Zusammenfassung:
In their comment, Delpeyrou and Bertoli (2024) make several critiques of our paper (Depetris-Chauvin et al., 2020). In this reply, we show that: (i) Their purported "errors" in data construction are in the vast majority of cases simply the result of different (and, we argue, superior) choices in implementing our identification strategy; (ii) Our specification choices regarding fixed effects are not only justified but are the correct ones for a proper apples-to-apples comparison; and most crucially, (iii) Our key results are essentially robust to their own proposed modifications. Far from overturning our findings, their critique further underscores their robustness.