Page 1 of 1

Discussant comments

Posted: Mon Jun 11, 2018 1:17 pm
by jdavalos
Do out-migration and remittances induce shifts to non-farm entrepreneurship among the left-behind? Evidence from Nepal

The effect of migration and remittances on transitioning from farm to non-farm entrepreneurship

Outcomes : Binary participation in non-farm entrepreneurship and as working hours (intensive and extensive margin)
• Why measurement error in working hours ? these are issues for independent variables, not dependent ones, unless they are systematic

Treatment : Dummy for receiving remittances of having a migrant abroad. (remittances ammount, here the measurement errors may play a role)

• You provide a nice discussion of the instruments issues and plan to introduce region specific controls to deal with this

• Using the remittances monetary value instead of the dummy for remittances may allow you to implement the Lewbel (2012) approach (In my oral presentation I’ve said working hours, that was a lapsus lingue, sorry for that)

https://www.stata.com/meeting/germany13 ... 3_baum.pdf

• Nice idea of interacting education with regional migration networks. The caveat is of course that education is likely to affect the main outcome. Better idea to calculate the migration network at the ethnic level.

• Also a nice idea to exploit the lagged growth of the destination as an instrument.

• Nice argumentation for LPM based on the preference for estimating marginal effects.

• Again, you can avoid the LPM criticisms by working with the working hours. In particular for the switching regression approach

• At the end you recall the possibility of using working hours from Amuedo-Dorantes and Pozo (2006) approach.