A CAUTIONARY NOTE ON USING (MARCH) CURRENT POPULATION SURVEY AND PANEL STUDY OF INCOME DYNAMICS DATA TO STUDY WORKER MOBILITY

Abstract

The monthly Current Population Survey (CPS), with its annual demographic March supplement, and the Panel Study of Income Dynamics (PSID) are the leading sources of data on worker reallocation across occupations, industries, and firms. Much of the active current research is based on these data. In this paper, we contrast these data sets as sources of data for measuring the dynamics of worker mobility. We find that (i) (March) CPS data are characterized by a substantial amount of noise when it comes to identifying occupational and industry switches; (ii) March CPS data provide a poor measure of annual occupational mobility and, instead, most likely measure mobility over a much shorter period; (iii) (the changes in) the procedure to impute missing data have a dramatic effect on the interpretation of the CPS data in, e.g., the trend in occupational mobility. The most important shortcomings of the PSID are the facts that (i) occupational and industry affiliation data are available in most years at an annual frequency; (ii) the PSID's sample, by design, excludes immigrants arriving in the United States after 1968; (iii) the Retrospective Occupation–Industry Files with reliable occupation and industry affiliation data are available only until 1980.