An equicorrelation Moulton factor in the presence of arbitrary intra-cluster correlation

Abstract

This note highlights the potential pitfalls of using an equicorrelation model to estimate standard errors when the true model has arbitrary intra-cluster correlation. It derives a generalized equicorrelation Moulton factor that quantifies the potential biases in standard errors for OLS estimators. As with the famous Moulton factor, the key role is not played by the correlation of the error terms but rather by the intra-correlation of the covariates themselves.