Does import competition worsen the gender gap? Evidence from matched employer-employee data

Using Italian matched employer-employeedata, I examine how accounting for unobserved worker or firm heterogeneity can impact estimates of import competition's impact on industry-level gender wage gaps, and how this can be driven by changes in the composition of female workers and firms within affected industries. First, in wage regressions, I find that import competition lowers women's wages relative to men, but only in specifications that include worker or firm fixed effects. Accounting for these sources of heterogeneity matters because: (1) women that earn low wages are more likely than men to change industries or leave the sample, and (2) firms that employ women are more likely to exit and shrink due to import competition. My findings illustrate how, using data or methods that do not account for worker and firm heterogeneity, researchers can conclude that import competition can improve gender equality, when in fact gender equality is worsened.