Bounding the heterogeneous effects of first childbirth on women’s income in the United States

I estimate a lower-bound measure of treatment-effect heterogeneity in the effect of first childbirth on women’s wage and salary income in the United States. Using data from the National Longitudinal Survey of Youth 1979 and a Changes-in-Changes framework with quantile regression, I recover the distribution of the individual-level treatment effect under cross-sectional rank invariance (CSRI). I use CSRI as a lower-bound approach to measuring this heterogeneity. I find that around 92% of mothers experience a strictly negative treatment effect, 6% have a treatment effect of zero, and 2% experience income gains. The mean effect is –$5,248, with a standard deviation of $6,208. About 23% of mothers report zero wage and salary income in the post-birth year, and most of them would have earned positive income absent childbirth. Notably, even under CSRI, the effect of childbirth on income remains substantially heterogeneous.