Panel Model with Multiplicative Measurement Errors

Abstract

The analysis of panel data is a common problem in economic research. Because panel data are often subject to measurement error, it is important to develop consistent statistical estimation techniques that take the measurement error into account. A related problem is given when statistical offices anonymize confidential panel data before publication. In this case, artificial “measurement” errors are often imposed on the data to prevent the disclosure of the identity of observations. Consequently, the anonymized data can be analyzed by using the same statistical techniques as those for measurement error models. While most articles in the literature deal with the analysis of additive measurement errors, this paper is concerned with the estimation of a panel data model when multiplicative error is present. Using the generalized method of moments (GMM), we construct consistent estimators of the parameters of the panel data model and compare them to traditional estimators that are based on the least squares principle.