« Back to Results

Gender in the Workplace

Paper Session

Saturday, Jan. 6, 2018 10:15 AM - 12:15 PM

Marriott Philadelphia Downtown, Grand Ballroom Salon J
Hosted By: American Economic Association
  • Chair: Ragan Petrie, Texas A&M University

Leave-taking and Labor Market Attachment Under California's Paid Family Leave Program: New Evidence From Administrative Data

Maya Rossin-Slater
University of California-Santa Barbara
Sarah Bana
University of California-Santa Barbara
Kelly Bedard
University of California-Santa Barbara


Leave-Taking and Labor Market Attachment under California’s Paid Family
Leave Program: New Evidence from Administrative Data∗
Sarah Bana, Kelly Bedard, and Maya Rossin-Slater
More than half of American mothers and over 90 percent of American fathers of infants
under the age of one are employed in the labor market.1 Yet the United States remains
the only OECD country without a national paid family leave (PFL) policy, and only 12
percent of private sector workers have access to PFL through their employers.2
In July 2004,
California enacted the first state-level PFL policy that provides six weeks of leave with 55
percent of usual pay replaced (currently up to a maximum weekly benefit of $1,137); since
then, three other states (New Jersey, Rhode Island, and New York) have followed suit. We
use detailed administrative data from the California Employment Development Department
on nearly 2 million PFL claims over 2005-2014 linked to individual-level quarterly earnings
data to provide novel insights about California’s first-in-the-nation experience with PFL.
Our analysis delivers four key take-aways. First, we can precisely document trends in CAPFL
take-up separately for bonding with a new child (hereafter, “bonding”) and caring for
an ill family member (hereafter, “caring”). Our calculations suggest that about 40 percent
(4.5 percent) of employed new mothers (fathers) made a bonding claim in 2005, while 47
percent (12 percent) of employed new mothers (fathers) made a bonding claim in 2014.
Second, we find that low earning men and women are less likely to take leave than their
more advantaged counterparts, consistent with survey reports that too little pay serves as a
barrier for taking leave (Fass, 2009) and with polls suggesting that awareness of the program
is lowest among low-income voters (DiCamillo and Field, 2015). We also show that individuals
in firms with fewer than 50 employees—who are not concurrently eligible for unpaid
job protected leave with health insurance through the Family and Medical Leave Act—are
substantially under-represented in the PFL claims data, which may reflect their reluctancy
∗We gratefully acknowledge funding from the California Employment Development Department (Agreement
No. M6102380).
1According to data from the 2005-2014 American Communities Survey (ACS) data, 55 percent of mothers
with youngest children less than 1 year old and 90 percent of fathers with youngest children less than 1 year
old were employed. Note that the ACS data only include parents who reside with their children.
2See: https://www.dol.gov/wb/PaidLeave/PaidLeave.htm.
to take paid leave without job protection or health insurance. Additionally, although there
are no major differences in leave-taking rates across the most common industries for women,
there are important discrepancies for men. Men in retail, transportation, and health industries
are the most over-represented in the PFL claims data, while men in the construction
industry are most under-represented.
Third, we document that nearly all women who take bonding leave take the full six weeks
that are provided. Among men, only 24 percent of claimants take all six weeks of bonding
leave. As for caring leave duration, 31 percent of men and 36 percent of women take the full
six weeks.
Fourth, the vast majority of both men and women remain in the labor force after taking
PFL, and more than half—53 percent of women and 68 percent of men—return to their
pre-leave employers.
We build on a small set of studies that uses survey data to analyze the effects of CA-PFL
(Rossin-Slater et al., 2013; Das and Polachek, 2015; Bartel et al., 2015; Baum and Ruhm,
2016; Stanczyk, 2016). Our analysis of administrative data can overcome several limitations
of these studies, which include small sample sizes, measurement error, non-response bias,
lack of panel data, and missing information on key variables such as PFL take-up and leave
duration. Although papers outside the U.S. have also used administrative data (see RossinSlater,
2017 for an overview), they analyze policies that offer leaves much longer than those
considered in U.S. policy discussions. If the impacts of leave are non-linear in duration
(Ruhm, 1998), then these estimates may have limited relevance for the current U.S. policy
climate. Moreover, the broader policy landscape likely matters for understanding the impacts
of PFL—a reform that expands PFL in a setting with subsidized child care and universal
health insurance is dramatically different from a setting where PFL may be introduced for
the first time and neither child care nor health insurance is guaranteed (as in the U.S. today).
Bartel, A., Rossin-Slater, M., Ruhm, C., Stearns, J. and Waldfogel, J. (2015).
Paid Family Leave, Fathers’ Leave-Taking, and Leave-Sharing in Dual-Earner Households.
Working Paper 21747, National Bureau of Economic Research.
Baum, C. L. and Ruhm, C. J. (2016). The effects of paid family leave in california on
parental leave-taking and labor market outcomes. Journal of Policy Analysis and Management,
35 (2), 333–356.
Das, T. and Polachek, S. W. (2015). Unanticipated effects of california’s paid family
leave program. Contemporary Economic Policy, 33 (4), 619–635.
DiCamillo, M. and Field, M. (2015). Just 36% of voters aware of state’s paid family
leave program. Tech. rep., Field Research Corporation.
Fass, S. (2009). Paid Leave in the States A Critical Support for Low-wage Workers and
Their Families. Tech. rep.
Rossin-Slater, M. (2017). Maternity and Family Leave Policy. Working Paper 23069,
National Bureau of Economic Research.
—, Ruhm, C. J. and Waldfogel, J. (2013). The effects of california’s paid family leave
program on mothers’ leave-taking and subsequent labor market outcomes. Journal of Policy
Analysis and Management, 32 (2), 224–245.
Ruhm, C. J. (1998). The economic consequences of parental leave mandates: lessons from
europe. Quarterly Journal of Economics, 113 (1), 285–317.
Stanczyk, A. (2016). Does paid family leave improve household economic security following
a birth? evidence from california, University of California at Berkeley, unpublished

The Gendered Effects of Career Concerns on Fertility

Nayoung Rim
University of Chicago
Kyung Park
Wellesley College


The Gendered Effects of Career Concerns on Fertility
Kyung Park
Wellesley College
Nayoung Rim
Harris School of Public Policy
University of Chicago
March 13, 2017
A growing literature reveals that the adverse effect of children on career advancement falls
disproportionately on women. This raises the possibility that women respond to career concerns by delaying
family formation more than men. Using a novel dataset on lawyers, we find females are less likely to have
their first-child before the promotion decision. This fertility gap is not explained away by gender-based
sorting or gender differences in marriage-timing and spousal occupation. Two channels drive our results:
women bear child-rearing costs and gender-specific promotion thresholds. This implies the focus on the
gender wage gap understates gender inequality in the labor market.
JEL Codes: J13, J71
A growing literature reveals that the adverse effect of children on career advancement falls
disproportionately on women (Wood, Corcoran, & Courant, 1993; Bertrand, Goldin, & Katz, 2010; Kleven,
Landais, & Sogaard, 2015; Adda, Dustmann, & Stevens, 2016). These findings are consistent with a
theoretical literature that suggests the female-male gap in earnings may reflect optimal responses to a
gender-specific comparative advantage in child-rearing rather than an employer's taste for discrimination
(Lazear & Rosen, 1990). However, far less attention is given to the possibility that, in equilibrium, gendered
effects of children on career advancement have important implications for fertility and family formation.
Our paper explores the extent to which women respond to gendered career concerns by delaying fertility
decisions more or less in comparison with men.
We use a novel dataset called the After the JD (AJD) study. AJD is a panel data set that follows a
nationally representative sample of lawyers throughout the first 12 years of their careers. Our focus on
lawyers is appealing due to the fact that the career trajectory for young attorneys is well-known and typically
standard across firms. For example, associate attorneys in large private firms are on a “partner-track” that
is analogous to “tenure-track” positions in academia. This facilitates analysis of fertility timing across the
“partner clock” when career concerns are conceivably high or low.
To mitigate concerns about gender-based sorting into occupations based on fertility preferences,
we leverage the richness of our data. The survey asks respondents to weigh the importance of various factors
in their decision to select into their chosen career. The questions cover a wide range of reasons, from
“medium-to-long-term earning potential” to “the potential to balance work and personal life” to “the
opportunity to do socially responsible work”. We use these survey questions to conduct factor analysis,
which assumes that latent preferences affect the lawyer's responses on the survey and uses the variation in
observed survey responses to find the common, underlying factors. Crucially, this allows us to control for
heterogeneous latent preferences, such as career ambition or fertility preferences, when we examine how
career concerns affect fertility decisions.
We find that female lawyers are less likely than males to start their family formation before the
partnership decision and more likely to start after the decision. Moreover, this gender difference in fertility
timing nearly doubles among lawyers who face greater career concerns at work. We identify these lawyers
by predicting their mid-career billed hours using intrinsic characteristics and classifying them as “highintensity
lawyers” if their predicted work-intensity is in the top quartile. The idea is that high-intensity
lawyers face greater career concerns because they have a greater chance of making partner, relative to their
predicted low-intensity peers. Constructed factor scores mitigate concerns about gender-based selection
into occupations. Additionally, we do not find a gender difference in completed fertility, which is consistent
with the argument that family-size preferences are not driving our results. Our results are also not driven
by a gender difference in marital timing or spousal occupation.
Two mechanisms drive our results. First, women face a greater career cost of having children.
Consistent with the literature, we find that this trade-off is influenced by spousal income, family-friendly
work conditions, and gender norms. A contribution of this paper is the second mechanism: the possibility
that firms have higher promotion thresholds for mothers because gender predicts child-related career
interruptions. We test this mechanism by comparing the gender difference in ability of equity partners who
were parents at the time of promotion. We find that mothers are significantly more likely to have
participated in general law review, held a judicial clerkship, and worked on more cases early in their career
than their male counterparts. Moreover, women are less likely to be promoted than men, even conditional
on ability proxies, billed hours, and caseload. These patterns indicate that the current focus on the gender
wage gap understates the level of inequality in the labor market.

Gender and Sorting in the On-Demand Economy

Zoe Cullen
Harvard University
John Humphries
University of Chicago
Bobak Pakzad-Hurson
Brown University


Extended Abstract
The literature on the male-female earnings gap consistently finds that the gap starts small but
grows over the life-cycle. The most common explanation for this is that women are skipped over
for promotions or fall behind due to maternity leave and larger time commitments to their families
(Goldin, 2014). Yet, using administrative data from an online platform for hiring local temporary
labor, we also find that wage discrepancies grow as individuals gain experience in the marketplace.
Given that the work on the platform is temporary, career concerns, such as promotion or maternity
leave, cannot explain these patterns. Moreover, we find that as individuals spend time on the
platform they increasingly specialize in tasks that follow traditional gender roles – a pattern at odds
with what a simple model of statistical discrimination would predict. This paper evaluates which
economic mechanisms account for dynamic sorting and a widening wage gap on the online platform,
and discusses how these findings may relate to the wage gap in the economy as a whole. We show
that differences in preferences and differences in skill are inconsistent with the gender dynamics we
observe, and we introduce a model of stigma surrounding a reputation in non-normative gender
roles. To test the predictions of this model, we take advantage of three unique aspects of the online
platform. First, the data include the salient set of jobs workers choose among, their asking wage,
and the employer’s selection among workers. These data allow us to see if women seek out different
jobs and if they are less likely to be selected conditional on applying.
Second, the platform matches workers to jobs in two ways. For half the jobs, workers submit
private bids and employers select from the list of applicants. In the second, employers post jobs for
a fixed price which is awarded

Using Econometrics to Reduce Gender Discrimination: Evidence From a Differences-in-Discontinuities Design

Giannina Vaccaro
University of California-Irvine


Despite governments’ efforts to reduce discrimination, gender wage discrimination persists in most countries. This paper estimates the effectiveness of an inexpensive anti-discrimination policy with an unprecedented design that directly targets the reduction of the unexplained wage gap. Introduced in Switzerland, the policy addresses firms with at least 50 workers and sanctioning is designed to affect only public contractors. By using a Diff-in-Disc design, a combination of an RDD and Diff-in-Diff strategy, this paper shows the effectiveness of this policy in reducing the unexplained wage gap for firms targeted by this regulation.
Francine Blau
Cornell University
Kevin Lang
Boston University
JEL Classifications
  • J1 - Demographic Economics
  • Z1 - Cultural Economics; Economic Sociology; Economic Anthropology