Table 5

[bkudrzycki]

• [x] also include reference for interaction
• [x] I would write baseline years of training (instead of experience)
• [x] did you control for trade.
• [x] Also include A7 in main text.
• [x] But why are numbers of observations in A7 larger than in Table 5?

[bkudrzycki]

Note - no difference if trade or firm FE are used. Neither affects the estimates. Idea: Move FE regressions to appendix, include regressions excluding DNA apps in main body

[bkudrzycki]

Also make sure that the text still matches the table