So I'm reading this recently published paper (2020): "Sources of Displaced Workers’ Long- Term Earnings Losses" (American Economic Review), and I'm puzzled why the authors included a "vector of gender, race, and education indicators, interacted with the worker’s age" in their fixed effects model (pages 3238-9). I thought all time-invariant variables drop out of the FE model (e.g. gender, race, education). In sum, the fixed effects eliminate all between-individuals variability (which may be contaminated by omitted variable bias) and leaves only the within-subject variability to analyze. So why the vector?
Thanks in advance for shedding light on this!
Note: if you can’t access the paper via the AER: https://www-aeaweb-org.ezp1.lib.umn.edu/articles?id=10.1257/aer.20180652
the authors have an earlier draft publicly available here: https://www-nber-org.ezp1.lib.umn.edu/papers/w24217
