Trends in Mothers’ Work Schedules in the United States, 1988–2019: Differences by Education, Family Structure, and Race and Ethnicity Open Access

Abstract

The latter half of the twentieth century saw dramatic increases in mothers’ labor force participation rates, accompanied by declining job quality and increasing labor market inequality. Despite evidence of growing labor market inequality in wages and benefits, less is known about how job quality changed with respect to work schedules. This study tests the hypothesis that mothers’ employment in jobs with nonstandard schedules increased between 1988 and 2019 and that such schedules are increasingly concentrated among mothers with lower education levels, single mothers, and mothers of color, who are overrepresented in low-wage jobs. We find that mothers’ employment in jobs with nonstandard schedules stayed relatively flat at 15% to 16%, and the prevalence of weekend work increased from 15% to 18%. Moreover, we find growing disparities in who works nonstandard schedules. The propensity to work such schedules increased among mothers with less than a college degree, single mothers living without other adults, and Black mothers relative to mothers with a college degree, married mothers, and White mothers, respectively. Additionally, mothers are more likely to work nonstandard schedules for involuntary reasons than before the Great Recession.

Introduction

During the latter half of the twentieth century, mothers’ labor force participation increased dramatically, while job quality declined and inequality in wages and benefits increased (Groshen and Holzer 2019; Kalleberg 2009, 2011). The growth of the service sector and the rise of the 24/7 economy presumes an increase in jobs with working hours outside the traditional Monday to Friday, 9 a.m. to 5 p.m., workweek (Presser 2003). The proliferation of “just-in-time” scheduling practices has made work schedules more unstable (i.e., work hours changing from week to week) and unpredictable (i.e., workers having little control or advance notice over when they work), particularly in low-wage jobs (Golden 2015; Lambert et al. 2014). Thus, as mothers’ labor force participation reached record-high levels in the late 1990s (Blau and Winkler 2017), the availability of high-quality jobs with standard daytime schedules was eroding—particularly for groups of mothers overrepresented in low-wage jobs, including mothers without a college degree, mothers of color, and single mothers (Ross and Bateman 2019; U.S. Bureau of Labor Statistics 2021). These trends suggest that mothers today are more likely to be working nonstandard schedules and that nonstandard schedules are increasingly concentrated among mothers with fewer social and economic resources.

Understanding whether mothers’ employment in jobs with nonstandard schedules has increased is important because of its negative implications for child and family well-being. Despite increases in recent decades in the amount of time fathers spend with their children, mothers are most often their child's primary caregiver (Sayer 2016), and mothers’ work hours and schedules shape how children and families spend their time (Han and Fox 2011; Han et al. 2010; Pilarz and Awkward-Rich 2024; Raley et al. 2012). Various types of maternal nonstandard work schedules—including working during evenings and nights and unstable and unpredictable hours—are associated with poorer maternal psychological well-being (Joshi and Bogen 2007; Strazdins et al. 2006), less time spent engaged in developmentally supportive activities (Han and Fox 2011; Rapoport and Le Bourdais 2008; Wight et al. 2008), and adverse child cognitive and behavioral outcomes (Dunifon et al. 2013; Han 2005, 2008; Han et al. 2010; Schneider and Harknett 2022; Walther and Pilarz 2024). The adverse effects of maternal nonstandard work schedules are evident across a range of child age groups (Li et al. 2014), including lower cognitive skills among preschool-age children (Han 2005) and more risky behaviors among adolescents (Han et al. 2010). However, these associations are more pronounced among children whose mothers have low education levels and children living in low-income and single-mother families (Han 2005, 2008; Han et al. 2010; Han and Waldfogel 2007; Joshi and Bogen 2007; Wang 2023).

Given that low-income and less educated workers, workers of color, and single parents are more likely to work nonstandard schedules than their counterparts (Enchautegui 2013; Golden 2015; Hepburn 2020; Lambert et al. 2014; Lambert et al. 2019; Presser and Ward 2011), mothers’ nonstandard work schedules could be contributing to growing inequality by family socioeconomic status (SES) or “diverging destinies” (McLanahan 2004; McLanahan and Jacobsen 2015). College-educated mothers are increasingly likely to have children within the context of marriage, greater family stability, and more social and economic resources than less educated mothers. Children in higher SES families receive greater investments of parents’ time and money (Kalil et al. 2016; Western et al. 2016) and experience higher quality nonparental care (Bassok et al. 2016; Flood et al. 2022; Ruzek et al. 2014). Children's school achievement is higher in families with higher SES, and these gaps have grown in recent decades (Reardon 2011). Moreover, there are persistent gaps in children's achievement by race and ethnicity (Conwell 2021; Reardon and Portilla 2016). Therefore, if nonstandard schedules are increasingly concentrated among socioeconomically disadvantaged and historically marginalized groups, these disparities have implications for growing inequality between families.

Although labor market trends suggest that the proportion of mothers working nonstandard schedules has increased over time, limited empirical evidence exists. In this study, we aim to describe trends in mothers’ work schedules over 31 years, focusing on differences by mothers’ education, family structure, and race and ethnicity. To shed light on how the reasons for mothers’ work schedules have changed over time, we also examine mothers’ self-reported reasons for working nonstandard schedules.

Background

Prevalence and Trends in Nonstandard Work Schedules

Nonstandard schedules refer to work hours that fall outside a traditional workweek (9 a.m. to 5 p.m. on weekdays). Although studies vary in their definitions of nonstandard schedules, they generally include working most work hours in the evening or at night, working weekends, and working schedules with nonregular timing, such as rotating, split, or irregular shifts that vary from week to week (Presser 2003; Wight et al. 2008). Irregular shifts include worker-determined and employer-determined variable schedules. The latter are often referred to as unstable or unpredictable schedules because workers typically have limited advance notice of when they will work (Lambert et al. 2014; Schneider and Harknett 2019). Approximately 20% to 28% of workers are employed in nonstandard schedules at any given time (Enchautegui 2013; Presser 2003), and more than 87% of workers are estimated to have worked a nonstandard schedule by age 39 (Presser and Ward 2011). Workers employed in service sector jobs, low-wage and less educated workers, younger workers, and workers of color are more likely to work nonstandard schedules than their counterparts (Enchautegui 2013; Golden 2015; Lambert et al. 2019; Presser 2003; Presser and Ward 2011). Most nonstandard schedule workers report involuntarily working their schedule because it is the nature or requirement of the job or because they could not get another job (Presser 2003).

Little is known about how the prevalence of nonstandard schedules in the United States has changed over time because prior studies have provided only cross-sectional point-in-time estimates. However, changes in the labor market in recent decades would suggest that the prevalence of nonstandard schedules has increased (Presser 2003). Since the late 1970s, employers have shifted more economic risk to workers (Hacker 2006; Kalleberg 2011). As a result, employees experience more instability and insecurity and less control over their schedules, especially in low-wage jobs (Henly et al. 2021; Howell and Kalleberg 2019). The proliferation of just-in-time scheduling practices allows employers to closely link consumer demand to workers’ schedules, resulting in irregular schedules that change week to week with little advance notice (Lambert 2008, 2014). Technological advances and the rise of online platforms suggest an increase in gig work or alternative work arrangements, which are also likely to require nonstandard and unpredictable hours subject to customer demand. Moreover, growth in the service sector—which has the highest concentration of jobs with nonstandard schedules—has outpaced growth in other sectors since the 1980s (Kalleberg 2009; Presser 2003).

Growing labor market inequality suggests that increases in nonstandard schedules are likely to be concentrated among workers with lower education levels and in low-wage jobs. Inequality in wages and benefits has risen dramatically since the late 1970s between workers with and without a college degree, suggesting that inequality in other aspects of job quality has increased as well (Groshen and Holzer 2019; Hamermesh 2003; Howell and Kalleberg 2019). Indeed, workers in lower paying jobs are more likely to work during nonstandard times and have little advance notice of their work schedules and less control over them relative to higher paid workers (Enchautegui 2013; Gerstel and Clawson 2018; Lambert et al. 2014). Furthermore, the Great Recession disproportionately impacted less educated workers. Unemployment and involuntary part-time work increased by a larger margin among less educated workers (Glauber 2017; Mattingly et al. 2011). Following the recession, employment rates did not recover for those without a college degree (Groshen and Holzer 2021). Just-in-time scheduling practices and irregular work schedules, which are concentrated in low-paying service occupations, also likely increased during the Great Recession (Finnigan 2018; Henly et al. 2021; Lambert et al. 2019). Together, this evidence suggests that nonstandard work schedules are increasingly concentrated among less educated workers and those overrepresented in low-wage jobs, including workers of color and single mothers (Ross and Bateman 2019; U.S. Bureau of Labor Statistics 2021).

Mothers’ Work Schedules and Implications for Family Inequality

Among families with children, both labor market and childcare contexts shape parents’—especially mothers’—propensity for working nonstandard schedules. Mothers are more likely to work nonstandard schedules than fathers, and mothers with young children not yet in school are more likely than those with older children to work nonstandard schedules (Enchautegui et al. 2015; Presser 2003). In some families, mothers might select nonstandard schedules to engage in tag-team parenting, arranging their work schedules to be nonoverlapping with their partners’ or close relatives’ schedules, because they prefer parental (or relative) care or cannot afford or access nonparental care (Boushey 2006; Hattery 2001; Presser 1988, 2003). Although tag-team parenting is especially likely among mothers with young children, mothers with school-age children might also select nonstandard schedules to be available during after-school hours (Pilarz and Awkward-Rich 2024). Yet, only a third or fewer mothers who work nonstandard schedules do so to secure better childcare arrangements, suggesting that this strategy is relatively uncommon (Presser 2003).

Changes in the childcare market and public policies since the 1990s have had important implications for mothers’ work schedules. Public funding for early care and education (ECE) for children aged 0–12 and the supply of center-based ECE have increased substantially since the 1990s, resulting in children's increased participation in center-based ECE programs (Bassok et al. 2011; Greenberg 2010; Magnuson et al. 2007; Phillips et al. 2018; Pilarz 2018). If some mothers shift their work hours to match center-based programs’ standard daytime schedules, this shift could contribute to a decline in maternal nonstandard schedules (Tekin 2007). Some mothers, particularly higher educated mothers in higher paying jobs (Gerstel and Clawson 2018), might have the flexibility to change their job or work schedule to use center-based ECE. However, for mothers with limited employment prospects and in low-wage jobs, it is more likely that nonstandard, unstable, and unpredictable schedules act as barriers to their use of center-based ECE and necessitate reliance on informal or multiple childcare arrangements (Carrillo et al. 2017; Harknett et al. 2022; Henly and Lambert 2005; Hepburn 2018; Pilarz et al. 2019). Among families with fewer social and economic resources, greater availability and affordability of center-based ECE is more likely to increase children's participation in center-based ECE (in combination with informal care) than to decrease mothers’ employment in nonstandard schedules.

Because of deteriorating work conditions in low-wage jobs, we expect that mothers who are more likely to work these jobs—single mothers, less educated mothers, and mothers of color—will be increasingly likely to work nonstandard schedules, especially for involuntary reasons. The 1996 welfare reform ended low-income and single mothers’ entitlement to cash assistance. This shift toward a work-based safety net suggests that low-income and single mothers cannot as easily opt out of the labor market or jobs with nonstandard schedules (Fox et al. 2012). Indeed, single mothers are more likely to report involuntarily working nonstandard schedules than married mothers (Enchautegui et al. 2015; Kim 2022). We expect that these changes will contribute to growing gaps in the prevalence of nonstandard work schedules between mothers with a college degree and those with less than a college degree, between single and married mothers, and between White mothers and mothers of color.

Limited empirical evidence exists on the prevalence of or trends in nonstandard work schedules among mothers. Estimates vary considerably across studies owing to differences in methods. Some studies have used time-diary data and defined nonstandard schedules as working most hours during nonstandard times on the diary day. This definition might exclude workers who regularly work nonstandard hours, especially those who work irregular shifts, but did not do so on the diary day. Other studies have used survey items asking about typical work schedules and working during nonstandard times (i.e., evenings, nights) and nonregular shifts (i.e., rotating, split, irregular). These survey items are subject to measurement error owing to respondents’ varying definitions of, for example, evening versus night shifts. Studies also differ on whether they included all mothers in their sample or restricted the sample to working mothers or low-income mothers. For example, using time-diary data from 2003–2004, Wight et al. (2008) found that 12% of mothers with children younger than 18 work a majority of their hours outside of 8 a.m. to 4 p.m. on weekdays. Using survey items on typical work schedules, Enchautegui et al. (2015) found that 27% of low-income, working mothers with children younger than 12 reported primarily working a nonstandard schedule (i.e., evenings, nights, irregular shifts) in 2010. It is unclear whether these estimates differ as a result of differences in methods or changes in the prevalence of work schedules over time.

A more recent study of mothers using data from two cross-sectional surveys found that mothers’ employment in jobs with variable and weekend schedules increased between 1990 and 2012 (Hepburn 2020). Among single mothers, employment in jobs with evening schedules increased as well. The study also found that having a college degree had a stronger protective effect from working evening, variable, and weekend schedules in 2012 than in 1990. However, it found mixed evidence with respect to mothers’ race and ethnicity, comparing non-Hispanic White mothers with all other mothers. Because the study used data from only two time points, it is not possible to draw conclusions about trends in mothers’ schedules or how the prevalence of nonstandard schedules changed during the latter years of the economic recovery from the Great Recession.

The Current Study

In this study, we aim to describe trends in mothers’ work schedules from 1988 to 2019, overall and by mothers’ education, family structure, and race and ethnicity. Given the changes in the labor market we described, including growth in the service sector and just-in-time scheduling practices, we expect that mothers’ employment in jobs with nonstandard schedules increased over this period, especially for irregular schedules. Given the growing inequality in the labor market and policy shifts toward a work-based safety net, we expect that increases in mothers’ employment in jobs with nonstandard schedules were concentrated among mothers without a college degree, unmarried mothers, and mothers of color. Thus, we expect to find growing gaps in rates of nonstandard schedules between mothers with higher versus lower education levels, between married and single mothers, and between White mothers and mothers of color. To shed light on mothers’ reasons for working nonstandard schedules, we examine trends in mothers’ self-reported reasons, both voluntary and involuntary, for working their schedules. We expect that more mothers involuntarily worked a nonstandard schedule over time. This study makes several contributions to our understanding of the changing prevalence of nonstandard work schedules among mothers, with implications for family inequality.

Methods

Data and Sample

This study used data from the Survey of Income and Program Participation (SIPP), a household-based survey that uses a continuous series of national panels. Each panel includes a nationally representative sample of households (sample size ranges from approximately 12,000 to 52,000 households) interviewed multiple times. In the 1984–2008 panels, households were interviewed approximately three times per year. The SIPP collected information about work schedules for up to two current jobs for all persons aged 15 years or older in the household who were employed sometime during the previous month. The work schedule questions were included with a rotating topical module, which was administered during one or multiple waves of a panel between 1988 and 2011. Beginning in 2014, the SIPP was redesigned to interview households once per year, and topical modules were eliminated. The work schedule questions were collected for up to seven jobs worked in the previous year.

Although the administration of the work schedule questions changed when the SIPP was redesigned in 2014, the same items about work hours and schedules were collected. We used data from the 1987–2020 SIPP panels, which cover 1988–2019.¹ To make the estimates as comparable as possible across SIPP panels, we used the work hours and schedules information provided for the job that the respondent worked for the greatest number of hours in the previous month; we refer to it as the primary job.² We restricted the analytic sample to women aged 18 or older, who were a designated parent of at least one child aged 0–17 years, and who had nonmissing data on variables included in our analytic models (N = 154,036). Our sample size ranged from 2,714 to 13,065 in each year of data available. We pooled all waves of data, treating each wave as a cross-sectional sample, and used SIPP weights to produce nationally representative estimates.

The composition of our sample changed over the observation period, consistent with demographic trends (Smock and Schwartz 2020). As shown in Table 1, between 1988 and 2019, the sample became more racially diverse, the percentage of mothers with a college degree increased, the percentage of married mothers decreased, and mothers’ age increased.³ Among employed mothers, the percentage employed in managerial and professional occupations and service occupations increased between 1988 and 2019. Meanwhile, the percentage employed in technical, sales, and administrative support occupations or as operators, fabricators, and laborers declined. A very small percentage of mothers were employed in the other occupations (farming, forestry, and fishing sector; or precision production, craft, and repair sector).⁴

Measures

Work Schedules

Work schedule information in the SIPP is captured in different ways. First, respondents were asked which of the following best describes their work schedule at their job during a typical workweek: regular daytime schedule, regular evening shift, regular night shift, rotating shift, split shift, or irregular schedule (changes day to day). We refer to this measure as the typical work schedule item. Second, respondents were asked to report the time of day they started and ended work on most days during a typical week, which provides information about the start and end times of work for a typical day in a typical week (rather than a weekly calendar with start and end times for each workday). We refer to these measures as the start and end times of work items. Third, respondents were asked to report the specific days of the week they worked during a typical week (including Saturdays and Sundays), which we refer to as the days of the week worked items.

We used this information to create a measure of work schedules in the respondent's primary job with the following categories: fixed day, fixed evening, fixed night, rotating/split/other non-daytime schedule, and irregular schedule. Because respondents might differ in their definitions of evening and night schedules, we used only the start and end times of work items (not the typical work schedule item) to create fixed day, evening, and night schedule categories. This approach is also consistent with Presser's method (2003). Respondents who worked more than 50% of their work hours between 8 a.m. and 4 p.m. were coded as working a fixed day schedule, those who worked more than 50% of their work hours between 4 p.m. and 12 a.m. were coded as working a fixed evening schedule, and those who worked more than 50% of their work hours between 12 a.m. and 8 a.m. were coded as working a fixed night schedule. Because the start and end times of work items do not capture variability in respondents’ schedules, we used only the typical work schedule item to measure rotating, split, and irregular schedules. If a respondent said they worked a rotating, split, irregular, or other schedule, they were coded as working that schedule regardless of their reported start and end times of work. We defined working any type of nonstandard schedule as working a fixed evening, fixed night, rotating/split/other nonstandard shift, or irregular schedule. Using the days of the week worked items, we measured weekend work in the respondent's primary job using another set of categorical variables for worked any weekend day (Saturday or Sunday) versus worked only weekdays.

Work Hours

We measured mothers’ total weekly work hours in their primary job by multiplying the number of hours they worked per day (using the start and end times of work items) by the number of days worked per week. We used the start and end times of work items in our measure of total weekly work hours because these items were consistently asked across all survey waves. We used our measure of total weekly work hours to code mothers who worked 1–34 hours per week as working part-time and those who worked 35 or more hours per week as working full-time.

Reasons for Working Nonstandard Schedules

We used an item that asked respondents to select the main reason for working this type of schedule. Voluntary reasons included better childcare arrangements, better pay, better arrangements for care of other family members, allows time for school, and “other voluntary reasons.” Involuntary reasons included could not get any other job, requirement of the job, and “other involuntary reasons.”

Analytic Approach

We first conducted descriptive analyses to examine trends from 1988 to 2019 in work schedules among all mothers. We used four primary work schedule measures. The first measure examines working any nonstandard schedule, working a standard daytime schedule only, and not working. The second measure examines mothers’ work schedules by their employment status: nonstandard full-time, nonstandard part-time, standard full-time, and standard part-time schedules, as well as not working. The third measure differentiates between specific types of nonstandard schedules: fixed evening, fixed night, rotating/split/other, and irregular schedules. Lastly, we examined weekend work by comparing mothers who work weekends versus weekdays only or not working.

To determine whether the prevalence of nonstandard schedules changed over time by maternal characteristics, we estimated multinomial logistic regression models predicting each categorical work schedule outcome from mothers’ demographic characteristics and interactions between year indicators and mothers’ race and ethnicity, education, and family structure. These models adjusted for other maternal and family characteristics that have been associated with mothers’ propensity to work or to work nonstandard schedules: the mother’s age (categorical variable for 18–24 years, 25–34 years, 35–44 years, or 45 years or older), the mother's number of children (categorical variable for one, two, or three or more own children younger than 18), and the mother's youngest child's age (categorical variable for ages 0–5 years, 6–12 years, or 13–17 years). We also controlled for the month of the year in which the survey was conducted. Because our analytic sample includes both working and nonworking mothers, we did not include family income (because mothers’ earnings contribute to family income) or mothers’ occupation (because it is not observed for mothers not currently working) in our analytic models. All models were weighted using the SIPP person-level weights to account for attrition and survey design. Standard errors were clustered at the state level to adjust for the potential nonindependence of observations due to the complex survey design.

To test our hypotheses about diverging trends in work schedules, we plotted the predicted probabilities of working nonstandard schedules from these models for each subgroup for each year in the study. We compared mothers without a college degree with those who had a college degree, single mothers with married mothers, and mothers of color with White mothers. We estimated the average marginal effects (i.e., average change in the probability of the outcomes attributed to a change in the independent variable) by mothers’ education, family structure, and race and ethnicity to determine whether the difference in the predicted probabilities between subgroups was statistically significant at p < .05. We show results for our primary work schedule measure (i.e., working any nonstandard schedule) in the manuscript and include results from alternative measures of work schedules in the online appendix. Finally, we descriptively examined trends in mothers’ reasons for working their schedules to determine how they changed overall and for each subgroup.

Results

Trends in Mothers’ Work Schedules

Results show that between 1988 and 2019, the percentage of all mothers who were employed in standard daytime schedules increased from 49.8% to 56.9% (see Figure 1, panel a).⁵ Meanwhile, the percentage working any type of nonstandard schedule remained relatively flat at roughly 16%, peaking at 17.9% in 1997 and declining to 13.4% in 2019. Among working mothers, the percentage working any nonstandard schedule ranged between approximately 21% and 25% during this period, reaching a low of 19.0% in 2019 (results not shown in figure). Consistent with trends in economic growth and contractions, the percentage of mothers not working declined precipitously during the economic boom of the late 1990s, increased during the Great Recession, and began declining again in 2013 as the economy recovered.

When considering mothers’ work hours (Figure 1, panel b), we found that the percentage of mothers working nonstandard schedules full-time (7.5% to 10.3%) remained relatively stable while the percentage working nonstandard schedules part-time declined slightly beginning in 2013 (from approximately 7% to 9% in 1988–2011 to approximately 5% to 6% in 2013–2019). That a greater proportion of mothers with nonstandard schedules worked full-time versus part-time schedules suggests that jobs with nonstandard schedules are not primarily secondary or supplemental jobs.

With respect to specific types of nonstandard schedules (Figure 1, panel c), irregular work schedules were the most common type, followed by evening, rotating or split, and night schedules. In line with our hypotheses, mothers’ employment in jobs with irregular schedules increased from the mid-1990s through 2011—peaking at 7.9% in 2010. However, in contrast to our expectations, mothers’ employment in these jobs then began declining to 5.6% in 2019. Mothers’ employment in jobs with evening schedules hovered around 4%, declining slightly during the Great Recession. Employment in rotating or split schedules remained relatively steady, at 3% to 4%. Mothers’ employment in night schedules hovered around 2%, declining slightly beginning in 2009 and reaching 1.5% in 2019.

The percentage of mothers working weekends increased from 14.6% to 17.7% between 1988 and 2019 (Figure 1, panel d). This increase was driven by more working mothers reporting that they worked both weekdays and weekends rather than weekends only (only 1% of mothers worked only on the weekends, and this rate has not changed over time; not shown in the figure).⁶ With the exception of an increase in weekend work, our results suggest relatively little change in mothers’ employment in jobs with nonstandard schedules over time.

To assess the sensitivity of our findings to our measure of nonstandard schedules, we conducted two additional analyses. First, we expanded our measure of nonstandard work schedules so that mothers who reported at least one hour of work outside of 8 a.m. to 5 p.m. were coded as working a fixed evening or night schedule. This coding led us to reclassify mothers who worked a small number of hours during nonstandard times from the regular daytime schedule category to working a nonstandard schedule. This analysis tells us whether our main findings might miss an increase in nonstandard work among mothers who worked a small amount of time in the early morning, in the evening, or at night. As expected, we found higher rates of nonstandard work schedules than with our main measure; however, trends were similar in that the rate of nonstandard schedules remained relatively steady at roughly 33% to 34%, ranging from 32.2% in 1988 to 33.8% in 2019 (see Figure A1; all figures and tables designated with an “A” are available in the online appendix). We prefer our main measure because it more accurately captures more extensive nonstandard schedules that are more likely to affect child and family outcomes, such as childcare arrangements (Pilarz et al. 2019).

Second, we created a combined measure of nonstandard schedules that includes mothers who worked during nonstandard times (i.e., evenings, nights, split, rotating, irregular schedules) and those who worked weekends. Using this combined measure, we found that the proportion of mothers who worked a nonstandard schedule was quite stable between 23% and 24%, ranging from 23.9% in 1988 to 22.7% in 2019 (see Figure A2). The analyses that follow use the disaggregated measures of nonstandard schedules and weekend work.

Trends in Nonstandard Schedules by Maternal Characteristics

Overall trends in mothers’ work schedules mask differences by mothers’ education, family structure, and race and ethnicity. Figures 2–4 show the predicted probabilities estimated from our interaction models for each year in the study and whether the gap between subgroups is statistically significant at p < .05. With respect to mothers’ education, gaps in nonstandard schedules between mothers with a college education and those without grew beginning in the late 1990s (see Figure 2). Before 1999, the prevalence of working a job with a nonstandard schedule was similar between mothers with and without a college degree. Beginning in 1999, mothers with a college education became less likely to work a nonstandard schedule relative to mothers without a college degree. For example, the percentage of college-educated mothers working nonstandard schedules declined from a peak of 18.7% in 1997 to 10.6% in 2019. By contrast, the percentage of mothers with a high school education working nonstandard schedules declined from 19.3% in 1997 to 14.9% in 2019 (see Figure 2, panel b). The gap between the two groups peaked at 6.5 percentage points in 2013 and declined thereafter, reaching 4.3 percentage points in 2019. When considering alternative measures of work schedules (see Figures A3–A5), we found similar trends for full-time nonstandard schedules, fixed evening or night schedules, and working weekends; however, rates of irregular work schedules were similar across education groups.

With respect to family structure, single mothers living without other adults became more likely to work nonstandard schedules than married mothers (Figure 3, panel b). In the early 1990s, married mothers were more likely to work nonstandard schedules than single mothers living without other adults, although the gap was statistically significant for only one of those years. This trend reversed in the late 1990s, when single mothers living without other adults became more likely to work nonstandard schedules than married mothers. This gap peaked at 6.1 percentage points in 2013. The elimination of the gap in 2019 could indicate that these trends were starting to reverse; however, the COVID-19 pandemic that began in 2020 likely altered existing trends in mothers’ schedules. Single mothers living with other adults were also more likely to work any nonstandard schedule than married mothers, but this gap remained relatively stable over time, peaking at 5.8 percentage points in 2015 (Figure 3, panel a). These patterns were consistent across all work schedule measures (see Figures A6 and A7).

Turning to differences by race and ethnicity,⁷ results show that Black mothers became as likely or more likely to work a nonstandard schedule relative to White mothers (Figure 4, panel a). White mothers were more likely to work nonstandard schedules than Black mothers before the late 1990s, by as much as 5.6 percentage points in 1991. This gap closed in the early 2000s and reemerged during the Great Recession period, when Black mothers’ employment was disproportionately impacted by the economic downturn. Black mothers became more likely to work nonstandard schedules than White mothers after 2013 (the difference peaking at 4.1 percentage points in 2018), although the gap between the two groups was not statistically significant. This pattern—increases in the probability of Black mothers working nonstandard schedules relative to White mothers—was evident across all alternative work schedule measures (see Figure A8). The gap between Black and White mothers in 2019 was statistically significant when we used indicators for working a full-time nonstandard schedule and working a fixed evening or night schedule. Hispanic mothers were generally less likely to work a nonstandard schedule than White mothers throughout the period (Figure 4, panel b), although the gap narrowed beginning in 2010. Mothers who identified as Asian or with other racial groups (i.e., Native Hawaiian, Pacific Islander, American Indian, Alaska Native, or multiple races) were equally likely to work a nonstandard schedule as White mothers during the study period (Figure 4, panel c). We found similar patterns when considering alternative measures of work schedules (see Figures A9 and A10).

Potential Explanations for Diverging Trends in Maternal Nonstandard Schedules

Is the relative increase in nonstandard work among mothers without a college degree (vs. those with a degree), single mothers living without other adults (vs. married mothers), and Black mothers (vs. White mothers) driven by these mothers being more likely to enter the labor force during the study period? Because our sample includes both working and nonworking mothers, increases in the prevalence of nonstandard schedules could be driven by increases in work. Supplemental analyses suggest that they are not. Analyses restricted to working mothers showed substantially larger gaps in nonstandard work schedules by mothers’ education. For example, in 2019, 25.0% of working mothers with a high school diploma versus 13.0% of working mothers with a college degree worked a nonstandard schedule, a gap of 12.0 percentage points (see Figure A11). Disparities in nonstandard schedules by family structure and race and ethnicity changed less when we restricted the sample to working mothers (see Figures A12 and A13). For example, at its peak in 2013, the gap between single mothers living without other adults and married mothers was 6.1 percentage points among all mothers and 4.5 percentage points among working mothers only. At its peak in 2018, the gap in nonstandard schedules between Black and White mothers was 4.1 percentage points among all mothers and 5.2 percentage points among working mothers.

Are disparities in maternal nonstandard work schedules driven by changes in occupations across subgroups? Mothers employed in the service sector consistently had the highest rates of nonstandard schedules (31% to 40%), whereas those in the managerial and professional specialty sector had the lowest (13% to 20%; Figure A14). Further, we observed different trends in occupations by mothers’ education, family structure, and race and ethnicity. For example, employment in service occupations increased among mothers without a college degree but not among college-educated mothers (results not shown). To determine the extent to which occupational trends explain growing disparities in nonstandard schedules, we estimated our main model restricted to the subsample of working mothers while controlling for occupation.⁸ We found that gaps in nonstandard schedules by mothers’ education were substantially smaller (e.g., in 2019, a 7.4-percentage-point gap between high school–educated and college-educated mothers; Figure A15). This finding suggests that increasing occupational segregation by education contributed to growing gaps in nonstandard schedules between college-educated and non-college-educated mothers. Gaps in nonstandard schedules by family structure and race and ethnicity also decreased but by a smaller magnitude (e.g., in 2018, a 4.5-percentage-point gap between Black and White mothers; Figures A16 and A17).

If changes in the childcare market and mothers’ childcare decisions contributed to trends in maternal nonstandard schedules, then we would expect to see diverging trends by the age of the mother’s youngest child because mothers with young children not yet in school are most impacted by childcare availability and cost. To examine this possibility, we added to our main model interactions between the age of the mother's youngest child (ages 0–5, 6–12, and 13–17) and year. Results showed no evidence that trends in maternal nonstandard schedules differed by the youngest child's age (see Figure A18), and results for the other subgroups were unchanged (results not shown). Among all mothers, rates of nonstandard schedules were similar across all groups by the youngest child's age. Among working mothers, those with young children aged 0–5 were consistently more likely to work nonstandard schedules than those with older children (results not shown).

Reasons for Working Nonstandard Schedules

Among working mothers, the proportion reporting an involuntary reason for working their schedule increased across all types of work schedules, with a notable shift beginning in 2009 (see Figure 5). In turn, the proportion of working mothers who reported voluntarily working their schedule because of better childcare arrangements (Figure 5, panel b) decreased over this period, particularly beginning in 2009. Mothers who worked a fixed daytime schedule (vs. any nonstandard schedule; see Figure 5, panel a) and mothers who worked full-time schedules (vs. part-time schedules; see Figure 5, panel c) were consistently more likely to report an involuntary reason. With respect to specific types of nonstandard schedules (Figure 5, panel d), mothers who worked nights were most likely to report an involuntary reason, whereas those who worked evenings were least likely to do so. However, beginning in 2009, involuntary reasons substantially increased among mothers working evenings, rotating/split, and irregular schedules. We found a similar pattern in weekend work: mothers who worked weekends were less likely to report an involuntary work schedule than those who worked only weekdays, but both groups became more likely to report an involuntary reason beginning in 2009 (results not shown).

We also examined mothers’ involuntary reasons for working their schedule among subgroups by mothers’ education, family structure, race and ethnicity, and the youngest child's age. The proportion of mothers reporting an involuntary nonstandard schedule rose in all subgroups beginning in 2009. However, mothers with lower education levels, single mothers (living with or without other adults), and Black and Hispanic mothers were more likely to report involuntarily working a nonstandard schedule than their counterparts (results not shown). For example, in 2019, 78% of mothers with a high school education reported involuntarily working a nonstandard schedule versus 68% of mothers with a college degree. Additionally, mothers whose youngest child was aged 13–17 were consistently more likely than mothers with younger children to report an involuntary nonstandard schedule (results not shown).

Discussion

Given that labor market inequality increased and the quality of low-wage jobs deteriorated over the past 30 years, understanding mothers’ employment in jobs that might be detrimental to their own and their family's well-being is critical for understanding inequality between families. This study examined how mothers’ employment in jobs with nonstandard schedules changed from 1988 to 2019 and the extent to which these changes differed by mothers’ education, family structure, and race and ethnicity. Findings show that between 1988 and 2019, mothers’ employment in jobs with nonstandard schedules remained relatively stable, except that weekend work increased modestly. Mothers’ employment rate and employment in standard daytime schedules increased over this period, suggesting that some mothers might be opting out of nonstandard schedules. However, we also found growing gaps in the propensity to work nonstandard schedules by mothers’ education, family structure, and race and ethnicity. These findings suggest that whereas college-educated, married, and White mothers might be opting out, mothers without a college degree, single mothers, and Black mothers might increasingly be working nonstandard schedules out of necessity.

This study makes several contributions to prior research. We estimated trends in mothers’ work schedules over 31 years using consistent measures of work schedules. Contrary to expectations, we found little evidence of increases in mothers’ employment in nonstandard schedules over time. Approximately 15% to 16% of mothers worked any type of nonstandard schedule during this period, whereas weekend work increased from approximately 15% to 18%. Our findings are similar to those of Hepburn (2020), who found increases in weekend work between 1990 and 2012 but not in evening or night schedules. As in prior research (Finnigan 2018; Hepburn 2020), we found increases in irregular schedules that peaked shortly after the Great Recession, but our findings show that mothers’ employment in irregular schedules declined beginning in 2013.

What might explain these patterns of relative stability in mothers’ employment in nonstandard schedules? Countervailing trends are likely at play. Mothers’ demographic characteristics have shifted toward both groups that are more likely to work nonstandard schedules—unmarried mothers and mothers of color—and groups that are less likely to work nonstandard schedules—older and college-educated mothers (Enchautegui 2013; Golden 2015; Lambert et al. 2019; Presser 2003; Presser and Ward 2011). Moreover, mothers’ occupations have changed substantially. Mothers’ employment in the managerial and professional sector, which typically involve higher paying jobs with the lowest rate of nonstandard schedules, nearly doubled over this period and increased only slightly in the service sector, which has the highest concentration of jobs with nonstandard schedules.

A key finding from this study is that trends in nonstandard schedules differ substantially by mothers’ education, family structure, and race and ethnicity. As seen in prior research (Enchautegui 2013; Golden 2015; Hepburn 2020; Lambert et al. 2019; Presser 2003; Presser and Ward 2011), we found that mothers with lower education levels, Black mothers, and single mothers are more likely to work nonstandard schedules than college-educated, White, and married mothers, respectively. However, our study shows that these gaps emerged or grew from 1988 to 2019. We found that having a college degree became a stronger protective factor against working a nonstandard schedule. We also found that single mothers living without other adults became more likely to work all types of nonstandard schedules than married mothers. By contrast, the propensity to work nonstandard schedules changed little among single mothers living with other adults. Single mothers living without other adults are particularly economically vulnerable because they likely cannot opt out of the labor force. Our findings highlight the need to consider this group separately from other unmarried mothers.

With respect to mothers’ race and ethnicity, we found that Black mothers became as or more likely to work nonstandard schedules—especially full-time nonstandard schedules and fixed evening or night schedules—relative to White mothers in recent years. We found less consistent evidence of changes in the propensity to work nonstandard schedules among other racial and ethnic groups, emphasizing the importance for future research to consider heterogeneity among mothers of color. Additionally, our study included data during the economic recovery from the Great Recession through 2019. These results suggest that gaps by mothers’ education, family structure, and race and ethnicity widened during the economic recovery period.

We considered several potential explanations for these diverging trends. Our findings suggest that increases in single mothers’ labor force participation over this period likely contributed to their increasing rates of nonstandard schedules relative to married mothers. However, changes in mothers’ labor force participation do not explain gaps by mothers’ education or race and ethnicity. Our findings also suggest that changes in occupations across subgroups contributed to diverging trends in nonstandard schedules by education, family structure, and race and ethnicity. In particular, the greater concentration of nonstandard schedules among mothers without a college degree is largely explained by their growing overrepresentation in service occupations. Despite substantial changes to the childcare market and ECE policies beginning in the 1990s, we found no differences in trends in nonstandard schedules among mothers with young children relative to mothers with school-age children. Taken together, this evidence suggests that labor market trends and conditions are primarily driving our findings. Nevertheless, future research should test additional possible explanations, including changes in ECE and safety net policies, such as the 1996 welfare reform and expansions in the Earned Income Tax Credit, which disproportionately affect low-income and single-mother families.

An important finding from our study is that mothers who worked nonstandard schedules became increasingly likely to do so for involuntary reasons. We found a notable increase in involuntary work schedules beginning during the Great Recession, particularly among mothers working part-time nonstandard schedules and those working evening, rotating, split, or irregular schedules. That mothers who work irregular schedules increasingly did so involuntarily suggests an increase in unstable and unpredictable schedules (Schneider and Harknett 2019). Our findings also show that mothers with lower education levels, unmarried mothers, and Black and Hispanic mothers were more likely to report working a nonstandard schedule involuntarily than their counterparts. These findings suggest a potential shift beginning during the Great Recession in mothers’ ability to secure jobs that best fit with their work–family demands or provide them with some control over or flexibility in their schedules (Henly et al. 2021), particularly among mothers from historically marginalized groups. Future research should consider how mothers’ employment in jobs with nonstandard and involuntary schedules has changed since the onset of the COVID-19 pandemic, which impacted working conditions across the board. The pandemic also created a new dimension of workplace bifurcation between workers with and without the ability to do their jobs remotely.

Our findings have implications for maternal, child, and family well-being. Adverse associations between maternal nonstandard work schedules and child and family well-being are generally stronger among children in low-income and single-mother families and those whose mother has lower education levels (Han 2005, 2008; Han et al. 2010; Han and Waldfogel 2007; Joshi and Bogen 2007; Wang 2023). Our findings suggest that children who are most vulnerable to the effects of maternal nonstandard schedules are increasingly likely to be exposed. These diverging trends in mothers’ work schedules could be contributing to inequality in child and family outcomes by mothers’ education, family structure, and race and ethnicity. Non-college-educated mothers and single mothers were more likely to work a nonstandard schedule involuntarily than their counterparts, which might help explain more adverse effects of nonstandard schedules among these groups. Additionally, Black and Hispanic mothers were more likely to work nonstandard schedules involuntarily than White mothers, highlighting the need for future research to consider heterogeneity in the effects of nonstandard schedules on children and families by race and ethnicity. We also found that mothers with older school-age children were more likely to work a nonstandard schedule involuntarily than those with younger children. This finding suggests that mothers’ reasons for working nonstandard schedules depend on their children's ages and that the adverse effects of nonstandard schedules might operate via different mechanisms for older school-age children than younger children. Future research should examine how the associations between maternal nonstandard work schedules and child and family well-being differ across demographic characteristics and how these associations have changed over time.

Our study focused on mothers’ work schedules because mothers are most often children's primary caregivers (Raley et al. 2012; Sayer 2016) and, therefore, their employment shapes how children and families spend their time. However, fathers’ work schedules also matter for child and family well-being (Han and Fox 2011; Han et al. 2010; Zilanawala and McMunn 2024) and are important for understanding the effects of mothers’ work schedules on children, such as the extent to which parents work nonoverlapping schedules (Boushey 2006; Hattery 2001). Future research should examine how fathers’ work schedules have changed overall and in relation to mothers’ schedules. Further, little is known about changes over time in work schedules among nonparental caregivers or childless families, for whom the timing of work is also critical for their well-being (Bolino et al. 2021; Schneider and Harknett 2019). Understanding whether inequality in work schedules has also increased among these groups is an important area of future research.

We note several limitations of this study. Some periods have a three- to four-year gap in the data because the SIPP did not field the work schedules topical module in consistent intervals. In particular, it would have been ideal to observe mothers’ work hours and schedules in 2006–2008, before and at the beginning of the Great Recession. Our measures of work schedules capture mothers’ typical work schedules but are not well-suited for capturing variability in work schedules from day to day or week to week. Although this feature is not unique to the SIPP (most national surveys use similar work schedule items), future surveys would benefit from including a weekly work calendar and other measures of variability in work hours and schedules (Fugiel and Lambert 2019; Lambert et al. 2014). Our measure of irregular work schedules did not distinguish between irregular work schedules that are employee-driven, which might indicate flexibility, and those that are employer-driven, which are likely to be experienced as unstable and unpredictable. Moreover, our measure of irregular schedules likely underestimated the prevalence of unpredictable schedules. Some mothers who reported a regular schedule might nevertheless experience unpredictability in their work schedules if they work on call or are asked to stay late or leave early at the last minute (Ananat et al. 2022; Lambert et al. 2014; Schneider and Harknett 2019). Future research should incorporate items that explicitly ask about workers’ input or control over their work schedules and about schedule unpredictability.

As inequality in wages and benefits between workers with and without a college degree has grown (Groshen and Holzer 2019; Howell and Kalleberg 2019), our findings provide evidence of growing inequality in another aspect of job quality—work schedules—by mothers’ education levels and by family structure and race and ethnicity. These findings call for labor market regulations to improve the quality of work schedules. In recent years, multiple localities and the state of Oregon have passed fair workweek laws that grant workers the right to request work schedules without retaliation and more stability, predictability, and input into their schedules. In Seattle and Emeryville, California, these regulations have reduced unpredictability in workers’ schedules, although their effects on the timing of work are unknown (Ananat et al. 2022; Harknett et al. 2021). By providing workers with more control over when they work, scheduling regulations also hold promise for reducing mothers’ involuntary nonstandard schedules and work schedule inequality.

Beyond reducing inequality in the quality of work schedules, policies that facilitate work–family balance could help mitigate the potentially harmful effects of nonstandard work schedules on parental and child well-being. A growing number of states and localities have begun mandating paid time off (e.g., paid family and medical leave and paid sick leave), as well as unpaid leave for caregivers to attend childcare and school activities (Bipartisan Policy Center 2024; Center for WorkLife Law 2023; National Partnership for Women and Families 2023). These policies could facilitate parents’ ability to attend to their children's needs and be more involved with their children even when they work nonstandard schedules. Further, increasing access to affordable, high-quality ECE, especially during nonstandard hours, would alleviate childcare challenges among parents who work nonstandard schedules and would support children's development (Henly and Adams 2018; Pilarz et al. 2019).

Acknowledgments

The authors gratefully acknowledge Kess Ballentine, Dylan Bellisle, Marcy Carlson, Heejung Chung, Julia Henly, and Christine Schwartz for helpful comments on prior versions of this manuscript. The authors are also grateful to Russell Dimond for technical assistance. This research was supported by the Eunice Kenney Shriver National Institute of Child Health and Human Development under award number K01HD104002. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Notes

References