IV-GMM and Dynamic Panel Data (DPD) estimation. Answer the following questions: (a) Why do we include the lagged dependent variable as a regressor in a panel data model? List two reasons (no examples). (b) Mention 3 reasons to use a large sample size for IV-GMM and/or DPD estimation? Explain briefly each one
(a) Occurrence of autocorrelation arising from model misspecification that can be reduced by including lagged dependent variable as a regressor.
Including lagged Dependent Variable helps to find the maximum variance and make other Dependent variable's effects less significant. By not including the lagged Dependent variable will lead to neglect to do variable bias and results might be unreliable.
(b)Using the large sample size leads to less scope for errors biases despite heterogeneity.
Get Answers For Free
Most questions answered within 1 hours.