Advanced micro panel models: theory and applications

Part 1: Micro Panels

Course description
The course aims to provide a comprehensive overview of the more advanced techniques for modeling panels of micro data, such as companies’ balance sheets and households’ surveys, where the individual dimension is much larger than the time dimension.
At the end of the course, students will acquire the set of basic tools and techniques which are necessary to understand the existing empirical literature based on micro panel data, to assess the strengths and weaknesses of an empirical paper, and to make their own empirical models. Therefore, this course is open to students interested in empirical issues, either because they plan to make empirical works as a part of their research project, or because they would like to be able to read with critical view the applied panel literature available in their area of specialization.

1. Static panels. Understanding the clustered data structure. Dealing with endogeneity, and measurement error problems. Methods: variance decomposition; handling unobserved heterogeneity; instrumental variables (IV) with panel models; unbalanced panels and selection bias.
Application: intangible capital and productivity of firms.
2. Dynamic panels. The problem of endogeneity, predetermines, Nickell’s bias; overfitting and weak instruments. Methods: alternative data transformations; first differences IV, GMM diff-level-sys estimators, bias-corrected LSDV.
Applications: corporate capital structure models; investment and uncertainty relationship (the Euler and the reduced-form approach).

Econometrics 1 and 3, LMEC, or knowledge of econometrics to a level of the first six chapters of Verbeek M., A guide to Modern Econometrics, 3rd ed, Wiley.

Teaching methods
Each lecture will be devoted to specific methodological aspects both from the theoretical and the applied point of view, using Stata package in class. Students can use their mobile PCs in class, as long as they have installed the Stata software. 

Assessment methods
Students can choose one of the following three: (1) Applied exercises based on a list of questions and data provided by the teacher; (2) preliminary inspection of their own data; (3) deepen a specific methodological aspect of the course. Assessment methods (2) and (3) assume that the student has started her research project.

Suggested textbooks (in increasing degree of difficulty):
Wooldridge Jeff M. (2003) Introductory Econometrics. A Modern Approach, Thomson South-Western, Ch. 13-14 (panel).
Verbeek Marno (2008) A guide to Modern Econometrics, Wiley, 3rd edition ch. 10 (panel)
Greene William H. ( 2002) Econometric Analysis, Prentice Hall, 5th edition, ch. 13 (panel) and 14 (until section 14.2.3, SUR).
Wooldridge Jeff M. (2002) Econometric Analysis of Cross-Section and panel Data, Cambridge Mass.: MIT Press, ch. 10 and 11 (panel data), ch 7 (SUR).