, ethnicity, and gender in the United States. However, the data necessary to detect possible discrimination and to act to counter … pictures to identify race, ethnicity, and gender. We show that one can use LinkedIn data to obtain reasonably reliable measures … of workforce demographic composition by race, ethnicity, and gender, based on validation exercises comparing estimates …

work, using recently-constructed matched employer-employee data at the establishment level. We explicitly measure the … importance of network effects for groups broken out by race, ethnicity, and various measures of skill, for networks generated by …

different gender, race, or ethnic groups and how they evolve as buyers obtain more information about seller quality. We consider … have a choice about how strongly to signal gender. We develop a new approach to identifying statistical discrimination … using doctors' choices about signaling their gender. We find evidence of statistical discrimination against female doctors …

effects could differ by race and gender, so implications for hourly earnings do not necessarily extend to overall earnings. We …