Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
This paper aims to isolate the ethnic gap on the labor market that can be attributed to ethnicity and not to differences in individual characteristics or residential location. Controlling for residential location is important as ethnic minorities often live in distressed neighborhoods. It is also challenging because spatial sorting is likely to differ across ethnicities because of labor- or housing-market discrimination. This paper shows that controlling for neighborhoods and observed individual characteristics fails to provide a consistent estimate for the component of the gap accountable to ethnicity only. However, under some assumptions, the quantity of interest is set identified even when heterogeneous sorting patterns across ethnicities are allowed for and the set estimate can still be informative. A two-step estimation method is presented and applied to explain the ethnic employment differential in France, between French individuals of North African ancestry and those with non-immigrant parents. Most of the gap is not due to differences in residential location or individual characteristics, but rather to ethnicity itself.