Abstract
A simple proof of the unbiasedness of L -norm and M-estimators is given for regression models.