P 31: only 5-11 percent of apprentices generate positive net benefits when productivity is left uncounted for - > which it shouldn’t. I think what the results show that firms do not use apprentices as an income stream as often argued (they also do not seem to keep them very long-you could also report on this) but they might have a positive effect on the firm given that the allowances are lower than wages -> again I would double check with Rubain(Bolli) whether it make sense to assume trained wage for the last two years? What happens if only untrained wage is used in model 3 (in any case the most realistic model to me)
P 31: only 5-11 percent of apprentices generate positive net benefits when productivity is left uncounted for - > which it shouldn’t. I think what the results show that firms do not use apprentices as an income stream as often argued (they also do not seem to keep them very long-you could also report on this) but they might have a positive effect on the firm given that the allowances are lower than wages -> again I would double check with Rubain(Bolli) whether it make sense to assume trained wage for the last two years? What happens if only untrained wage is used in model 3 (in any case the most realistic model to me)