofs or derivations. We will examine practical issues that arise with implementation, such as inference, efficiency, small sample biases, hypothesis testing, or under-identification. We will also discuss important technical issues pertaining to the workings of multi-dimensional optimization that lies at the heart of the aforementioned estimators. Finally, we will apply these techniques in real time using prepared data. Although MATLAB will be the software used for this illustration, reference to other software and programming languages will be made. We encourage participants to have their laptops with them so as to follow the practical application on their own computers. The course will consist of two 3-hour sessions on Tuesday 28 March 2017. Session 1: 10.00 - 13.00 Generalized Method of Moments, Minimum Distance, Inference Session 2: 14.00 - 17:00 Method of Simulated Moments, Indirect Inference, Technical Issues, Practical Application Lunch and refreshments/coffee will be provided. A detailed program will be circulated to confirmed participants.