“…Of those studies, their estimates are often infected by having small sample sizes in each category of household migration status, especially in the category of households with migrant mother and nonmigrant father, and thus are prone to lack statistical power and stability. 3 An alternative strategy is to focus only on paternal migration, which is the preferred strategy when using data from countries where fathers are much more likely to migrate than mothers (Antman, 2011(Antman, , 2012Creighton, Park, & Teruel, 2009;Kuhn, 2006;Lee & Park, 2010;Nobles, 2011;Yabiku & Agadjanian, 2017). Even for this refined group, the results are mixed.…”