Codes and Frequencies
OCC1995 is a modified version of the 1995 revised Standard Occupational Classification-based codes (SOC), offering researchers a consistent long-term classification of occupations from 1969 to present. All original occupation information is stored in OCC. For more information on all of the occupational schemes that were used in previous NHIS surveys, please refer to the "User Note on Occupation and Industry Variables." Please see the "Universe" tab for more detailed information on changes in the universe across survey samples.
Occupation coding in NHIS surveys underwent its two biggest changes in 1983, when the NHIS switched from using the Classified Index of Occupation to using the 1977 SOC, and in 2004, after the 2000 SOC was released. The introduction of the 1977 SOC and the 2000 SOC contributed to significant discontinuities in occupation coding in the NHIS history. The IPUMS NHIS staff chose the 1995 revised SOC as the standard because it falls in-between the years when the 1977 SOC and the 2000 SOC were implemented in the NHIS.
The construction of OCC1995 is based on the work done by the IPUMS-USA project, which harmonized Census Bureau occupation classifications from 1950 to present, using the 1990 Census Bureau occupational classification scheme as a standard. Based on a series of technical papers and crosswalks that provide detailed analyses on how the occupation coding scheme for each census year differed from the scheme used in the previous census year, the IPUMS staff traced the proportion of each occupation as it broke down into more specific occupations or as it was combined with others into a more aggregated occupation.
The IHIS followed procedures established by IPUMS-USA staff to create a version of the occupation codes that is comparable, to the extent possible, to the 1995 revised SOC that was used in NHIS surveys from 1995 to 2003. However, the IHIS staff was not able to follow the exact procedures employed by IPUMS-USA staff because the NHIS has released far less detail on occupation codes on the public use files than the Census or American Community Survey. For example, in the 2004 sample, the NHIS occupation code "59" (Communications equipment operators) combined multiple occupational classifications that were distinct in the Census occupational classifications, including code "5010" for switchboard operators, code "5020" for telephone operators, and code "5030" for all other communications equipment operators.
Changes in the universe across years limit the comparability of this variable.
From 1969 to 1996, the universe included sample persons aged 17 and older (1969-1981 samples) and sample persons aged 18 and older (1982-1996 samples) who were in the labor force in the past 2 weeks. According to the Interviewer's Manual, being in the labor force means that the person is currently working, is temporarily absent from a job, or is looking for work. From 1997 to 2003, the question was asked of sample persons aged 18 and older who were working at a paying job last week, were with a job/business but not at work, or were working at a non-paying job last week. The universe for the 2004-forward samples is similar to that of the 1997-2003 samples, except that people who had ever worked are also included in the universe. Users wishing to increase the comparability of OCC1995 should use WORKEV to exclude persons who have ever worked and are not currently employed.
- 1969-1981: Persons age 17+ who were in the labor force (working, temporarily absent from a job, or looking for work) during the past 2 weeks.
- 1982-1996: Persons age 18+ who were in the labor force (working, temporarily absent from a job, or looking for work) during the past 2 weeks.
- 1997-2003: Sample adults aged 18 and older who were working at a paying job last week, with a job or business but not at work, working at a non-paying job last week
- 2004-2017: Sample adults aged 18 and older who were working at a paying job last week, with a job or business but not at work, working at a non-paying job last week, or who had ever worked