Explanatory variables

Post Reply
John Cockburn
Posts: 9
Joined: Sat Jan 27, 2018 7:48 pm
Institutional affiliation*: Partnership for Economic Policy
Contact:

Explanatory variables

Post by John Cockburn » Mon Jun 11, 2018 11:20 am

I wonder if the explanatory variables could be reconsidered to better capture and test the theoretical hypotheses. For the income channel, do you expect income from remittances to have different effects relative to income from other sources? If so, would you want to also include a total (or other) income variable in addition to the remittance income variable to be able to test if their coefficients are different? If not, would you get a more robust estimate by using total income as the explanatory variable and then using observed remittances and the estimated regression to predict how additional income from remittances will affect child schooling?

For the absence channel, you postulate that migration will increase the domestic work burden of children or deprive them of adult support and encouragement. You use the number of migrants. Presumably, the impact depends on the household size composition. Having one migrant from a family of two will have a bigger impact than from a family of ten. So would you want to explore including a household size variable, interaction terms or the share of migrants? I am not sure the right solution, but encourage some thought on this.

I would assume larger households will be more likely to engage in migration, as they can more easily spare them. If, by simply using number of migrants as the explanatory variable, you may conclude that it has little impact as only large households engage. But then can you recommend that all households should be encourage to engage in migration? Children in smaller households may suffer.

Also, the sex of the migrants will presumably have effects. Female migrants may increase the child domestic work burden more than male migrants, and may also have differential effects on the support and encouragement children receive for schooling. If this can be differentiated, you could test these hypotheses.

Finally, the number of remaining children in the household (to share the domestic work burden) will likely affect the impacts of migration.

chanhang
Posts: 5
Joined: Mon Jun 04, 2018 1:05 am
Contact:

Re: Explanatory variables

Post by chanhang » Tue Jun 12, 2018 9:40 am

Hi John,
Thank you very much for your comments.

Post Reply

Return to “Migration, Remittances and Child Schooling in Rural Cambodia – PMMA 20209”

Who is online

Users browsing this forum: No registered users and 2 guests