Ethnicity, Language, and Workplace Segregation : Evidence from a New Matched Employer-Employee Data Set

We describe the construction and assessment of a new matched employer-employee data set (the Decennial Employer-Employee Dataset, or DEED) that we have undertaken as a part of a broad research agenda to study segregation in the U.S. labor market. In this paper we examine the role of segregation by Hispanic ethnicity and language proficiency, contributing new, previously unavailable descriptive information on segregation along these lines, and evidence on the wage premia or penalties associated with this segregation. The DEED is much larger and more representative across regional and industry dimensions than previous matched data sets for the United States, and improvements along both of these dimensions are essential to isolating the importance of segregation by language and ethnicity in the workplace. Our empirical results reveal considerable segregation by Hispanic ethnicity and by English language proficiency. We find that Hispanic workers, but not white workers, suffer wage penalties from employment in a workplace with a large share of Hispanic workers, and even more so a large share of Hispanic workers with poor English language proficiency. In addition, we find that segregation of Hispanic workers among other Hispanics with similar English language proficiency does not reduce the penalties associated with poor own language skills