Social work is often described as a female-dominated profession. That numerical majority can make gender inequality inside the occupation less visible, particularly when inequality is measured only through hourly pay. A new study of social workers in China suggests that the more revealing differences may appear elsewhere: in working time, monthly earnings and access to organizational authority.
Using a national survey of 6,776 social workers, researchers found that mothers in the employed analytic sample worked fewer hours per week, earned less per month and had substantially lower odds of holding management or supervisory positions than unmarried childless men. Their adjusted hourly wage difference, however, was smaller and not statistically significant.
That combination matters because it changes the question from whether mothers are simply paid less for each hour worked to whether parenthood is associated with different opportunities to accumulate hours, earnings and authority inside an occupation where women are already the majority.
A female-majority profession can still have unequal authority
The study, published in Scientific Reports on 27 September 2026, examines gender and parenthood inequality within Chinese social work. The setting is analytically useful because social work is a profession in which women are strongly represented, yet representation alone does not guarantee equal career outcomes.
Research on the motherhood penalty has often concentrated on wages. Yet a worker’s economic position is produced by more than an hourly rate. Weekly working time affects monthly income. Management and supervisory roles can shape promotion, influence and future earnings. Parenthood may therefore leave a weak signal in hourly wages while still producing a stronger pattern across these other dimensions.
The researchers designed their analysis around that broader idea. Instead of treating pay as a single outcome, they examined hourly wage, weekly hours, monthly salary and two measures of organizational authority.
How the researchers compared 6,102 employed social workers
The data came from the 2019 China Social Work Longitudinal Study, which included 6,776 social workers. For the labour-outcome models, the researchers used a common analytic sample of 6,102 respondents with the required information. Using the same sample across outcomes made it easier to compare patterns without changes in sample composition driving apparent differences.
Participants were classified into six gender, marriage and parenthood groups. Unmarried childless men served as the reference category. The analysis then estimated ordinary least squares models for log hourly wage, weekly working hours and log monthly salary, while logistic regression was used for management and supervisory roles.
The adjusted models accounted for a range of individual and employment characteristics intended to separate family-status patterns from other observable differences among workers. The authors also ran robustness checks, including winsorized specifications designed to reduce the influence of extreme values and models using single-year age controls.
This remains an observational, cross-sectional analysis. The models can identify adjusted associations in the 2019 workforce, but they cannot show that becoming a parent caused a particular worker’s hours, salary or authority to change.
The clearest gap appeared in working time and monthly salary
After adjustment, mothers worked 1.31 fewer hours per week than unmarried childless men, a difference that was statistically significant at p < .001. In an occupation where salaries can depend partly on time worked and opportunities to take on demanding roles, even a modest weekly gap can accumulate across a month or year.
Monthly salary showed a corresponding disadvantage. The coefficient for mothers was -0.066 on the log salary scale, again statistically significant at p < .001. A log-point coefficient of this size represents a modest but meaningful earnings difference after the model’s included characteristics are taken into account.
Hourly wages told a different story. The adjusted coefficient for mothers was -0.024 and was not statistically significant, with p = .240. In other words, the study did not find clear evidence that mothers received a lower hourly wage than the reference group once the included controls were applied.
This divergence is central to the paper. Looking only at hourly wages could lead to the conclusion that there is little parenthood-related inequality. Looking at total working time and monthly earnings reveals a more consistent disadvantage.
Who mothers were compared with changed part of the wage story
The researchers did not rely only on the unmarried childless male reference category. Direct comparisons among female groups showed that mothers did not have a statistically significant hourly-wage difference relative to unmarried childless women. A difference did appear relative to married childless women.
That nuance prevents the results from being reduced to a simple claim that motherhood lowers hourly pay. Marriage, parenthood, gender and career stage overlap, and the apparent size of a gap depends partly on which comparison group is used.
The hourly-wage results also showed some sensitivity in robustness checks using winsorization and more detailed age specifications. The broader pattern in working time, monthly earnings and organizational position was more consistent than the hourly-wage result.
The authority gap was larger than the hourly-pay gap
The most striking findings concerned organizational authority. Mothers had 38% lower odds of occupying a management role than unmarried childless men, corresponding to an odds ratio of 0.62. Their odds of holding a supervisory role were 48% lower, with an odds ratio of 0.52.
These outcomes capture something wages alone cannot. Management and supervisory positions influence who makes decisions, allocates work, mentors colleagues and gains experience that may support later promotion. A profession can therefore have a large female workforce while still distributing authority unevenly within that workforce.
The pattern for fathers was notably different. Fathers did not differ significantly from married childless men in the same way. This asymmetry is consistent with the broader idea that parenthood can interact with gendered expectations about paid work and caregiving, although the cross-sectional design cannot establish the mechanism responsible for the observed differences.
Why fewer hours can matter even when hourly pay looks similar
Hourly wages are attractive to researchers because they appear to compare like with like. But hourly rates can miss inequalities that operate through access to working time. If one group works fewer paid hours, monthly and annual earnings can diverge even when hourly rates are similar.
Working time can also interact with promotion. Management positions may demand longer or less predictable schedules. Workers carrying greater unpaid caregiving responsibilities may be less able to accept those demands, may choose jobs with more predictable hours, or may be evaluated differently by organizations. The present study cannot distinguish among those possibilities, but its multidimensional design shows why they matter.
The findings therefore point toward a form of inequality that is distributed across several connected outcomes rather than concentrated in a single wage coefficient. Mothers’ hourly wage was not clearly lower, yet their working time, monthly salary and organizational authority were all less favourable relative to the male reference group.
The results should not be read as a causal motherhood penalty
The study has several important limitations. Most importantly, the data are cross-sectional. Researchers observed respondents at one point in 2019 rather than following the same workers before and after marriage or childbirth. Differences between groups may therefore reflect selection, career stage, preferences, workplace conditions or unmeasured characteristics as well as parenthood itself.
The sample also contains people who were currently employed in social work. It cannot capture workers who left employment or moved out of the occupation because of family responsibilities. If parenthood affects whether people remain in social work at all, an employed-worker sample may understate part of the broader inequality.
Nor can the study determine why mothers worked fewer hours or were less likely to hold authority. Reduced hours could reflect personal preferences, constrained choices, organizational practices, caregiving demands or combinations of these factors. Management gaps could arise from promotion processes, job selection, tenure pathways or workplace expectations. The statistical associations identify where disparities exist, not the causal chain that produced them.
The research is also specific to Chinese social work and to the institutional conditions captured in 2019. Social work systems, family policy, childcare access, employment contracts and promotion structures differ across countries and occupations. The exact coefficients should not be assumed to generalise universally.
What employers and policymakers can take from the study
The practical implication is that organizations assessing gender equality may need to look beyond equal hourly pay. A workplace can show relatively small adjusted differences in hourly rates while still producing substantial gaps in monthly earnings and leadership access.
For social-service organizations, that means monitoring who receives management and supervisory opportunities, how working hours are structured, whether leadership tracks require schedules that systematically disadvantage caregivers, and whether flexible arrangements preserve or weaken promotion prospects.
The findings also complicate the assumption that female representation is itself evidence of workplace equality. Women can constitute the numerical majority of an occupation while mothers within that occupation remain less represented in positions of authority.
Future longitudinal research could follow social workers across marriage, childbirth and career transitions to test whether hours, earnings and authority change within the same individuals. Linking employment histories to childcare arrangements, organizational policies and promotion decisions would also help distinguish voluntary changes from constrained ones.
For now, the study’s strongest conclusion is narrower but still important: among currently employed Chinese social workers, parenthood-related inequality was more consistent in working time, monthly earnings and organizational position than in hourly wage rates. Measuring only the hourly rate would have missed much of that pattern.
Source Information
Study: Gender, parenthood, pay, working time, and organizational authority among social workers in China
Authors: Congde Xu, Yang Wang, Jinlin Pang and Zhen Li
Journal: Scientific Reports
Published: 27 September 2026
DOI: 10.1038/s41598-026-69343-3
Study type: Cross-sectional analysis of the 2019 China Social Work Longitudinal Study using OLS and logistic regression models
Sample: 6,776 surveyed social workers, with 6,102 respondents in the common labour-outcome analytic sample
