Symmetrically Normalized Instrumental-Variable Estimation Using Panel Data

Abstract

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __In this paper we discuss the estimation of panel data models with sequential moment restrictions using symmetrically normalized GMM estimators. These estimators are asymptotically equivalent to standard GMM but are invariant to normalization and tend to have a smaller finite sample bias.They also have a very different behaviour compared to standard GMM when the instruments are poor. We study the properties of SN-GMM estimators in relation to GMM, minimum distance and pseudo maximum likelihood estimators for various versions of the AR(1) model with individual effects by mean of simulations. The emphasis is not in assessing the value of enforcing particular restrictions in the model; rather, we wish to evaluate the effects in small samples of using alternative estimating criteria that produce asymptotically equivalent estimators for fixed T and large N. Finally, as an empírical illustration, we estimate by SN-GMM employment and wage equations using panels of UK and Spanish firms.

Initial conditions and moment restrictions in dynamic panel data models

“…In dynamic panel data models where the autoregressive parameter is moderately large and the number of time series observations is moderately small, the widely used linear generalised method of moments (GMM) estimator obtained after first differencing has been found to have large finite sample bias and poor precision in simulation studies (see Alonso-Borrego and Arellano, 1996). Lagged levels of the series provide weak instruments for first differences in this case.…”

The Effect of Board Directors from Countries with Different Genetic Diversity Levels on Corporate Performance

“…In both the regressions of risk-adjusted returns and Tobin's q, the economic significance strengthens somewhat as we include additional controls. The coefficient on the deviation of diversity in model 4 equals approximately 59 and shows that an increase in the deviation of diversity by 0.016 points (equal to a one standard deviation 12 Applying LIML to the model in differences is more rarely used but is a consistent estimator for dynamic panel data models (see Baltagi, 2005, pp.153-155), In fact, Alonso-Borrego and Arellano (1999) compare the GMM and LIML methods using simulations. Monte Carlo and empirical results show that the GMM can exhibit large biases when the instruments are poor, whereas LIML remains essentially unbiased.…”

The Impact of a Common Currency on East Asian Production Networks and China's Exports Behavior

“…Alonso-Borrego and Arellano (1996) and Blundell and Bond (1998) show that the first-differenced GMM estimators are weakly identified when the instruments are weak in the sense that they have a low correlation with the included endogenous variables. The estimators can be seriously downward biased in two important cases.…”