Parental Work Hours in Childhood Obesity : Evidence Using Instrumental Variables Related to Sibling School Eligibility

This study exploits plausibly exogenous variation from the youngest sibling's school eligibility to estimate the effects of parental work on the weight outcomes of older children in the household. Data come from the 1979 cohort of the National Longitudinal Survey of Youth linked to the Child and Young Adult Supplement. We first show that mothers' work hours increase gradually as the age of the youngest child rises, whereas mothers' spouses' work hours exhibit a discontinuous jump at kindergarten eligibility. Leveraging these insights, we develop an instrumental variables model that shows that parents' work hours lead to larger increases in children's BMI z-scores and probabilities of being overweight and obese than those identified in previous studies. We find no evidence that the impacts of maternal and paternal work are different. Subsample analyses find that the effects are concentrated among advantaged households, as measured by an index involving education, race, and mother's marital status