In a recently published article (open access) in the Journal of Marriage and Family, Anette Fasang and I focused on parenthood wage gaps in the United States. Parenthood wage gaps are the percentage difference in average wages between parents and childless individuals and are a standard measure of labour market advantage or disadvantage for parents. Findings generally point to motherhood wage penalties – mothers earn less than childless women – and fatherhood wage premiums – fathers earn more than childless men.
In this study, Anette and I wanted to not only assess how parenthood wage gaps vary across individual lives for different gender and race groups in the United States. To this end, we integrated intersectional and life course approaches to understanding social inequalities.
Intersectionality was originally introduced in the context of anti-discrimination legislation, but has been adopted in the social sciences to think about complex inequalities. One could think of it like so: Gender power dynamics put men in more advantaged positions compared to women. Similarly, racial power dynamics put whites in more advantaged positions compared to Blacks and Hispanics. However, it is important to note that gender power dynamics operate within racial groups and racial power dynamics operate within gender groups. Therefore, white men are in a more privileged position than Black and Hispanic men, while Black and Hispanic women are in a more disadvantaged position that white women. Intersectional approaches to inequality can – and have been – extended to also look at social class differences and much more than what we do in this study. The intersectional lens helps us avoid naturalising white and/or male privilege as a reference and focuses our attention on the unique experience of each gender by race group.
The life course approach highlights the importance of different aspects of time for social inequality. The short- and long-term consequences of parenthood – positive or negative – differ by the age of parents, but also by the historical context in which men and women become parents. For example, a common conception in US society is that entering parenthood at an early age will have permanent negative effects on individuals, especially women. In contrast, entering parenthood later is seen to be less risky. The possibilities to combine work and parenthood change across the life course and are shaped by historically specific educational systems, local labour markets and child care provision, all of which differ by gender and race. The life course approach points our attention to three aspects of parenthood wage gaps: their magnitude (how big are parenthood wage gaps), their timing (when do parenthood wage gaps emerge and peak), and their persistence (over what age range do parenthood wage gaps persist).
Parenthood Wage Gaps across the Life Course
We argue that the magnitude, timing and persistence of parenthood wage gaps depends on the gender- and race-specific interplay of labour market dynamics, and family demographics over the life course (Figure 1).
Labour market dynamics are central to generating parenthood wage gaps. Traditional labour market explanations for parenthood wage gaps fall into labour supply and demand side mechanisms. On the supply side more or less productive and work oriented individuals might select into parenthood. Parents could also become less productive than childless individuals and less able to fulfil the ideal worker norm of reliability, flexibility and working long hours. Demand side explanations include positive or negative discrimination of parents in terms of hiring, firing, wages and promotions.
Figure 1: Parenthood Wage Gaps across the Life Course
Family demographics beyond childbirth itself impact parenthood wage gaps. Of course, parenthood wage gaps occur when entering parenthood, but their magnitude, timing and persistence depend on the timing of birth in the life course and increase with the number of children. For mothers, lost work experience, parenting demands that conflict with work, and negative employer discrimination increase with the number of children. For fathers, having more children might additionally motivate them to achieve higher incomes and reinforce positive employer discrimination. In addition, relationship dynamics alter parenthood penalties through the presence or absence of a partner with whom paid work and unpaid child care can be shared. Both motherhood penalties and fatherhood premiums will be largest among married individuals with a traditional household division of labour. Fathers in these households benefit from their spouses’ unpaid household labour, while women forego own labour income.
It is how labour market dynamics and family demographics interact, i.e. their interplay across the life course, which generates intersectional differences in the magnitude, timing, and persistence of parenthood wage gaps. For example, individuals’ intentions to have children and their relationship status can affect their selection into more or less family friendly and financially rewarding occupations. To-be fathers tend to enter into higher paying and more desirable occupations, while to-be mothers enter into lower paying and less desirable occupations. Following childbirth, women’s probability of upward occupational mobility and higher paying positions continues to decrease the longer they remain out of the labour force.
To develop hypotheses or expectations on how parenthood wage gaps evolve over the life course for different gender and race groups, we now need to take a deeper look at the unique labour market and family experiences of Black, Hispanic, and white men and women. It is important to say here that we are not examining how or why Black, Hispanic, and white men and women tend to find themselves in different labour market segments and have different family experiences, although we might speculate. Rather we are going to concentrate on what labour market and family experiences might mean for parenthood wage gaps.
Parenthood Wage Gaps across the Life Course by Gender & Race
The race-specific occupational segments that men and women navigate channel labour market dynamics relevant for parenthood wage gaps. To illustrate our main argument, we consider two occupational groups in particular: managerial and professional occupations and service occupations (see Figure 2). Figure 2 shows the percent of Black, Hispanic, and white men and women that are employed in managerial or service occupations, the age range of when those groups have their first child, and the compression or distribution of hourly wages within managerial or service occupations. When we say that a group of people in a certain labour market segment have highly compressed wages, then we mean that most everyone earns the same. In contrast, low wage compression means that some have very high wages, while others earn much less.
Figure 2: Employment and Wage Compression in Managerial and Service Occupations across the Life Course by Gender and Race (see article for more information)
As can be seen in Figure 2, Black and Hispanic men and especially women are found in household and non-protective service occupations. Low skilled manual and service occupations where Blacks and Hispanics are concentrated are not only characterised by low wages, but also by high wage compression, flat wage growth and few chances for internal or external upward mobility. In contrast, high skilled managerial and professional occupations where whites are over-represented provide high wages, low wage compression, steeper wage growth and opportunities for upward mobility over the life course. Especially minority women are concentrated in non-manual, poorly paid, insecure service occupations, while white women on average hold more prestigious and economically rewarding occupations compared to minority women and men. Residential segregation into distinct local labour markets may reinforce the processes that concentrate white men and women in skill intensive professional non-manual occupations, and Blacks and Hispanics in lower-skilled manual and service occupations.
Family demographic characteristics related to household dynamics reinforce the occupational factors that suppress motherhood wage penalties for Black and Hispanic women. Historically motherhood and paid work have not been mutually exclusive, especially in Black households where families have often resisted institutionalised inequalities. Minority mothers are more active on the labour market compared to white mothers, partially because extended family members support mothers’ motivation and productivity for paid work to a greater extent, for example through the provision of informal childcare.
However, black women can also rely to a lesser extent on a partner for income security due to black men’s lower earnings capacity compared to white men and higher family instability over the life course among blacks. Because black men and women’s wage trajectories over the life course are much flatter on a lower level, usually both spouses work full-time to make ends meet. In addition, Black women are less likely to be married at any time point in the life course compared to white women. As a result, gender divisions of labour will tend to be more equal, which leads to lower motherhood penalties and lower fatherhood premiums. In contrast, norms for female homemaking have been stronger among whites, corresponding to a greater increase in white women’s hours of housework upon entering motherhood.
Lower earnings capacities and higher income volatility among Black men compared to white men has been cited as one explanation for small or non-existent Black fatherhood premiums. Black men may be seen as less committed breadwinners and more irresponsible fathers than Hispanic and especially white fathers and thus benefit less from positive employer discrimination. In addition, the division of labour in Black households is more equal and therefore Black men may not increase their work productivity as much as Hispanic and white men in male-breadwinner female-homemaker households upon entering fatherhood. We expect medium motherhood penalties and fatherhood premiums for Hispanics, due to their intermediate position on the labour market, because they are more likely to be in married relationships as parents, and hold more traditional values that foster male breadwinner divisions of labour compared to Black men and women. We therefore hypothesise that:
H1a: The motherhood wage penalty for each parity will be largest for white women, followed by Hispanic women, and smallest for Black women.
H1b: The fatherhood wage premium for each parity will be largest for white men, followed by Hispanic men, and smallest for Black men.
Occupational sorting by race and gender evolves over the life course, partly in response to family demographic events as outlined above. The percentage of white men and women in professional and managerial occupations with high wages and low wage compression grows drastically with age (see Figure 2). In contrast, the growth of Blacks and Hispanics in high skilled occupations is less pronounced and flattens earlier in the life course. Therefore, Black and Hispanic men and women are entering parenthood within low skilled and low paying occupations with little opportunity for wage differences to emerge and persist. In contrast, white men and women are entering parenthood later when they are more likely to hold managerial and professional occupations that provide less compressed wages, and therefore more leverage for parenthood wage gaps to unfold and persist compared to minority men and women.
In addition, family formation events of Black and Hispanic men and women tend to be more tightly compressed in their early 20s, but spread out across a later period of the life course for white men and women (see Figure 2). Black and Hispanic women of our study cohorts enter parenthood much earlier than white women. Men’s fertility is similar to women, but men of all racial backgrounds enter parenthood somewhat later.
Due to on average earlier births among Blacks and Hispanics, any penalties that occur, are likely to occur early in life. For whites we also expect low penalties for early births, but higher penalties and premiums for births that occur later in life, when substantial wage growth has occurred for a large group of (childless) whites and they have accumulated skills that depreciate during child-care related interruptions. We therefore hypothesise that:
H2a: The motherhood wage penalty for white women for each parity will peak later in the life course compared to Black and Hispanic women.
H2b: The fatherhood wage premium for white men for each parity will peak later in the life course compared to Black and Hispanic men.
The gender- and race-specific interplay between labour market dynamics and family demographics elaborated above also has implications for intersectional differences in the persistency of wage gaps. Lower occupational mobility and flatter wage trajectories are coupled with a higher likelihood to live in dual breadwinner households for minority men and women compared to whites. This is even more the case for Blacks than for Hispanics. As family formation is contracted in the early life course of minority men and women, any parenthood wage penalties that emerge early in the life course are likely short-lived. In contrast, white men and women are concentrated in occupations with lower wage compression and steeper wage growth, more likely to opt for male breadwinner household divisions of labour, and form families over a longer period later in the life course. Therefore, we hypothesise that:
H3a: Motherhood wage penalties extend over a longer age range for each parity for white women compared to Black women with Hispanic women taking an intermediate position.
H3b: Fatherhood premiums will extend over a longer age range for each parity for white men compared to Black men with Hispanic men taking an intermediate position.
Data & Methods
We of course used sophisticated data and methods to estimate life course and race differences in motherhood and fatherhood wage gaps. If this brief overview is too brief, then please take a look at the article where everything is described in much more detail. We use the 1979 and 1997 National Longitudinal Surveys of Youth (NLSY79 and NLSY97). Our data covers the years 1979 – 2003, when the respondents were age 20-40. They are included in the analyses, if they reported to work at least one hour in the last year.
Consistent with most previous literature our dependent variable is the natural logarithm of the hourly wage for the main job reported by the respondent. Based on the fertility histories provided by the NLSYs, we use the number of biological children to assess the impact of each additional biological child separately. We also include analyses that account for differences for a wide range of individual characteristics, such as years of education, marital status, occupational group, and residential region. Parenthood wage gaps that do not account for differences in individual characteristics are called unadjusted estimates, while those that account for differences are called adjusted estimates.
We estimate parenthood wage gaps using fixed effects regressions with an interaction between age and the number of children as a categorical variable of “one child”, “two children”, and “three or more children” each compared to the reference category of childless men or women of a specific racial background at a given age. We exclude Asian and Native Americans from our white category by selecting only respondents of European, Canadian, or US American ancestry with additional information from the NLSYs.
Motherhood Penalties over the Life Course by Parity
The top panels of Figure 3 show age-specific unadjusted motherhood wage gaps with 95 percent confidence intervals for Black (left panel), Hispanic (middle panel) and white (right panel) women. Within each panel, the estimated wage gap for having one, two, or three and more children compared to being childless is represented with the yellow, orange and red line, respectively. The bottom panel of Figure 3 displays adjusted motherhood wage gaps in the same setting controlling human capital factors and all other characteristics.
Figure 3: Motherhood wage gaps, ages 20-40 by race and number of children, unadjusted (top) and adjusted (bottom)
As can be seen in Figure 3, unadjusted motherhood wage penalties were largest for all mothers between ages 30 and 35. At age 30, white women with one child earned roughly 11 percent less than childless white women, while Black and Hispanic women with one child earned around 15 percent less than their peers. However, these one-child motherhood wage penalties dissipated by age 40. For white women, both mothers of two and three or more children showed significant unadjusted motherhood penalties that lasted until age 40. Moreover, in line with hypotheses 1a and 3a, the unadjusted motherhood penalty for Black women of three and more children peaked at a lower level of 20 percent and is concentrated in a brief age window around 30, compared to a higher, but equally age concentrated peak of 25 percent for Hispanic women. In contrast, the unadjusted motherhood penalty reached almost 30 percent for white women of three and more children in their late 20s and persisted until age 40.
In line with expectations and previous research adjusted motherhood penalties were substantially lower, primarily driven by differential human capital endowment of Black, Hispanic and white childless women and mothers. Note that the differential human capital endowment reflects their actual life course experience and by “controlling” for these factors we create a hypothetical estimate of what their situation would be like, if they had the same human capital as white women. Black mothers only experienced a significant adjusted motherhood penalty of about 9 to 13 percent in a few years around age 30, irrespective of the number of children. Adjusted penalties for Hispanic mothers were similar to Black mothers, but we found no statistically significant penalty for Hispanic women with three or more children. For white women of only one or two children we also no longer found a significant adjusted motherhood penalty. Only for white women of three and more children was there a significant and enduring motherhood penalty of about 15 percent that peaked in the late 20s and persisted until age 40, which was not accounted for by human capital or demographic factors.
In sum, the findings partially support our hypothesis 1a that the motherhood wage penalty for each parity will be largest for white women, followed by Hispanic women, and smallest for Black women. Where we found statistically significant penalties, only white women with three or more children experience considerably larger penalties that Black and Hispanic women. We found little support for our hypothesis 2a, which predicted that the largest motherhood wage penalty for white women for each parity will be located later in the life course compared to the largest motherhood wage penalty for Black and Hispanic women. Rather, we found that motherhood wage penalties emerge around age 25 and peak around age 30 for all women. Our results are clearly in line with our hypothesis 3a and demonstrate that motherhood wage penalties are most persistent and extend over the longest age range for white women, from age 25-40, compared to Black and Hispanic women.
Fatherhood Premiums over the Life Course by the Number of Children
Analogous to Figure 3, Figure 4 shows unadjusted (bottom) and adjusted (top) fatherhood premiums for Black, Hispanic and white men by the number of children (see also the online supplement). For men of all racial backgrounds unadjusted fatherhood premiums were concentrated in very early adulthood, dissipated quickly and were statistically insignificant by age 25. The unadjusted wage premium for white fathers with two children between age 20 and 25 notably surpassed Black and Hispanic fathers with two children. White men with two children earned over 45 percent more than childless white men. In contrast, Hispanic fathers with two children earned up to 27 percent more than childless Hispanic men, and Black fathers with two children earned roughly 15 percent more. However, the unadjusted wage premiums across racial groups became similar within a few years and dissipated by age 30 for Black and white fathers and by age 35 for Hispanic fathers. Once adjusted for human capital and other factors, only white fathers with one or two children earned more than childless white men, between 8 and 12 percent around age 20. After age 24, there were no statistically significant fatherhood wage premiums and no differences across racial groups.
Figure 4: Fatherhood wage gaps, ages 20-45 by race and number of children, unadjusted (top) and adjusted (bottom)
Unadjusted fatherhood premiums were largest for white men and considerably lower for Black and Hispanic fathers. However, adjusted fatherhood premiums in early adulthood were similar in magnitude for all racial groups irrespective of the number of children. Mapping fatherhood premiums across the life course put them into perspective in important ways. The fact that they were concentrated exclusively in very early adulthood suggest that they emerge more as an artefact of differential selection into fatherhood, higher education and employment in this specific brief life course stage. Contrary to (unadjusted) motherhood penalties, there was no enduring fatherhood premium over the life course that coherently varied with the number of children.
In sum, unadjusted fatherhood premiums are in line with hypothesis 1b, which posited that the fatherhood wage premium for each parity will be largest for white men, followed by Hispanic men, and smallest for Black men. We find no support for hypothesis 2b, that the largest fatherhood wage premium for white men for each parity will be located later in the life course compared to the largest fatherhood wage premium for Black and Hispanic men. On the contrary, all unadjusted fatherhood wage premiums were located early in the life course. However, similar to our findings for motherhood wage penalties, results were clearly in line with hypothesis 3b that fatherhood premiums are most persistent and extend over a longer age range for white men compared to Black men with Hispanic men taking an intermediate position.
Later life consequences of parenthood wage gaps for earnings growth and wealth accumulation depend on how they are located in the life course. In this study, our goal was to describe intersectional differences in parenthood wage gaps by gender and race and their magnitude, timing and persistence across the life course depending on the number of children.
In terms of magnitude, our findings show the largest wage penalties for white mothers of many children and the largest fatherhood premiums for white men of many children. However, we find important gender and race differences in terms of life course timing and persistency. Interestingly, we find that when motherhood wage penalties and fatherhood wage premiums emerge in the life course is similar for all racial groups. However, parenthood wage gaps are limited to brief life course stages for minority men and women, but are longer-lived for white men and women. For Black and Hispanic women, motherhood wage penalties are concentrated around age 30, but do not diminish for white women. Similarly, substantial fatherhood premiums are evident longer in the early life courses of white and Hispanic men than in the lives of Black men.
Concerning racialized gender inequalities over the life course, our findings indicate that inequalities between white mothers and fathers are large early in the life course when young white men reap the benefits of fatherhood, and continue to increase as women bear the burden of motherhood. Inequalities between minority men and women seem to develop in a similar way with two important distinctions. First, differences between Black men and women, and Hispanic men and women, are smaller to begin with and increase to a lesser extent when compared to white men and women. This is because of smaller fatherhood premiums early in the life course and smaller motherhood penalties later in the life course for minority men and women. A second important distinction is that inequalities between minority men and women are shorter lived than for white men and women, mostly because motherhood penalties for Black and Hispanic women are less persistent than for white women.
Our theoretical framework largely offers consistent accounts for the gender- and race-specific differences in the magnitude, timing, and persistency of life course parenthood wage gaps that we find. Motherhood wage penalties are largest for all racial groups when childless women’s wages are growing faster than mothers’ wages. However, only among white women do childless women continue to experience faster wage growth than mothers after age 35. These women are likely employed and being promoted in occupations characterised by high skill levels and opportunities for high wage growth. Our results point to slightly higher wage growth among childless men compared to young fathers within each racial group. Our results for fathers correspond with the arguments that the fatherhood wage premium and men’s marriage premium is primarily a story of selection. These two family demographic events, marriage and fatherhood, tend to occur in a life course stage characterised by complex selection into education, employment and fatherhood, and men with high earnings potential tend to select into fatherhood and marriage.
Our theoretical framework is promising to examine life course variation in parenthood wage gaps in contexts beyond gender- and race-specific differences in the US. For example, differences among first- and second-generation migrants and native born men and women in European countries are likely also the result of a sub-group specific interplay between labour market dynamics and family demographics over the life course. A comparative perspective could also highlight an important driver of both labour market dynamics and family demographics that we were unable to incorporate in this single country study: the welfare state.