Unemployment Duration and Subsequent Wages: Evidence from Bunching

with Jonathan Mansfield and David Slichter

We study the effect of unemployment duration on subsequent wages, using a bunching design to account for selection bias. We find that unemployment duration reduces subsequent wages by about 4.4% on average among workers who experience unemployment, corresponding to approximately 0.28% per week. Per-week losses are similar across short and long durations. Wage penalties vary by age and are larger for workers from occupations with higher occupational switching rates. This pattern is consistent with workers accepting lower-quality matches that make less use of their occupation-specific human capital.

Read the paper (PDF)

Does the Immigrant STEM Advantage Erode with Assimilation?

with Arigna Guha

We ask whether exposure to US schooling moves childhood immigrants’ college major choices toward the native-born benchmark, using variation in age at arrival. We separate the access margin (college completion) from the choice margin (STEM major conditional on a BA) and find that the two move in opposite directions: Each additional year of US exposure raises BA attainment but lowers the probability of a STEM major among graduates by 0.54 percentage points. Childhood immigrants thus converge toward their native-born counterparts, eroding the immigrant STEM advantage.

Draft coming soon