U.S. flag

An official website of the United States government

Skip Header

Family Pies: Effects of Family Resource Shares in Childhood on Adult Earnings

Written by:
Working Paper Number CED-WP-2022-005


Effects of birth order on various outcomes have been extensively studied in both life sciences and social sciences. The present analysis exploits variation in household resources per child due to family size and parental earnings changes over time to estimate the effect of family earnings shares in early life on adult earnings. The analysis focuses on biological siblings in two-parent households in the Survey of Income and Program Participation linked to administrative data on earnings. Controlling for observable individual and parental characteristics, the baseline effects of earnings shares in childhood on adult earnings are positive and statistically significant across specifications. A 10 percent increase in childhood share of parental earnings (in real US dollars) implies an increase in own earnings as an adult between 1.3 percent and 2 percent. In the preferred specification that includes family fixed effects, the estimated earnings effects vary based on functional form as well as the childhood age ranges used. In general, these estimated effects are either negative (e.g., higher share of parental earnings in childhood implies lower earnings as an adult) or statistically indifferent from zero. Across models there is some evidence of differential effects for younger siblings – particularly second-born children. Additional analysis shows varying effects comparing paternal earnings effects to maternal earnings effects. The results of this study highlight the importance of controlling for within-family factors in analyzing long-term effects of parental earnings in childhood as well as the potential sensitivity of intergenerational earnings effects to sample, variable, or model-related decisions.

Page Last Revised - January 13, 2023
Is this page helpful?
Thumbs Up Image Yes Thumbs Down Image No
255 characters maximum 255 characters maximum reached
Thank you for your feedback.
Comments or suggestions?


Back to Header