The fixed-effects vector decomposition estimator is a method used to estimate time-invariant variables in panel data models with unit effects. Researchers found that this estimator is equivalent to a standard instrumental variables approach under certain conditions. The estimator gives accurate estimates for time-varying variables but underestimates standard errors. It becomes inconsistent when time-invariant variables are endogenous. The original Monte Carlo evidence on this estimator's performance was found to be incorrect. An alternative shrinkage estimator is recommended for better results unless the endogeneity issue is minor or there are no suitable instruments available.