EERC Methodological Seminars Series
ECONOMETRIC SEQUENCE
Panel Data Analysis
Instructor: Stanislav Anatolyev,
New Economic School,
This seminar is the third part of the econometric sequence. The seminar focused on panel data techniques, a methodology that makes an analysis of a large number of objects through short periods in time. It can be particularly relevant for studying dynamic changes in household behavior, enterprise productivity, investment, regional growth, etc. Studying these methods requires good knowledge of linear and matrix algebra. Beside a chapter in Greene's book, several applied papers will be assigned as readings.
ORGANIZATION
Along with lectures, there will be separate computer sessions. The statistical working tool is Econometric Views. The students are encouraged to work in groups. The instructor will be available during office hours.
LITERATURE
Greene, William (2000) Econometric Analysis, 4th edition, chapter 14
Balestra, Pietro (1996) Fixed Effects Models and Fixed Coefficients Models. Chapter 3 in: The Econometrics of Panel Data, edited by L. Matyas and P. Sevestre, Kluwer Academic Publishers.
Matyas, Laszlo (1996) Error Components Models. Chapter 4 in: The Econometrics of Panel Data, edited by L. Matyas and P. Sevestre, Kluwer Academic Publishers.
Owusu-Gyapong,
Anthony (1986) Alternative Estimating Techniques for Panel Data on Strike
Activity. Review of Economics and
Statistics 68, 526-531.
Lichtenberg, Frank (1988) Estimation of the
Internal Adjustment Costs Model Using Longitudinal Establishment Data. Review of Economics and Statistics 70,
421-430.
Sevestre, Patrick and Alain Trognon (1996) Dynamic Linear Models. Chapter 7 in: The Econometrics of Panel Data, edited by L. Matyas and P. Sevestre, Kluwer Academic Publishers.
Holtz-Eakin,
Hsiao, Cheng (1996) Logit and Probit Models. Chapter 16 in: The Econometrics of Panel Data, edited by L. Matyas and P. Sevestre, Kluwer Academic Publishers.
SYLLABUS
Classical linear regression: review
Structure of panel data: cross-sectional dimension and time-series dimension
Balanced and unbalanced panels, panels and pseudo-panels
Error component models: one-way and two-way
Fixed or random effects?
The fixed effects model: Least Squares Dummy Variables (LSDV) estimation
Testing for fixed effects
The random effects model: Generalized Least Squares (GLS) estimation
The Between and Within estimators
The fixed effects model: Least Squares Dummy Variables (LSDV) estimation
Testing for fixed effects
The random effects model: Generalized Least Squares (GLS) estimation
Tests for poolability
Tests for individual and time effects
The Hausman specification test
Serial correlation in error component models
Dynamic fixed effects model: bias of LSDV estimator
Dynamic random effects model: instrumental variables estimation
Vector Autoregressions: review
Panel stationary VAR with individual effects
Panel stationary VAR with individual effects and shocks to a common factor
Panel nonstationary VAR with individual effects
Binary dependent variable models: review
Fixed effects logit: conditional maximum likelihood estimation
Fixed effects logit: semiparametric estimation
Random effects probit