Econometric Methods for Cross-sectional and Panel Data
Full course description
The main topics of the course are (1) unobserved effects models for panel data, (2) probit and logit models for discrete choice, (3) tobit and related censored regression models, (4) models dealing with sample selectivity, and (5) the estimation of average treatment effects (a.k.a. policy impact evaluation). Dynamic extensions of the models are considered when feasible. Estimation and testing methods are applied in a number of empirical assignments and their properties are investigated.
- Thorough understanding of the most frequently used econometric models and methods for the analysis of panel data, categorical choice and limited dependent variables.
- Some practice in the application of the methods, the interpretation of the models, and the evaluation of inferences.
- The experience of conducting a theoretical, experimental and/or empirical investigation of the methods.
- Calculus, matrix algebra, probability, mathematical statistics, asymptotic theory, linear statistical models.
- Familiarity with statistical software like Stata and Gauss, Matlab, or R.
- Econometric methods at the level of Greene (2008) or Davidson & MacKinnon (2004), ideally as in courses Econometric Methods I (EBC2111), and Econometric Methods II (EBC2120).
The course is intended for students in the Econometrics Master programme as well as others with a comparable background and motivation. FLUENCY IN MATRIX ALGEBRA AND IN ASYMPTOTIC THEORY is necessary.
An advanced level of English.
- Cameron, A.C. and P.K. Trivedi (2005): Microeconometrics, Methods and Applications, Cambridge University Press 2005. ISBN 978-0521-84805-3.
- Wooldridge, J.M. (2010): Econometric Analysis of Cross Section and Panel Data, Second Edition. MIT Press, Cambridge, MA, 2010, 2nd ed., ISBN 0-978-0-262-23258-6.
These references will be supplemented with a reading list of journal articles and book chapters.