IZA DP No. 10558: The Implicit Costs of Motherhood over the Lifecycle: Cross-Cohort Evidence from Administrative Longitudinal Data
Published as "The Implicit Costs of Motherhood over the Lifecycle: Cross‐Cohort Evidence from Administrative Longitudinal Data" Southern Economic Journal 84 (3), 716-733
The explicit costs of raising a child have grown over the past several decades. Less well understood are the implicit costs of having a child, and how they have changed over time. In this paper we use longitudinal administrative data from over 70,000 individuals in the Synthetic SIPP Beta to examine the earnings gap between mothers and non-mothers over the lifecycle and between cohorts. We observe women who never have children beginning to out earn women who will have children during their 20s. Gaps increase monotonically over the lifecycle, and decrease monotonically between cohorts from age 26 onwards. In our oldest cohort, lifetime gaps approach $350,000 by age 62. Cumulative labor market experience profiles show similar patterns, with experience gaps between mothers and non-mothers generally increasing over the lifecycle and decreasing between cohorts. We decompose this cumulative gap in earnings (up to age 43) into portions attributable to time spent out of the labor force, differing levels of education, years of marriage and a number of demographic controls. We find that this gap between mothers and non-mothers declines from around $220,000 for women born in the late 1940s to around $160,000 for women born in the late 1960s. Over 80% of the change in this gap can be explained by variables in our model, with changes in labor force participation by far the best explanation for the declining gap. Comparing our oldest cohort as they approach retirement to the projected lifecycle behavior of the 1965 cohort, we find that the earnings gap is estimated to drop from $350,000 (observed) to $282,000 (expected) and that the experience gap drops from 3.7 to 2.1 years. We also explore the intensive margin costs of having a child. A decomposition of earnings gaps between mothers of one child and mothers of two children also controls for age at first birth. Here, we find a decline in the gap from around $78,000 for our oldest cohorts to around $37,000 for our youngest cohorts. Our model explains a smaller share of the intensive margin decline. Changes in absences from the labor market again explain a large amount of the decline, while differences in age at first birth widen the gap.