An official website of the United States government
OCSP surveys have a full or limited industrial scope. Most surveys use the full scope which covers establishments employing 50 workers or more in goods producing industries (mining, construction, and manufacturing); service producing industries (transportation, communications, electric, gas, and sanitary services; wholesale trade; retail trade; finance, insurance, and real estate; and service industries); and State and local governments. Private households, agriculture, the Federal Government, and the self-employed are excluded from the survey coverage. Limited scope surveys do not cover governments, mining and construction establishments, and their coverage of services is limited (health and education are the most notable exclusions). The limited scope surveys are a result of 1) our earlier Area Wage Survey Program, and are still used in some localities because of resource limitations, and 2) ESA selected areas, where the limited scope fits their needs. Limited scope surveys not selected by ESA will become full scope surveys with new samples starting in 1996 and 1997.
Sample design
OCSP samples are designed according to generally accepted probability sampling practices. Sampling frames are developed primarily from lists of establishments provided by State Unemployment Insurance Agencies. Sample sizes are scientifically determined, depending on the size of the universe, the diversity of the occupations to be studied, the relative dispersion of earnings among establishments, the distribution of establishments by size, and the planned levels of presentation of survey findings.Sampling errors are not uniform for the occupations studied, since, for each occupation, the dispersion of earnings and frequency of occurrence among establishments differs. OCSP samples are designed so that published average earnings will not differ substantially from the average that would result if all establishments, rather than a sample, were surveyed. We use the following standards in designing our samples and in assessing the publishability of the data: white-collar jobs — relative standard error must not exceed 5%; blue-collar and service jobs, relative standard error must not exceed 7.5%.
Data Collection
Data are obtained primarily by personal visits of the Bureau's field economists to a sample of establishments. To conserve funds, data are also collected by telephone when conditions in the establishment warrant. Data are collected for occupations from the following employment groups: Professional and administrative; technical and protective service; clerical; maintenance and toolroom; and material movement and custodial. Occupational classification is based on a uniform set of job descriptions.
Confidentiality
We collect OCSP data under a pledge of confidentiality. We promise that individually identifiable data will be accessible only to authorized persons (BLS employees) and will be used only for statistical purposes. For this reason, we screen the survey estimates to ensure that data for individual establishments are not disclosed.
Last Modified Date: October 16, 2001