Does Overwork Attenuate the Motherhood Earnings Penalty among Full-Time Workers?

Abstract

This study examines whether working long hours alters the motherhood earnings penalty in the context of the United States. The author uses data from the National Longitudinal Survey of Youth (1979–2014) to model the annual earnings penalty mothers incur per child in the United States. The results support that working long hours (50+ hours per week) reduces the negative effect of motherhood on earnings for white women. Once we control for human capital and labour supply, however, there is no difference in the effect of children on earnings between full-time workers and overworkers. For Black full-time workers and overworkers, having an additional child has little effect on earnings. The findings suggest that although overwork appears to attenuate the earnings penalty for white mothers, white mothers who work long hours exhibit a smaller penalty because they already have high levels of human capital and supply a great amount of labour.

Keywords

earnings full-time work motherhood penalty overwork work–family conflict working long hours

Introduction

Mothers in North America and Europe tend to be paid less than childless women who have the same qualifications and experience (Budig and England, 2001; Cooke, 2014; Fuller and Hirsh, 2018; Gangl and Ziefle, 2009). Given that the majority of women become mothers at some point in their life, this ‘motherhood penalty’, defined as lower pay associated with motherhood status, can deter the progress toward gender equity in the labour market (for a similar argument, see Correll et al., 2007).

Studies have shown that contexts, such as occupational characteristics and earnings level, can modify the motherhood penalty (Buchman and McDaniel, 2016; Cooke, 2014; England et al., 2016; Fuller and Hirsh, 2018; Yu and Kuo, 2017). Yet, little is understood about how work hours, especially long work hours (50 hours or more weekly), alter the motherhood penalty. Overwork may have complex implications for the motherhood penalty by amplifying workplace rewards and, simultaneously, exacerbating workplace challenges for mothers who assume a disproportionate share of family responsibilities (Blair-Loy, 2003; Casper and Bianchi, 2002; Cha and Weeden, 2014; Stone, 2007).

Although some studies have noted that the difference in the motherhood penalty between full-time and part-time workers is not significant (i.e. Budig and England, 2001), the primary interest in work hours in the existing literature is generally limited to the assumption that reducing work hours after a woman becomes a mother or has an additional child has confounding effects on the motherhood penalty. Researchers have used work hours or part-time employment status as control variables for labour supply or mother-friendly work conditions to parcel out the motherhood penalty from wage reductions as a result of changes in work hours after childbirth (England et al., 2016; Yu and Kuo, 2017). Although this conventional assumption and methodological approach reveal the motherhood penalty net of work hours, such practices focus little on the variance of the motherhood penalty across work hours, and especially how long work hours can condition the motherhood penalty.

As an exception, Weeden et al. (2016) reported a closing wage gap between childless women and mothers who work long hours during the 2010s in the United States. Although this study reveals an important trend in the motherhood wage penalty among women who work long hours, its analytical models and dataset structure do not substantiate a causal inference for the association between within-person variations in motherhood status and wages.

This study focuses on the United States for two reasons. First, working long hours is more prevalent in the United States than in many Western countries (Burger, 2018), in part because of the weak regulation of work hours and strong employer power over employees’ work arrangements (Berg et al., 2014). Second, although American mothers are less likely to overwork than American childless women, at 6% and 9% in 2010, respectively (Weeden et al., 2016), the overall incidence of overwork among women has increased over the past decades (Cha and Weeden, 2014), and the gap between childless women and mothers has shrunk substantially among younger cohorts of professional women (Percheski, 2008), suggesting that the gap may be narrowing. Thus, understanding how overwork moderates the motherhood penalty in the United States has become increasingly important.

The primary motivation for this study is to reconsider the role of work hours in the motherhood penalty. In particular, this study asks whether working long hours alters the negative effect of motherhood on earnings. Drawing on gender, work and family research, this study presents competing hypotheses for whether working long hours can buffer or amplify the negative effect of motherhood on earnings. Using data from the National Longitudinal Survey of Youth (NLSY) and fixed effect models, we test whether within-person change in the number of children affects annual earnings in overwork, in contrast to full-time work, to provide a better causal inference on the motherhood penalty. This study advances the overwork literature and theories of the motherhood penalty and illuminates the complex implications of overwork for the intersection between motherhood and work.

Theoretical framework

The motherhood penalty

One strand of studies has shown that mothers’ distinctive qualifications and commitment play a key role in explaining why mothers get paid less than childless women. As educational attainment is positively associated with childlessness and delayed childbearing (Casper and Bianchi, 2002), motherhood predicts fewer years in school. In addition, although there is little difference in work effort and career motivation between childless women and mothers (Kmec, 2011; Pas et al., 2011), mothers tend to reduce work hours and forgo gaining more work experience because of a disproportionate share of housework compared to their spouses (Bianchi et al., 2012; Craig and Mullan, 2010). Some researchers further posit that mothers choose ‘mother-friendly’ benefits over demanding, high-paying jobs to combine their work and family responsibilities (e.g. Becker, 1985), but empirically, these ‘compensating differentials’ explain little of the motherhood penalty (Budig and England, 2001). Whereas human capital and labour supply can mediate a substantial portion of the motherhood penalty (Budig and England, 2001; Gangl and Ziefle, 2009), it is possible that the motherhood penalty may be attributable to unobserved characteristics of mothers, such as job productivity (Budig and England, 2001).

Another strand of research posits that these differences in qualifications and commitment between mothers and childless women do not fully explain the motherhood penalty. One field experiment study has shown that mothers tend to be more disadvantaged in the hiring and promoting process than childless women who have the same qualifications and who exhibit the same amount of effort and productivity (Correll et al., 2007). Similarly, qualitative studies have revealed that pervasive scepticism about mothers’ commitment reinforces the belief that mothers should quit their jobs (Neely, 2020; Turco, 2010). These arguments suggest that discriminatory practice against mothers at the workplace can result in the motherhood penalty.

As mentioned earlier, studies have shown that the motherhood penalty varies across contexts (Buchman and McDaniel, 2016; Cooke, 2014; England et al., 2016; Fuller and Hirsh, 2018; Yu and Kuo, 2017). These studies have theorized that work contexts, such as earnings levels and job autonomy, can affect the intensity of the work–family conflict (Cooke, 2014; Fuller and Hirsh, 2018; Yu and Kuo, 2017), the wage penalty of forgone work experiences (England et al., 2016) and the level of employer discrimination against mothers (Yu and Kuo, 2017). Expanding these studies, this article develops hypotheses to explain how overwork can alter the motherhood penalty.

Working long hours and the motherhood penalty

Several forces drive employees to work long hours. As job evaluation processes often treat long work hours as evidence of commitment to work, especially in workplaces where productivity is difficult to measure, employees are pressured to be physically present in their workplace for long hours (Sharone, 2004). In addition, as technology makes work more permeable to non-work life, employees who have some degree of autonomy to choose the location of work are expected to stay nearly always connected to work (Blair-Loy, 2009; Mullan and Wajcman, 2019). With this backdrop, cultural and institutional expectations pressure many full-time employees to work longer than their preferred work hours (Reynolds, 2003; Stier and Lewin-Epstein, 2003).

In this study, I build two competing hypotheses for whether long work hours alter the motherhood penalty. On the one hand, working long hours may attenuate the negative effect of motherhood on earnings. Studies have shown that working long hours can increase the economic reward (Cha and Weeden, 2014) and workplace support (Sturges, 2013). As mothers who engage in overwork may increase employers’ expectations of mothers’ commitment and competence, overwork may reduce the negative effect of motherhood on earnings.

In addition to signalling commitment to work, overwork may buffer the negative effect of motherhood on earnings by reducing the burden of unpaid work. Compared with mothers who work full-time, mothers who engage in overwork tend to spend less time on housework (Hook, 2017), potentially by outsourcing housework (Kornrich and Roberts, 2018). Mothers who work long hours may also have a particularly strong preference to ‘opt out’ of housework (Killewald, 2011). A relatively lighter domestic burden can help overworking mothers enhance productivity at work by lessening the potential interference of family responsibilities with work responsibilities.

Furthermore, overwork may appear to attenuate mother-specific disadvantages at work because of highly selected characteristics of overworking mothers. Given that overwork is more common in professional/managerial occupations in the United States (Cha and Weeden, 2014), overworking mothers tend to have a particularly high level of human capital, such as educational attainment and work experience, compared to mothers who work less. In addition, the expectations that come along with working long hours may prevent overworking mothers from reducing job tasks according to their family responsibilities and instead lead them to quit their jobs entirely (Cech and Blair-Loy, 2019; Stone, 2007). As a result, women who reduce work hours after childbearing are less likely to engage in overwork. Thus, the gap in work hours between childless women and mothers who continue to overwork may be smaller among those who engage in overwork than among those who work full-time.

In sum, overwork can signal commitment to work, reduce unpaid work and exclude mothers who are likely to experience a substantial penalty. Based on these arguments, I make the following extra-effort hypothesis for the buffering effect of overwork on the motherhood earnings penalty:

Hypothesis 1: Working long hours reduces the magnitude of the negative effect of motherhood on earnings.

On the other hand, working long hours may amplify the negative effect of motherhood on earnings. Whereas people who work long hours generally have substantial difficulties with combining work and personal life (Bianchi and Milkie, 2010), overworking mothers may experience particularly intensive work–family conflict because of a disproportionate burden of family responsibilities (Bianchi et al., 2012) and lack of spouse’s support (Craig and Mullan, 2011). In addition, in poorly paid workplaces, overworking mothers may have severe work–family conflict because they have little control over their work schedule and are often unable to purchase domestic services (Gerstel and Clawson, 2015). In competitive and highly paid workplaces, the work interference with family life can lead to a severe career penalty (see also England et al., 2016) as employees are expected to be immersed in work (Blair-Loy, 2003; Sharone, 2004). As the extreme work–family conflict experienced by mothers who work long hours may reinforce employers’ discrimination against mothers and reduce their work productivity, these arguments suggest that working long hours can increase the magnitude of the negative effect of motherhood on earnings.

Furthermore, overwork may reinforce the negative stereotype of highly committed mothers. Benard and Correll (2010) have shown that highly achieving mothers tend to be evaluated as having lower interpersonal skills than equally achieving childless women, providing evidence that people tend to disapprove of competent and committed mothers. As overworking mothers may be subject to an additional burden from such ‘normative discrimination’, overworking mothers may be penalized in workplace interpersonal relationships and job evaluation. These arguments allow me to construct the alternative extra-burden hypothesis on the amplifying effect of overwork on the negative effect of motherhood on earnings (for a similar argument, see Cha, 2013):

Hypothesis 2: Working long hours increases the magnitude of the negative effect of motherhood on earnings.

This study focuses on full-time workers and overworkers. Part-time workers are less likely to experience work–family conflict due to their relatively less intense work demands (Voydanoff, 2004). Because the hypotheses developed about the motherhood penalty among overworkers draw on the risk of substantial work–family conflict and expectations inherent in the ideal worker norm, this study limits the scope of analysis to overworkers compared to full-time workers.

Studies have documented that the motherhood penalty is less pronounced among Black women than white women (England et al., 2016; Glauber, 2007). Van Winkle and Fasang (2020) argued that this is in part because Black women are concentrated in low-paying jobs where the wage differential across workers is small and earnings growth over the life course is sparse. Given the racial difference in the intersection between parenthood and work, this study examines how overwork alters the motherhood earnings penalty, separately for white and Black women.

Data, variables and methods

Data and sample

This study uses the NLSY 79 from 1979 to 2014. The waves from 1979 to 1994 were collected annually and the waves from 1994 to 2014 were collected biannually. These data are nationally representative of the baby boomer cohort who were aged 14 to 22 in 1979 and became aged 49 to 57 in 2014. As these data provide longitudinal information on employment and family characteristics from the beginning to the end of the reproductive career, it is widely used in studies on the intersection between parenthood and labour outcomes in the United States (e.g. England et al., 2016; Florian, 2018). Despite this advantage, women with more children and highly educated women in the sample were more likely to drop out of the survey. As these women may have the larger motherhood penalty (Budig and England, 2001; England et al., 2016), the estimates for the motherhood penalty might be underestimated. This study limits the sample to women who are employed, work for at least 35 hours per week and have earnings in at least two waves.

I used multiple imputation methods for independent variables with more than 5% missing values. Using the chained equation method, I imputed 667 missing values for the percentage of females in a given occupation and spouses’ work hours in various years. Among the variables that were not imputed, I found that item non-response is associated with several observed characteristics, including number of children and marital status. As the missing values from item non-response became a relatively small portion of the sample after the imputation, I listwise deleted 1370 (4.5%) and 798 (5.3%) person-year observations for white and Black women, respectively, with missing values on other variables.

The analyses for Black and white women were conducted separately. The oversampling structure of the NLSY and lower attrition rate of Black women allow me to have sufficient observations in the sample (Aughinbaugh et al., 2017). Hispanic women were not included because the sample size of overworkers is relatively small.

The final sample is 30,115 person-year observations from 3112 white women and 14,907 person-year observations from 1361 Black women. This sample includes both childless women and mothers. As people often change work hours across their life course, some respondents are included in both the full-time work and overwork samples in the main analysis.

Variables

The dependent variable is annual earnings. Annual earnings allow us to address the general economic status of employees (Glauber, 2008). Whereas some studies have measured the motherhood penalty using hourly wages (Budig and England, 2001; England et al., 2016), other studies have measured the motherhood and fatherhood penalty with annual earnings (Budig et al., 2016; Glauber, 2008). Although hourly wages measure ‘standardized compensation per hour on the job’, they may not comprehensively capture the association between work hours and monetary compensation among salaried workers. For example, the hourly wage of salaried workers, which is conventionally calculated by dividing annual earnings by annual work hours in the NLSY, can be lowered by either an increase in work hours or a decrease in annual earnings. If salaried workers reduce their work hours after having a child and retain the same pay, this would increase their hourly wage. Given that such standardized compensation per hour can complicate an estimation of the within-person motherhood penalty among salaried workers, this study used annual earnings. Annual earnings are adjusted to 2014 US dollars and top-/bottom-coded at the 95th and 1st percentile, respectively. This study uses the natural log of annual earnings to reduce skewness.

The motherhood penalty is estimated by the coefficient of motherhood for annual earnings. This study measures motherhood with the number of children ever born to or adopted by the respondent for the main results. I used a continuous variable for the number of children instead of dummy variables for having one child and having two or more children because dummy variables produce large standard errors for the overworker sub-sample. A supplementary analysis with dummy variables, however, leads to similar substantive conclusions (online supplementary material: Table A1).

Work hours per week are calculated by dividing the number of hours worked over the course of the previous calendar year by the number of weeks worked in the same time frame. In accordance with previous research (Burger, 2018; Cha, 2013), full-time work is defined as 35 or more hours but less than 50 hours per week and overwork is defined as 50 or more hours per week. The results remain largely similar when overwork is defined at higher thresholds (55 or more hours). When overwork is defined at lower thresholds (48 and 45 or more hours), the difference in the motherhood penalty between full-time workers and overworkers is not statistically significant.

Control variables include basic demographic characteristics, human capital, labour supply and job characteristics. Basic demographic characteristics include age, marital status (never married, married, or divorced), spouse’s full-time employment status (an indicator that a spouse works 35 or more hours per week), region (East, West, South, or North), and urban residence (an indicator for urban residence).

Following previous studies (England et al., 2016; Yu and Kuo, 2017), human capital, which indicates job qualifications and work experience, is measured by the number of years of schooling, years of seniority and years of work experience. Having a child may limit the accumulation of mothers’ human capital and thus reduce the earnings (Budig and England, 2001). Labour supply, which reflects work effort and commitment, is measured by weekly work hours and the number of weeks worked in the past year. Weekly work hours are included as a continuous variable to control for the potential difference in work hours between childless women and mothers among full-time workers and overworkers, separately. Years of seniority, years of work experience and weekly work hours are top- and bottom-coded at the 95th percentile and 1st percentile, respectively.

Job characteristics include percentage of female workers in the occupation, employment in the public sector and professional/managerial occupations. The percentage of female workers in a given occupation is obtained from the Current Population Studies and combined with the data collected from the NLSY through an occupation code crosswalk (Meyer and Osborne, 2005). Public sector refers to being employed in the government sector and is included as a dummy variable. Working in a professional or managerial occupation is included as a dummy variable.

Whereas most correlations between these variables are weak, several correlations (i.e. years of seniority and years of work experience) are high. However, variance inflation factors (VIF) show no substantial multicollinearity problems.

Methods

Following previous studies on the motherhood penalty (England et al., 2016; Yu and Kuo, 2017), this study uses person-fixed effect models to explore whether within-person change in motherhood status affects earnings and how the within-person motherhood penalty varies across work hour status. Fixed effect models have advantages in drawing a causal inference for the motherhood penalty because they reduce selection bias from time-invariant unobserved characteristics (Allison, 2009). I used dummy variables for years to adjust for the secular trend.

The main models are run separately for full-time workers and overworkers in order to investigate whether the effects of within-person change in the number of children on earnings vary by work hour status. In each group, the baseline models demonstrate the motherhood penalties with only the basic demographics included. Similarly to the previous studies on the motherhood penalty (i.e. Budig and England, 2001), I investigate how the baseline motherhood penalty changes according to the progressive adjustment for human capital, labour supply and job characteristics.

The supplementary models estimate the difference in the motherhood penalty between full-time workers and overworkers with the interaction term between number of children and work hour status (online supplementary material: Table A2). Whereas the main model is primarily interested in the difference in the motherhood penalty between mothers who remain in full-time work and mothers who remain in overwork, these results may not capture how within-individual transition in work hour status (e.g. change from overwork to full-time work) can affect the motherhood penalty. As mothers’ work hours may substantially vary across the life course (Killewald and Zhuo, 2019), the main models may have limited implications for the role of long work hours in the motherhood penalty. To address this issue, I further investigate how within-individual differences in work hour status moderate the effect of number of children on earnings.

As both the number of children and overwork are time-varying characteristics, the interaction effect between these variables was estimated by a ‘double-demeaned’ estimator, which demeans the product of the two demeaned variables. Whereas the product of two variables or demeaned two variables produces consistent estimators for the interaction effect between one time-invariant variable and one time-varying variable in fixed effect models, such an approach does not accurately estimate the interaction effect between two time-varying variables because it does not tease apart unobserved time-invariant effects from time-varying effects (Giesselmann and Schmidt-Catran, 2018). Although a ‘double-demean’ estimator is less efficient than the standard interaction estimators, it prevents bias from unobserved unit-specific components (Giesselmann and Schmidt-Catran, 2018).

Results

Table 1 shows that among white and Black women, overworking mothers have several distinct characteristics from mothers who work full-time. Annual earnings are higher per year among overworking mothers than mothers who work full-time. Overworking mothers report more years of schooling and years of work experience than mothers who work full-time. Among white women, overworking mothers tend to have more years of seniority than mothers who work full-time. Overworking mothers also exhibit more weekly work hours and weeks worked in the past year than mothers who work full-time. Overworking mothers are also more likely to be in professional occupations and are less likely to be in female-dominated occupations than mothers who work full-time.

Table 1.

Descriptive Statistics on Childless Women and Mothers, National Longitudinal Survey of Youth 1979 -2014.

	White women				Black women
	Full-time work		Overwork		Full-time work		Overwork
	Childless	Mothers	Childless	Mothers	Childless	Mothers	Childless	Mothers
	M(SD)	M(SD)	M(SD)	M(SD)	M(SD)	M(SD)	M(SD)	M(SD)
Number of person-year observations	13,216	13,200	1760	1939	3939	9020	634	1314
Number of person observations	2343	1502	667	797	678	685	212	490
Log of annual earnings (US$2014)^a,b,e,f,h	10.09	10.11	10.40	10.42	9.9	9.95	10.43	10.32
	(0.81)	(0.90)	(0.89)	(0.92)	(0.9)	(0.94)	(0.88)	(0.85)
Annual earnings (US$2014)^{a,b,c,d,e,f,h}	30,544.88	32,653.92	42,742.61	45,230.54	28,323.6	28,282.7	43,536.8	38,942.5
	(18,906.66)	(22,088.85)	(25,577.85)	(30,202.49)	(19,059.6)	(18,611.9)	(24,933.7)	(24,060.7)
Number of children^b,f	−	1.87	−	2.02	−	2.09	−	2.23
	−	(0.91)	−	(0.91)	−	(1.06)	−	(1.17)
Basic demographics
Age^{a,b,c,d,e,f,g,h}	27.04	35.21	30.94	39.25	29.06	35.05	33.26	38.94
	(7.35)	(9.49)	(8.18)	(9.19)	(8.41)	(9.28)	(8.77)	(8.76)
Marital status (Ref.: never married)
Married^{a,b,c,d,f,g,h}	0.37	0.72	0.29	0.67	0.21	0.40	0.20	0.37
	(0.48)	(0.45)	(0.46)	(0.47)	(0.41)	(0.49)	(0.40)	(0.48)
Divorced^{a,b,c,d,e,f,g,h}	0.09	0.23	0.15	0.30	0.11	0.28	0.15	0.37
	(0.29)	(0.42)	(0.36)	(0.46)	(0.31)	(0.45)	(0.35)	(0.48)
Spouse’s full-time work^{a,b,c,d,f,g,h}	0.37	0.70	0.30	0.64	0.20	0.39	0.21	0.36
	(0.48)	(0.46)	(0.46)	(0.48)	(0.40)	(0.49)	(0.41)	(0.48)
Region (Ref.: East)
North Central^b,c,d,g,h	0.28	0.31	0.26	0.33	0.14	0.18	0.11	0.19
	(0.45)	(0.46)	(0.44)	(0.47)	(0.34)	(0.38)	(0.31)	(0.39)
South^c,d,g	0.32	0.39	0.32	0.41	0.61	0.64	0.64	0.64
	(0.47)	(0.49)	(0.47)	(0.49)	(0.49)	(0.48)	(0.48)	(0.48)
West^a,d,f,g,h	0.16	0.16	0.19	0.16	0.08	0.05	0.10	0.06
	(0.37)	(0.37)	(0.39)	(0.37)	(0.27)	(0.21)	(0.29)	(0.25)
Urban^c,d,e,f,g,h	0.76	0.65	0.77	0.66	0.86	0.83	0.90	0.85
	(0.43)	(0.48)	(0.42)	(0.48)	(0.35)	(0.38)	(0.30)	(0.36)
Human capital
Years of schooling^{a,b,c,d,e,f,g,h}	13.66	12.96	14.82	13.81	13.68	13.05	14.89	13.55
	(2.16)	(2.17)	(2.53)	(2.71)	(2.01)	(1.86)	(2.54)	(2.06)
Years of seniority^a,b,c,d,e,g	3.19	5.42	4.01	5.87	3.99	5.31	5.61	5.59
	(3.90)	(5.88)	(5.06)	(6.46)	(4.89)	(5.94)	(6.14)	(6.11)
Years of work experience^{a,b,c,d,e,f,g,h}	8.03	13.69	11.39	17.03	8.79	12.12	12.97	15.98
	(6.75)	(8.76)	(7.63)	(8.93)	(7.51)	(8.29)	(8.45)	(8.62)
Labour supply
Weekly work hours^{a,b,c,d,e,f,h}	40.63	40.50	54.80	55.34	45.27	45.43	48.38	47.68
	(3.10)	(2.89)	(4.29)	(5.00)	(12.48)	(12.37)	(9.77)	(10.25)
Weeks worked in the past year^a,b,c,e,f	46.70	45.87	47.26	47.10	40.39	40.32	55.97	56.49
	(10.70)	(11.90)	(10.62)	(10.62)	(2.93)	(2.71)	(4.78)	(5.18)
Job characteristics
Professional occupations^a,b,e,f,g,h	0.32	0.31	0.51	0.51	0.25	0.22	0.48	0.36
	(0.47)	(0.46)	(0.50)	(0.50)	(0.43)	(0.41)	(0.50)	(0.48)
Public^a,b,c,d,e,h	0.13	0.17	0.17	0.21	0.24	0.25	0.30	0.25
	(0.34)	(0.37)	(0.37)	(0.41)	(0.43)	(0.43)	(0.46)	(0.44)
% females in occupation^{a,b,c,e,f,g,h}	63.23	64.17	56.68	56.23	65.91	64.45	60.71	60.88
	(27.67)	(26.72)	(26.24)	(26.58)	(26.54)	(26.44)	(26.80)	(26.43)

Notes: Data given as means and standard deviations (in parentheses) of the variables. The NLSY 79 oversampled Black respondents.

The difference between full-time workers and overworkers among white childless women is statistically significant (p < 0.05).

The difference between full-time workers and overworkers among white mothers is statistically significant (p < 0.05).

The difference between childless women and mothers among white full-time workers is statistically significant (p < 0.05).

The difference between childless women and mothers among white overworkers is statistically significant (p < 0.05).

The difference between full-time workers and overworkers among Black childless women is statistically significant (p < 0.05).

The difference between full-time workers and overworkers among Black mothers is statistically significant (p < 0.05).

The difference between childless women and mothers among Black full-time workers is statistically significant (p < 0.05).

The difference between childless women and mothers among Black overworkers is statistically significant (p < 0.05).

Overworking mothers are distinguished from overworking childless women in many ways. Overworking mothers are older, partially because many overworking childless women eventually have children over time. Overworking mothers also tend to have fewer years of schooling but more years of work experience than overworking childless women. Among white women, overworking mothers report more years of tenure. Overworking mothers report more weekly work hours than overworking childless women.

Table A3 (online supplementary material) shows the motherhood earnings penalty among Black full-time workers and overworkers. The coefficient for the number of children represents the impact of each additional child on annual earnings. Although each additional child may decrease earnings to some extent among Black women, each additional child does not have a significant effect on earnings either in full-time work or overwork. For this reason, further comparison for full-time workers and overworkers is limited to white women whose motherhood penalty is more substantial.

In Table 2, results show that per child penalties are not statistically significant among white overworkers (Model 1). The absence of per child penalties among white overworkers contrasts with significant child penalties among white full-time workers. In full-time work, each additional child reduces earnings by 11.3% on average. The significant reduction in earnings is consistent with previous findings on the wage penalty (i.e. Budig and England, 2001).

Table 2.

Fixed effects for the motherhood earnings penalty among white women, by work hour (N = 26,416 full-time work, 3699 overwork).

	Model 1		Model 2		Model 3		Model 4
	Full-time work	Overwork	Full-time work	Overwork	Full-time work	Overwork	Full-time work	Overwork
Number of children	−0.12***	−0.05^a	−0.09***	−0.05	−0.04***	−0.05	−0.04***	−0.05
	(0.01)	(0.03)	(0.01)	(0.03)	(0.01)	(0.03)	(0.01)	(0.03)
Basic demographics	Included	Included	Included	Included	Included	Included	Included	Included
Human capital
Years of schooling			0.14***	0.10**	0.08***	0.06*	0.08***	0.06*
			(0.01)	(0.03)	(0.01)	(0.03)	(0.01)	(0.03)
Years of tenure			0.02***	0.01***	0.01***	0.01	0.01***	0.005
			(0.002)	(0.003)	(0.001)	(0.003)	(0.001)	(0.003)
Years of experience			0.03***	−0.01	0.03***	0.001	0.03***	0.003
			(0.01)	(0.02)	(0.004)	(0.015)	(0.004)	(0.015)
Labour supply
Weekly work hours					0.02***	−0.00004	0.02***	−0.0003
					(0.001)	(0.00257)	(0.001)	(0.0025)
Annual work weeks					0.04***	0.03***	0.04***	0.03***
					(0.0006)	(0.002)	(0.001)	(0.002)
Job characteristics
Public sector							−0.01	−0.03
							(0.02)	(0.05)
% females in occupation							−0.001***	−0.0005
							(0.0002)	(0.001)
Professional occupations							0.07***	0.09**
							(0.01)	(0.03)
Constant	9.38***	8.16***	7.79***	6.64***	6.25***	6.49***	6.45***	6.65***
	(0.26)	(0.65)	(0.28)	(0.77)	(0.22)	(0.66)	(0.22)	(0.65)

Notes: Basic demographics include age, marital status, spouse’s full-time work, region and urban residence. Dummy variables for years are included in all the models. Weekly work hour adjusts for the difference in work hours between childless women and mothers among full-time workers and overworkers, separately. ^aThe coefficient is significantly larger from the coefficient for full-time workers (one-tailed test, p < 0.05).

p < 0.05, **p < 0.01, ***p < 0.001.

The difference in the coefficients of the number of children for earnings between white full-time workers and overworkers indicates the difference in the motherhood earning penalties between white full-time workers and overworkers. The test on comparisons of the coefficients (Clogg et al., 1995) shows that, when basic demographics are held constant, per child penalties in full-time work are statistically greater than per child penalties in overwork among white women. This finding provides evidence to support the idea that working long hours reduces the negative effect of motherhood on earnings among white women, thus supporting the extra-effort hypothesis.

Once human capital is adjusted for in Model 2, the gap in the motherhood earnings penalty between white full-time workers and overworkers is reduced to statistical non-significance. The gap in per child penalties between white full-time workers and overworkers narrows due to a considerable decrease in per child penalties for full-time workers after adjusting for human capital. Controlling for human capital reduces the motherhood penalty to 8.6% in full-time work. In contrast to the motherhood penalty in full-time work, adjusting for human capital does not change the statistically insignificant motherhood penalty in overwork. This finding suggests that although human capital explains a good portion of the motherhood penalty for white full-time workers, it does not alter the coefficient for the number of children among white overworkers. This may reflect the fact that mothers who select into overwork tend to have fairly high levels of schooling, tenure and work experience.

Model 3 includes weekly work hours and annual work weeks to capture how labour supply explains the motherhood earnings penalty. In this model, the gap in the motherhood penalty between white full-time workers and overworkers becomes even smaller. This further reduction in the gap in per child penalties between white full-time workers and overworkers is driven by the drastic decline in the motherhood penalty for white full-time workers. For full-time work, adjusting for labour supply substantially reduces the motherhood penalty to 4%, although the motherhood penalty remains robust (Model 3). Whereas labour supply captures a key mechanism of the motherhood penalty in full-time work, it does not alter the statistically insignificant motherhood penalty for white overworkers. This might be due to the fact that overworking mothers do not differ from overworking childless women in the number of weeks worked per year. This model introduces work hours as a continuous variable to rule out the potential differences between childless women and mothers in each work hour group. The continuous variable for work hours is bounded between 35 and 49 for full-time workers and 50 and 68 (top-coded) for overworkers.

Including job characteristics in the analysis does little to explain the motherhood penalty for both full-time work and overwork. The difference in the effects of each job characteristic on earnings between full-time work and overwork is not statistically significant. Consistent with Budig and England (2001), this suggests that there is little evidence to support that the motherhood penalty is associated with privileged occupations or mother-friendly job characteristics.

Supplementary analyses

Considering that the intensity of overwork is often greater in professional occupations (Moen et al., 2013), these results may mask the occupational contexts of overwork that can potentially reshape the motherhood penalty. I found no evidence that professional occupations lessen or exacerbate the motherhood penalty in overwork (online supplementary material: Table A4).

As previously mentioned, the models in Table A2 estimate the variance of the motherhood penalty with the interaction effect between number of children and overwork. Whereas the main models reveal whether the effect of change in the number of children on earnings differs between full-time workers and overworkers, these supplementary models show whether the effect of change in the number of children on earnings decreases when persons change their work hour status. Whereas the coefficient for the main effect of number of children is negative, results show that the interaction effect between number of children and overwork is not statistically significant in all models. Consistent with the main results, this finding suggests that we have little evidence that within-individual transition from full-time work to overwork significantly changes the magnitude of the motherhood penalty.

Discussion and conclusion

This study asked whether overwork attenuates the motherhood earnings penalty. This study provides evidence to support the idea that working long hours buffers the motherhood earnings penalty among white women. Expanding the argument for the benefits of overwork for workplace reward (Cha and Weeden, 2014; Sturges, 2013), this finding illuminates that intensive time commitment to work may reward mothers by offsetting mother-specific disadvantages at work. Perhaps, working long hours may increase job performance among mothers by allowing them to complete demanding tasks. Another possibility is that overwork among mothers may have a positive impact on job evaluation by reducing their supervisors and colleagues’ scepticism about mothers’ productivity. This study did not fully examine these potential different mechanisms due to a lack of information in the dataset. Future studies should explore how overwork affects employee- and employer-related factors of the motherhood penalty.

Another main finding is that the buffering effect of overwork against the motherhood earnings penalty becomes non-significant when we adjust for human capital and labour supply. This reduction in the buffering effect suggests that working long hours appears to attenuate the motherhood penalty, potentially, because overworking mothers tend to have fairly high levels of human capital and labour supply. In other words, overwork itself may not be protective; its buffering effect is largely associated with the positive selection on human capital and labour supply. This is consistent with the idea that working long hours tends to pull mothers out of their careers and only highly selected mothers remain able to engage in overwork (Cech and Blair-Loy, 2019; Stone, 2007). The findings further show that the selection mechanism of overwork may inflate the benefit of overwork in mitigating mother-specific disadvantages by excluding mothers who are likely to experience a severe penalty.

Importantly, the findings do not imply that mothers in demanding jobs do not face any significant mother-specific challenges at work. The buffering effect of overwork against the motherhood penalty is not incompatible with substantial difficulties in balancing work and family responsibilities among mothers who have demanding careers. Rather, working long hours, which is often a mixture of long paid work hours, extra work without notice and time pressure, can have a considerable impact on work–family conflict (Maume and Houston, 2001; Roeters et al., 2010).

These findings are limited to white women. Consistent with previous studies (England et al., 2016; Glauber, 2007), Black women are less likely to experience the motherhood earnings penalty in both full-time work and overwork, and there is no difference in the motherhood penalty between full-time workers and overworkers among Black women. While a recent study argued that poor access to highly rewarding jobs may explain why Black women report the small motherhood penalty (Van Winkle and Fasang, 2020), some studies suggest that the racial difference in the motherhood penalty may be attributable to racial differences in the work–family interface. Despite the structural disadvantages in access to higher education and high-paying jobs, Black mothers are more likely to stay in full-time work and return quickly to work after childbirth than white mothers (Florian, 2018). Compared to white women, Black women are more likely to be financially independent and rely on caregiving support from kin and community members (Dow, 2016). Such high propensity to attach to the labour market after childbearing may potentially prevent Black mothers from earnings penalties.

Given that working long hours is increasingly common in some affluent countries (Burger, 2018) and exacerbates work–family conflict (Stier et al., 2012), it may be a significant work context in other countries, too, that conditions mother-specific disadvantages at the workplace. However, as working conditions of overworkers and regulations of overtime substantially vary across countries (Burger, 2018; Gornick and Meyers, 2003), additional research is needed to investigate how well findings apply elsewhere.

This study advances our understanding of the role of working long hours in the motherhood penalty. Whereas Weeden et al. (2016) have documented the changing wage gap between overworking childless women and mothers, this study provides more robust causal evidence of the motherhood penalty than do cross-sectional analyses by controlling for unobserved time-invariant individual characteristics (such as generalized productivity). However, this study could not rule out unobserved time-varying individual characteristics. Future studies should explore how overwork can reshape the motherhood penalty by using other methods, such as survey experiments, that allow for better control of unobserved characteristics.

This study enriches our understanding of the heterogeneity of the motherhood penalty across work contexts by theorizing the link between work demand and the gendered work–family interface. This study developed the argument that overwork may enable highly selected mothers to signal additional work commitment and to reduce domestic labour. Whereas previous studies have noted that there is little difference in the motherhood penalty between part-time and full-time workers (Budig and England, 2001; Budig and Hodges, 2010), this study reveals a more nuanced role of work hours in the motherhood earnings penalty by focusing on the distinction between overworkers and other full-time workers.

Previous studies have explored how overwork has long excluded mothers (Cha, 2013; Stone, 2007), yet this study shifts the focus to the systematic differences in the motherhood penalty between overworking mothers and mothers who work full-time. Given that working long hours is increasingly common among mothers (Percheski, 2008; Weeden et al., 2016), the finding may imply that more mothers are unlikely to experience the motherhood penalty. As mothers’ time commitment to work is still limited by their disproportionate share of housework and childcare responsibilities (Bianchi and Milkie, 2010), however, it would be too simplistic to conclude that working long hours would be beneficial for mothers in overcoming mother-specific disadvantages at the workplace. Indeed, as overworkers are expected to be unconstrained by domestic tasks, overwork can attenuate the benefits of workplace family-friendly policies for facilitating mothers’ employment and reinforce workplace expectations and practices that pose challenges to mothers’ careers (Lindsay and Maher, 2014; Niemistö et al., 2021; Stone, 2007; Wright, 2014). As working long hours is at best limited to those who can pay its severe cost on family life and well-being (Bianchi and Milkie, 2010; Voydanoff, 2004), I call attention to the broader implications of long work hours in mothers’ disadvantages at work.

Supplemental Material

sj-docx-1-wes-10.1177_09500170211041293 – Supplemental material for Does Overwork Attenuate the Motherhood Earnings Penalty among Full-Time Workers?

Supplemental material, sj-docx-1-wes-10.1177_09500170211041293 for Does Overwork Attenuate the Motherhood Earnings Penalty among Full-Time Workers? by Eunjeong Paek in Work, Employment and Society

Footnotes

Acknowledgements

An earlier version of this work was presented at the 2019 Annual Meeting of the Population Association of America, Austin, TX. Thank you Jennifer Hook for guidance and comments throughout many revisions of this article. Thank you Lynne Casper, Sarah Thébaud, Jennifer Candipan, Sandra Florian, Maria Adamson and the anonymous reviewers for helpful comments. Thank you to the participants of the USC Sociology Colloquium and the UCSB Gender & Sexuality Workshops for helpful comments and suggestions. Thank you Alli Coritz, Latoya Council, Yael Findler, Mary Ippolito, Briana Jex and Lizette Solorzano for your help in polishing the draft.

Funding

The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by the USC Korean Studies Fellowship.

ORCID iD

Eunjeong Paek

Supplementary material

The supplementary material is available online with the article.

Eunjeong Paek is a PhD candidate in the Department of Sociology at the University of Southern California. Her dissertation examines how changing labour markets affect family-friendly policies with a focus on technological advances, union decline and demographic shifts in the labour force. She also studies the stratification of women’s employment from cross-national perspectives, the relationship between family-friendly policies and gendered workplaces, and the intersection of family transition and labour outcomes.

References

Allison

(2009) Fixed Effects Regression Models. Thousand Oaks, CA: SAGE.

Aughinbaugh

Pierret

Rothstein

(2017) Attrition and Its Implications in the National Longitudinal Survey of Youth 1979. Washington, DC: Bureau of Labor Statistics.

Becker

(1985) Human capital, effort, and the sexual division of labor. Journal of Labor Economics 3(1): 33–58.

Benard

Correll

(2010) Normative discrimination and the motherhood penalty. Gender & Society 24(5): 616–646.

Berg

Bosch

Charest

(2014) Working-time configurations: a framework for analyzing diversity across countries. Industrial and Labor Relations Review 67(3): 805–837.

Bianchi

Milkie

(2010) Work and family research in the first decade of the 21st century. Journal of Marriage and Family 72(3): 705–725.

Bianchi

Sayer

Milkie

, et al. (2012) Housework: who did, does or will do it, and how much does it matter? Social Forces 91(1): 55–63.

Blair-Loy

(2003) Competing Devotions. Cambridge, MA: Harvard University Press.

Blair-Loy

(2009) Work without end? Scheduling flexibility and work-to-family conflict among stockbrokers. Work and Occupations 36(4): 279–317.

10.

Buchman

McDaniel

(2016) Motherhood and the wages of women in professional occupations. RSF: The Russell Sage Foundation Journal of the Social Sciences 2(4): 128–150.

11.

Budig

England

(2001) The wage penalty for motherhood. American Sociological Review 66(2): 204–225.

12.

Budig

Hodges

(2010) Differences in disadvantage: variation in the motherhood penalty across white women’s earnings distribution. American Sociological Review 75(5): 705–728.

13.

Budig

Misra

Boeckmann

(2016) Work–family policy trade-offs for mothers? Unpacking the cross-national variation in motherhood earnings penalties. Work and Occupations 43(2): 119–177.

14.

Burger

(2018) Extreme work hours in Western Europe and North America: diverging trends since the 1970s. Socio-Economic Review 18(4): 1–23.

15.

Casper

Bianchi

(2002) Change and Continuity in the American Family. Thousand Oaks, CA: SAGE.

16.

Cech

Blair-Loy

(2019) The changing career trajectories of new parents in STEM. Proceedings of the National Academy of Sciences of the United States of America 116(10): 4182–4187.

17.

Cha

(2013) Overwork and the persistence of gender segregation in occupations. Gender & Society 27(2): 158–184.

18.

Cha

Weeden

(2014) Overwork and the slow convergence in the gender gap in wages. American Sociological Review 79(3): 457–484.

19.

Clogg

Petkova

Haritou

(1995) Statistical methods for comparing regression coefficients between models. American Journal of Sociology 100(5): 1261–1293.

20.

Cooke

(2014) Gendered parenthood penalties and premiums across the earnings distribution in Australia, the United Kingdom, and the United States. European Sociological Review 30(3): 360–372.

21.

Correll

Benard

Paik

(2007) Getting a job: is there a motherhood penalty? American Journal of Sociology 112(5): 1297–1339.

22.

Craig

Mullan

(2010) Parenthood, gender and work–family time in the United States, Australia, Italy, France, and Denmark. Journal of Marriage and Family 72(5): 1344–1361.

23.

Craig

Mullan

(2011) How mothers and fathers share childcare: a cross-national time-use comparison. American Sociological Review 76(6): 834–861.

24.

Dow

(2016) Integrated motherhood: beyond hegemonic ideologies of motherhood. Journal of Marriage and Family 78(1): 180–196.

25.

England

Bearak

Budig

, et al. (2016) Do highly paid, highly skilled women experience the largest motherhood penalty? American Sociological Review 81(6): 1161–1189.

26.

Florian

(2018) Motherhood and employment among whites, Hispanics, and blacks: a life course approach. Journal of Marriage and Family 80(1): 134–149.

27.

Fuller

Hirsh

(2018) ‘Family-friendly’ jobs and motherhood pay penalties: the impact of flexible work arrangements across the educational spectrum. Work and Occupations 46(1): 1–42.

28.

Gangl

Ziefle

(2009) Motherhood, labor force behavior, and women’s career: an empirical assessment of the wage penalty for motherhood in Britain, Germany, and the United States. Demography 46(2): 341–369.

29.

Gerstel

Clawson

(2015) Class advantage and the gender divide: flexibility on the job and at home. American Journal of Sociology 120(2): 395–431.

30.

Giesselmann

Schmidt-Catran

(2018) Interactions in Fixed Effects Regression Models. Berlin: German Institute for Economic Research, DIW Berlin.

31.

Glauber

(2007) Marriage and the motherhood wage penalty among African Americans, Hispanics, and Whites. Journal of Marriage and Family 69(4): 951–961.

32.

Glauber

(2008) Race and gender in families and at work. Gender & Society 22(1): 8–30.

33.

Gornick

Meyers

(2003) Families That Work. New York: Russel Sage Foundation.

34.

Hook

(2017) Women’s housework: new tests of time and money. Journal of Marriage and Family 79(1): 179–198.

35.

Killewald

(2011) Opting out and buying out: wives’ earnings and housework time. Journal of Marriage and Family 73(2): 459–471.

36.

Killewald

Zhuo

(2019) U.S. mothers’ long-term employment patterns. Demography 56(1): 285–320.

37.

Kmec

(2011) Are motherhood penalties and fatherhood bonuses warranted? Comparing pro-work behaviors and conditions of mothers, fathers, and non-parents. Social Science Research 40(2): 444–459.

38.

Kornrich

Roberts

(2018) Household income, women’s earnings, and spending on household services, 1980–2010. Journal of Marriage and Family 80(1): 150–165.

39.

Lindsay

Maher

(2014) The intersections of work time and care time: nurses’ and builders’ family time economies. Work, Employment and Society 28(2): 189–205.

40.

Maume

Houston

(2001) Job segregation and gender differences in work–family spillover among white-collar workers. Journal of Family and Economic Issues 22(2): 171–189.

41.

Meyer

Osborne

(2005) Proposed category system for 1960–2000 Census occupations. BLS Working Papers. Washington, DC: U.S. Department of Labor, Bureau of Labor Statistics.

42.

Moen

Lam

Ammons

, et al. (2013) Time work by overworked professionals: strategies in response to the stress of higher status. Work and Occupations 40(2): 79–114.

43.

Mullan

Wajcman

(2019) Have mobile devices changed working patterns in the 21st century? A time-diary analysis of work extension in the UK. Work, Employment and Society 33(1): 3–20.

44.

Neely

(2020) The portfolio ideal worker: insecurity and inequality in the new economy. Qualitative Sociology 43(2): 271–296.

45.

Niemistö

Hearn

Kehn

, et al. (2021) Motherhood 2.0: slow progress for career women and motherhood within the ‘Finnish Dream’. Work, Employment and Society 35(4): 696–715.

46.

Pas

Peters

Eisinga

, et al. (2011) Explaining career motivation among female doctors in the Netherlands: the effects of children, views on motherhood and work–home cultures. Work, Employment and Society 25(3): 487–505.

47.

Percheski

(2008) Opting out? Cohort differences in professional women’s employment rates from 1960 to 2005. American Sociological Review 73(3): 497–517.

48.

Reynolds

(2003) You can’t always get the hours you want: work hours in the U.S. Social Forces 81(4): 1171–1199.

49.

Roeters

Lippe

Van Der Kluwer

(2010) Work characteristics and parent–child relationship quality: the mediating role of temporal involvement. Journal of Marriage and Family 72(5): 1317–1328.

50.

Sharone

(2004) Engineering overwork: bell-curve management at a high-tech firm. In: Epstein

Kalleberg

(eds) Fighting for Time. New York: Russell Sage Foundation, 191–218.

51.

Stier

Lewin-Epstein

(2003) Time to work: a comparative analysis of preferences for working hours. Work and Occupations 30(3): 302–326.

52.

Stier

Lewin-Epstein

Braun

(2012) Work–family conflict in comparative perspective: the role of social policies. Research in Social Stratification and Mobility 30(3): 265–279.

53.

Stone

(2007) Opting Out? Why Women Really Quit Careers and Head Home. Berkeley, CA; Los Angeles, CA: University of California Press.

54.

Sturges

(2013) A matter of time: young professionals’ experiences of long work hours. Work, Employment and Society 27(2): 343–359.

55.

Turco

(2010) Cultural foundations of tokenism: evidence from the leveraged buyout industry. American Sociological Review 75(6): 894–913.

56.

Van Winkle

Fasang

(2020) Parenthood wage gaps across the life course: a comparison by gender and race. Journal of Marriage and Family 82(5): 1515–1533.

57.

Voydanoff

(2004) Demands and resources on the effects of work conflict and facilitation. Journal of Marriage and Family 66(1): 398–412.

58.

Weeden

Cha

Buccac

(2016) Long work hours, part-time work, and trends in the gender gap in pay, the motherhood wage penalty, and the fatherhood wage premium. RSF: The Russell Sage Foundation Journal of the Social Sciences 2(4): 71–102.

59.

Wright

(2014) Gender, sexuality and male-dominated work: the intersection of long-hours working and domestic life. Work, Employment and Society 28(6): 985–1002.

60.

Kuo

(2017) The motherhood wage penalty by work conditions: how do occupational characteristics hinder or empower mothers? American Sociological Review 82(4): 744–769.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.04 MB