08 disclosure masking techniques
Disclosure Masking Techniques
EXPO often employs partial and full masking methods to deal with non-disclosable data in macro aggregates. These are applied in both “primary” and “secondary” disclosure checking processes. Primary disclosure masking is based upon the method formerly employed by BLS, requiring at least three firms (i.e., different U-I account numbers), with no firm comprising 80% or more of the third month employment.1
Secondary disclosure is applied when multiple levels of aggregation are employed, and precisely one of the lower-level (i.e., more digits in the NAICS code) aggregates is found to be non-disclosable, which would force the higher-level (i.e., fewer NAICS digits) aggregate to be masked as non-disclosable. Instead of masking ever higher industry levels, the secondary masking will mark the next smallest macro cell in the lower-level aggregate as non-disclosable (even though that total in and of itself would be disclosable), so that there are now two lower-level cells identified as non-disclosable, allowing for the release of the higher-level aggregate (since someone can no longer calculate the missing lower-level aggregate). These procedures are described in more detail with examples in the paragraphs that follow.
1 This assumes that the 3/80 rule is in effect. However, the P-percent process could be used instead and is described later in this document.
Related Links
