A revised socioeconomic index of occupational status

Abstract

Duncan's Socioeconomic Index (SEI), a widely used indicator of occupational ranking, is based on education and income data from the 1950 census. The major purpose of this paper is to offer a more contemporary version of this index. There are several reasons for doing so. Not only has the occupational classificatory scheme been altered, but the educational and economic characteristics of the American labor force and of specific occupational groups have changed since 1950. The two decades may also have seen a shift in the relations between the educational and economic attributes of an occupational grouping and its social standing or prestige. Second, the construction of the original SEI rested on the characteristics of the male labor force, rather than those of the total labor force. Third, in the process of updating the index, we illustrate how certain arbitrary decisions (dictated by data limitations) in the construction of the Duncan SEI served to vest the socioeconomic index with some artifactual properties. In the production of an updated version of the socioeconomic index, we use three approaches. First, we experiment with differing measures of the income and educational criteria. Second, we reconstruct the dependent variable, occupational standing, to provide a better approximation of the prestige measure used by Duncan (1961). Third, we consider the attributes of both the male and total labor forces in generating contemporary indexes of occupational status. We also compare the performance of the new socioeconomic indexes in models of occupational attainment against the performance of the original Duncan index and subsequent occupational prestige measures. The paper appends new socioeconomic indexes for detailed occupational titles based on the 1970 census classification of occupations.