Ensuring data security in interrelated tabular data

Several service agencies including government organizations are engaged in publishing tabular economic data. On the one hand, it is desirable to publish as much data as possible, for use by the public. On the other hand, it is necessary to ensure that confidential data is not divulged. Cell suppression is a popular method that is used for this purpose. Earlier research has focused on cell suppression at the level of individual tables. We present a method that considers interrelationships between tables while attempting to ensure data confidentiality. The method is general enough to be applied to tables with different geometrical structures, and accommodates different cell suppression procedures at the table level.<
>