Working Paper: NBER ID: w27618
Authors: Gordon B. Dahl; Danolof Rooth; Anders Stenberg
Abstract: This paper estimates family spillovers in high school major choice in Sweden, where admission to oversubscribed majors is determined based on GPA. Using a regression discontinuity design, we find large sibling and intergenerational spillovers that depend on the gender mix of a dyad. Same-gender siblings copy one another, while younger brothers recoil from older sister’s choices. Fathers and mothers influence sons, but not their daughters, except when a mother majors in the male-dominated program of Engineering. Back of the envelope calculations reveal these within family spillovers have sizable implications for the gender composition of majors.
Keywords: family spillovers; high school majors; gender dynamics; regression discontinuity design
JEL Codes: I21; J24
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| younger siblings choose same major as older siblings (C92) | younger siblings are more likely to choose the same major (C92) |
| older brothers choose engineering (C92) | younger brothers are more likely to choose engineering (C92) |
| older sisters choose same major (Y80) | younger sisters are more likely to choose the same major (Y80) |
| older sisters choose major (Y80) | younger brothers are less likely to choose the same major (D29) |
| father is enrolled in major (I29) | sons are more likely to choose the same major (D29) |
| father completes major (J12) | sons are more likely to choose the same major (D29) |
| father influences daughters (J12) | daughters are less likely to choose the same major (J79) |