Motherhood Penalties in the U.S., 1986–2014

• DOI: http://doi.org/10.1111/jomf.12543
• Date: 01 April 2019
• Author: Marta Murray‐Close,Eunjung Jee,Joya Misra
• Published date: 01 April 2019

E J University of Massachusetts Amherst

J M University of Massachusetts Amherst∗

M M-C U.S. Census Bureau∗∗

Motherhood Penalties in the U.S., 1986–2014

Objective: This article aims to determine

whether the wage penalty for motherhood

changed in the United States between 1986

and 2014 and to assess the relative impact

of education and experience on the penalty

over time.

Background: Mothers earn less than childless

women. Although mothers’ educational levels

and labor market experience have increased

over time, most studies do not analyze the moth-

erhood penalty over time.

Method: This article uses ordinary least squares

and xed effects regressions to estimate the

motherhood penalty in three time periods

(1986–1995, 1996–2004, and 2006–2014) and

compares the penalty across time periods. The

data used in the analysis are from the U.S.Panel

Study of Income Dynamics, one of the only

nationally representative data sets that contains

a measure of actual labor market experience.

Results: The motherhood penalty remains

quite stable over time and may have worsened

for mothers with one child. The gross gap in pay

Department of Economics, University of Massachusetts

Amherst, Crotty Hall, 412 North Pleasant St., Amherst, MA

01002.

∗Department of Sociology, Universityof Massachusetts

Amherst, Thompson Hall, 200 Hicks Way,Amherst, MA

01003.

∗∗Social, Economic, and Housing Statistics Division, U.S.

Census Bureau, 4600 Silver Hill Rd, Suitland, MD 20746

(marta.murray.close@census.gov).

Key Words: cohort, families and work, gender,labor market,

motherhood, wages.

between childless women and mothers of two

or more children has narrowed only because

mothers’ have increased their investments

in human capital, such as education and work-

force experience. Ordinary least squares

and xed effects models provide similar results.

Conclusion: Mothers’ human capital investment

may not be enough to lower the motherhood

wage penalty in the United States.

Previous research has found that mothers earn

less than childless women (Budig & England,

2001; Budig & Hodges, 2010, 2014; Waldfogel,

1997). Lower wages for mothers make a sub-

stantial difference to many families. Mothers

working full-time earn approximately 71% of

what fathers earn, even as more than 70% of

families rely on income from a working mother

(Williams-Baron, Anderson, & Hegewisch,

2017). Although women’s educational lev-

els and engagement in the labor market have

changed during the past several decades, most

studies do not analyze variation in the mother-

hood penalty over time. We know surprisingly

little about how the labor market status of

mothers has evolved or whether the role of

motherhood in shaping labor market outcomes

for women has changed.

One previous study to examine changes in

the motherhood penalty in the United States

over time found that, controlling for differences

in demographic characteristics, educational

attainment, labor market experience, and sta-

ble personal characteristics, the penalty was

statistically unchanged between 1975 to 1985

434 Journal of Marriage and Family 81 (April 2019): 434–449

DOI:10.1111/jomf.12543

Motherhood Penalties in the U.S., 1986–2014 435

and 1986 to 1998 (Avellar & Smock, 2003).

Other recent studies found some reduction in

the penalty over time in the United States,

without controlling for labor market experience

(Buchmann & McDaniel, 2016; Glauber, 2018;

Pal & Waldfogel, 2016; Weeden, Cha, & Bucca,

2016). In contrast, a study using Norwegian data

found that, controlling for differences in demo-

graphic characteristics, educational attainment,

and potential labor market experience, and

comparing workers in the same occupation with

the same employer, the motherhood penalty in

Norway declined substantially during the same

period. The authors of the second study attribute

the improvement in the relative wages of Nor-

wegian mothers to an expansion in work–family

policies between 1979 and 1996 (Petersen,

Penner, & Høgsnes, 2014).

There are a number of reasons to think that

the motherhood penalty in the United States

may have persisted beyond the late 1990s.

This country has not committed to expand-

ing work–family policies to anywhere near

the extent of most other wealthy countries,

although there has been some limited expan-

sion at the state level (Barnett, 2010; Meyers

& Gornick, 2005; Rossin-Slater, Ruhm, &

Waldfogel, 2013). For example, there is no

federal commitment to paid parental leave, and

there is limited commitment to publicly funded

preschool relative to other wealthy countries.

This may mean that U.S. mothers may nd it

more difcult to balance work and care, which

could harm their experience, productivity, and

subsequent wages. In addition, progress toward

gender equality seems to have stalled in the late

1990s. Although the ratio of women’s to men’s

earnings rose steeply from 62% in 1979 to 77%

in 1993, it has hovered around 81% since 2005

(U.S. Department of Labor, 2017).

On the other hand, even absent extensive

work-family policies, there are reasons to think

that the motherhood penalty may have declined.

Women’s labor force participation rate has risen

dramatically, from less than 33% in 1948 to 57%

in 2015, as men’s labor force participation rate

has fallen from more than 87% to 69% (U.S.

Department of Labor, 2017). Women’s labor

force participation rate is particularly high dur-

ing their childbearing years, between the ages of

25 and 54, and substantially lower at younger

and older ages. Indeed, the labor force partic-

ipation of mothers of children aged younger

than 18 has risen from 47% in 1975 to 70% in

2015 and is thus higher than for other groups

of women (U.S. Department of Labor, 2017).

This may mean that recent cohorts of mothers

have higher levels of experience than previous

cohorts, which may lead them to earn higher

wages. Employers may also have revised their

expectations about the relative productivity of

mothers and childless women and may be less

inclined than in the past to discriminate against

mothers in hiring, promotion, and pay.

This article uses data from the Panel Study

of Income Dynamics (PSID; https://psidonline

.isr.umich.edu/), one of the only nationally rep-

resentative data sets that contains a measure of

actual labor market experience, to examine the

evolution of the motherhood penalty in recent

years. We estimate the wage gap between moth-

ers and childless women for the following three

time periods: 1986 to 1995, 1996 to 2004, and

2006 to 2014. The earliest time period in our

study is roughly comparable to the latest time

period in Avellar and Smock (2003). The more

recent time periods allow us to consider how

the motherhood wage penalty has continued to

evolve over time.

Although much previous research has focused

on the portion of the motherhood penalty that

might plausibly be attributed to discrimina-

tion in wage setting, we take a broader view.

Discrimination is just one of several channels

through which motherhood may reduce wages,

and analysts concerned with the full extent of

mother’s economic disadvantage must consider

wage losses due to discrimination alongside

losses from other sources, including apparently

voluntary choices of mothers to accommodate

children’s needs at the expense of their own

human capital development and earnings.

In keeping with this view, we use a series

of regressions to examine the magnitude and

evolution of both the total motherhood penalty

and the penalty net of education and labor mar-

ket experience. First, we use pooled ordinary

least squares (OLS) regressions to estimate the

wage gap between mothers and childless women

controlling only for demographic differences.

The resulting estimates quantify wage inequal-

ity between mothers and childless women from

all sources, including any effect of motherhood

on women’s investments in their human capi-

tal. Next, we estimate the wage gap controlling

for demographics and measures of human cap-

ital, including educational attainment and labor

market experience. This second set of estimates