## Introduction

The gender wage gap in the United States between employed men and women sharply declined throughout the 1970s and 1980s, followed by a slowing of the convergence of men’s and women’s wages since the late 1990s (Blau and Kahn 2017). In concert with the reduction and then stagnation of the gender wage gap, women and mothers entered the labor force in high numbers from the 1960s through the 1980s. Since, there has been a leveling off such that the gender gap in labor force participation has remained steady at about 13% to 15% since the mid-1990s, with working- age men’s labor force participation hovering around 90% and women’s being about 75% (FRED 2017). The associations between work experience and attachment, short-term employment interruptions, and wages have been well documented in extant literature, suggesting that women’s lost work experience is a culprit for the persistent gender wage gap (Aisenbrey and Bruckner 2008; Alon and Haberfeld 2007; Blau and Kahn 2017; Cha and Weeden 2014).

Recent demographic and sociological research has established that aggregate labor force participation rates mask the movement of individuals into different labor force statuses throughout their careers—a common occurrence in the modern labor market. Scholarship has found substantial variation in employment patterns within individuals over the life course, predicted by gender, race, ethnicity, and social class (Damaske and Frech 2016; Halpern-Manners et al. 2015; Killewald and Zhuo 2019; Lu et al. 2017; Percheski 2018). Meanwhile, economic scholars have examined the effects of temporary employment lapses on wages (Albrecht et al. 1999; Arulampalam 2001; Eriksson and Rooth 2014; Gangl 2006). Building on these foundational studies and integrating these areas of literature, we focus in this article on how long-term employment patterns affect subsequent wages. We examine how both the timing and level of employment in early to midcareer affect wages, considering cumulative employment rather than short-term employment effects.

Insofar as long-term employment is associated with subsequent wages, scholarship predicts that employment levels across career stages may be associated with different family formations (Aisenbrey and Fasang 2017; García-Manglano 2015; Killewald and Zhuo 2019; Percheski 2018) and work characteristics (Abendroth et al. 2014; Aisenbrey et al. 2009; Cha 2010; Hodges and Budig 2010; Jacobs and Gerson 2001; Landivar 2014; Stone and Lovejoy 2019; Williams and Boushey 2010)—factors that may themselves contribute to wage levels. We thus examine whether and how the effects of employment pathways on wages are explained by individuals’ family decisions (marriage and childbirth) and work characteristics (occupation and intensity of work when employed).

Further, given that employment pathways are differently distributed by gender—with men being more likely than women to hold steady, sustained employment throughout their careers—theories on gender and employment argue that the relative wage payoffs or penalties to following certain pathways could themselves be gendered. We build on literature finding that labor force attachment is a contributor of the gender wage gap (Blau and Kahn 2017; Spivey 2005) to ask to what extent gender wage gaps vary within long-term employment pathways.

To assess these questions, we use data from the National Longitudinal Survey of Youth (NLSY) 1979. We begin by developing six employment trajectories that emerge empirically from the 1979–2014 samples of U.S. men and women aged 22–50. Women are overrepresented in late-entry and midcareer-interruption trajectories, and men are overrepresented in a steady high employment trajectory. We find that Black men and individuals who experienced poverty are more likely to follow intermittent or nonsteady pathways, lending evidence that the intersections of race, class background, and gender are associated with employment throughout the life course.

In the second analytical section, we examine wage variation in later career stages associated with following certain employment paths. We find that low work attachment trajectories have large negative effects on wages, relative to steady high employment. These trajectory effects are explained largely by work characteristics and, to a lesser extent, by family formation patterns. Further, following these long-term employment pathways has gendered consequences. Men experience lower wage premiums in nonsteady employment pathways than they do in steady high employment, aligning with predictions from theories of norm violation for men and fathers. Women thus face a double bind in employment, with a trade-off between economic prosperity and gender equity: women who “become more like men” and remain steadily attached to work experience a larger gender wage gap, but those who are less attached incur large wage penalties. Our study contributes to theories and scholarship on life course employment from an intersectional perspective, on wage inequality, and on gendered labor market outcomes.

## Employment Over the Life Course: Heterogeneity by Gender, Race and Ethnicity, and Class

Life course perspectives suggest that the timing and order of key events matter throughout the aging process (Elder et al. 2003). A life course approach focuses on how transitions between states have social dynamics, allowing researchers to focus on trajectories or sequences rather than snapshot perspectives of individuals at a single point in time (Elder et al. 2003; Frytak et al. 2003; Macmillan and Copher 2005). In the United States, scholars have found that gender, parental status, race/ethnicity, and social class background predict employment patterns independently and intersectionally (Aisenbrey and Fasang 2017; García-Manglano 2015; Halpern-Manners et al. 2015; Lu et al. 2017). These findings are generally consistent with life course theories of demographic status (dis)advantage: individuals who have relatively advantaged demographic statuses or who grew up in advantaged contexts are more able than individuals who are less advantaged to secure stable employment opportunities (Bielby and Bielby 1992; Fernandez-Mateo 2009; O’Rand 1996).

A primary reason why women incur employment lapses is to care for their children; a substantial minority of women scale back or drop out of employment upon motherhood (Abendroth et al. 2014; England et al. 2004; García-Manglano 2015; Hynes and Clarkberg 2005; Lu et al. 2017; Percheski 2018). With respect to race and ethnicity, Black and Hispanic women are more likely than White women to continue employment after childbirth, net of socioeconomic status (Killewald and Zhuo 2019; Lu et al. 2017). Social class has a nonuniform effect on women’s employment pathways: low-wage workers without schedule control may use employment lapses to cover childcare gaps (Landivar 2014, 2017; Percheski 2018; Williams and Boushey 2010), whereas mothers in inflexible and high-intensity professional jobs and mothers with spouses who overwork (work more than 50 hours per week) may “opt out”1 of work to care for children (Cha 2010, 2013; Gerstel and Clawson 2014, 2018; Han and Moen 1999; Roehling et al. 2001; Stone 2007).

Men’s long-term employment has been studied to a lesser extent, and studies suggest that men are more likely than women to hold steady high levels of employment throughout their careers (Aisenbrey and Fasang 2017; Halpern-Manners et al. 2015; Han and Moen 1999) and that most men have a stable income during working ages (Frech and Damaske 2019). Yet, men’s employment is not uniformly high: social class and race/ethnicity shape employment prospects. Less-educated men and those from lower-class backgrounds tend to be less able to secure consistent stable employment, in part because of inconsistent work availability (Aisenbrey and Fasang 2017; Kalleberg 2000). With respect to race, Black men and Black fathers have been shown not to sustain high-intensity employment at the same levels of White men overall as well as White fathers (Aisenbrey and Fasang 2017; Glauber and Gozjolko 2011).

In addition to demographic statuses, such as gender, race, and social class, predicting employment pathways, human capital and gender beliefs may also affect employment, all else being equal. A human capital perspective argues that individuals who have higher wage potential resulting from investment in education should have higher sustained rates of employment throughout their careers because of increased opportunity costs for exiting work (Blau and Kahn 2017; Lundberg and Rose 2000). Furthermore, gender attitudes may produce effects on employment paths even beyond material circumstances. This theoretical perspective suggests that those with more traditional gender beliefs may make employment decisions to maintain male breadwinner and female caretaker expectations (Damaske and Frech 2016; Glauber 2008).

## Contribution and Predictions: Employment Trajectories

Our study tests the integration of the aforementioned theoretical perspectives on a sample of men and women. Given the dearth of studies on men’s long-term employment, scholarship on the independent and intersectional effects of gender, race and ethnicity, and social class within the context of long-term employment pathways has been limited (for exceptions, see Aisenbrey and Fasang 2017; Alon and Haberfeld 2007; Halpern-Manners et al. 2015). We thus investigate variation in employment patterns both within and between men and women of different racial/ethnic and social class backgrounds.

Existing literature has found considerable variation in long-term employment outcomes (Aisenbrey and Fasang 2017; Damaske and Frech 2016; Killewald and Zhuo 2019). In line with this scholarship, we expect to find at least four employment trajectories. First, we anticipate at least one pathway of steady employment: sustained employment throughout individuals’ careers, with few or no interruptions. Second, we expect to find at least one trajectory of delayed employment: low levels of employment early in individuals’ careers (associated with having children at young ages or extended education or training), followed by increased employment rates toward the later portions of their careers. Third, we also expect to find a trajectory of midcareer reduced employment: beginning with high employment levels but then cutting back for a substantial period in midcareer as family demands ramp up, and returning to high levels of employment in late career. Finally, we expect to find one or more trajectories associated with early reductions in employment: stable high employment levels that start decreasing in the late career, resulting from family demands or reduced workplace opportunities (Flippen and Tienda 2000).

## Theoretical Perspectives on Wages Payoffs and Penalties of Employment Trajectories

Long-term employment trajectories are expected to be associated with subsequent wages. Three theoretical perspectives predict this relationship: skill deterioration, family formation, and work opportunities. After we discuss these theoretical orientations and how they relate to variation in employment pathways, we then turn to how gender may differentially affect the returns to following unsteady employment trajectories. The gender wage gap literature does not offer clear predictions for how gendered payoffs and penalties to long-term employment may vary by class and race/ethnicity. Employment attachment has been posited to partially explain within-gender racial variation in wages (see Alon and Haberfeld 2007), but the gender theories that we discuss shortly are not as clearly connected to intersections with race/ethnicity or class. We thus focus on gender in this section, but we later discuss intersections with race/ethnicity and class in the empirical analysis.

First, the skill deterioration perspective emphasizes that steady high employment is associated with consistent maintenance and development of skills—suggesting that relative to sustained employment, all other trajectories will produce wage penalties. The largest wage penalties are predicted to occur for the trajectories with the longest periods of nonemployment because individuals in these paths are simultaneously losing existing skills and falling behind on building up-to-date skills (Becker 1962). Prior studies have provided support for this idea, finding that a bout of unemployment or time out of the labor force is associated with reduced subsequent wages (Albrecht et al. 1999; Arulampalam 2001; Gangl and Ziefle 2009; Hotchkiss and Pitts 2007; Jacobsen and Levin 1995).

Next, the family formation perspective predicts that the reduced attachment trajectories associated with childrearing or marriage may produce wage penalties. The logic of this argument aligns with both supply- and demand-side explanations. On the supply side, time out of work for family reasons can change individuals’ work/family preferences: for example, they may develop preferences for reduced workloads and accept lower wages for work (Stone and Lovejoy 2019). On the demand side, employers may view family-related lapses as indications of lower commitment levels, thus offering lower wages to employees with such circumstances (Evertsson 2016; Theunissen et al. 2011; Weisshaar 2018). The family formation perspective would predict particularly low wages for pathways with work interruptions aligning with family formation timing.

Finally, the work opportunities perspective suggests that employment paths are associated with different occupational groups and may vary by work intensity as well. Again, both supply- and demand-side perspectives may be at play here: lost experience and time out of work may change individuals’ preferences for jobs/occupations or work intensity, or employers may offer fewer opportunities to those who have had long periods of nonemployment. Existing literature has documented that despite varying in terms of working conditions, benefits, and opportunities, part-time work is generally underpaid (Webber and Williams 2008; Weeden et al. 2016), but overwork is associated with wage premiums (Cha and Weeden 2014). Insofar as sustained employment is associated with overwork and intermittent trajectories are correlated with part-time employment, work intensity may explain some of the negative association between non–steady high paths and wages. Further, some occupations may themselves be associated with employment interruptions and reduced wages—for example, occupations that follow contract or seasonal work, such as construction and some service jobs (Kalleberg 2000).

## Gender Variation in Trajectories’ Effects on Wages

The aforementioned perspectives predict that less attached trajectories will lead to reduced wages, but we ask how the gender wage gap will vary within employment trajectories. It is well established that in the United States, there exists (1) a gender wage gap, such that women earn less than men, on average; and (2) a motherhood wage penalty/fatherhood wage premium, such that, on average, mothers earn less than both childless women and fathers, and fathers earn more than childless men (Blau and Kahn 2017; Budig and England 2001; Gough and Noonan 2013; Hodges and Budig 2010). Black and Hispanic men earn less, on average, than non-Hispanic White men. Within racial/ethnic groups, the extent of gender wage gaps varies: compared with the same racial/ethnic male reference group, non-Hispanic White women incur slightly higher gender wage gaps than Black and Hispanic women (Hegewisch and Williams-Baron 2018).

Work experience is one of many contributing factors of the gender wage gap, with women and mothers who reduce employment losing out on wage growth through lost work experience (England et al. 2016; Gangl and Ziefle 2009; Patterson et al. 2017). Work interruptions not only directly affect wages but also affect occupational sex segregation (Blau and Kahn 2017), within-job mobility, and career advancement (Aisenbrey et al. 2009), which in turn can affect wages.

Although existing literature has examined the contribution of labor force attachment on the aggregate gender wage gap, less is known about the differential returns to long-term employment paths by gender. This is an important topic to study, given the implicit counterfactual that arises from the lost experience explanation behind the gender wage gap: if women were more steadily employed, the gender wage gap could be reduced. We explore three theoretical predictions for how long-term employment trajectories could have differently gendered wage associations: (1) skill deterioration theories, suggesting no gender difference for within-trajectory wages; (2) theories of work commitment, suggesting that women will face greater penalties for pursuing intermittent employment than men; and (3) male breadwinner and ideal worker norm-violation theories, predicting that men will face greater penalties than women when following nonsteady employment pathways. Here we highlight the processes that yield these competing predictions.

The skill deterioration perspective is ostensibly gender-neutral, under the assumption that there are no gender differences in skill-deterioration rates (Becker 1983). These theories thus predict that there will be wage penalties for reduced labor force attachment but that these penalties will be similar for both men and women (Hypothesis 1 (H1)).

Drawing from aforementioned work and family explanations, theories on work-family decisions and gendered workplace evaluations have opposing predictions for how men and women may experience different wages even while following similar long-term career pathways. Because women are more likely than men to take long-term lapses—particularly to care for children or family—employers are likely more familiar with women and mothers having intermittent employment than they are for men. Scholars have documented that many employers hold assumptions about women’s (and mothers’) competence and commitment to work (Avellar and Smock 2003; Correll et al. 2007; Ridgeway and Correll 2004; Weisshaar 2018).

From a demand-side perspective, employers’ beliefs about gender, motherhood, and competence could be exacerbated when women follow nonsteady employment paths, contributing to larger gender wage gaps among those who follow intermittent employment trajectories. This prediction is consistent with research suggesting that high-wage or high–occupational status mothers experience the largest relative motherhood wage penalty, in part because of the sharp wage declines associated with even short employment lapses (England et al. 2016; Landivar 2017). On the supply side, women in nonsteady employment could make employment (or family) decisions that contribute to reduced wages (Bass 2015; Stone and Hernandez 2013).

This perspective of commitment perception penalties and supply-side family decisions predicts that women will face greater penalties than men for taking intermittent employment paths (Hypothesis 2 (H2)).

Yet, normative expectations for men and fathers to uphold “breadwinner” and “ideal worker” norms—that is, providing for partners and families and working continuously and at high intensity levels (Coltrane et al. 2013; Fuegen et al. 2004; Hodges and Budig 2010; Pedulla 2016; Rudman and Mescher 2013; Weisshaar 2018)—could lead to greater penalties for men who do experience nonstandard employment paths. Norm violation literature suggests that those who are most expected to uphold a norm face larger penalties for violating said norm than others (e.g., Rudman and Mescher 2013). From the demand side, employers may thus evaluate men who have intermittent work paths more harshly than women (Weisshaar 2018). Supply-side processes could also be at play: men who face pressure to provide for their families or think they will be perceived as less normatively masculine might accept lower wages in order to secure any job rather than continuing to be out of work (Brescoll and Uhlmann 2005; Vandello et al. 2013). Therefore, ideal worker and breadwinner norm literatures suggest Hypothesis 3 (H3): men may experience greater wage penalties than women for breaking from steady high employment paths, especially in paths that are more typically associated with lapses among women with caretaking responsibilities.

## Analytic Approach: Employment Trajectories

### Data

To document employment trajectories over the life course, we use data from the National Longitudinal Survey of Youth (NLSY) 1979, which followed one cohort of individuals starting in 1979. Respondents were born between 1959 and 1964, at the tail end of the Baby Boomer generation. Respondents were interviewed annually through 1994 and biennially from 1996 to 2014. We exclude subsamples dropped by NLSY (2,700 respondents), those who reported no wages between ages 45 and 50 (3,241 respondents), and those with missing data on covariates (see the online appendix, section 3). The final sample includes 6,254 respondents: 3,135 men and 3,119 women.2 We examine employment trajectories between ages 22 and 50. This timing allows for most respondents to complete education before we begin the trajectories and continues to pre–retirement age employment.

The dependent variable used to identify the most common employment trajectories is the percentage of weeks within each year that a respondent is employed, based on the weekly labor force status arrays. We use a purposefully broad definition of employment and include any type of employment in our outcome measure—full-time, part-time, private, government, and self-employment—to capture groups who are marginally employed or who transition into different types of employment.3

### Methods and Measures

To identify the employment trajectories that the cohort followed between ages 22 to 50, we use group-based trajectory models (Jones and Nagin 2013; Nagin 2005). These models identify a set of predicted employment trajectories and the probabilities that individuals from the sample are a member of each of the trajectories (Nagin 2005). Group-based trajectory models are appropriate for modeling longitudinal paths because individuals move in and out of work throughout their careers, and we combine men and women to allow trajectories to emerge from the full sample.

Group-based trajectory models can fit any number of groups and functional forms, and we use an empirical approach to select the group trajectories. The selection of the number of groups and functional forms is based on three criteria: the model fit (BIC), the averaged predicted probabilities of group membership (APP), and the size of the smallest group (minimum of 5% of the sample). We follow Nagin’s (2005) suggestions for choosing a model with the best fit and with probabilities of assignment of at least .70. After testing three to nine groups with cubic age terms, we opt for the six-group model. Each model has an APP higher than .77, and increasing the number of groups improves the BIC in all cases (see Table A1, online appendix). Therefore, the six-group model selection is based on finding the most parsimonious model that shows common employment trajectories of both men and women. Next, we test whether different functional forms—including linear and quadratic age terms—produce a better fit for the six-group model, and we find that the model with cubic terms has the best fit.

After the heterogeneity of individual-level employment trajectories is reduced to a set of group trajectories, we conduct multinomial logistic risk models to estimate how membership in a certain group trajectory varies as a function of key time-invariant predictors measured at the start of the trajectories (Nagin 2005). The explanatory variables included as predictors of the employment trajectories are gender (49.1% women in the sample); race and ethnicity (non-Black/non-Hispanic4 (79.3%), non-Hispanic Black (14.2%), and Hispanic (6.6%)); poverty status ever reported by age 22 (14.7%); interactions among gender, race/ethnicity, and poverty status; a gender attitudes index measured in 1979 (mean = 2.7, SD = 0.74), in which a higher score means more traditional attitudes; and years of education at age 22 (mean = 12.5, SD = 1.78).5

## Analytic Approach: The Effects of Trajectories on Wages

### Measures

In the second analysis, the outcome variable for which we compare the effect of following different employment trajectories is logged hourly wages at ages 45 to 50 (mean = 3.09, SD = 0.69). The annual wages/salary reported are divided by the hours worked each year and adjusted for inflation to be comparable with 2014 U.S. dollars. Because of data inconsistency across wages at low and high levels, all hourly wages below $1 and above$200 are assigned as missing.6 We construct a mean annual hourly wages variable for each respondent aged 45–50 to minimize variation in annual wages.7 In addition to the same covariates used in the risk models, we add measures in the wage models representing family formation predictions (number of children and marital status at age 50 (never married, currently married with spouse present, and other)),8 measures of work opportunities (occupation at the detailed level, using the occupation held for the longest period during ages 45–50, and whether respondents worked part-time or more than 50 hours per week (overwork) for the longest tenure job).

### Methods

To analyze the long-term effects that different trajectories have on wages, we conducted multiple ordinary least squares (OLS) regression models using the trajectory groups as explanatory variables and the logged average hourly wage between ages 45 and 50 as the outcome. We first present a model with no control variables, just the trajectories, to assess the total effect of trajectories on logged wages. We then test how the trajectory effects change as we add different sets of individual, work, and family variables. Finally, we test the gendered payoff theories by interacting gender with trajectory and estimating average hourly wages for men and women following each trajectory.

## Results: Employment Trajectories Across the Life Course

Figure 1 shows the six employment trajectory groups identified by the group-based trajectory model, evidencing substantial heterogeneity. In Table 1, we classify each group regarding the type and shape of the trajectory and describe their levels of attachment to work by age.

We identify two groups that follow a relatively steady trajectory. The largest group (42.1%) is the most steadily attached to work across the life course. We call this the steady high attachment (SH) trajectory; 52.8% of men and 31.1% of women follow this trajectory. The steady medium-high attachment group (SMH, 25.5%) is also steadily employed: in their 20s, individuals in this group are employed at about 80% of the year, with increasing employment levels throughout later ages; 24.5% of men and 26.4% of women follow the SMH path.

We find four groups with rates of employment that vary substantially across the life course. Two groups follow an increasing attachment to work pattern across the life course. The late entry group (LE, 11.4%; 7.0% of men and 15.9% of women in the sample) shows increasing employment throughout the 30s, toward high levels of employment in the 40s. The low to late entry group (LLE, 4.8%; 2.9% of men and 6.9% of women) has low employment rates during their 20s and 30s and increases in their 40s but never reaches high employment levels.

The final two groups have the greatest variation in employment. The early exit group (EE, 8.6%; 8.6% of men and 8.5% of women) follows somewhat of a bell-shaped pattern, with medium to high levels of employment until the late 30s, followed by a steady decline in employment. The medium temporary low group (MTL, 7.6%; 4.1% men and 11.2% women) has medium levels of employment in the early career, followed by a temporary decline in employment and subsequent increase in later ages.

Table 2 shows the descriptive statistics across each group, split by gender. As predicted, women are underrepresented in the steady-high trajectory but are overrepresented in the late entry, low to late entry, and medium temporary low paths. About equal proportions of men and women follow the steady medium high and early exit trajectories.

We find that respondents in the steady high (SH) trajectory are the most advantaged on multiple characteristics: this group is composed of disproportionately non-Black, non-Hispanic men who have not reported being in poverty and who have the highest average years of education. Individuals in this group have later family formations compared with other trajectories, and have stable employment pathways with fewer cumulative jobs, lapses, and months out of work. A relatively higher proportion of individuals in the SH group work in managerial/professional occupations.

The steady medium-high (SMH) trajectory group is relatively comparable to the full sample in terms of respondent characteristics and family formation. This trajectory has more lapses than the overall average, but these tend to be short lapses, resulting in more total jobs.

The late entry (LE) trajectory group is predominantly women (68.7%) and contains a relatively large proportion of Black and Hispanic men and of individuals who reported being in poverty by age 22. Women in this group tend to form a family earlier than in other trajectories.

The early exit (EE) trajectory has an overrepresentation of Black and Hispanic respondents as well as respondents reporting poverty. Men and women in this group have relatively few children, on average, and a large proportion of men have not married by ages 45–50. Both women and men in service and construction, production, transportation, and other occupations are overrepresented in the group.

The medium temporary low (MTL) trajectory group has the largest proportion of women (72.3%) and a relatively high proportion of Black men and men who reported being in poverty by age 22. Men also have relatively low average years of education. In this group, women have their first birth at an older age and have more children than most of the other groups, but men tend to have their first child at younger ages. This group has relatively more jobs and work lapses across all groups, and most individuals have worked part-time at least once between ages 45 and 50. Additionally, the MLT group has a relatively high proportion of women working in service and men working in construction, production, and transportation occupations.

Finally, the low to late entry (LLE) trajectory group has a large proportion of women (69.7%) and the highest proportion of Hispanic and Black individuals, particularly of Black men, across all groups. Women in this group have the lowest average years of education. Individuals tend to form a family early and have more children than average. They also have relatively few jobs and spend relatively long periods out of work. Individuals working in service occupations are overrepresented in this group, and women are particularly likely to have worked part-time between ages 45 and 50.

In sum, a large proportion of the sample follows a steady high or medium-high attachment to work over the life course, but one-third of the sample does not experience either of these high attachment trajectories. Although the proportion of those in nonsteady employment trajectories is quite high (42.5%) among women, we also find that for one-fifth of men, a steady work trajectory does not capture their work experience over the life course—challenging the idea of a steady and intense work trajectory as the norm.

### Predictors of the Employment Trajectory Groups

Table 3 shows the results from the multinomial logit risk models predicting group membership, displaying coefficients in log odds. The relative likelihood of following any non-SH trajectory compared with following the steady high trajectory is higher for women than for men, as indicated by the significant positive main effect of gender across all trajectories. The gender effect is particularly large for the late entry, low to late entry, and medium temporary low groups. We also find a positive and significant effect for following unstable or less attached trajectories for Black respondents (relative to White respondents) for all trajectories compared with the SH trajectory, and those who report poverty experience are at increased likelihood of following four of the five non-SH paths relative to those who have not reported poverty experience. Hispanic respondents are more likely than White respondents to belong to the early exit, relative to the SH group. Years of education consistently negatively predicts following non-SH pathways compared with SH pathways, and more traditional gender attitudes positively predict two of the nonsteady pathways: late entry and early exit.

The interaction terms between gender and race/ethnicity show negative effects, which are consistently significant for Black × women. Note that because these interaction effects must be added to the main effects of woman and Black, this does not mean that Black women are less likely to follow unsteady paths than White women. Instead, we find that Black women’s likelihood of following nonsteady paths relative to the SH trajectory is similar to that of White and Hispanic women, and Black men are more likely than White men to follow nonsteady paths relative to the SH trajectory. We do not find significant interaction effects for Hispanic × woman.

The two-way interactions with poverty and gender, and poverty and race/ethnicity, and also the three-way interaction with gender, race/ethnicity, and poverty are not statistically significant. This may be in part due to the necessary sample selection criteria for the second analysis: we exclude individuals who are not working at any level or report no wages at ages 45–50, which could be related to class background. Smaller sample sizes in the three-way interaction may also contribute to nonsignificant effects; we thus remove the interactions with poverty status in the subsequent empirical analyses.

In sum, women of each of the racial/ethnic groups and Black men are more likely than White men to step off the SH work path—evidencing how gender and racial/ethnic statuses and their intersections predict employment trajectories over the life course. Poverty status also contributes to nonsteady employment as a main effect, operating similarly for men and women of different racial/ethnic groups. These findings are in line with previous research documenting that demographic statuses (gender, race, and poverty), along with human capital (education) and gender beliefs, are associated with long-term career pathways. In the next section, we assess how following these different trajectories impacts wages later in careers.

## Results: The Long-Term Effects of Employment on Wages

From Model 1 in Table 4, we find that all trajectories have significant negative effects on wages at ages 45 to 50 compared with steady high employment. Relative to following the steady high (SH) employment trajectory, steady medium high incurs approximately a 29% wage penalty (= 1 – exp(–.341)). Late entry and early exit incur wage penalties relative to SH of approximately 39% and 45%, respectively. The largest effects occur for medium temporary low and low to late entry, which are associated with wage penalties relative to SH by about 55% and 53%, respectively. These findings suggest that employment trajectories and nonsteady employment are important predictors of wages later in life, consistent with the skill deterioration theory.

In Model 2, we add individual characteristics, including gender, race/ethnicity, poverty and education at age 22, and gender attitudes. These measures reduce the trajectory effects somewhat, although all nonsteady trajectories remain significant and negative. In this model, there is a gender wage gap of approximately 24% (= 1 – exp(–.280)). There is also a wage penalty for Black respondents, relative to White respondents, of approximately 17%. Poverty, years of education, and gender attitudes are significant predictors in the directions we would expect.

In Models 3–5, we add family and work measures to test the relative impact of each on the association between wages and trajectories. Model 3 adds the marital status and number of children by age 50. Although these measures have some independent effects, the trajectory coefficients remain largely stable from Model 2 compared with Model 3.

Model 4 removes the family variables and adds work characteristics: whether the respondent worked part-time or overworked, and their occupation at ages 45–50 included as a series of dichotomous variables. Here the trajectory effects are attenuated substantially, suggesting that much of the initial trajectory penalties can be attributed to differences in occupation and work hours. From Model 4, the largest trajectory penalty continues to be for the medium temporary low trajectory, which is predicted to produce a penalty of about 29.7% relative to SH. This is about a 30% reduction in the effect relative to the Model 2 estimates. Model 5 combines the family and work measures into the same model, and we see similar effects of trajectories across Models 4 and 5, again suggesting that work measures are doing much of the explanatory power, and family measures change estimates by only a small amount.

In Model 6, we add interaction effects with gender and race/ethnicity, and with gender and number of children. This model thus tests for whether a motherhood penalty or racial differences in gender effects are associated with further reductions in the trajectories’ explanatory power. We find a significant positive effect of woman × Black, suggesting a reduced gender wage gap among Black respondents but no significant interaction effect with woman × Hispanic. There is evidence of a motherhood penalty occurring: women who have two or three children experience lower wages than women with no children. Including these interactions in the model reduces the trajectories’ effects by a small amount. This suggests that even after we control for work trajectories, mothers experience a wage penalty.

Altogether, from Oaxaca-Blinder decomposition models based on Model 6 (presented in the online appendix, Table A3), we find that work measures explain a sizable portion of the trajectory penalties relative to SH (ranging from about 25.4% for late entry to 36.4% for steady medium-high and medium temporary low). Family measures, including the gender × children interaction, explain 1.9% (medium temporary low) to 9.4% (steady medium-high) of the trajectory penalties, and individual characteristics explain 17.3% (steady medium-high) to 34.7% (late entry). These findings provide strong support for the work opportunities/constraints theories and persisting status (dis)advantage theories, with weaker (although positive) support for family formation processes mediating the effects of trajectories on wages.

### The Gendered Wage Payoffs to Trajectories

In Table 5 and Fig. 2, we present estimates of the predicted hourly wages for men and women within each trajectory. These estimates are derived from Model 7 in Table 4, which includes the measures of work, family, and individual characteristics, along with a gender × trajectory interaction. This model also controls for the individual’s difference between their duration of nonemployment and the trajectory’s average nonemployment level, which adjusts for minor within-trajectory variation by gender in duration of nonemployment. Table 5 presents men’s and women’s average wages (displayed graphically in Fig. 2), the within-trajectory gender wage gap (women’s wages as a percentage of men’s wages), and the relative wage penalty (percentage) compared with the SH trajectory.

As Table 5 and Fig. 2 display, within-trajectory gender wage gaps are smaller in non–steady high trajectories than in the steady high (SH) path. Women earn about 78.9% of men’s wages in the SH trajectory; this gap reduces to 83.2% to 86.7% for the early exit, late entry, and steady medium-high trajectories. The medium temporary low and low to late entry paths—the paths with the least work attachment—have no significant gender wage gap. These results are confirmed in Model 7, Table 4: the woman × trajectory interaction effects are positive and in most cases statistically significant, indicating that women face lower wage penalties in the non-SH trajectories.

These findings suggest that men have larger relative penalties than women for following non–steady high trajectories, in line with the breadwinner/ideal worker violation theories (H3). To illustrate this, we display men’s wages as a percentage of men’s steady high wages and women’s wages as a percentage of women’s steady high wages in Table 5. Within each of the non-SH trajectories, men incur greater penalties relative to the same-gender SH group than do women, but note that men also have a higher earning ceiling in the SH trajectory. Men in the low to late entry group have about 60.7% of the wages that SH men are predicted to earn. In contrast, women’s lowest wage penalty is in the medium temporary low group, with a wage penalty compared with SH women of about 74.1%.

In Table A2 (Model 9) of the online appendix, we present a model with the interaction of gender, race/ethnicity, and trajectory. We find no significant three-way interaction effects, suggesting that gender wage inequality by race/ethnicity does not vary significantly across trajectories. These interactions, however, correspond with small sample sizes, which may limit detection of effects.

These findings demonstrate the complexity of how gender and employment pathways over the life course operate to produce inequalities. Women’s exclusion from SH pathways and their relative overrepresentation in the late entry, medium temporary low, and low to late entry groups means that women in these groups earn 59.5% to 67% of men’s SH wages. In this sense, women who follow the SH pathway can secure the highest wages (78.9% of men’s SH wages). Yet, when inequality within employment trajectories is considered, there is evidence of a double bind for women: the within-gender wage gaps are lowest in the medium temporary low and low to late entry groups, which are the pathways with the lowest average wages for both men and women. These findings suggest no clear path for both gender equity and economic security: gender inequality is reduced in more unsteady trajectories, but this near-equity comes at the cost of the ability to secure high wages.

## Conclusion and Discussion

In this article, we examine variation in labor force participation in the United States throughout the early to midcareers of the tail end of the Baby Boomer cohort. Using longitudinal data from the NLSY 1979 study, we document six employment trajectories for individuals aged 22–50. We find that relatively advantaged demographic and social groups have higher access to following a steady high employment path: women across racial and ethnic groups, Black men, respondents with less education, and those who reported being in poverty by age 22 are relatively more likely than their respective counterparts to follow non–steady high employment paths.

Next, we examine the effects of the six employment trajectories on wages later in the career (at ages 45–50). We demonstrate that non–steady high employment pathways have significantly reduced wages relative to steady high employment, and that work characteristics (occupation and part-time/overwork status) explain a substantial portion of these effects, whereas family formation patterns explain less. These findings build on literature documenting the relationship between intermittency and career opportunities—the consequences of intermittency on future outcomes, such as occupational prestige (Aisenbrey and Fasang 2017; Aisenbrey et al. 2009) as well as the impacts on job quality and access to steady, good jobs in subsequent employment (Kalleberg 2011; Nelson and Smith 1999). Our findings suggest that these reduced career opportunities, in addition to skill deterioration processes, help explain the association between intermittency and wages.

Finally, we test for gender variation in the relative wage payoffs or costs of following each employment trajectory. We find that although the steady high attachment path has the highest wage payoff overall, it also has the largest within-trajectory gender wage gap. The within-trajectory gender wage gap is reduced among the least attached trajectories. In other words, men experience a smaller wage premium relative to women in less attached work trajectories than they do when highly attached to employment. These findings support our third gender hypothesis (H3) corresponding to ideal worker norms: men experience lower wage premiums when they follow non–steady high attachment paths. These trajectories are commonly associated with women who experience lapses to care for family, and men who follow them may be perceived as violating the breadwinner and ideal worker norm ideals, leading to wage penalties (Rudman and Mescher 2013). In supplementary models, we did not find significant interaction effects with gender, trajectories, and race/ethnicity or poverty status, suggesting that wage inequality by race/ethnicity and poverty does not vary significantly across trajectories. This is in line with a model showing that race, ethnicity, and early-life poverty lead to selection into employment paths and affect wages in the aggregate, but there are no clear additional wage payoffs or penalties for racial/ethnic and class groups within these paths.

This research leaves some questions unanswered, which also inspire future research directions. First, the NLSY 1979 is a cohort-based study, meaning that findings could change if a different cohort were examined. For instance, this NLSY cohort experienced the stagnation of men’s wages in the United States during recent decades (Wessel 2015), and more recent cohorts will have experienced the Great Recession, which could normalize employment lapses. Further, it would be worthwhile to test whether trajectories associated with different types of lapses (e.g., family care lapses, or involuntary or voluntary spells of unemployment) produce different penalties. These measures would provide a strong test of skill deterioration models: do all reasons for employment lapses incur the same wage penalties? We find that the effects of trajectories remain even after we control for total duration of nonemployment (see online appendix, Table A2, Model 8), suggesting that a pure skill deterioration model does not account for our findings, nor does the skill deterioration model explain gender differences in wages within trajectories. This indicates that trajectory characteristics in terms of timing and intensity of nonemployment matter in addition to amount of nonemployment. Next, the sample excludes individuals who left work entirely prior to age 45 and did not report any wages, meaning that there could exist a trajectory with relatively low employment levels beyond those identified here. This unobserved trajectory may be associated with race/ethnicity or class, in addition to gender, and a low attachment trajectory may be worth studying in future research. Furthermore, we identify a trajectory of early exit from the labor force that lacks attention in previous research; understanding the reasons behind this early decline in employment is important, particularly considering the wage penalties associated with it.

Additionally, country context could play a large role in how intermittent employment and gender norms are understood. In countries with different institutional support for parents (e.g., in terms of maternity or paternity leave policies), labor market structures, gender norms, or aggregate gender wage gaps than in the United States, we may see different patterns than those found here. Specifically, we would expect that both institutional and cultural support for mothers would contribute to the gender wage gap within employment trajectories (Boeckmann et al. 2015) and that variation in labor market access would affect who follows particular paths (e.g., Aisenbrey and Fasang 2017; Lyness et al. 2012).

This research takes an important step toward assessing the predictors of wage-based inequality and gender inequality in wages. Although steady high employment is associated with higher wages than other, less steady pathways, our results imply that steady high employment is most beneficial for men. We suggest that it is not the case that if women’s employment paths became more like men’s, gender wage inequality would be eliminated. Given this finding, policies may benefit from considering how to change cultural gender expectations that constrain employment opportunities throughout the life course. For instance, the extent of occupational gender segregation and the lower rate of returns to experience in women-dominated occupations (Aisenbrey and Bruckner 2008; Blau et al. 2013; England 2005) seem to have larger effects on gender wage differences than pure human capital explanations. Policies that directly address these processes—including comparable worth wage policies for similar work across different jobs (England 1992) or setting industry wage recommendations based on experience—may be more fruitful than encouraging women to work more and harder. Workplace policies that encourage both mothers and fathers to participate equally in childcare could also reduce the commitment concerns employers may have for mothers as well as the stigmas associated with men violating ideal worker norms (Gangl and Ziefle 2015; Rivera and Tilcsik 2016). In the end, tackling these processes through multiple sources—policy, workplaces, and culture—could be the most effective way of interrupting these persistent gendered inequalities.

## Acknowledgments

We are grateful to Michael Rosenfeld, Koji Chavez, Ariela Schachter, Ted Mouw, and Demography reviewers and editors for comments on earlier versions of the manuscript. The NLSY79 survey is sponsored and directed by the U.S. Bureau of Labor Statistics and conducted by the Center for Human Resource Research at The Ohio State University. Interviews are conducted by the National Opinion Research Center at the University of Chicago. Both authors contributed equally to this work.

## Notes

1

“Opt out” is in quotation marks because scholarship suggests that leaving work for family is not always voluntary (Stone 2007).

2

We apply NLSY custom weights, which adjust for attrition in the longitudinal sample (see https://www.nlsinfo.org/weights/nlsy79).

3

Some studies on long-term employment have used a measure of hours worked rather than employment (e.g., Damaske and Frech 2016). We deviate from this measure to study labor force attachment variation rather than intensity of employment. Although both measures lend unique benefits, the measure that we use allows us to focus on variation in employment versus nonemployment rather than on levels of work.

4

NLSY uses “Black/Non-Hispanic,” “Hispanic,” and “Non-Black, Non-Hispanic.” We sometimes refer to this last group as “White,” but we recognize that it could include additional racial/ethnic groups as well. See https://www.nlsinfo.org/content/cohorts/nlsy79/topical-guide/household/race-ethnicity-immigration-data.

5

These are weighted averages.

6

Hours worked per week are top-coded at 100 per week (5,200 per year), with 7,000+ coded as missing. Our results are robust to these wage and hours coding decisions; models are available upon request.

7

Results are robust if the dependent variable measures the wages at the oldest working age from ages 45 to 50.

8

See section 3 of the online appendix for information on missing data.

## References

Abendroth, A-K, Huffman, M. L., & Treas, J. (
2014
).
The parity penalty in life course perspective: Motherhood and occupational status in 13 European countries
.
American Sociological Review
,
79
,
993
1014
. 10.1177/0003122414545986
Aisenbrey, S., & Bruckner, H. (
2008
).
Occupational aspirations and the gender gap in wages
.
European Sociological Review
,
24
,
633
649
. 10.1093/esr/jcn024
Aisenbrey, S., Evertsson, M., & Grunow, D. (
2009
).
Is there a career penalty for mothers’ time out? A comparison of Germany, Sweden and the United States
.
Social Forces
,
88
,
573
605
. 10.1353/sof.0.0252
Aisenbrey, S., & Fasang, A. (
2017
).
The interplay of work and family trajectories over the life course: Germany and the United States in comparison
.
American Journal of Sociology
,
122
,
1448
1484
. 10.1086/691128
Albrecht, J. W., Edin, P-A, Sundstrom, M., & Vroman, S. B. (
1999
).
Career interruptions and subsequent earnings: A reexamination using Swedish data
.
Journal of Human Resources
,
34
,
294
311
. 10.2307/146347
Alon, S., & Haberfeld, Y. (
2007
).
Labor force attachment and the evolving wage gap between White, Black, and Hispanic young women
.
Work and Occupations
,
34
,
369
398
. 10.1177/0730888407307247
Arulampalam, W. (
2001
).
Is unemployment really scarring? Effects of unemployment experiences on wages
.
Economic Journal
,
111
,
585
606
. 10.1111/1468-0297.00664
Avellar, S., & Smock, P. J. (
2003
).
Has the price of motherhood declined over time? A cross-cohort comparison of the motherhood wage penalty
.
Journal of Marriage and Family
,
65
,
597
607
. 10.1111/j.1741-3737.2003.00597.x
Bass, B. C. (
2015
).
Preparing for parenthood?: Gender, aspirations, and the reproduction of labor market inequality
.
Gender & Society
,
29
,
362
385
. 10.1177/0891243214546936
Becker, G. S. (
1962
).
Investment in human capital: A theoretical analysis
.
Journal of Political Economy
,
70
(
5, Part 2
),
9
49
. 10.1086/258724
Becker, G. S. (
1983
).
Human capital: A theoretical and empirical analysis, with special reference to education
(2nd ed.).
Chicago, IL
:
The University of Chicago Press
.
Bielby, W. T., & Bielby, D. D. (
1992
).
Cumulative versus continuous disadvantage in an unstructured labor market: Gender differences in the careers of television writers
.
Work and Occupations
,
19
,
366
386
. 10.1177/0730888492019004003
Blau, F. D., Brummund, P., & Liu, AY-H (
2013
).
Trends in occupational segregation by gender 1970–2009: Adjusting for the impact of changes in the occupational coding system
.
Demography
,
50
,
471
492
. 10.1007/s13524-012-0151-7
Blau, F. D., & Kahn, L. M. (
2017
).
The gender wage gap: Extent, trends, and explanations
.
Journal of Economic Literature
,
55
,
789
865
. 10.1257/jel.20160995
Boeckmann, I., Misra, J., & Budig, M. J. (
2015
).
Cultural and institutional factors shaping mothers’ employment and working hours in postindustrial countries
.
Social Forces
,
93
,
1301
1333
. 10.1093/sf/sou119
Brescoll, V. L., & Uhlmann, E. L. (
2005
).
.
Psychology of Women Quarterly
,
29
,
436
445
. 10.1111/j.1471-6402.2005.00244.x
Budig, M. J., & England, P. (
2001
).
The wage penalty for motherhood
.
American Sociological Review
,
66
,
204
225
. 10.2307/2657415
Cha, Y. (
2010
).
Reinforcing separate spheres: The effect of spousal overwork on men’s and women’s employment in dual-earner households
.
American Sociological Review
,
75
,
303
329
. 10.1177/0003122410365307
Cha, Y. (
2013
).
Overwork and the persistence of gender segregation in occupations
.
Gender & Society
,
27
,
158
184
. 10.1177/0891243212470510
Cha, Y., & Weeden, K. A. (
2014
).
Overwork and the slow convergence in the gender gap in wages
.
American Sociological Review
,
79
,
457
484
. 10.1177/0003122414528936
Coltrane, S., Miller, E. C., DeHaan, T., & Stewart, L. (
2013
).
Fathers and the flexibility stigma
.
Journal of Social Issues
,
69
,
279
302
. 10.1111/josi.12015
Correll, S. J., Benard, S., & Paik, I. (
2007
).
Getting a job: Is there a motherhood penalty?
.
American Journal of Sociology
,
112
,
1297
1339
. 10.1086/511799
Damaske, S., & Frech, A. (
2016
).
Women’s work pathways across the life course
.
Demography
,
53
,
365
391
. 10.1007/s13524-016-0464-z
Elder, G. H., Johnson, M. K., & Crosnoe, R. (
2003
).
The emergence and development of life course theory
. In Mortimer, J. T., & Shanahan, M. J. (Eds.),
Handbook of the life course
(pp.
3
19
).
Boston, MA
:
Springer US
.
England, P. (
1992
).
Comparable worth: Theories and evidence
.
Piscataway, NJ
:
Transaction Publishers
.
England, P. (
2005
).
Gender inequality in labor markets: The role of motherhood and segregation
.
Social Politics: International Studies in Gender, State & Society
,
12
,
264
288
. 10.1093/sp/jxi014
England, P., Bearak, J., Budig, M. J., & Hodges, M. J. (
2016
).
Do highly paid, highly skilled women experience the largest motherhood penalty?
.
American Sociological Review
,
81
,
1161
1189
. 10.1177/0003122416673598
England, P., Garcia-Beaulieu, C., & Ross, M. (
2004
).
Women’s employment among Blacks, Whites, and three groups of Latinas: Do more privileged women have higher employment?
Gender & Society
,
18
,
494
509
.
Eriksson, S., & Rooth, D-O (
2014
).
Do employers use unemployment as a sorting criterion when hiring? Evidence from a field experiment
.
American Economic Review
,
104
,
1014
1039
. 10.1257/aer.104.3.1014
Evertsson, M. (
2016
).
Parental leave and careers: Women’s and men’s wages after parental leave in Sweden
.
,
29
,
26
40
. 10.1016/j.alcr.2016.02.002
Fernandez-Mateo, I. (
2009
).
Cumulative gender disadvantage in contract employment
.
American Journal of Sociology
,
114
,
871
923
. 10.1086/595941
Flippen, C., & Tienda, M. (
2000
).
Pathways to retirement: Patterns of labor force participation and labor market exit among the pre-retirement population by race, Hispanic origin, and sex
.
Journals of Gerontology, Series B: Psychological Sciences and Social Sciences
,
55
,
S14
S27
. 10.1093/geronb/55.1.S14
Frech, A., & Damaske, S. (
2019
).
Men’s income trajectories and physical and mental health at midlife
.
American Journal of Sociology
,
124
,
1372
1412
. 10.1086/702775
FRED
. (
2017
,
May
11
).
Good news for women in the labor force?
Frytak, J. R., Harley, C. R., & Finch, M. D. (
2003
).
Socioeconomic status and health over the life course
. In Mortimer, J. T., & Shanahan, M. J. (Eds.),
Handbook of the life course
(pp.
623
643
).
Boston, MA
:
Springer US
.
Fuegen, K., Biernat, M., Haines, E., & Deaux, K. (
2004
).
Mothers and fathers in the workplace: How gender and parental status influence judgments of job-related competence
.
Journal of Social Issues
,
60
,
737
754
. 10.1111/j.0022-4537.2004.00383.x
Gangl, M. (
2006
).
Scar effects of unemployment: An assessment of institutional complementarities
.
American Sociological Review
,
71
,
986
1013
. 10.1177/000312240607100606
Gangl, M., & Ziefle, A. (
2009
).
Motherhood, labor force behavior, and women’s careers: An empirical assessment of the wage penalty for motherhood in Britain, Germany, and the United States
.
Demography
,
46
,
341
369
. 10.1353/dem.0.0056
Gangl, M., & Ziefle, A. (
2015
).
The making of a good woman: Extended parental leave entitlements and mothers’ work commitment in Germany
.
American Journal of Sociology
,
121
,
511
563
. 10.1086/682419
García-Manglano, J. (
2015
).
Opting out and leaning in: The life course employment profiles of early baby boom women in the United States
.
Demography
,
52
,
1961
1993
. 10.1007/s13524-015-0438-6
Gerstel, N., & Clawson, D. (
2014
).
Class advantage and the gender divide: Flexibility on the job and at home
.
American Journal of Sociology
,
120
,
395
431
. 10.1086/678270
Gerstel, N., & Clawson, D. (
2018
).
Control over time: Employers, workers, and families shaping work schedules
.
Annual Review of Sociology
,
44
,
77
97
. 10.1146/annurev-soc-073117-041400
Glauber, R. (
2008
).
Race and gender in families and at work: The fatherhood wage premium
.
Gender & Society
,
22
,
8
30
. 10.1177/0891243207311593
Glauber, R., & Gozjolko, K. L. (
2011
).
Do traditional fathers always work more? Gender ideology, race, and parenthood
.
Journal of Marriage and Family
,
73
,
1133
1148
. 10.1111/j.1741-3737.2011.00870.x
Gough, M., & Noonan, M. (
2013
).
A review of the motherhood wage penalty in the United States
.
Sociology Compass
,
7
,
328
342
. 10.1111/soc4.12031
Halpern-Manners, A., Warren, J. R., Raymo, J. M., & Nicholson, D. A. (
2015
).
The impact of work and family life histories on economic well-being at older ages
.
Social Forces
,
93
,
1369
1396
. 10.1093/sf/sov005
Han, S-K, & Moen, P. (
1999
).
Work and family over time: A life course approach
.
Annals of the American Academy of Political and Social Science
,
562
,
98
110
. 10.1177/000271629956200107
Hegewisch, A., & Williams-Baron, W. (
2018
).
The gender wage gap: 2017 earnings differences by race and ethnicity
(IWPR Fact Sheet Report #C464).
Washington, DC
:
Institute for Women’s Policy Research
Hodges, M. J., & Budig, M. J. (
2010
).
Who gets the daddy bonus? Organizational hegemonic masculinity and the impact of fatherhood on earnings
.
Gender & Society
,
24
,
717
745
. 10.1177/0891243210386729
Hotchkiss, J. L., & Pitts, M. M. (
2007
).
The role of labor market intermittency in explaining gender wage differentials
.
American Economic Review: Papers & Proceedings
,
97
,
417
421
. 10.1257/aer.97.2.417
Hynes, K., & Clarkberg, M. (
2005
).
Women’s employment patterns during early parenthood: A group-based trajectory analysis
.
Journal of Marriage and Family
,
67
,
222
239
. 10.1111/j.0022-2445.2005.00017.x
Jacobs, J. A., & Gerson, K. (
2001
).
Overworked individuals or overworked families? Explaining trends in work, leisure, and family time
.
Work and Occupations
,
28
,
40
63
. 10.1177/0730888401028001004
Jacobsen, J. P., & Levin, L. M. (
1995
).
Effects of intermittent labor force attachment on women’s earnings
.
Monthly Labor Review
,
118
(
9
),
14
19
.
Jones, B. L., & Nagin, D. S. (
2013
).
A note on a Stata plugin for estimating group-based trajectory models
.
Sociological Methods & Research
,
42
,
608
613
. 10.1177/0049124113503141
Kalleberg, A. L. (
2000
).
Nonstandard employment relations: Part-time, temporary and contract work
.
Annual Review of Sociology
,
26
,
341
365
. 10.1146/annurev.soc.26.1.341
Kalleberg, A. L. (
2011
).
Good jobs, bad jobs: The rise of polarized and precarious employment systems in the United States, 1970s to 2000s
.
New York, NY
:
Russell Sage Foundation
.
Killewald, A., & Zhuo, X. (
2019
).
U.S. mothers’ long-term employment patterns
.
Demography
,
56
,
285
320
. 10.1007/s13524-018-0745-9
Landivar, L. C. (
2014
).
Opting out, scaling back, or business-as-usual? An occupational assessment of women’s employment
.
Sociological Forum
,
29
,
189
214
. 10.1111/socf.12075
Landivar, L. C. (
2017
).
Mothers at work: Who opts out?
.
Boulder, CO
:
Lynne Rienner Publishers
.
Lu, Y., Wang, JS-H, & Han, W-J (
2017
).
Women’s short-term employment trajectories following birth: Patterns, determinants, and variations by race/ethnicity and nativity
.
Demography
,
54
,
93
118
. 10.1007/s13524-016-0541-3
Lundberg, S., & Rose, E. (
2000
).
Parenthood and the earnings of married men and women
.
Labour Economics
,
7
,
689
710
. 10.1016/S0927-5371(00)00020-8
Lyness, K. S., Gornick, J. C., Stone, P., & Grotto, A. R. (
2012
).
It’s all about control: Worker control over schedule and hours in cross-national context
.
American Sociological Review
,
77
,
1023
1049
. 10.1177/0003122412465331
Macmillan, R., & Copher, R. (
2005
).
Families in the life course: Interdependency of roles, role configurations, and pathways
.
Journal of Marriage and Family
,
67
,
858
879
. 10.1111/j.1741-3737.2005.00180.x
Nagin, D. S. (
2005
).
Group-based modeling of development
.
Cambridge, MA
:
Harvard University Press
.
Nelson, M. K., & Smith, J. (
1999
).
Working hard and making do: Surviving in small town America
.
Berkeley
:
University of California Press
.
O’Rand, A. M. (
1996
).
The precious and the precocious: Understanding cumulative disadvantage and cumulative advantage over the life course
.
Gerontologist
,
36
,
230
238
. 10.1093/geront/36.2.230
Patterson, S. E., Damaske, S., & Sheroff, C. (
2017
).
Gender and the MBA: Differences in career trajectories, institutional support, and outcomes
.
Gender & Society
,
31
,
310
332
. 10.1177/0891243217703630
Pedulla, D. S. (
2016
).
Penalized or protected? Gender and the consequences of nonstandard and mismatched employment histories
.
American Sociological Review
,
81
,
262
289
. 10.1177/0003122416630982
Percheski, C. (
2018
).
Marriage, family structure, and maternal employment trajectories
.
Social Forces
,
96
,
1211
1242
. 10.1093/sf/sox094
Ridgeway, C. L., & Correll, S. J. (
2004
).
Motherhood as a status characteristic
.
Journal of Social Issues
,
60
,
683
700
. 10.1111/j.0022-4537.2004.00380.x
Rivera, L. A., & Tilcsik, A. (
2016
).
Class advantage, commitment penalty: The gendered effect of social class signals in an elite labor market
.
American Sociological Review
,
81
,
1097
1131
. 10.1177/0003122416668154
Roehling, P. V., Roehling, M. V., & Moen, P. (
2001
).
The relationship between work-life policies and practices and employee loyalty: A life course perspective
.
Journal of Family and Economic Issues
,
22
,
141
170
. 10.1023/A:1016630229628
Rudman, L. A., & Mescher, K. (
2013
).
Penalizing men who request a family leave: Is flexibility stigma a femininity stigma?
.
Journal of Social Issues
,
69
,
322
340
. 10.1111/josi.12017
Spivey, C. (
2005
).
Time off at what price? The effects of career interruptions on earnings
.
Industrial and Labor Relations Review
,
59
,
119
140
. 10.1177/001979390505900107
Stone, P. (
2007
).
Opting out? Why women really quit careers and head home
.
Berkeley
:
University of California Press
.
Stone, P., & Hernandez, L. A. (
2013
).
The all-or-nothing workplace: Flexibility stigma and “opting out” among professional-managerial women
.
Journal of Social Issues
,
69
,
235
256
. 10.1111/josi.12013
Stone, P., & Lovejoy, M. (
2019
).
Opting back in: What really happens when mothers go back to work
.
Oakland
:
University of California Press
.
Theunissen, G., Verbruggen, M., Forrier, A., & Sels, L. (
2011
).
Career sidestep, wage setback? The impact of different types of employment interruptions on wages
.
Gender, Work and Organization
,
18
(
Suppl. 1
),
e110
e131
. 10.1111/j.1468-0432.2009.00471.x
Vandello, J. A., Hettinger, V. E., Bosson, J. K., & Siddiqi, J. (
2013
).
When equal isn’t really equal: The masculine dilemma of seeking work flexibility
.
Journal of Social Issues
,
69
,
303
321
. 10.1111/josi.12016
Webber, G., & Williams, C. (
2008
).
Mothers in “good” and “bad” part-time jobs: Different problems, same results
.
Gender & Society
,
22
,
752
777
. 10.1177/0891243208325698
Weeden, K. A., Cha, Y., & Bucca, M. (
2016
).
Long work hours, part-time work, and trends in the gender gap in pay, the motherhood wage penalty, and the fatherhood wage premium
.
Russell Sage Foundation Journal of the Social Sciences
,
2
(
4
),
71
102
. 10.7758/rsf.2016.2.4.03
Weisshaar, K. (
2018
).
From opt out to blocked out: The challenges for labor market re-entry after family-related employment lapses
.
American Sociological Review
,
83
,
34
60
. 10.1177/0003122417752355
Wessel, D. (
2015
,
September
18
).
The typical male U.S. worker earned less in 2014 than in 1973
Williams, J. C., & Boushey, H. (
2010
).
The three faces of work-family conflict
(Work-Life-Law report).
Washington, DC
:
Center for American Progress, UC Hastings College of the Law
.

## Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.