Econometrics on Panel Data, 2009 Fall

Contents

Part I: Micro Panel Data (Weak Cross Sectionally Dependent Panel)

1. Introduction:

I. Difference among Cross sectional and Time series regressions

II. Where does panel gain comes from?

2. Pooling Panel and Random Effects (Estimation)

I. Pooling Regression

II. Random Effects Model

3. Fixed Effects (Estimation)

I. Eyeball Approach

II. Common Time Effects

III. Dummy Variable (on Coefficients) Approach

IV. Dynamic Panel Regressions

4. Asymptotic Distribution and Testing

I. Standard Model with POLS and Random Effects

II. Selection of Panel Length: Large N or Large T?

III. Covariance Matrix Estimation

IV. Testing

5. Dynamic Panel Regression (Issues and Problems)

I. Source of Serial Correlation

II. Modeling Dynamic Panel Regression

III. Inconsistency of LSDV Estimator

IV. Inconsistency of Pooled OLS Estimator

V. Asymptotic Distribution of LSDV Estimator

6. Dynamic Panel Regressions (Solutions)

I. GMM/IV Estimator

II. First Difference MLE

III. X-Differencing

7. Endogeneity and Control Variables

8. Limited Dependent Variable

9. Conclusion and Summaries

Part II: Macro Panel Data (Strong Cross Sectionally Dependent Panel)

1. Introduction

2. How to Model Cross Section Dependence

I. Spatial Correlation

II. Cholesky Decomposition

III. Factor Analysis

3. Estimation of the Number of Common Factors

I. Static Factor

II. Dynamic Factor

4. Estimation of Common Factors

5. Factor Augmented Panel Regression

6. Panel Unit Root Test under Cross Section Dependence

7. TBA

8. TBA

9. TBA

10. TBA

11. TBA