States and aid agencies prefer supply side programs–more schools, better teachers, textbooks or lunches.
But will parents send their kids to school if it aren’t jobs to enter?
We examine the effects of a roughly simultaneous supply-side influence (the Female Stipend Program) and a major demand-side influence (the expansion of the garment industry) on girls schooling in Bangladesh in the 1980’s and 1990’s.
…We ï¬nd that the arrival of garment jobs increases schooling for younger girls: a ten percent increase in garment jobs corresponds to an 1.35 percentage point percent increase in the probability that a 5-year-old girl is in school.
There is a zero average effect for older girls, some of whom likely drop out of school to take the jobs right away.
We identify effects of the FSP with a regression discontinuity and do not ï¬nd a statistically signiï¬cant effect of the program.
A new paper from Rachel Heath and Mushfik Mobarak.
Many other excellent papers from the NEUDC conference (the Cannes of development economics) here.
4 Responses
Thank you for your question. We didn’t include the main effect of a village garment because we had sibling fixed effects — and since we couldn’t include data on enrollment for families moving across villages (since we wouldn’t know whether these are garment villages or not), the indicator for garment village doesn’t vary within family.
Fascinating paper. (Maybe) trivial econometrics question: Isn’t their main specification in equation (1) missing the own “garment village_{ivft}” term? I think it’s absence introduces bias but perhaps there’s something I’ve missed.
“the Cannes of development economics”
Does that mean that there’s booing?
Maybe more girls would want to go to school if they were allowed to mingle with boys?
But then the parents wouldn’t. What a dilemma.