Panel Data Econometrics

Aims of the course

The objective of the course is to provide necessary training in statistic and econometrics at advanced level for independent research in economics.

Course syllabus

1. Static linear panel data models
1.a. Omitted variable problem
1.b. Assumptions about the unobserved effects
and explanatory variables
1.c. Fixed effects model
1.d. First-difference and least squares dummy variables estimator
1.e. Random effects estimator
1.f. Comparison of estimators
1.g. Extensions with IV
1.h. Two-way high-dimensional fixed effects
2. Dynamic linear panel data models
2.a. Nickel’s bias of fixed effects estimator
2.b. Inconsistency of first-differenced model
2.c. Solutions to the problem
2.d. Anderson-Hsiao (1982) estimator
2.e. Generalized method of moment
2.f Arellano-Bond (1991) estimator
2.g. Blundel-Bond (1998) estimator
3. Binary choice models
3.a. Linear probability model
3.b. Index models for binary response: Probit
and logit
3.c. Maximum likelihood estimation
3.d. Testing in binary response index models
3.e. Pooled probit and logit model in panel
3.f. Unobserved effects in panel data
3.g. Dynamic unobserved effects models and
other extensions
4. Multinomial and ordered response models
4.a Multinomial logit,
4.b. Mixed logit
4.c. Multinomial panel data methods
4.d. Ordered logit and probit models
5. Selection models and estimation of average treatment effects
5.a. Tobit model
5.b. Heckman selection model
5.c. Dynamic Heckman
5.d. Propensity score methods
5.e. Matching methods and treatment effects
5.f. Regression discontinuity design

Course director(s)