How to do Xtabond2: An Introduction to Difference and System GMM in Stata

Abstract

, are increasingly popular. Both are general estimators designed for situations with "small T , large N " panels, meaning few time periods and many individuals; independent variables that are not strictly exogenous, meaning they are correlated with past and possibly current realizations of the error; fixed effects; and heteroskedasticity and autocorrelation within individuals. This pedagogic article first introduces linear generalized method of moments. Then it describes how limited time span and potential for fixed effects and endogenous regressors drive the design of the estimators of interest, offering Stata-based examples along the way. Next it describes how to apply these estimators with xtabond2. It also explains how to perform the Arellano-Bond test for autocorrelation in a panel after other Stata commands, using abar. The article concludes with some tips for proper use.

The Role of Reciprocation in Social Network Formation, with an Application to Blogging

“…In addition, for several bloggers, these variables have high persistence, thus causing an additional weak-instruments concern. 9 Nevertheless, by relying on various specification of system generalized method of moments (GMM) (Arellano and Bover, 1995;Blundell and Bond, 1998;Roodman, 2009a), and by paying attention to weak identification (Bobba and Coviello, 2007) and instruments proliferation (Roodman, 2009b), we obtain results in line with theoretical predictions.…”

‘No man is an island entire of itself.’ The hidden effect of peers on physical activity

“…The potential presence of such serial correlation is explored by testing the null hypothesis of no second order serial correlation in differences in , to check for first-order serial correlation in the levels of . This is because of the overlapping time periods for the levels of the residuals that are implied (Roodman, 2009). The diagnostic statistic AR(2) is used to address the null hypothesis.…”

Can alcohol make you happy? A subjective wellbeing approach

“…We here use the cluster-robust sandwich estimator to calculate standard errors, which is robust to moment-to-moment correlations in the error term. While dynamic (lagged) FE models violate the assumption that the error is uncorrelated with the regressors, this bias becomes ignorable where the number of time periods is moderately large (Roodman, 2009); a sensitivity analysis restricts the sample to those providing>=20 relevant observations.…”