Statistical Problems In BLS Compensation Surveys When Collected Establishment Data Differs From The Assigned Data

Susan R. Black, Lawrence R. Ernst, and Jason Tehonica


The National Compensation Survey (NCS) is the new BLS survey for measuring employee wages by skill level. The sample design for the NCS includes the selection of a sample of establishments and then the selection of a sample of occupations within each selected establishment. Special sampling and weighting procedures are required for these stages of selection when the data collected are not what was originally intended. The issues discussed include: collection from a unit that is other than the assigned unit, as for example when establishments merge; situations where, because of respondent or interviewer burden, subsampling of physical locations that comprise an assigned establishment or employees within an occupation are required; and a single central selection of occupations that is done for certain large companies instead of a separate selection for each sample establishment within the company.