The Bureau of the Census is looking into new methods of disclosure limitation for use with establishment tabular data. Currently we use a strategy that suppresses a cell in a table if the publication of that cell could potentially lead to the disclosure of an individual respondent's data. As an alternative to cell suppression that would allow us to publish more data and to fulfill more requests for special tabulations, we are experimenting with adding noise to our underlying microdata. By perturbing each respondent's data, we can provide protection to individual respondents without having to suppress cell totals.
While adding noise is a much less complicated and time-consuming procedure than cell suppression, the question remains as to the utility of the data after noise is added. To preserve the quality of aggregate estimates that would not normally be at risk for disclosure, we tested the option of forcing estimates at certain levels of aggregation to equal their true values before the addition of noise. Interior table cells were then raked to these aggregate cells. In this paper we discuss the advantages and disadvantages of adding noise to microdata as compared to cell suppression, and we describe the results of using noise and raking with the Research and Development survey.
