With the help of Monte Carlo studies, we conclude that, overall, the bias of the naive estimators when multiplicative measurement error and blanking are considered as a disclosure limitation technique is substantial and does not decrease with larger sample sizes-the naive use of such an anonymized dataset is not advisable. We show that IPW-SIMEX and MAT-SIMEX estimators perform very well in particularly reducing the bias. In other words, we show that noise multiplication combined with blanking as a masking procedure does not necessarily lead to a severe reduction in the estimation quality. Even if the statements made in this paper have to be conditioned on the data generating process of the MC experiments, we get expertise on how appropriate estimation techniques allowing for the consequences of disclosure limitation techniques lead to consistent estimates of the true parameter of interest.
