Editors' introduction

Some research papers selected for publication on the topic of Generalized Method of Moments (GMM) and its recent developments, including information-theoretic approaches, many moments, and non-parametric instrumental variables, are discussed. The paper by Arellano, Hansen, and Sentana presents a testing method for the null hypothesis that a model is not identified by a set of moment restrictions. Hall, Han, and Boldea study the estimation and tests of structural changes in linear regressions with endogenous regressors. The paper by Antoine and Renault addresses more generally the issue of efficient GMM estimation when identification goes through several pieces of information with different rates of convergence. Anatolyev's paper considers a linear regression with homoskedastic error and studies testing when the number of regressors grows proportionately to the sample size. Florens and Simoni consider a non-parametric regression model where the regressors are endogenous and instruments are used to identify the unknown function.