Is There Really an Export Wage Premium? A Case Study of Los Angeles Using Matched Employee-Employer Data

Abstract

This article investigates the effects of exporting on wages, specifically the claim that workers are paid higher wages if they are employed in manufacturing plants that export vis-à-vis plants that do not. Past research on U.S. plants has supported the existence of an export wage premium, though European studies dispute those results, calling for more care in econometric investigation to control for worker characteristics. The authors answer this call by developing a matched employee-employer data set linking worker characteristics from the one-in-six long form of the Decennial Household Census to manufacturing establishment data from the Longitudinal Research Database. Analysis focuses on 1990 and 2000 data for the Los Angeles Consolidated Metropolitan Statistical Area. Results confirm that the average wage in manufacturing plants that export is larger. However, after controlling for worker characteristics such as age, gender, education, race, and nationality, the export wage premium vanishes. That is, when comparing workers with similar characteristics, there is no wage difference between exporting and nonexporting plants. These results concord with recent findings from Europe and elsewhere.