The dynamic panel estimators developed by Arellano and Bond (1991), Arellano and Bover (1995) and Blundell and Bond(1998) are best for situations where there are panels with small time periods (T) and large cross sections (N); the dependentvariable is dynamic (depends on its past values); independent variables are not strictly exogenous (they are correlated withpast and possibly current realizations of the error); and there are time invariant individual fixed effects and heteroscedasticityand autocorrelation within individuals but not across them. The Arellano-Bond estimation starts by transforming all regressors,usually by differencing, and uses the generalized method of moments (GMM), and is called a difference GMM. The systemGMM estimator combines the standard set of equations in first-difference with a suitable lagged level as instruments, and anadditional set of equations in levels with suitably lagged first differences as instruments. Generally, linear difference and systemGMM estimators have one and two step variants. The two step variant uses residuals from the one-step estimates and is asymptotically more efficient than the one-step.
đang được dịch, vui lòng đợi..
