Is Two Too Many? Parity and Mothers' Labor Force Exit

• Date: 01 April 2019
• DOI: http://doi.org/10.1111/jomf.12533
• Published date: 01 April 2019
• Author: Catherine Doren

C D University of Wisconsin–Madison

Is Two Too Many? Parity and Mothers’ Labor

Force Exit

Objective: How do women’s chances of labor

force exit vary by the number of children they

have?

Background: Conventional wisdom suggests

that there may be a tipping point at the second

child when women are particularly likely to

leave. Women who only have one child, by

contrast, are thought to be uniquely unlikely

to exit.

Method: Using data from the nationally repre-

sentative 1979 to 2012 waves of the National

Longitudinal Survey of Youth1979 (https:// www

.nlsinfo.org/content/ cohorts/ nlsy79), event his-

tory methods estimate the likelihood of labor

force exit as women progress across parity

transitions.

Results: The results show no evidence for a

tipping point around the birth of second chil-

dren. Women are instead most likely to leave the

labor force when they are pregnant with their

rst child, and each subsequent child is asso-

ciated with a smaller increase in the probabil-

ity of exit. In addition, women who only ever

have one child are less likely to leave the labor

force than those who have more children, and

these differences arise as early as their pregnan-

cies with their rst children. College-educated

women who only have one child are especially

unlikely to exit.

University of Wisconsin–Madison, 3473 SewellSocial

Sciences, 1180 Observatory Drive, Madison, WI 53706

(cdoren@ssc.wisc.edu).

Key Words: event history analysis, labor force participa-

tion, maternal employment, National Longitudinal Survey of

Youth, transitionto parenthood.

Conclusion: The ndings thus do not support the

second child tipping point hypothesis, but they

emphasize the importance of completed parity

and the transition to motherhood for mothers’

labor force behavior.

I

Although labor force participation among moth-

ers has become increasingly common since

the 1960s, many women still leave the labor

force, at least temporarily, upon having children

(Boushey, 2008; Cotter, England, & Hermsen,

2007). Popular concern about highly educated

mothers “opting out” of the labor force reached a

high point in the 2000s and remains a persistent

theme (Belkin, 2003; Cohen, 2017; Kuperberg

& Stone, 2008; Warner, 2013). Although these

accounts do not explicitly focus on the number

of children women have, parity is often cited as

a factor when women discuss what pushed them

out of the labor force. Conventional wisdom

suggests that two children “constitute a tipping

point in favor of domesticity” (Stone, 2007,

p. 120) and thus the risk of labor force exit may

be particularly high around women’s second

births. From increasing the time commitment

and logistical complexity of parenting to raising

child-care costs beyond women’s level of pay,

second births are often discussed as a key reason

for mothers’ labor force exit (Abrams, 2001;

Hewlett, 2007; Hirshman, 2006; Stone, 2007).

Relatedly, a second popular notion suggests

that women who have multiple children by

the end of their childbearing years may have

systematically different labor force behaviors

than women who have only have one, evenprior

Journal of Marriage and Family 81 (April 2019): 327–344 327

DOI:10.1111/jomf.12533

328 Journal of Marriage and Family

to having their later births, bringing attention

to the importance of completed family size in

addition to current parity (Lehrer & Nerlove,

1986; Stone, 2007). This would indicate that

some underlying factors simultaneously predict

completed family size and mothers’ labor force

behaviors.

Observations from qualitative literature and

popular accounts related to the tipping point at

two children have been picked up by the media

and are widespread (e.g., Abrams, 2001; Belkin,

2013; Casserly, 2011; Hewlett, 2007; Hirsh-

man, 2006; Houston & Marks, 2005; Levine,

2010; Stone, 2007; Summerskill & Ryle, 2001),

even inspiring “how to” guides for navigat-

ing the challenges of second children (Abrams,

2001) and recommendations to avoid them alto-

gether to maintain a successful career (Hir-

shman, 2006). Despite this discourse, little is

known from nationally representative studies

about whether second births indeed represent

a tipping point in women’s labor force partici-

pation. The limited number of studies cited as

supporting the tipping point hypothesis suffer

from methodological limitations such as small

samples sizes (on the order of around 50 to a

few hundred women) and selective samples, for

example, samples that only contain women who

have already left the labor force or have multi-

ple children (Abrams, 2001; Houston & Marks,

2005; Stone, 2007). Changes in attitudes toward

or intentions regarding work rather than actual

labor force exits are also sometimes cited as sup-

porting this hypothesis (Abrams, 2001; Hous-

ton & Marks, 2005). It is thus unclear whether

the two-child tipping point hypothesis holds in a

large-scale, nationally representative sample.

The vast majority of prior research that has

used nationally representative data to examine

mothers’ patterns of labor force participation

has focused on motherhood as a single state

(Boushey, 2008; Cotter et al., 2007; Damaske &

Frech, 2016; Evertsson, 2013; Percheski, 2008),

which averages across mothers of different pari-

ties and thus obscures variation in the likelihood

that mothers continue working after each subse-

quent birth. Indeed, previous research has found

that as women have more children, their labor

force participation, occupational status, earn-

ings, and wages decline (Abendroth, Huffman,

& Treas, 2014; Angrist & Evans, 1996; Stone,

2007). However, this prior literature has not

examined labor force exit by tracing women’s

labor force trajectories across their births, a

topic of considerable scholarly and public inter-

est, making it difcult to assess whether there

is in fact a tipping point at the second child.

Past work has instead primarily considered the

relationship of current parity with labor market

outcomes without attention to women’s behav-

iors at previous parities. Women who leave the

labor force, even for periods as short as only a

few months, experience downward mobility in

occupational status, long-term penalties in earn-

ings and wages, and lower likelihoods of later

employment (Aisenbrey, Evertsson, & Grunow,

2009; Arun, Arun, & Borooah, 2004; Hewlett,

2007; Lovejoy & Stone, 2012). Given these last-

ing consequences, it is important to understand

how each successive child affectswhether moth-

ers drop out of paid work.

Insights about the relationship between higher

order births and women’s labor force exit can be

drawn from interview-based studies of mothers,

many of whom continue working after their rst

child only to leave on having a second or third

(Abrams, 2001; Hewlett, 2007; Stone, 2007).

These ndings support the conventional view

about second children, revealing that although

the transition to motherhood is indeed impor-

tant for shaping women’s work and family lives,

higher order births, especially second children,

are instrumental in mothers’ accounts of their

turn toward domesticity. A limitation of these

studies is that they often sample only profes-

sional mothers who have left the labor force.

Thus, their ndings may not generalize to less

advantaged women and may only describe pat-

terns among women who have already left the

labor force rather than all mothers.

Consistent with these accounts, analyses

of the causal effects of family size on labor

force participation grounded in the human

capital tradition nd that additional children

have a negative effect on women’s employ-

ment (Angrist & Evans, 1996; Chun & Oh,

2002; Cools, Markussen, & Strøm, 2017;

Lehrer & Nerlove, 1986). Studies taking this

perspective, however, give little theoretical

attention to variation across specic parity

transitions and have not assessed whether the

effects of second children are especially large.

They typically use methods relying on random

increases to family size, often in the form of

twins or sibling sex composition. Their nd-

ings can therefore be interpreted as the effects

of an additional child rather than identifying