With occupation data from the 2000 census now available, it is an appropriate time to analyze occupational employment trends over the 20th century. The shift from a workforce composed mostly of manual workers to one comprising mostly white-collar and service workers is generally known. This article reveals just how radical that shift has been. It also shows that many of the projected employment changes over the 2004-14 period (1) are continuations of trends that began in the previous century.
The article analyzes changes in occupational staffing patterns--occupations and occupation groups as a percent of total employment in the economy--rather than numeric changes. (2) This methodology indexes employment growth to the average for all occupations over the period. Occupations and occupational groups growing faster than average appear as an increasing proportion of total employment, those growing as fast as average as a constant percent, and slower growing or declining ones as a declining percent. (3) For clarity, however, numeric employment data also are given.
Data and methodology
Occupational data presented in this article are from decennial censuses, adjusted by the Inteted Public Use Microdata Series (IPUMS) from the University of Minnesota's Minnesota Population Center. (4) Every census taken in the 20th century used a different system to classify occupations, so data between censuses are not necessarily comparable. IPUMS used the 1950 Index of Occupations and Industries to impose an occupational scheme on data from each census. Because of definitional changes and because some occupations in the 1950 index were components of broader occupations in other years, it was difficult to determine some decade-to-decade employment changes. That is, while the broad trends shown for larger occupation groups and many individual occupations are believed to be relatively accurate, some decade-to-decade changes may reflect data comparability problems between surveys rather than indicating actual changes in employment. (5) Nevertheless, data estimates are shown to the closest thousand; readers should be aware that actual employment may have been somewhat different.
The 1950 census classified all workers into 269 occupation categories, hereafter referred to as occupations; (6) the same census also gives employment estimates for each occupation, in its effort to create a consistent time series, IPUMS reduced the number of occupations to 230. The 1950 census arranged all occupations into 11 major groups, as shown in chart 1, but, with a few exceptions, no subgroups--all occupations were just listed alphabetically. (7) To better analyze growth patterns within these 11 major groups, this article classifies the majority of occupations into subgroups, closely corresponding to 2000 Standard Occupational Classification (soc) major or minor groups. (8)
Some 1950 occupation and group names are gender specific or differ in other ways from those in current use, and their coverage of occupations also may differ. In addition, in 1950, some occupations were classified into major groups different from those they were classified into in 2000. For example, cashiers, judged a sales occupation in 2000, constituted a clerical occupation in 1950, and the category of farmers and farm managers, which formed a...