Abstrakt
There is much evidence that migration of a parent affects the educational performance of children left behind (CLB). Nevertheless, there is no agreement on the direction of the impact.
In this paper, we use Armenian school data and report evidence of a negative impact of parental seasonal migration on the educational performance of CLB. We employ a different approach than those used in the prior literature by (i) using the intensity of seasonal migration (the number of times the parent migrated) instead of a binary variable (whether the parent migrated or not) and (ii) the number of children entering first grade whose parent is a seasonal migrant as an instrument for the intensity of seasonal migration.
We find that seasonal migration negatively affects the educational performance of CLB, and that it mainly affects boys, there is no significant impact on girls. Additionally, we find that using a zero-one dummy for migration as prior studies have done upwardly biases the IV estimate by approximately a factor of three, while our intensity measure yields more accurate results.