In employment data tables, categories are sometimes marked with an asterisk (*) or “ND” to indicate nondisclosable or suppressed information.
This is because of the Bureau of Labor Statistics’ disclosure rules for the Quarterly Census of Employment and Wages, which aim to protect the confidentiality of specific employers’ numbers and identity when releasing these figures. Data are typically suppressed in small geographic areas, an industry dominated by a single employer, or where one segment of government dominates (but information on federal employees is fully disclosable). This is because if the pool is small enough, it may be possible to distinguish the results of a single or handful of entities.
There are two types of data omission. Primary suppression is required when an employer's identity or data can be directly inferred from the numbers. Primary suppression in a category is determined by a BLS formula based on the number of establishments, the total employment, the number of employers, and the contribution of the largest employers to total wages and jobs.
Secondary suppression is necessary when looking at certain figures may make it possible to infer the value or identity of other withheld employment. In that case, both categories must be withheld to protect anonymity.
For more on the BLS’s QCEW data methods, see the frequently asked questions at http://www.bls.gov/cew/questions-and-answers.htm.