Econometric Analysis of Panel Data Dynamic Panel Data

Econometric Analysis of Panel Data • Dynamic Panel Data Analysis – First Difference Model – Instrumental Variables Method – Generalized Method of Moments – Arellano-Bond-Bover Estimator

Dynamic Panel Data Analysis • The Model

Dynamic Panel Data Analysis • The Model

Dynamic Panel Data Analysis • Model Assumptions: First-Difference Model

Dynamic Panel Data Analysis • Instrumental Variables

Dynamic Panel Data Analysis • Instrumental Variables

Dynamic Panel Data Analysis • Instrumental Variables Method

Dynamic Panel Data Model • Instrumental Variables Estimation • HAC Variance-Covariance Matrix

Dynamic Panel Data Model • Arellano-Bond Estimator is based on Generalized Method of Moments (GMM) – With a proper choice of a p. d. f. weighting matrix γ

Arellano-Bond Estimator

Arellano-Bond Estimator • Notations

Arellano-Bond Estimator • First-step GMM Estimator

Arellano-Bond Estimator • Second-step GMM Estimator

Example: Returns to Schooling • Cornwell and Rupert Model (1988) • Data (575 individuals over 7 years) – Dependent Variable yit: • LWAGE = log of wage – Explanatory Variables xit: • Time-Variant Variables x 1 it: – EXP = work experience WKS = weeks worked endogenous OCC = occupation, 1 if blue collar, IND = 1 if manufacturing industry SOUTH = 1 if resides in south SMSA = 1 if resides in a city (SMSA) MS = 1 if married UNION = 1 if wage set by union contract • Time-Invariant Variables x 2 i: – ED = years of education endogenous FEM = 1 if female BLK = 1 if individual is black