thesis gender gap

An official website of the United States government

Official websites use .gov A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

• Publications
• Account settings
• Advanced Search
• Journal List

The Gender Pay Gap: Income Inequality Over Life Course – A Multilevel Analysis

Lisa toczek, richard peter.

• Author information
• Article notes
• Copyright and License information

Edited by: Andrzej Klimczuk , Warsaw School of Economics, Poland

Reviewed by: Kristin Farrants , Karolinska Institutet (KI), Sweden

Maria Bohdalova , Comenius University in Bratislava, Slovakia

*Correspondence: Lisa Toczek, [email protected]

This article was submitted to Work, Employment and Organizations, a section of the journal Frontiers in Sociology

Received 2021 Nov 15; Accepted 2021 Dec 8; Collection date 2021.

This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

The gender pay gap has been observed for decades, and still exists. Due to a life course perspective, gender differences in income are analyzed over a period of 24 years. Therefore, this study aims to investigate income trajectories and the differences regarding men and women. Moreover, the study examines how human capital determinants, occupational positions and factors that accumulate disadvantages over time contribute to the explanation of the GPG in Germany. Therefore, this study aims to contribute to a better understanding of the GPG over the life course. The data are based on the German cohort study lidA (living at work), which links survey data individually with employment register data. Based on social security data, the income of men and women over time are analyzed using a multilevel analysis. The results show that the GPG exists in Germany over the life course: men have a higher daily average income per year than women. In addition, the income developments of men rise more sharply than those of women over time. Moreover, even after controlling for factors potentially explaining the GPG like education, work experience, occupational status or unemployment episodes the GPG persists. Concluding, further research is required that covers additional factors like individual behavior or information about the labor market structure for a better understanding of the GPG.

Keywords: employment biographies, gender inequality, income trajectories, LidA-study, growth curve analysis, life course perspective, trajectories of labor market factors

1 Introduction

In the European Union (EU) in 2019, women’s average gross hourly earnings were 14.1% below the earnings of men ( Eurostat, 2021a ). The gender pay gap (GPG) has existed for decades and still remains to date. According to Eurostat GPG statistics, the key priorities of gender policies are to reduce the wage differences between men and women at both the EU and national levels ( Eurostat, 2021a ). Nevertheless, the careers of men and women differ considerably in the labor market, with women being paid less than men ( Arulampalam et al., 2005 ; Radl, 2013 ; Boll et al., 2017 ). A report from the European Parliament in 2015 about gender equality assessed Germany’s performance in that field as mediocre. The federal government in Germany has already improved laws that focus on gender equality ( Botsch, 2015 ). Regarding Germany, in 2019 the earning difference between men and women were found to be 19.2% ( Eurostat, 2021a ). The reasons behind gender income inequality are complex and have multidimensional explanations.

1.1 Determinants of the GPG

The early 1990s represented a turning point for the participation of women in the labor market ( Botsch, 2015 ). In previous years, women’s participation rate in the workforce has strongly increased, from 51.9% in the year 1980 (West Germany) to 74.9% in 2019 ( OECD, 2021 ). This upward trend represents the increase of women working at older ages ( Sackmann, 2018 ). However, the gender income inequality remains. Different explaining factors of the GPG were found in previous research: patterns of employment, access to education and interruptions in the careers of men and women.

Although there are nearly equal numbers of men and women in the labor market, when considering women’s careers, various gender-specific barriers are occurring. The working patterns were found to have a relevant impact on the GPG in previous research. Atypical employment is increasing and this result in an expansion of the low-wage sector, which mainly affects women in Germany ( Botsch, 2015 ). Additionally, labor market integration of women has mainly been in jobs that provide few working hours and low wages ( Botsch, 2015 ). Moreover, part-time employment represents a common employment type in Germany, which is more frequent among women – as various studies have demonstrated – and explains the GPG significantly ( Boll et al., 2017 ; Ponthieux and Meurs, 2015 ; Boll and Leppin, 2015 ). In addition, the part-time employment occurs more often in occupations characterized by a high proportion of women and low wages ( Matteazzi et al., 2018 ; Boll and Leppin, 2015 ; Hasselhorn, 2020 ; Manzoni et al., 2014 ). Another employment type with few working hours and low pay is a special form of part-time work: marginal work. Marginal work is defined as earnings up to 450 Euros per month or up to 5.400 Euros annually. Also, it is also more common among women than among men ( Botsch, 2015 ; Broughton et al., 2016 ). The marginal part-time work has increased in nearly all EU countries, especially in Germany where it can be found to be above the EU average ( Broughton et al., 2016 ). Besides the working time, occupational status influences the wage differences of men and women. Female-dominated occupational sectors are characterized by lower wages compared to male-dominated ones ( Brynin and Perales, 2016 ). Additionally, in women-dominant industries, remunerations are less attractive and it often entails low-status work in sectors like retail, caregiving or education ( Boll and Leppin, 2015 ; Hasselhorn, 2020 ; Matteazzi et al., 2018 ; Brynin and Perales, 2016 ). Hence, working patterns such as the amount of working time or the occupational status are crucial determinants that contribute to explaining the GPG in Germany ( Blau and Kahn, 2017 ; Boll et al., 2017 ).

The access to education and vocational training are important factors, that influence the GPG. Both influence a first access to the labor market and are considered to be ‘door openers’ for the working life ( Manzoni et al., 2014 ). In Germany, education represents a largely stable variable over time, i.e. only few individuals increase their first educational attainment. Education influences the careers of men and women and can be seen as important an determinant of future earnings ( Boll et al., 2017 ; Bovens and Wille, 2017 ). Although women’s educational attainment caught up with those of men’s in recent years, for men, a higher qualification was still rewarded more than for women ( Botsch, 2015 ; Boll et al., 2017 ). Moreover, in previous research the impact of education on the GPG was not found to be consistent with different influences for men than for women ( Aisenbrey and Bruckner, 2008 ; Ponthieux and Meurs, 2015 ). Manzoni et al. (2014) found out, that the effect of education on career developments were dependent of their particular educational levels. In addition, regardless of the women’s educational catching-up in the last years, looking at older cohorts – born between 1950 and 1964 – women had a lower average level of education than men ( Boll et al., 2017 ).

An increasing GPG over time can also be the result of interruptions in careers, which are found more often for women than for men ( Eurostat, 2021a ; Boll and Leppin, 2015 ). Previous research of Boll and Leppin (2015) has identified explanations for the GPG in Germany by analyzing data from the German Socio-Economic Panel (SOEP) in 2011. They demonstrated that the amount of time spent in actual work was lower for women than for men. Therefore, women gain less work experience than their male counterparts ( Boll and Leppin, 2015 ). Career interruptions not only impact the accumulation of work experience but also the scope of future work. Especially in the period of family formation higher rates of part-time employment among women can be observed ( Boll et al., 2017 ; Ponthieux and Meurs, 2015 ). Moreover, work-life interruptions such as raising children or caring for family members have a major impact on the employment development and are more likely to appear for women than for men ( Ponthieux and Meurs, 2015 ). Although the employment rate of mothers has increased in recent years in Germany, it is still considerably lower than that of fathers ( Federal Statistical Office, 2021 ). Hence, taking care of children is still attributed to mothers, to the detriment of their careers ( Botsch, 2015 ). A recent study, however, found sizable wage differences between men and women who were not parents, refuting the assumption that the GPG applies only to parents ( Joshi et al., 2020 ). Other interruptions in the working lives of men and women are caused by unemployment. Azmat et al. (2006) found that in Germany, transition rates from employment to unemployment were higher for women than for men. Career interruptions have lasting negative effects on women’s wages. Therefore, it can be useful to examine unemployment when analyzing gender inequality in the labor market ( Eurostat, 2021b ).

1.2 Theoretical Background

1.2.1 human capital model.

In previous research, economic theories had been applied to explain the income differences of men and women. Two essential factors could be found: qualification and discrimination. The human capital model claims that qualifications with greater investments can be directly related to higher wages of men and women. The earnings are assumed to be based on skills and abilities that are required through education and vocational training, and work experience ( Grybaitė, 2006 ; Lips, 2013 ; Blau and Kahn, 2007 ). Educational attainment of women has caught up in recent years ( Botsch, 2015 ). However, women’s investments in qualifications were still not equally rewarded as those of men. Therefore, the expected narrowing of the GPG was not confirmed in earlier research ( Boll et al., 2017 ; Lips, 2013 ). Another determinant of the human capital model is work experience. Labor market experience contributes to a large extent to the gender inequality in earnings ( Sierminska et al., 2010 ). Hence, work experience influences the wages of men and women. On the one hand, interruptions due to family life lower especially women’s labor market experience compared to men. On the other hand, part-time employment is more frequent among women with fewer working hours and therefore less work experience. The lesser accumulation of work experience leads to lower human capital and lower earnings for women compared with men ( Blau and Kahn, 2007 ; Mincer and Polachek, 1974 ). Nonetheless, the association of work experience and income is more complex. Regarding the wages of men and women the influence of occupation itself also needs to be considered ( Lips, 2013 ). In the paper of Polachek (1981) different occupations over the careers of men and women were explained by different labor force participation over lifetime. Referring to the human capital model, it is argued that women more likely expect discontinuous employment. Therefore, women choose occupations with fewer penalties for interruptions ( Polachek, 1981 ). However, it should be questioned if working in specific occupations can be defined as a simple choice ( Lips, 2013 ). Besides, part-time employment is found to be more frequent among women, which ultimately leads to few working hours and hence low earnings ( Botsch, 2015 ; Ponthieux and Meurs, 2015 ; Boll et al., 2017 ). Though different working hours cannot be defined as a simple choice either ( Lips, 2013 ).

Earlier criticism about the human capital model discussed that the wage differences of men and women cannot only be explained by the qualification and the labor market experience ( Grybaitė, 2006 ; Lips, 2013 ). Another theoretical approach explaining the GPG refers to labor market discriminations, which effect occupations and wages ( Boll et al., 2017 ; Grybaitė, 2006 ). On the one hand, occupational sex segregation can be associated with income differences of men and women. The different occupational allocation in the labor market of men and women are defined as allocative discrimination ( Petersen and Morgan, 1995 ). In addition, occupations in female-dominated sectors are mostly characterized by low-wages compared to more male-dominated occupations ( Brynin and Perales, 2016 ). On the other hand, even with equal occupational positions and skill requirements women mostly earn less than men, this refers to the valuative discrimination ( Petersen and Morgan, 1995 ). Even within female-dominated jobs a certain discrimination exists, with men being paid more than women for the same occupation. Additionally, employment sectors with a large number of female workers are more likely to be associated with less prestige and lower earnings ( Lips, 2013 ). Achatz et al. (2005) analyzed the GPG with an employer-employee database in Germany. The authors examined the discrimination in the allocation of jobs, differences in productivity-, and firm-related characteristics. They found out that in occupational groups within companies, the wages decreased with a higher share of women in a group. Additionally, a higher proportion of women in a groups resulted in a higher wage loss for women than for men ( Achatz et al., 2005 ).

Although relevant criticism of the human capital model exists, its determinants are still found to be important in explaining the wage differences of men and women ( Boll et al., 2017 ). Nonetheless, income differences of men and women can still be found even with the same investments in human capital. The reason for this could be the occupational discrimination of women ( Brynin and Perales, 2016 ; Achatz et al., 2005 ; Lips, 2013 ). Therefore, the occupational positions can be associated as a relevant factor of the GPG.

1.2.2 Life Course Approach

Besides economic theories, there are other theoretical approaches of explaining the GPG. One of them focusses on the accumulation of disadvantages over the life course: the ‘cumulative advantage/disadvantage theory’ by Dannefer (2003) . It also involves social inequalities which can expand over time. The employment histories of men and women evolve over their working lives and during different career stages, advantages and disadvantages can accumulate. First, this life course perspective considers and underlines the dynamic approach of how factors shape each individual life course. Secondly, it can contribute to explain the different income trajectories of men and women over their working lives ( Doren and Lin, 2019 ; Dannefer, 2003 ; Härkönen et al., 2016 ; Manzoni et al., 2014 ; Barone and Schizzerotto, 2011 ).

The importance of the life course perspective was underlined by some earlier studies. They demonstrated that certain conditions in adolescence or early work-life affected future careers of men and women. Visser et al. (2016) found evidence for an accumulation of disadvantages in the labor market over working life, in particular for the lower educated. The cohort study SHARE had assessed economic and social changes over the life course in numerous European countries in several publications ( Börsch-Supan et al., 2013 ). Overall, education and vocational training, occupational positions and income illustrate parts of the social structure which in turn can demonstrate gender inequality in the labor market ( Boll and Leppin, 2015 ; Hasselhorn, 2020 ; Du Prel et al., 2019 ). Moreover, family events and labor market processes repeatedly affect one another over the life course. The work-family trajectories have consequences on employment outcomes such as earnings ( Aisenbrey and Fasang, 2017 ; Jalovaara and Fasang, 2019 ). Furthermore, the income differences of men and women are not steady but tend to be lower at the beginning of employment and increase with age ( Goldin, 2014 ; Eurostat, 2021a ). Therefore, careers should not be analyzed in a single snapshot, but with a more appropriate life course approach that takes into account factors that influences the wages of men and women over time.

1.3 Aim and Hypotheses

The aim of the present study is to examine income trajectories and to investigate the income differences of men and women over their life course. We are interested in how human capital determinants, occupational positions and the accumulation of disadvantages over time contribute to the explanation of the GPG from a life course perspective.

Focusing on older German employees, our study includes 24 years of their careers and considers possible cumulative disadvantages of women in the labor market compared to those of men. In contrast to Polachek (1981) , who analyzed the GPG as a unit over lifetime, we used a life course approach in regard to the theory of cumulative disadvantages of Dannefer (2003) . Accordingly, we analyze explaining factors of the GPG not only in a single snapshot but over the working careers of men and women. Life course data based on register data and characteristics of employment biographies with information on a daily basis are two additional important and valuable advantages of our study. Existing studies rarely have this information in the form of life course data and when they do, the data is either self-reported and retrospective including possible recall bias, or based on register data which was only collected on a yearly basis. We expect to find differences in the income of men and women over a period of time with overall higher, and more increasing earnings of men than of women.

Hypothesis 1 (H1): The differences of income trajectories throughout working life is expected to demonstrate more income over time among men than among women.

Education and vocational training, and work experience are human capital determinants. They have influence on the earnings of men and women. Although previous research estimated additional important factors contributing to the GPG, human capital capabilities continue to be relevant in explaining the wage differences of men and women ( Blau and Kahn, 2007 ; Boll et al., 2017 ). In our life course approach, we control for human capital determinants due to the information about education and vocational training, and work experience via the amount of working time (full-/part-time) for each year. We expect to find a strong influence of both determinants on the wages of men and women in Germany.

Hypothesis 2 (H2): The income differences between men and women can be explained by determinants of the human capital model.

Previous research found out that factors such as occupational status had an impact on the income differences of men and women ( Blau and Kahn, 2007 ; Boll et al., 2017 ). For a better understanding and explanation of the GPG, gender differences regarding occupational positions must be included to human capital determinants ( Boll et al., 2017 ). We assume that men and women can be found in different occupations, measured via occupational status, and these explain a substantial part of the wage differences between men and women.

Hypothesis 3 (H3): The occupational status of men and women can contribute to the explanation of the GPG.

The life-course approach acknowledges time as an important influence on the wages of men and women. Income differences of men and women can change over time and career stages, while the GPG was found to be lower at the beginning of the employment career and widened with age ( Goldin, 2014 ). Hence, the earning differences between men and women tend to be higher for older employees ( Eurostat, 2021a ; Federal Statistical Office, 2016 ). To account for the influence of age, we additionally included the age of each person in our analysis. Another factor that changes over time and contribute to explain the GPG is part-time work. In general, part-time work result in a disadvantage in pay compared to full-time employment ( Ponthieux and Meurs, 2015 ). However, explanations of the GPG due to different amount of part-time work need to include a special form of part-time work: marginal work. Marginal employment conditions are characterized by low wages and high job insecurities. Also discontinuous employment due to unemployment are characterized by job insecurities and affect the low-paid sector – therefore mainly women ( Botsch, 2015 ). Besides the human capital determinants and occupational positions as important factors explaining the GPG, the region of employment influences the wages of men and women and can also change over the career stages. Evidence from the Federal Statistical Office of Germany in 2014 noticed a divergence of the GPG trend in the formerly separated parts of Germany. The GPG among employees was wider in the Western part (24%) compared to the Eastern part of Germany, where it was found to be 9% ( Federal Statistical Office, 2016 ). Therefore, to examine income differences, the amount of less advantaged employment such as marginal work or periods of unemployment throughout the careers of men and women needs to be considered, as well as the region of employment and the age of a person.

Hypothesis 4 (H4): Factors of the living environment such as regional factors, and social disadvantage work conditions such as marginal work or unemployment, contribute to the income difference between men and women.

Our study about the GPG in Germany adds to earlier research in different ways. First, the accumulation of inequalities over the life course of men and women is known, but only few studies exist that focus on income through life course approach. We can analyze factors that influence the GPG over the careers of men and women due to the availability of social security data with daily information of each person. Besides the wages of men and women, the data additionally contains time-varying information about occupational status, working time and unemployment breaks. Therefore, we use longitudinal data of the German baby-boomers which allow us to measure changes of factors explaining the GPG over time. Second, a relevant contribution of our study is that we can consider different factors contributing to the explanation of the GPG through a life course perspective. The few studies focusing on the GPG over life course included either only determinants of the human capital model ( Joshi et al., 2020 ) or factors of occupational careers ( Moore, 2018 ). Some research included both aspects but had other disadvantages, such as Monti et al. (2020) , who could not analyze temporal evolution of the GPG with the data available. Moreover, previous research on the GPG in Germany could not trace vertical occupational segregation due to missing information of part-time workers, included only data of West Germany and used merely accumulated earnings over time ( Boll et al., 2017 ). Nonetheless, previous research demonstrated the need of analyzing the GPG via life course approach with which the accumulation of advantages and disadvantages for both, men and women, can be considered. Third, due to the usage of a multilevel framework we can examine income trajectories simultaneously at an individual and at a time-related level. Moreover, the influences of time-invariant and time-varying factors can be analyzed regarding differences in earnings of men and women. Hence, the multilevel approach examines income changes between and also within individuals. Furthermore, it acknowledges the importance of the life course perspective with including time as a factor in the model. A recent study also used growth curve modelling to explain gender inequality in the US. However, gender inequality measured through gender earnings was analyzed only across education and race without considering other variables explaining the GPG ( Doren and Lin, 2019 ). To our knowledge, there exists no research on the GPG that covers several essential determinants, hence we aim to fill those research gaps with our study.

2 Materials and Methods

The data were obtained from the cohort study lidA (living at work). The lidA sample includes two cohorts of employees (born in 1959 and in 1965) and was drawn randomly from social security data. LidA combines two major sources of information – register data of social insurance and questionnaire data derived from a survey. The survey was conducted in two waves, 2011 (t 0 ) and 2014 (t 1 ) ( Hasselhorn et al., 2014 ). The ethics commission of the University of Wuppertal approved the study.

In Germany, the social insurance system assists people in case of an emergency such as unemployment, illness, retirement, or nursing care. Employees have to make a contribution to the system depending on their income – except of civil servants or self-employed ( Federal Agency for Civic Education, 2021 ). In our analyses, we included men and women in Germany who participated in the baseline (2011) and in the follow-up (2014), were employed during both waves and subjected to social security contributions. We only included persons who agreed via written consent to the linkage of the survey data to their social security data. Thus, our sample for analysis included 3,338 individuals ( Figure 1 ).

Decision tree – inclusion and exclusion criteria in the sample for analysis.

2.2 Measurements

The social security data of the Institute for Employment Research of the German Federal Employment Agency is based on employers’ reports. The so-called “Integrated Employment Biographies” (IEB) or register data comprises information about individual employment; that is, type of employment, occupational status, episodes of unemployment and income with information about age, gender and education and vocational training. The IEB data are retrieved from employers’ yearly reports submitted to the social security authority ( Hasselhorn et al., 2014 ). The information of the register data was available on a daily basis and contained yearly information from 1993 to 2017 for each person. However, the IEB data contain missing details, especially regarding information that is not directly relevant for social security data and therefore, not of the highest priority for employers’ reports. This is particularly true for data on gender and education and vocational training. As our sample participants consented to the linkage of IEB with questionnaire data, we were able to impute the missing information on these variables with the help of the survey data. All time-varying information in the IEB is coded to the day. Our data have a multilevel structure with time of measurements (Level 1) being nested within individuals (Level 2) and defined as follows.

2.2.1 Level 1 Variables

In our analysis the variable time was based on information about the year of measurement. The starting point represents 1993 and was coded with zero. The outcome variable income was calculated from the IEB data as nominal wages in Euros (€). As time-varying variable, it can be defined as the average daily income per year of each person whose work contributes to social security and/or marginal employment. Information about the work experience due to working time was available for jobs that require social security contribution. To draw this information from the IEB data, the time-varying variable working time was computed with three different types: full- and part-time, part-time, and full-time. The data on occupational status were based on the International Standard of Classification of Occupations 2008 (ISCO-08). This time-varying variable contained information on the occupational status of each job that a person has held over the years. For the multilevel analysis, ISCO-08 was transformed from the German classification KldB 2010 (classification of occupations 2010) of the register data. ISCO-08 is structured according to the skill level and specialization of jobs, which are grouped into four hierarchical levels. Occupational status in our study was defined by the 10 major groups (level one of the classifications ISCO-08), without the group of armed forces who did not appear in our data. Therefore, the nine groups were analyzed: elementary occupations; plant and machine operators and assemblers; craft and related trades workers; skilled agricultural, forestry and fishery workers; services and sales workers; clerical support workers; technicians and associate professionals; professionals; and managers ( International Labour Office, 2012 ). Moreover, information about the number of episodes of marginal work could also be drawn from the register data. Marginal work was defined due to having at least one marginal employment per year. The time periods (episodes) of every marginal employment were counted and added up yearly. Furthermore, the duration of unemployment as time-varying variable was calculated due to information of the register data about the days of unemployment per year. In the register data unemployment is defined as being unemployed or unable to work for up to 42 days, excluding those with sickness absence benefits or disability pensions. The IEB data also provided information on the region of employment, which represents the area in which a company is located (East Germany and West Germany). This time-varying variable was available for each person over the years. A description of the Level 1 characteristics of our sample is provided in Table 2 using the last available information (2017) from the IEB data.

Characteristics of Level 1 variables a for men (n = 1,552) and women (n = 1,786).

M mean; SD standard deviation.

The database of the variable is provided by IEB, data in 2017.

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

2.2.2 Level 2 Variables

Information about the time-invariant variable education and vocational training was assessed from the survey data in 2011 (baseline). Education and vocational achievements of the sample were grouped in: low, intermediate and high education and vocational training (see Supplementary Table S1 ). The time-invariant variable gender had missing values in the register data. Therefore, we imputed the missing data using information of the survey data. The variable was coded 0 = female and 1 = male. Also based on the survey data, we included the time-invariant variable year of birth with measurements of 1959 and 1965 in the analysis. The characteristics of the Level 2 variables are displayed in Table 1 .

Characteristics of the Level 2 variables a for men (n = 1,552) and women (n = 1,786).

The database of the variable is provided by survey data in 2011.

2.3 Statistical Analysis

The characteristics of our sample are displayed in Table 1 and Table 2 . Statistical analyses were performed using either Cramer’s V or by unpaired two sample t -test for numeric variables. Regarding the multilevel analysis, we used a so-called growth curve analysis. It demonstrates a multilevel approach for longitudinal data that model growth or decline over time. For this purpose, all daily information in the IEB were transformed into data on a yearly basis. Level 1 (year of measurements) represents the intraindividual change with time-varying variables. Interindividual changes are determined with time-invariant variables on Level 2 (individuals). Therefore, time of measurements predictors was nested within individuals. We applied a random intercept and slope model, which assumed variations in intercept and slope of individuals over time ( Singer and Willett, 2003 ; Rabe-Hesketh and Skrondal, 2012 ; Hosoya et al., 2014 ). Besides the Level 1 and Level 2 predictors, the cross-level interaction of gender*time interaction was constituted to analyze differences in income slopes of men and women over time ( Rabe-Hesketh and Skrondal, 2012 ).

Level 1 of the two-level growth model is presented below ( Eq. (1) ). y i j measures the income trajectory y for individual i at time j . True initial income for each person is represented with β 0 i . The slope of the individual change trajectory demonstrates β i j . T I M E i j stands for the measure of assessment at time j for individual i (Level 1 predictor). The residual or random error, specific to time and the individual is demonstrated by ε i j .

Eq. 2 and 3 represent the submodels of the Level 2. Eq. 2 defines the intercept γ 00 for individual i with the intercept of z i (illustrating a Level 2 predictor) and residual in the intercept v 0 i . The slope at Level 2 is represented in Eq. 3 with γ 10 and the slope error v 1 i . The effect γ 11 provides information on the extent to which the effect of the Level 1 predictor ( T I M E i j ) varies depending on the Level 2 predictor ( z i ).

To test our hypotheses, we calculated the influence of different variables with adjusting various predictors stepwise into the multilevel analysis. First, we estimated an unconditional means model which describes the outcome variation only and not its change over time (model 1). The next preliminary step was calculating the intraclass correlation coefficient (ICC) of this model 1. It identifies and partitions the two components: within- and between-person variance. The ICC estimates the proportion of total variation of the outcome y that lies between persons ( Singer and Willett, 2003 ). In the next model (model 2), we calculated an unconditional growth curve model which included time as predictor on Level 1. In model 3, the GCA was controlled for gender and time as well as the interaction of both variables. Model 4 was additionally adjusted for human capital determinant: education and vocational training, and working time. The GCA of model 5 was controlled for occupational status. The last model included year of birth, number of episodes of marginal work, duration of unemployment and region of employment (model 6 – fully adjusted model).

In Table 5 , the indices of the Akaike’s Information Criterion (AIC) were used to compare models and explore the best model fit ( Singer and Willett, 2003 ; Rabe-Hesketh and Skrondal, 2012 ). The statistical analyses were performed with IBM SPSS 25.

Goodness-of-fit statistics of the GCA.

AIC Akaike’s Information Criterion.

Unconditional means model. Not displayed in detail.

Unconditional growth model. Not displayed in detail.

Model controlled for gender, year and the interaction gender*year. See Table 3 .

Model additionally adjusted for education and vocational training, and working time. See Table 4 .

Model additionally adjusted for occupational status. See Table 4 .

Model additionally adjusted for year of birth, number of episodes of marginal work, duration of unemployment (days per year) and region of employment. See Table 4 .

3.1 Descriptive

Characteristics of Level 2 variables stratified by gender are displayed in Table 1 . 1,552 men and 1,786 women were included in the analyses. It is observed that women significantly differ from men in education and vocational training. Women were less likely than men to have both low and high levels of education and vocational training.

The characteristics of Level 1 variable are represented in Table 2 . Men and women differ significantly in their occupational positions. Also, men had a higher average daily income than women. Part-time jobs are more likely among women as compared to men, who are more likely to be represented in full-time jobs. Moreover, the numbers of episodes of marginal work differ significantly between men and women.

Figure 2 displays the income trajectories over the observation period (1993–2017) among men and women. In 24 years, average daily income per year increased for both. However, men have a higher average income over their life course than women. Over time, a steeper growth of the average daily income per year can be observed for men, compared to the income development of women.

Income trajectories of men and women.

3.2 Growth Curve Analysis

Results of the multilevel analyses with average daily income per year as dependent variable concerning H1 are presented in Table 3 . The ICC of the unconditional means model (model 1) demonstrates that 74% of the total variability in income can be attributed to differences between persons and 26% to the differences within persons. Adding time as a predictor in the multilevel analysis (model 2), the variance components on Level 1 become smaller. Concluding that time accounts for 68% (from 607.34 to 197.12) of the within-person variance in average income. On Level 2, time explains 40% of the variance between persons (interindividual). However, there can be still found significant unexplained results in both levels which suggests that predictors on both levels should be further included. The GCA in model 3 was adjusted for gender (with women as reference group) and the interaction gender*time. The results show a significant effect of gender on the average income over time. The starting place (intercept) lies at 41.74€ with an incremental growth per year of 1.76€. However, regarding women as reference group, men have a higher average income. The significant interaction term also indicates different income development of men and women over time – with men having higher average income trajectory than women. As expected, no relevant change can be found in the within-person variance due to the adding of the Level 2 variable: gender. The variance on Level 2, however, become less concluding that gender accounts for 26% of the variance between persons. Overall, we can verify H1 with these results.

Growth curve models 1 to 3: Estimates of average daily income per year.

L1 = Level 1; L2 = Level 2.

Unconditional Means Model.

Unconditional Growth Model.

Model controlled for gender, time, and interaction gender by time.

Results of the GCA with average daily income per year as the dependent variable controlled by determinants of the human capital model are presented in Table 4 (model 4). In addition to the multilevel analysis of model 3, model 4 is also adjusted for: education and vocational training, and working time. The results show that the average income is found to be significantly higher for full-time workers and higher educated. There is a social gradient for income regarding education and vocational training – with decreasing levels of education, the income also reduces. People who are working full-time have a higher average income than those who work part-time or full- and part-time. The effect of gender is found to be significant with less average income of women compared to men. Moreover, the income development of men and women over time is still significantly different, with more income growth over time for men than for women. The results of the variance components demonstrate that human capital determinants are explaining 16% of the variance within person and 25% of the variance between persons. However, on both levels there can be still found significant variance and additional variables need to be considered. Our hypothesis 2 can be partially confirmed.

Growth curve models 4 to 6: Estimates of average daily income per year.

Model additionally adjusted for education and vocational training, and working time.

Model additionally adjusted for occupational status.

Model additionally adjusted for year of birth, marginal work, duration of unemployment and region of employment.

Model 5 ( Table 4 ) embeds occupational status to the analysis to find out the contribution of the occupational positions on the earning differences of men and women. Significant differences in the daily average income for each occupational group can be identified. The reference group is represented with the highest occupational group ‘manager’. In nearly all other occupations, manager had the highest average income, except of ‘technicians and associate professionals’. Moreover, the effects of occupational status on income are significant for all ISCO groups except for professionals. However, compared to education and vocational training, occupational status trends are less clear, and a social gradient cannot be identified. The estimated of the fixed effect of gender persists and stays the same, concluding that the occupational position of a person could not influence the effect of gender on income. The increase of income over time can be still found to be significant higher for men than for women. Moreover, including the Level 1 variable, occupational position cannot explain a substantial part of the within-person variance. We can identify occupational positions as significant predictor of the income, but a relevant contribution to explain the GPG cannot be observed. Therefore, we cannot approve hypothesis 3.

The results of investigating the influence of factors of the living environment are presented in Table 4 (model 6). Those, who are born earlier (1959) are found to have a higher average daily income, compared to those born in 1965. Having at least one marginal employment per year influences the average daily income negatively, as does having more unemployed days. Furthermore, average income is influenced by the region of employment, being lower in East Germany than in West Germany. The estimate of gender become a little less, but the average income and the development of income over time still substantially differs between men and women. The factors of living environment account for 10% of the variance between persons. We can only partially accept hypothesis 4.

3.3 Goodness of Fit

Table 5 displays the goodness of fit statistics for the different models of the GCA. The AIC is computed to find the best model fit. Considering the different indices of AIC, model 6 has the best fit.

4 Discussion

This study aimed to examine the income differences of men and women over their life course. We investigated how different factors can explain the GPG over time. Even after extensive control for human capital determinants, occupational factors and various factors of the living environment, the effect of gender on the average daily income persisted. Moreover, the average income development was found to be higher for men compared to women.

The accumulation of inequalities over time can be seen in the difference between men’s and women’s wages. Over the period of 24 years, our results showed that the income development of men increased more compared to women – the GPG widened with time. Due to the availability of life course data, we could consider cumulative disadvantages regarding the earnings of men and women. Moreover, the results of the variance componence also showed the importance of including time to explain the GPG ( Table 3 , model 2). Therefore, we can verify our first hypothesis. The steeper incline of income for men compared to women over time substantiates the presence of GPG in Germany. Goldin (2014) also found a small GPG when people enter the labor market and a widening gap with age. Our findings are also in line with information from the Federal Statistical Office (2016) and Eurostat (2021a) who used representative data and not use cohort specific data of the German working population.

The second hypothesis assumed that human capital determinants (education and work experience) can explain the GPG. The effects of education and vocational training on daily average income significantly differed in our results ( Table 4 , model 4). Findings of Bovens and Wille (2017) also demonstrated that the level of a person’s education determines the income level. Our results also support the previous finding, that education is most often a requirement for the achievement of a certain desired financial situation ( Du Prel et al., 2019 ). Our results also showed that the average income significantly differed considering working time. Full-time workers had higher average income, while men were more likely to work full-time compared to women. Earlier research also showed that part-time work was more frequent among women than among men ( Boll and Leppin, 2015 ; Matteazzi et al., 2018 ; Eurostat, 2021a ). After adjusting for human capital determinants, the unexplained variance was still substantial and the effect of gender remained significant. Hence, H2 can only partially be accepted.

In our third hypothesis, we assumed that the gender differences in occupational position can explain the GPG. We demonstrated that the average income differed according to the occupational status of a person. This is in line with previous findings of Blau and Kahn (2001) who assumed occupation to be an important factor of the financial status of a person. After controlling for occupational status, the effect of gender could still be found to be significant. We cannot accept H3 and therefore cannot confirm results of earlier studies ( Blau and Kahn, 2007 ; Boll et al., 2017 ). In contrast to the results of education and vocational training, we did not observe a clear social gradient of occupational status and income in our analyses. One explanation could be the classification of the occupational status. The ISCO classification is structured hierarchically on four levels. The construction is based on skill level and specialization. In our study, we used the major group structure (level one) with 10 different occupational groups. Using ISCO at level one (major groups) cannot be interpreted as a strict hierarchical order of occupations; instead, it can be considered more of a summary information on occupational status regarding skill level. Moreover, we were only able to generate the major groups of the register data and therefore cannot provide more detailed information about the occupational status. However, ISCO is applied in our study for the purpose of international comparability ( International Labour Office, 2012 ).

The accumulation of disadvantages over time could also be found in our results after controlling for factors such as unemployment or marginal employment. Having (at least one) marginal employment per year influenced the income negatively. We found that discontinuities in employment and interruptions such as unemployment also had a significant negative effect. Average income decreased when the number of days per year of unemployment increased. Furthermore, controlling for the region of employment, people in East Germany had lower daily average income compared to those in West Germany. Regarding the difference between men and women, previous findings also suggested a wider GPG in West Germany than in East Germany ( Federal Statistical Office, 2016 ). However, the GPG in West and East Germany should be compared with caution due to different societal models in the past. Moreover, different labour market characteristics and different infrastructure of childcare facilities lead to a lower GPG in East Germany than in West Germany ( Federal Ministry for Family Affairs, Senior Citizens, Women and Youth, 2020 ). The year of birth was included to eliminate cohort effects, and it was found to influence average income. Men and women born earlier (1959) had higher income than those born in 1965. The fact that they are older and have worked longer in the labor market could be an explanation. The significant effects of gender on the average income and the income trajectories remained after adjusting for these factors. Therefore, hypothesis 4 can only be partially confirmed.

4.1 Strengths and Limitations

Our study has limitations concerning the generalizability of our results due to the database. Our sample includes employees of two age groups (1959 and 1965) in Germany, who are subjected to social security. Thus, the generalizability or extension of the findings to self-employed people, civil servants and other age groups may be limited. The GPG differs considerably between the EU members. The GPG in Germany is one of the widest in the EU, with 19.2% in 2019. Netherlands and Sweden are two EU countries with similar employment rates, but still have lower GPGs with 14.6 and 11.8% ( Eurostat, 2021a ). Efforts to promote gender equality in politics in Germany are limited compared to other EU members. Women are still underrepresented, not only in the political but also in the economic area. Moreover family policy needs to further support full-time employment of women and working mothers ( Andersson et al., 2014 ; Botsch, 2015 ). Therefore, the transfer of our results to other countries should be made with caution. There are some other limitations regarding the IEB data. Information about occupational careers exist from the beginning (1975), but only for persons born in West Germany. Information about people born in East Germany was not available for the period before 1993. Hence, to counteract the systematic bias, we defined 1993 as a cut-off point, when people were either 28 or 34 years old. Additionally, we adjusted our analyses for the region of employment (East/West Germany). Furthermore, information about the marginal work and duration of unemployment were only available from 1999 onwards. Due to the composition of the IEB data, we could not include people who were unwell for long periods of time. Only persons who were unable to work for less than 42 days were included in the data. Regarding the income development of women in our study, Figure 2 shows a decrease between 1997 and 1999. Being in their thirties (32–40 years) and having to raise children at that time can be one possible explanation. Regarding family formation, in 1993 the average age of a mother at birth was 28.4 years ( Federal Statistical Office, 2020 ). At the beginning of our analysis (1993) the average age of both cohorts in the study (28 years; 34 years) is similar to the average age of a mother during that time – especially for the younger cohort. However, our data do not cover information about persons on parental leave or homemakers. Due to the lack of information in the IEB data, implications of family life contributing to a difference in pay for women cannot be included in our analysis. Furthermore, Joshi et al. (2020) could not find a GPG only for parents but also for men and women without children. Therefore, the issue of wage differences between men and women is relevant either way.

Besides these restrictions, our study exhibits several strengths. The study population is highly representative for German employees subject to social insurance contributions, born in 1959 and 1965 and is, therefore, characterized by a high external validity ( Schröder et al., 2013 ). Moreover, the IEB data itself and the nature of the data that the IEB provides, are one important strength of this study. The register data is not subject to possible recall bias. This is a relevant advantage compared to most previous studies that used self-reported data. In addition, the availability of information on a daily basis regarding many variables can be seen as another strength of the study. As a result, income trajectories could be calculated more precisely, compared to many previous studies. Furthermore, in Germany, income is used to calculate the amount of social benefit accruing to each person and therefore represents highly valid information. A further major advantage of our study is represented in our long observation period of 24 years. Only a few studies have applied the life course approach to examine the complexity of the GPG. Our life course data contain various information about employment characteristics which are relevant for the GPG and of high data quality.

Our results showed, even after controlling for relevant factors, that the GPG still persisted. There exist some explanations of the GPG regarding different behaviors of men and women in wage negotiations, which further influence different income developments ( Boll and Leppin, 2015 ). Also, structural disadvantages in the labor market can be a factor explaining the GPG. Individual behavior and labor market structures are not represented in our register data. We can only extract information that is relevant for social security contribution. Nonetheless, previous research of Blau and Kahn (2017) found a larger and more slowly decreasing GPG in the US at the top compared to other levels of the wage distribution. This ‘glass ceiling effect’ describes the reduced career opportunities of women compared to men due to frequent denial of access to leadership positions. Consequently, gender inequality can be found to be greater at the top of the wage distribution. Among European countries, previous studies have found this “glass ceiling effect” in Germany as well ( Arulampalam et al., 2005 ; Boll and Leppin, 2015 ; Huffman et al., 2017 ). However, recent results of Boll et al. (2017) could not confirm the glass ceiling effect in West Germany, thus further research is needed.

5 Conclusion

The gender pay inequalities in the German labor market from a life course perspective exist. Our results demonstrated that human capital determinants continue to be important in explaining the GPG over time. Furthermore, factors of working disadvantages such as marginal work or unemployment are important when trying to explain the income differences of men and women. For further research the availability of more work data over the life course with matching individual data would help to understand the GPG even better.

Acknowledgments

We gratefully acknowledge the support of two staff members of the University Ulm. We would like to thank Gaurav Berry for his support of the data preparation and Diego Montano for his feedback on the statistical analysis.

Data Availability Statement

The datasets presented in this article are not readily available because the study data contain social security information. Due to legal regulations in Germany, it is not permitted to share data with social security information. Requests to access the datasets should be directed to [email protected] .

Ethics Statement

The studies involving human participants were reviewed and approved by the ethics commission of the University of Wuppertal. The patients/participants provided their written informed consent to participate in this study.

Author Contributions

LT substantially contributed to the statistical analysis and interpretation of the data, and wrote the manuscript. HB discussed the results and provided critical comments on the manuscript. RP contributed to the obtaining of the funding, interpreting the data, and critically revised the manuscript for important aspects. All authors read and approved the final manuscript.

This work was supported by the German Research Foundation (DFG), grant number 393153877.

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s Note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary Material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fsoc.2021.815376/full#supplementary-material .

• Achatz J., Gartner H., Glück T. (2005). Bonus Oder Bias? Koelner Z.Soziol.u.Soz.Psychol 57, 466–493. 10.1007/s11577-005-0185-6 [ DOI ] [ Google Scholar ]
• Aisenbrey S., Bruckner H. (2008). Occupational Aspirations and the Gender Gap in Wages. Eur. Sociological Rev. 24, 633–649. 10.1093/esr/jcn024 [ DOI ] [ Google Scholar ]
• Aisenbrey S., Fasang A. (2017). The Interplay of Work and Family Trajectories over the Life Course: Germany and the United States in Comparison. Am. J. Sociol. 122, 1448–1484. 10.1086/691128 [ DOI ] [ Google Scholar ]
• Andersson G., Kreyenfeld M., Mika T. (2014). Welfare State Context, Female Labour-Market Attachment and Childbearing in Germany and Denmark. J. Pop Res. 31, 287–316. 10.1007/s12546-014-9135-3 [ DOI ] [ Google Scholar ]
• Arulampalam W., Booth A. L., Bryan M. L. (2005). Is There a Glass Ceiling over Europe? Exploring the Gender Pay gap across the Wages Distribution. Colchester, EX: ISER Working Paper Series, 25. [ Google Scholar ]
• Azmat G., Güell M., Manning A. (2006). Gender Gaps in Unemployment Rates in OECD Countries. J. Labor Econ. 24, 1–37. 10.1086/497817 [ DOI ] [ Google Scholar ]
• Barone C., Schizzerotto A. (2011). Introduction. Eur. Societies 13, 331–345. 10.1080/14616696.2011.568248 [ DOI ] [ Google Scholar ]
• Blau F. D., Kahn L. M. (2007). The Gender Pay Gap. AMP 21, 7–23. 10.5465/amp.2007.24286161 [ DOI ] [ Google Scholar ]
• Blau F. D., Kahn L. M. (2017). The Gender Wage Gap: Extent, Trends, and Explanations. J. Econ. Lit. 55, 789–865. 10.1257/jel.20160995 [ DOI ] [ Google Scholar ]
• Blau F., Kahn L. (2001). Understanding International Differences in the Gender Pay Gap. Ithaca, NY: NBER Working Paper 8200. Available at: https://www.nber.org/papers/w8200 (Accessed November 30, 2021). [ Google Scholar ]
• Boll C., Jahn M., Lagemann A. (2017). The Gender Lifetime Earnings gap: Exploring Gendered Pay from the Life Course Perspective. Hamburg, Germany: HWWI Research Paper, 179. [ Google Scholar ]
• Boll C., Leppin J. S. (2015). Die geschlechtsspezifische Lohnlücke in Deutschland: Umfang, Ursachen und Interpretation. Wirtschaftsdienst 95, 249–254. 10.1007/s10273-015-1814-y [ DOI ] [ Google Scholar ]
• Börsch-Supan A., Brandt M., Hunkler C., Kneip T., Korbmacher J., Malter F., et al. (2013). Data Resource Profile: the Survey of Health, Ageing and Retirement in Europe (SHARE). Int. J. Epidemiol. 42, 992–1001. 10.1093/ije/dyt088 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Botsch E. (2015). The Policy on Gender Equality in Germany. Available at: https://www.europarl.europa.eu/RegData/etudes/IDAN/2015/510025/IPOL_IDA(2015)510025_EN.pdf (Accessed November 25, 2020).
• Bovens M., Wille A. (2017). Education as a Cleavage. Oxford University Press. [ Google Scholar ]
• Broughton A., Green M., Rickard C., Swift S., Eichhorst W., Tobsch V., et al. (2016). Precarious Employment in Europe: Patterns, Trends and Policy Strategy. Europarl. Available at: https://www.europarl.europa.eu/RegData/etudes/STUD/2016/587285/IPOL_STU(2016)587285_EN.pdf . [ Google Scholar ]
• Brynin M., Perales F. (2016). Gender Wage Inequality: The De-gendering of the Occupational Structure. Eur. Sociol. Rev. 32, 162–174. 10.1093/esr/jcv092 [ DOI ] [ Google Scholar ]
• Dannefer D. (2003). Cumulative Advantage/disadvantage and the Life Course: Cross-Fertilizing Age and Social Science Theory. J. Gerontol. B Psychol. Sci. Soc. Sci. 58, S327–S337. 10.1093/geronb/58.6.s327 [ DOI ] [ PubMed ] [ Google Scholar ]
• Doren C., Lin K. Y. (2019). Diverging Trajectories or Parallel Pathways? an Intersectional and Life Course Approach to the Gender Earnings Gap by Race and Education. Socius 5, 237802311987381–23. 10.1177/2378023119873816 [ DOI ] [ Google Scholar ]
• Du Prel J. B., Schrettenbrunner C., Hasselhorn H. M. (2019). Vertical and Horizontal Social Inequality and Motivation for Early Retirement. Z. Gerontol. Geriatr. 52, 3–13. 10.1007/s00391-018-1450-4 [ DOI ] [ PubMed ] [ Google Scholar ]
• Eurostat (2021a). Statistics Explained: Gender Pay gap Statistics. Available at: https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Gender_pay_gap_statistics (Accessed December 01, 2021).
• Eurostat (2021b). Statistics Explained: Gender Statistics. Available at: https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Gender_statistics#Earnings (Accessed November 30, 2021).
• Federal Agency for Civic Education (2021). Sozialversicherung [Social Insurance]. Available at: https://www.bpb.de/nachschlagen/lexika/das-junge-politik-lexikon/321149/sozialversicherung (Accessed November 30, 2021).
• Federal Ministry for Family Affairs, Senior Citizens, Women and Youth (2020). The Road to Equal Pay for Women and Men. Available at: https://www.bmfsfj.de/bmfsfj/meta/en/publications-en/the-road-to-equal-pay-for-women-and-men-161370 (Accessed December 01, 2021).
• Federal Statistical Office (2020). Durchschnittliches Alter der Mutter bei der Geburt ihrer lebend geborenen Kinder: Deutschland, Jahre, Familienstand der Eltern [Average age of the mother at the birth of her children born alive: Germany, years, marital status of parents]. Available at: https://www-genesis.destatis.de/genesis/online (Accessed March 17, 2021).
• Federal Statistical Office (2021). Three in Four Mothers in Germany Were in Employment in 2019. Available at: https://www.destatis.de/EN/Press/2021/03/PE21_N017_13.html (Accessed March 16, 2021).
• Federal Statistical Office (2016). Unbereinigter Verdienstunterschied nach persönlichen Merkmalen im Jahr 2014 [Gender Pay Gap by personal characteristics in 2014 (unadjusted)]. Available at: https://www.destatis.de/DE/Themen/Arbeit/Verdienste/Verdienste-Verdienstunterschiede/Tabellen/gpg-persoenlich.html (Accessed March 16, 2020).
• Goldin C. (2014). A Grand Gender Convergence: Its Last Chapter. Am. Econ. Rev. 104, 1091–1119. 10.1257/aer.104.4.1091 [ DOI ] [ Google Scholar ]
• Grybaitė V. (2006). Analysis of Theoretical Approaches to Gender Pay gap. J. Business Econ. Manag. 7, 85–91. 10.3846/16111699.2006.9636127 [ DOI ] [ Google Scholar ]
• Härkönen J., Manzoni A., Bihagen E. (2016). Gender Inequalities in Occupational Prestige across the Working Life: An Analysis of the Careers of West Germans and Swedes Born from the 1920s to the 1970s. Adv. Life course Res. 29, 41–51. 10.1016/j.alcr.2016.01.001 [ DOI ] [ Google Scholar ]
• Hasselhorn H. M., Peter R., Rauch A., Schröder H., Swart E., Bender S., et al. (2014). Cohort Profile: the lidA Cohort Study-A German Cohort Study on Work, Age, Health and Work Participation. Int. J. Epidemiol. 43, 1736–49. 10.1093/ije/dyu021 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Hasselhorn H. M. (2020). “Social Inequality in the Transition from Work to Retirement,” in Handbook of Socioeconomic Determinants of Occupational Health. Editor Theorell T. (Cham: Springer International Publishing; ), 1–26. 10.1007/978-3-030-05031-3_32-1 [ DOI ] [ Google Scholar ]
• Hosoya G., Koch T., Eid M. (2014). Längsschnittdaten und Mehrebenenanalyse. Köln Z. Soziol 66, 189–218. 10.1007/s11577-014-0262-9 [ DOI ] [ Google Scholar ]
• Huffman M. L., King J., Reichelt M. (2017). Equality for Whom? Organizational Policies and the Gender Gap across the German Earnings Distribution. ILR Rev. 70, 16–41. 10.1177/0019793916673974 [ DOI ] [ Google Scholar ]
• International Labour Office (2012). International Standard Classification of Occupations: ISCO-08. Geneva: International Labour Organization. [ Google Scholar ]
• Jalovaara M., Fasang A. E. (2019). Family Life Courses, Gender, and Mid-life Earnings. Eur. Sociol. Rev. 36, 159–178. 10.1093/esr/jcz057 [ DOI ] [ Google Scholar ]
• Joshi H., Bryson A., Wilkinson D., Ward K. (2020). The Gender gap in Wages over the Life Course: Evidence from a British Cohort Born in 1958. Gend. Work Organ. 28, 397–415. 10.1111/gwao.12580 [ DOI ] [ Google Scholar ]
• Lips H. M. (2013). The Gender Pay Gap: Challenging the Rationalizations. Perceived Equity, Discrimination, and the Limits of Human Capital Models. Sex Roles 68, 169–185. 10.1007/s11199-012-0165-z [ DOI ] [ Google Scholar ]
• Manzoni A., Harkonen J., Mayer K. U. (2014). Moving on? A Growth-Curve Analysis of Occupational Attainment and Career Progression Patterns in West Germany. Social Forces 92, 1285–1312. 10.1093/sf/sou002 [ DOI ] [ Google Scholar ]
• Matteazzi E., Pailhé A., Solaz A. (2018). Part-time Employment, the Gender Wage gap and the Role of Wage-Setting Institutions: Evidence from 11 European Countries. Eur. J. Ind. Relations 24, 221–241. 10.1177/0959680117738857 [ DOI ] [ Google Scholar ]
• Mincer J., Polachek S. (1974). Family Investments in Human Capital: Earnings of Women. J. Polit. Economy 82, S76–S108. 10.1086/260293 [ DOI ] [ Google Scholar ]
• Monti H., Stinson M., Zehr L. (2020). How Long Do Early Career Decisions Follow Women? the Impact of Employer History on the Gender Wage Gap. J. Labor Res. 41, 189–232. 10.1007/s12122-020-09300-9 [ DOI ] [ Google Scholar ]
• Moore T. S. (2018). Occupational Career Change and Gender Wage Inequality. Work and Occupations 45, 82–121. 10.1177/0730888417742691 [ DOI ] [ Google Scholar ]
• OECD (2021). OECD Statistics. Available at: http://stats.oecd.org/ (Accessed March 16, 2021).
• Petersen T., Morgan L. A. (1995). Separate and Unequal: Occupation-Establishment Sex Segregation and the Gender Wage Gap. Am. J. Sociol. 101, 329–365. 10.1086/230727 [ DOI ] [ Google Scholar ]
• Polachek S. W. (1981). Occupational Self-Selection: A Human Capital Approach to Sex Differences in Occupational Structure. Rev. Econ. Stat. 63, 60. 10.2307/1924218 [ DOI ] [ Google Scholar ]
• Ponthieux S., Meurs D. (2015). “Gender Inequality,” in Handbook of Income Distribution: Volume 2. Editors Atkinson A. B., Bourguignon F. (Burlington: Elsevier Science; ), 981–1146. 10.1016/b978-0-444-59428-0.00013-8 [ DOI ] [ Google Scholar ]
• Rabe-Hesketh S., Skrondal A. (2012). Multilevel and Longitudinal Modeling Using Stata. 3rd ed.. College Station, Tex: Stata Press. [ Google Scholar ]
• Radl J. (2013). Labour Market Exit and Social Stratification in Western Europe: The Effects of Social Class and Gender on the Timing of Retirement. Eur. Sociological Rev. 29, 654–668. 10.1093/esr/jcs045 [ DOI ] [ Google Scholar ]
• Sackmann R. (2018). “Demographie als Herausforderung für die Soziologie,” in Handbuch Soziologie des Alter(n)s. Editors Schroeter K. R., Vogel C., Künemund H. (Wiesbaden: Springer VS; ), 1–23. 10.1007/978-3-658-09630-4_5-1 [ DOI ] [ Google Scholar ]
• Schröder H., Kersting A., Gilberg R., Steinwede J. (2013). Methodenbericht zur Haupterhebung lidA - leben in der Arbeit [Methodology Report of the main survey of lidA]. Available at: http://doku.iab.de/fdz/reporte/2013/MR_01-13.pdf (Accessed March 15, 2020).
• Sierminska E. M., Frick J. R., Grabka M. M. (2010). Examining the Gender Wealth gap. Oxford Econ. Pap. 62, 669–690. 10.1093/oep/gpq007 [ DOI ] [ Google Scholar ]
• Singer J. D., Willett J. B. (2003). Applied Longitudinal Data Analysis: Modeling Change and Event Occurrence. Oxford: Oxford Univ. Press. [ Google Scholar ]
• Visser M., Gesthuizen M., Kraaykamp G., Wolbers M. H. J. (2016). Inequality Among Older Workers in the Netherlands: A Life Course and Social Stratification Perspective on Early Retirement. Eur. Sociol. Rev. 32, 370–382. 10.1093/esr/jcw013 [ DOI ] [ Google Scholar ]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

• View on publisher site
• PDF (873.2 KB)
• Collections

Similar articles

Cited by other articles, links to ncbi databases.

• Download .nbib .nbib
• Format: AMA APA MLA NLM

Add to Collections

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here .

Loading metrics

Open Access

Peer-reviewed

Research Article

The persistence of pay inequality: The gender pay gap in an anonymous online labor market

Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Supervision, Writing – original draft, Writing – review & editing

* E-mail: [email protected] (LL); [email protected] (LB)

Affiliation Department of Psychology, Lander College, Flushing, New York, United States of America

Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Writing – original draft, Writing – review & editing

Affiliation Department of Computer Science, Lander College, Flushing, New York, United States of America

Roles Formal analysis, Writing – original draft, Writing – review & editing

Affiliation Department of Health Policy & Management, Mailman School of Public Health, Columbia University, New York, New York, United States of America

Roles Conceptualization, Writing – review & editing

Affiliation Department of Clinical Psychology, Columbia University, New York, New York, United States of America

Roles Formal analysis

Affiliation Department of Computer Science, Stern College for Women, New York, New York, United States of America

Roles Conceptualization, Methodology, Writing – original draft, Writing – review & editing

Affiliation Department of Epidemiology, Mailman School of Public Health, Columbia University New York, New York, United States of America

• Leib Litman, 
• Jonathan Robinson, 
• Zohn Rosen, 
• Cheskie Rosenzweig, 
• Joshua Waxman, 
• Lisa M. Bates

• Published: February 21, 2020
• https://doi.org/10.1371/journal.pone.0229383
• Reader Comments

Studies of the gender pay gap are seldom able to simultaneously account for the range of alternative putative mechanisms underlying it. Using CloudResearch, an online microtask platform connecting employers to workers who perform research-related tasks, we examine whether gender pay discrepancies are still evident in a labor market characterized by anonymity, relatively homogeneous work, and flexibility. For 22,271 Mechanical Turk workers who participated in nearly 5 million tasks, we analyze hourly earnings by gender, controlling for key covariates which have been shown previously to lead to differential pay for men and women. On average, women’s hourly earnings were 10.5% lower than men’s. Several factors contributed to the gender pay gap, including the tendency for women to select tasks that have a lower advertised hourly pay. This study provides evidence that gender pay gaps can arise despite the absence of overt discrimination, labor segregation, and inflexible work arrangements, even after experience, education, and other human capital factors are controlled for. Findings highlight the need to examine other possible causes of the gender pay gap. Potential strategies for reducing the pay gap on online labor markets are also discussed.

Citation: Litman L, Robinson J, Rosen Z, Rosenzweig C, Waxman J, Bates LM (2020) The persistence of pay inequality: The gender pay gap in an anonymous online labor market. PLoS ONE 15(2): e0229383. https://doi.org/10.1371/journal.pone.0229383

Editor: Luís A. Nunes Amaral, Northwestern University, UNITED STATES

Received: March 5, 2019; Accepted: February 5, 2020; Published: February 21, 2020

Copyright: © 2020 Litman et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: Due to the sensitive nature of some of the data, and the terms of service of the websites used during data collection (including CloudResearch and MTurk), CloudResearch cannot release the full data set to make it publically available. The data are on CloudResearch's Sequel servers located at Queens College in the city of New York. CloudResearch makes data available to be accessed by researchers for replication purposes, on the CloudResearch premises, in the same way the data were accessed and analysed by the authors of this manuscript. The contact person at CloudResearch who can help researchers access the data set is Tzvi Abberbock, who can be reached at [email protected] .

Funding: The authors received no specific funding for this work.

Competing interests: We have read the journal's policy and the authors of this manuscript have the following potential competing interest: Several of the authors are employed at Cloud Research (previously TurkPrime), the database from which the data were queried. This does not alter our adherence to PLOS ONE policies on sharing data and materials.

Introduction

The gender pay gap, the disparity in earnings between male and female workers, has been the focus of empirical research in the US for decades, as well as legislative and executive action under the Obama administration [ 1 , 2 ]. Trends dating back to the 1960s show a long period in which women’s earnings were approximately 60% of their male counterparts, followed by increases in women’s earnings starting in the 1980s, which began to narrow, but not close, the gap which persists today [ 3 ]. More recent data from 2014 show that overall, the median weekly earnings of women working full time were 79–83% of what men earned [ 4 – 9 ].

The extensive literature seeking to explain the gender pay gap and its trajectory over time in traditional labor markets suggests it is a function of multiple structural and individual-level processes that reflect both the near-term and cumulative effects of gender relations and roles over the life course. Broadly speaking, the drivers of the gender pay gap can be categorized as: 1) human capital or productivity factors such as education, skills, and workforce experience; 2) industry or occupational segregation, which some estimates suggest accounts for approximately half of the pay gap; 3) gender-specific temporal flexibility constraints which can affect promotions and remuneration; and finally, 4) gender discrimination operating in hiring, promotion, task assignment, and/or compensation. The latter mechanism is often estimated by inference as a function of unexplained residual effects of gender on payment after accounting for other factors, an approach which is most persuasive in studies of narrowly restricted populations of workers such as lawyers [ 10 ] and academics of specific disciplines [ 11 ]. A recent estimate suggests this unexplained gender difference in earnings can account for approximately 40% of the pay gap [ 3 ]. However, more direct estimations of discriminatory processes are also available from experimental evidence, including field audit and lab-based studies [ 12 – 14 ]. Finally, gender pay gaps have also been attributed to differential discrimination encountered by men and women on the basis of parental status, often known as the ‘motherhood penalty’ [ 15 ].

Non-traditional ‘gig economy’ labor markets and the gender pay gap

In recent years there has been a dramatic rise in nontraditional ‘gig economy’ labor markets, which entail independent workers hired for single projects or tasks often on a short-term basis with minimal contractual engagement. “Microtask” platforms such as Amazon Mechanical Turk (MTurk) and Crowdflower have become a major sector of the gig economy, offering a source of easily accessible supplementary income through performance of small tasks online at a time and place convenient to the worker. Available tasks can range from categorizing receipts to transcription and proofreading services, and are posted online by the prospective employer. Workers registered with the platform then elect to perform the advertised tasks and receive compensation upon completion of satisfactory work [ 16 ]. An estimated 0.4% of US adults are currently receiving income from such platforms each month [ 17 ], and microtask work is a growing sector of the service economy in the United States [ 18 ]. Although still relatively small, these emerging labor market environments provide a unique opportunity to investigate the gender pay gap in ways not possible within traditional labor markets, due to features (described below) that allow researchers to simultaneously account for multiple putative mechanisms thought to underlie the pay gap.

The present study utilizes the Amazon Mechanical Turk (MTurk) platform as a case study to examine whether a gender pay gap remains evident when the main causes of the pay gap identified in the literature do not apply or can be accounted for in a single investigation. MTurk is an online microtask platform that connects employers (‘requesters’) to employees (‘workers’) who perform jobs called “Human Intelligence Tasks” (HITs). The platform allows requesters to post tasks on a dashboard with a short description of the HIT, the compensation being offered, and the time the HIT is expected to take. When complete, the requester either approves or rejects the work based on quality. If approved, payment is quickly accessible to workers. The gender of workers who complete these HITs is not known to the requesters, but was accessible to researchers for the present study (along with other sociodemographic information and pay rates) based on metadata collected through CloudResearch (formerly TurkPrime), a platform commonly used to conduct social and behavioral research on MTurk [ 19 ].

Evaluating pay rates of workers on MTurk requires estimating the pay per hour of each task that a worker accepts which can then be averaged together. All HITs posted on MTurk through CloudResearch display how much a HIT pays and an estimated time that it takes for that HIT to be completed. Workers use this information to determine what the corresponding hourly pay rate of a task is likely to be, and much of our analysis of the gender pay gap is based on this advertised pay rate of all completed surveys. We also calculate an estimate of the gender pay gap based on actual completion times to examine potential differences in task completion speed, which we refer to as estimated actual wages (see Methods section for details).

Previous studies have found that both task completion time and the selection of tasks influences the gender pay gap in at least some gig economy markets. For example, a gender pay gap was observed among Uber drivers, with men consistently earning higher pay than women [ 20 ]. Some of the contributing factors to this pay gap include that male Uber drivers selected different tasks than female drivers, including being more willing to work at night and to work in neighborhoods that were perceived to be more dangerous. Male drivers were also likely to drive faster than their female counterparts. These findings show that person-level factors like task selection, and speed can influence the gender pay gap within gig economy markets.

MTurk is uniquely suited to examine the gender pay gap because it is possible to account simultaneously for multiple structural and individual-level factors that have been shown to produce pay gaps. These include discrimination, work heterogeneity (leading to occupational segregation), and job flexibility, as well as human capital factors such as experience and education.

Discrimination.

When employers post their HITs on MTurk they have no way of knowing the demographic characteristics of the workers who accept those tasks, including their gender. While MTurk allows for selective recruitment of specific demographic groups, the MTurk tasks examined in this study are exclusively open to all workers, independent of their gender or other demographic characteristics. Therefore, features of the worker’s identity that might be the basis for discrimination cannot factor into an employer’s decision-making regarding hiring or pay.

Task heterogeneity.

Another factor making MTurk uniquely suited for the examination of the gender pay gap is the relative homogeneity of tasks performed by the workers, minimizing the potential influence of gender differences in the type of work pursued on earnings and the pay gap. Work on the MTurk platform consists mostly of short tasks such as 10–15 minute surveys and categorization tasks. In addition, the only information that workers have available to them to choose tasks, other than pay, is the tasks’ titles and descriptions. We additionally classified tasks based on similarity and accounted for possible task heterogeneity effects in our analyses.

Job flexibility.

MTurk is not characterized by the same inflexibilities as are often encountered in traditional labor markets. Workers can work at any time of the day or day of the week. This increased flexibility may be expected to provide more opportunities for participation in this labor market for those who are otherwise constrained by family or other obligations.

Human capital factors.

It is possible that the more experienced workers could learn over time how to identify higher paying tasks by virtue of, for example, identifying qualities of tasks that can be completed more quickly than the advertised required time estimate. Further, if experience is correlated with gender, it could contribute to a gender pay gap and thus needs to be controlled for. Using CloudResearch metadata, we are able to account for experience on the platform. Additionally, we account for multiple sociodemographic variables, including age, marital status, parental status, education, income (from all sources), and race using the sociodemographic data available through CloudResearch.

Expected gender pay gap findings on MTurk

Due to the aforementioned factors that are unique to the MTurk marketplace–e.g., anonymity, self-selection into tasks, relative homogeneity of the tasks performed, and flexible work scheduling–we did not expect a gender pay gap to be evident on the platform to the same extent as in traditional labor markets. However, potential gender differences in task selection and completion speed, which have implications for earnings, merit further consideration. For example, though we expect the relative homogeneity of the MTurk tasks to minimize gender differences in task selection that could mimic occupational segregation, we do account for potential subtle residual differences in tasks that could differentially attract male and female workers and indirectly lead to pay differentials if those tasks that are preferentially selected by men pay a higher rate. To do this we categorize all tasks based on their descriptions using K-clustering and add the clusters as covariates to our models. In addition, we separately examine the gender pay gap within each topic-cluster.

In addition, if workers who are experienced on the platform are better able to find higher paying HITs, and if experience is correlated with gender, it may lead to gender differences in earnings. Theoretically, other factors that may vary with gender could also influence task selection. Previous studies of the pay gap in traditional markets indicate that reservation wages, defined as the pay threshold at which a person is willing to accept work, may be lower among women with children compared to women without, and to that of men as well [ 21 ]. Thus, if women on MTurk are more likely to have young children than men, they may be more willing to accept available work even if it pays relatively poorly. Other factors such as income, education level, and age may similarly influence reservation wages if they are associated with opportunities to find work outside of microtask platforms. To the extent that these demographics correlate with gender they may give rise to a gender pay gap. Therefore we consider age, experience on MTurk, education, income, marital status, and parental status as covariates in our models.

Task completion speed may vary by gender for several reasons, including potential gender differences in past experience on the platform. We examine the estimated actual pay gap per hour based on HIT payment and estimated actual completion time to examine the effects of completion speed on the wage gap. We also examine the gender pay gap based on advertised pay rates, which are not dependent on completion speed and more directly measure how gender differences in task selection can lead to a pay gap. Below, we explain how these were calculated based on meta-data from CloudResearch.

To summarize, the overall goal of the present study was to explore whether gender pay differentials arise within a unique, non-traditional and anonymous online labor market, where known drivers of the gender pay gap either do not apply or can be accounted for statistically.

Materials and methods

Amazon mechanical turk and cloudresearch..

Started in 2005, the original purpose of the Amazon Mechanical Turk (MTurk) platform was to allow requesters to crowdsource tasks that could not easily be handled by existing technological solutions such as receipt copying, image categorization, and website testing. As of 2010, researchers increasingly began using MTurk for a wide variety of research tasks in the social, behavioral, and medical sciences, and it is currently used by thousands of academic researchers across hundreds of academic departments [ 22 ]. These research-related HITs are typically listed on the platform in generic terms such as, “Ten-minute social science study,” or “A study about public opinion attitudes.”

Because MTurk was not originally designed solely for research purposes, its interface is not optimized for some scientific applications. For this reason, third party add-on toolkits have been created that offer critical research tools for scientific use. One such platform, CloudResearch (formerly TurkPrime), allows requesters to manage multiple research functions, such as applying sampling criteria and facilitating longitudinal studies, through a link to their MTurk account. CloudResearch’s functionality has been described extensively elsewhere [ 19 ]. While the demographic characteristics of workers are not available to MTurk requesters, we were able to retroactively identify the gender and other demographic characteristics of workers through the CloudResearch platform. CloudResearch also facilitates access to data for each HIT, including pay, estimated length, and title.

The study was an analysis of previously collected metadata, which were analyzed anonymously. We complied with the terms of service for all data collected from CloudResearch, and MTurk. The approving institutional review board for this study was IntegReview.

Analytic sample.

We analyzed the nearly 5 million tasks completed during an 18-month period between January 2016 and June 2017 by 12,312 female and 9,959 male workers who had complete data on key demographic characteristics. To be included in the analysis a HIT had to be fully completed, not just accepted, by the worker, and had to be accepted (paid for) by the requester. Although the vast majority of HITs were open to both males and females, a small percentage of HITs are intended for a specific gender. Because our goal was to exclusively analyze HITs for which the requesters did not know the gender of workers, we excluded any HITs using gender-specific inclusion or exclusion criteria from the analyses. In addition, we removed from the analysis any HITs that were part of follow-up studies in which it would be possible for the requester to know the gender of the worker from the prior data collection. Finally, where possible, CloudResearch tracks demographic information on workers across multiple HITs over time. To minimize misclassification of gender, we excluded the 0.3% of assignments for which gender was unknown with at least 95% consistency across HITs.

The main exposure variable is worker gender and the outcome variables are estimated actual hourly pay accrued through completing HITs, and advertised hourly pay for completed HITs. Estimated actual hourly wages are based on the estimated length in minutes and compensation in dollars per HIT as posted on the dashboard by the requester. We refer to actual pay as estimated because sometimes people work multiple assignments at the same time (which is allowed on the platform), or may simultaneously perform other unrelated activities and therefore not work on the HIT the entire time the task is open. We also considered several covariates to approximate human capital factors that could potentially influence earnings on this platform, including marital status, education, household income, number of children, race/ethnicity, age, and experience (number of HITs previously completed). Additional covariates included task length, task cluster (see below), and the serial order with which workers accepted the HIT in order to account for potential differences in HIT acceptance speed that may relate to the pay gap.

Database and analytic approach.

Data were exported from CloudResearch’s database into Stata in long-form format to represent each task on a single row. For the purposes of this paper, we use “HIT” and “study” interchangeably to refer to a study put up on the MTurk dashboard which aims to collect data from multiple participants. A HIT or study consist of multiple “assignments” which is a single task completed by a single participant. Columns represented variables such as demographic information, payment, and estimated HIT length. Column variables also included unique IDs for workers, HITs (a single study posted by a requester), and requesters, allowing for a multi-level modeling analytic approach with assignments nested within workers. Individual assignments (a single task completed by a single worker) were the unit of analysis for all models.

Linear regression models were used to calculate the gender pay gap using two dependent variables 1) women’s estimated actual earnings relative to men’s and 2) women’s selection of tasks based on advertised earnings relative to men’s. We first examined the actual pay model, to see the gender pay gap when including an estimate of task completion speed, and then adjusted this model for advertised hourly pay to determine if and to what extent a propensity for men to select more remunerative tasks was evident and driving any observed gender pay gap. We additionally ran separate models using women’s advertised earnings relative to men’s as the dependent variable to examine task selection effects more directly. The fully adjusted models controlled for the human capital-related covariates, excluding household income and education which were balanced across genders. These models also tested for interactions between gender and each of the covariates by adding individual interaction terms to the adjusted model. To control for within-worker clustering, Huber-White standard error corrections were used in all models.

Cluster analysis.

To explore the potential influence of any residual task heterogeneity and gender preference for specific task type as the cause of the gender pay gap, we use K-means clustering analysis (seed = 0) to categorize the types of tasks into clusters based on the descriptions that workers use to choose the tasks they perform. We excluded from this clustering any tasks which contained certain gendered words (such as “male”, “female”, etc.) and any tasks which had fewer than 30 respondents. We stripped out all punctuation, symbols and digits from the titles, so as to remove any reference to estimated compensation or duration. The features we clustered on were the presence or absence of 5,140 distinct words that appeared across all titles. We then present the distribution of tasks across these clusters as well as average pay by gender and the gender pay gap within each cluster.

The demographics of the analytic sample are presented in Table 1 . Men and women completed comparable numbers of tasks during the study period; 2,396,978 (48.6%) for men and 2,539,229 (51.4%) for women.

• PPT PowerPoint slide
• PNG larger image
• TIFF original image

https://doi.org/10.1371/journal.pone.0229383.t001

In Table 2 we measure the differences in remuneration between genders, and then decompose any observed pay gap into task completion speed, task selection, and then demographic and structural factors. Model 1 shows the unadjusted regression model of gender differences in estimated actual pay, and indicates that, on average, tasks completed by women paid 60 (10.5%) cents less per hour compared to tasks completed by men (t = 17.4, p < .0001), with the mean estimated actual pay across genders being $5.70 per hour.

https://doi.org/10.1371/journal.pone.0229383.t002

In Model 2, adjusting for advertised hourly pay, the gender pay gap dropped to 46 cents indicating that 14 cents of the pay gap is attributable to gender differences in the selection of tasks (t = 8.6, p < .0001). Finally, after the inclusion of covariates and their interactions in Model 3, the gender pay differential was further attenuated to 32 cents (t = 6.7, p < .0001). The remaining 32 cent difference (56.6%) in earnings is inferred to be attributable to gender differences in HIT completion speed.

Task selection analyses

Although completion speed appears to account for a significant portion of the pay gap, of particular interest are gender differences in task selection. Beyond structural factors such as education, household composition and completion speed, task selection accounts for a meaningful portion of the gender pay gap. As a reminder, the pay rate and expected completion time are posted for every HIT, so why women would select less remunerative tasks on average than men do is an important question to explore. In the next section of the paper we perform a set of analyses to examine factors that could account for this observed gender difference in task selection.

Advertised hourly pay.

To examine gender differences in task selection, we used linear regression to directly examine whether the advertised hourly pay differed for tasks accepted by male and female workers. We first ran a simple model ( Table 3 ; Model 3A) on the full dataset of 4.93 million HITs, with gender as the predictor and advertised hourly pay as the outcome including no other covariates. The unadjusted regression results (Model 4) shown in Table 3 , indicates that, summed across all clusters and demographic groups, tasks completed by women were advertised as paying 28 cents (95% CI: $0.25-$0.31) less per hour (5.8%) compared to tasks completed by men (t = 21.8, p < .0001).

https://doi.org/10.1371/journal.pone.0229383.t003

Model 5 examines whether the remuneration differences for tasks selected by men and women remains significant in the presence of multiple covariates included in the previous model and their interactions. The advertised pay differential for tasks selected by women compared to men was attenuated to 21 cents (4.3%), and remained statistically significant (t = 9.9, p < .0001). This estimate closely corresponded to the inferred influence of task selection reported in Table 2 . Tests of gender by covariate interactions were significant only in the cases of age and marital status; the pay differential in tasks selected by men and women decreased with age and was more pronounced among single versus currently or previously married women.

To further examine what factors may account for the observed gender differences in task selection we plotted the observed pay gap within demographic and other covariate groups. Table 4 shows the distribution of tasks completed by men and women, as well as mean earnings and the pay gap across all demographic groups, based on the advertised (not actual) hourly pay for HITs selected (hereafter referred to as “advertised hourly pay” and the “advertised pay gap”). The average task was advertised to pay $4.88 per hour (95% CI $4.69, $5.10).

https://doi.org/10.1371/journal.pone.0229383.t004

The pattern across demographic characteristics shows that the advertised hourly pay gap between genders is pervasive. Notably, a significant advertised gender pay gap is evident in every level of each covariate considered in Table 4 , but more pronounced among some subgroups of workers. For example, the advertised pay gap was highest among the youngest workers ($0.31 per hour for workers age 18–29), and decreased linearly with age, declining to $0.13 per hour among workers age 60+. Advertised houry gender pay gaps were evident across all levels of education and income considered.

To further examine the potential influence of human capital factors on the advertised hourly pay gap, Table 5 presents the average advertised pay for selected tasks by level of experience on the CloudResearch platform. Workers were grouped into 4 experience levels, based on the number of prior HITs completed: Those who completed fewer than 100 HITs, between 100 and 500 HITs, between 500 and 1,000 HITs, and more than 1,000 HITs. A significant gender difference in advertised hourly pay was observed within each of these four experience groups. The advertised hourly pay for tasks selected by both male and female workers increased with experience, while the gender pay gap decreases. There was some evidence that male workers have more cumulative experience with the platform: 43% of male workers had the highest level of experience (previously completing 1,001–10,000 HITs) compared to only 33% of women.

https://doi.org/10.1371/journal.pone.0229383.t005

Table 5 also explores the influence of task heterogeneity upon HIT selection and the gender gap in advertised hourly pay. K-means clustering was used to group HITs into 20 clusters initially based on the presence or absence of 5,140 distinct words appearing in HIT titles. Clusters with fewer than 50,000 completed tasks were then excluded from analysis. This resulted in 13 clusters which accounted for 94.3% of submitted work assignments (HITs).

The themes of all clusters as well as the average hourly advertised pay for men and women within each cluster are presented in the second panel of Table 5 . The clusters included categories such as Games, Decision making, Product evaluation, Psychology studies, and Short Surveys. We did not observe a gender preference for any of the clusters. Specifically, for every cluster, the proportion of males was no smaller than 46.6% (consistent with the slightly lower proportion of males on the platform, see Table 1 ) and no larger than 50.2%. As shown in Table 5 , the gender pay gap was observed within each of the clusters. These results suggest that residual task heterogeneity, a proxy for occupational segregation, is not likely to contribute to a gender pay gap in this market.

Task length was defined as the advertised estimated duration of a HIT. Table 6 presents the advertised hourly gender pay gaps for five categories of HIT length, which ranged from a few minutes to over 1 hour. Again, a significant advertised hourly gender pay gap was observed in each category.

https://doi.org/10.1371/journal.pone.0229383.t006

Finally, we conducted additional supplementary analyses to determine if other plausible factors such as HIT timing could account for the gender pay gap. We explored temporal factors including hour of the day and day of the week. Each completed task was grouped based on the hour and day in which it was completed. A significant advertised gender pay gap was observed within each of the 24 hours of the day and for every day of the week demonstrating that HIT timing could not account for the observed gender gap (results available in Supplementary Materials).

In this study we examined the gender pay gap on an anonymous online platform across an 18-month period, during which close to five million tasks were completed by over 20,000 unique workers. Due to factors that are unique to the Mechanical Turk online marketplace–such as anonymity, self-selection into tasks, relative homogeneity of the tasks performed, and flexible work scheduling–we did not expect earnings to differ by gender on this platform. However, contrary to our expectations, a robust and persistent gender pay gap was observed.

The average estimated actual pay on MTurk over the course of the examined time period was $5.70 per hour, with the gender pay differential being 10.5%. Importantly, gig economy platforms differ from more traditional labor markets in that hourly pay largely depends on the speed with which tasks are completed. For this reason, an analysis of gender differences in actual earned pay will be affected by gender differences in task completion speed. Unfortunately, we were not able to directly measure the speed with which workers complete tasks and account for this factor in our analysis. This is because workers have the ability to accept multiple HITs at the same time and multiple HITs can sit dormant in a queue, waiting for workers to begin to work on them. Therefore, the actual time that many workers spend working on tasks is likely less than what is indicated in the metadata available. For this reason, the estimated average actual hourly rate of $5.70 is likely an underestimate and the gender gap in actual pay cannot be precisely measured. We infer however, by the residual gender pay gap after accounting for other factors, that as much as 57% (or $.32) of the pay differential may be attributable to task completion speed. There are multiple plausible explanations for gender differences in task completion speed. For example, women may be more meticulous at performing tasks and, thus, may take longer at completing them. There may also be a skill factor related to men’s greater experience on the platform (see Table 5 ), such that men may be faster on average at completing tasks than women.

However, our findings also revealed another component of a gender pay gap on this platform–gender differences in the selection of tasks based on their advertised pay. Because the speed with which workers complete tasks does not impact these estimates, we conducted extensive analyses to try to explain this gender gap and the reasons why women appear on average to be selecting tasks that pay less compared to men. These results pertaining to the advertised gender pay gap constitute the main focus of this study and the discussion that follows.

The overall advertised hourly pay was $4.88. The gender pay gap in the advertised hourly pay was $0.28, or 5.8% of the advertised pay. Once a gender earnings differential was observed based on advertised pay, we expected to fully explain it by controlling for key structural and individual-level covariates. The covariates that we examined included experience, age, income, education, family composition, race, number of children, task length, the speed of accepting a task, and thirteen types of subtasks. We additionally examined the time of day and day of the week as potential explanatory factors. Again, contrary to our expectations, we observed that the pay gap persisted even after these potential confounders were controlled for. Indeed, separate analyses that examined the advertised pay gap within each subcategory of the covariates showed that the pay gap is ubiquitous, and persisted within each of the ninety sub-groups examined. These findings allows us to rule out multiple mechanisms that are known drivers of the pay gap in traditional labor markets and other gig economy marketplaces. To our knowledge this is the only study that has observed a pay gap across such diverse categories of workers and conditions, in an anonymous marketplace, while simultaneously controlling for virtually all variables that are traditionally implicated as causes of the gender pay gap.

Individual-level factors

Individual-level factors such as parental status and family composition are a common source of the gender pay gap in traditional labor markets [ 15 ] . Single mothers have previously been shown to have lower reservation wages compared to other men and women [ 21 ]. In traditional labor markets lower reservation wages lead single mothers to be willing to accept lower-paying work, contributing to a larger gender pay gap in this group. This pattern may extend to gig economy markets, in which single mothers may look to online labor markets as a source of supplementary income to help take care of their children, potentially leading them to become less discriminating in their choice of tasks and more willing to work for lower pay. Since female MTurk workers are 20% more likely than men to have children (see Table 1 ), it was critical to examine whether the gender pay gap may be driven by factors associated with family composition.

An examination of the advertised gender pay gap among individuals who differed in their marital and parental status showed that while married workers and those with children are indeed willing to work for lower pay (suggesting that family circumstances do affect reservation wages and may thus affect the willingness of online workers to accept lower-paying online tasks), women’s hourly pay is consistently lower than men’s within both single and married subgroups of workers, and among workers who do and do not have children. Indeed, contrary to expectations, the advertised gender pay gap was highest among those workers who are single, and among those who do not have any children. This observation shows that it is not possible for parental and family status to account for the observed pay gap in the present study, since it is precisely among unmarried individuals and those without children that the largest pay gap is observed.

Age was another factor that we considered to potentially explain the gender pay gap. In the present sample, the hourly pay of older individuals is substantially lower than that of younger workers; and women on the platform are five years older on average compared to men (see Table 1 ). However, having examined the gender pay gap separately within five different age cohorts we found that the largest pay gap occurs in the two youngest cohort groups: those between 18 and 29, and between 30 and 39 years of age. These are also the largest cohorts, responsible for 64% of completed work in total.

Younger workers are also most likely to have never been married or to not have any children. Thus, taken together, the results of the subgroup analyses are consistent in showing that the largest pay gap does not emerge from factors relating to parental, family, or age-related person-level factors. Similar patterns were found for race, education, and income. Specifically, a significant gender pay gap was observed within each subgroup of every one of these variables, showing that person-level factors relating to demographics are not driving the pay gap on this platform.

Experience is a factor that has an influence on the pay gap in both traditional and gig economy labor markets [ 20 ] . As noted above, experienced workers may be faster and more efficient at completing tasks in this platform, but also potentially more savvy at selecting more remunerative tasks compared to less experienced workers if, for example, they are better at selecting tasks that will take less time to complete than estimated on the dashboard [ 20 ]. On MTurk, men are overall more experienced than women. However, experience does not account for the gender gap in advertised pay in the present study. Inexperienced workers comprise the vast majority of the Mechanical Turk workforce, accounting for 67% of all completed tasks (see Table 5 ). Yet within this inexperienced group, there is a consistent male earning advantage based on the advertised pay for tasks performed. Further, controlling for the effect of experience in our models has a minimal effect on attenuating the gender pay gap.

Task heterogeneity

Another important source of the gender pay gap in both traditional and gig economy labor markets is task heterogeneity. In traditional labor markets men are disproportionately represented in lucrative fields, such as those in the tech sector [ 23 ]. While the workspace within MTurk is relatively homogeneous compared to the traditional labor market, there is still some variety in the kinds of tasks that are available, and men and women may have been expected to have preferences that influence choices among these.

To examine whether there is a gender preference for specific tasks, we systematically analyzed the textual descriptions of all tasks included in this study. These textual descriptions were available for all workers to examine on their dashboards, along with information about pay. The clustering algorithm revealed thirteen categories of tasks such as games, decision making, several different kinds of survey tasks, and psychology studies.We did not observe any evidence of gender preference for any of the task types. Within each of the thirteen clusters the distribution of tasks was approximately equally split between men and women. Thus, there is no evidence that women as a group have an overall preference for specific tasks compared to men. Critically, the gender pay gap was also observed within each one of these thirteen clusters.

Another potential source of heterogeneity is task length. Based on traditional labor markets, one plausible hypothesis about what may drive women’s preferences for specific tasks is that women may select tasks that differ in their duration. For example, women may be more likely to use the platform for supplemental income, while men may be more likely to work on HITs as their primary income source. Women may thus select shorter tasks relative to their male counterparts. If the shorter tasks pay less money, this would result in what appears to be a gender pay gap.

However, we did not observe gender differences in task selection based on task duration. For example, having divided tasks into their advertised length, the tasks are preferred equally by men and women. Furthermore, the shorter tasks’ hourly pay is substantially higher on average compared to longer tasks.

Additional evidence that scheduling factors do not drive the gender pay gap is that it was observed within all hourly and daily intervals (See S1 and S2 Tables in Appendix). These data are consistent with the results presented above regarding personal level factors, showing that the majority of male and female Mechanical Turk workers are single, young, and have no children. Thus, while in traditional labor markets task heterogeneity and labor segmentation is often driven by family and other life circumstances, the cohort examined in this study does not appear to be affected by these factors.

Practical implications of a gender pay gap on online platforms for social and behavioral science research

The present findings have important implications for online participant recruitment in the social and behavioral sciences, and also have theoretical implications for understanding the mechanisms that give rise to the gender pay gap. The last ten years have seen a revolution in data collection practices in the social and behavioral sciences, as laboratory-based data collection has slowly and steadily been moving online [ 16 , 24 ]. Mechanical Turk is by far the most widely used source of human participants online, with thousands of published peer-reviewed papers utilizing Mechanical Turk to recruit at least some of their human participants [ 25 ]. The present findings suggest both a challenge and an opportunity for researchers utilizing online platforms for participant recruitment. Our findings clearly reveal for the first time that sampling research participants on anonymous online platforms tends to produce gender pay inequities, and that this happens independent of demographics or type of task. While it is not clear from our findings what the exact cause of this inequity is, what is clear is that the online sampling environment produces similar gender pay inequities as those observed in other more traditional labor markets, after controlling for relevant covariates.

This finding is inherently surprising since many mechanisms that are known to produce the gender pay gap in traditional labor markets are not at play in online microtasks environments. Regardless of what the generative mechanisms of the gender pay gap on online microtask platforms might be, researchers may wish to consider whether changes in their sampling practices may produce more equitable pay outcomes. Unlike traditional labor markets, online data collection platforms have built-in tools that can allow researchers to easily fix gender pay inequities. Researchers can simply utilize gender quotas, for example, to fix the ratio of male and female participants that they recruit. These simple fixes in sampling practices will not only produce more equitable pay outcomes but are also most likely advantageous for reducing sampling bias due to gender being correlated with pay. Thus, while our results point to a ubiquitous discrepancy in pay between men and women on online microtask platforms, such inequities have relatively easy fixes on online gig economy marketplaces such as MTurk, compared to traditional labor markets where gender-based pay inequities have often remained intractable.

Other gig economy markets

As discussed in the introduction, a gender wage gap has been demonstrated on Uber, a gig economy transportation marketplace [ 20 ], where men earn approximately 7% more than women. However, unlike in the present study, the gender wage gap on Uber was fully explained by three factors; a) driving speed predicted higher wages, with men driving faster than women, b) men were more likely than women to drive in congested locations which resulted in better pay, c) experience working for Uber predicted higher wages, with men being more experienced. Thus, contrary to our findings, the gender wage gap in gig economy markets studied thus far are fully explained by task heterogeneity, experience, and task completion speed. To our knowledge, the results presented in the present study are the first to show that the gender wage gap can emerge independent of these factors.

Generalizability

Every labor market is characterized by a unique population of workers that are almost by definition not a representation of the general population outside of that labor market. Likewise, Mechanical Turk is characterized by a unique population of workers that is known to differ from the general population in several ways. Mechanical Turk workers are younger, better educated, less likely to be married or have children, less likely to be religious, and more likely to have a lower income compared to the general United States population [ 24 ]. The goal of the present study was not to uncover universal mechanisms that generate the gender pay gap across all labor markets and demographic groups. Rather, the goal was to examine a highly unique labor environment, characterized by factors that should make this labor market immune to the emergence of a gender pay gap.

Previous theories accounting for the pay gap have identified specific generating mechanisms relating to structural and personal factors, in addition to discrimination, as playing a role in the emergence of the gender pay gap. This study examined the work of over 20,000 individuals completing over 5 million tasks, under conditions where standard mechanisms that generate the gender pay gap have been controlled for. Nevertheless, a gender pay gap emerged in this environment, which cannot be accounted for by structural factors, demographic background, task preferences, or discrimination. Thus, these results reveal that the gender pay gap can emerge—in at least some labor markets—in which discrimination is absent and other key factors are accounted for. These results show that factors which have been identified to date as giving rise to the gender pay gap are not sufficient to explain the pay gap in at least some labor markets.

Potential mechanisms

While we cannot know from the results of this study what the actual mechanism is that generates the gender pay gap on online platforms, we suggest that it may be coming from outside of the platform. The particular characteristics of this labor market—such as anonymity, relative task homogeneity, and flexibility—suggest that, everything else being equal, women working in this platform have a greater propensity to choose less remunerative opportunities relative to men. It may be that these choices are driven by women having a lower reservation wage compared to men [ 21 , 26 ]. Previous research among student populations and in traditional labor markets has shown that women report lower pay or reward expectations than men [ 27 – 29 ]. Lower pay expectations among women are attributed to justifiable anticipation of differential returns to labor due to factors such as gender discrimination and/or a systematic psychological bias toward pessimism relative to an overly optimistic propensity among men [ 30 ].

Our results show that even if the bias of employers is removed by hiding the gender of workers as happens on MTurk, it seems that women may select lower paying opportunities themselves because their lower reservation wage influences the types of tasks they are willing to work on. It may be that women do this because cumulative experiences of pervasive discrimination lead women to undervalue their labor. In turn, women’s experiences with earning lower pay compared to men on traditional labor markets may lower women’s pay expectations on gig economy markets. Thus, consistent with these lowered expectations, women lower their reservation wages and may thus be more likely than men to settle for lower paying tasks.

More broadly, gender norms, psychological attributes, and non-cognitive skills, have recently become the subject of investigation as a potential source for the gender pay gap [ 3 ], and the present findings indicate the importance of such mechanisms being further explored, particularly in the context of task selection. More research will be required to explore the potential psychological and antecedent structural mechanisms underlying differential task selection and expectations of compensation for time spent on microtask platforms, with potential relevance to the gender pay gap in traditional labor markets as well. What these results do show is that pay discrepancies can emerge despite the absence of discrimination in at least some circumstances. These results should be of particular interest for researchers who may wish to see a more equitable online labor market for academic research, and also suggest that novel and heretofore unexplored mechanisms may be at play in generating these pay discrepancies.

A final note about framing: we are aware that explanations of the gender pay gap that invoke elements of women’s agency and, more specifically, “choices” risk both; a) diminishing or distracting from important structural factors, and b) “naturalizing” the status quo of gender inequality [ 30 ] . As Connor and Fiske (2019) argue, causal attributions for the gender pay gap to “unconstrained choices” by women, common as part of human capital explanations, may have the effect, intended or otherwise, of reinforcing system-justifying ideologies that serve to perpetuate inequality. By explicitly locating women’s economic decision making on the MTurk platform in the broader context of inegalitarian gender norms and labor market experiences outside of it (as above), we seek to distance our interpretation of our findings from implicit endorsement of traditional gender roles and economic arrangements and to promote further investigation of how the observed gender pay gap in this niche of the gig economy may reflect both broader gender inequalities and opportunities for structural remedies.

Supporting information

S1 table. distribution of hits, average pays, and gender pay gaps by hour of day..

https://doi.org/10.1371/journal.pone.0229383.s001

S2 Table. Distribution of HITs, average pays, and gender pay gaps by day of the week.

https://doi.org/10.1371/journal.pone.0229383.s002

• 1. United States Equal Employment Opportunity Commission, Lily Ledbetter Fair Pay Act of 2009 (2009), available at https://www.eeoc.gov/eeoc/publications/brochure- equal_pay_and_ledbetter_act.cfm, accessed on 11/12/2018.
• 2. United States Department of Labor (DOL), Office of Federal Contract Compliance Programs (OFCCP), Pay Transparency Nondiscrimination Provision, available at https://www.dol.gov/ofccp/PayTransparencyNondiscrimination.html , accessed on 11/12/2018.
• View Article
• Google Scholar
• 4. United States Department of Labor (DOL), Bureau of Labor Statistics (BLS) (2016) Women’s earning 83 percent of men’s, but vary by occupation. TED Econ Dly , available at https://www.bls.gov/opub/ted/2016/womens-earnings-83-percent-of-mens-but-vary-by-occupation.htm , accessed on 11/12/2018.
• 5. Davis A (2015) Women still earn less than men across the board (Economic Policy Institute, 2015), available at http://www.epi.org/publication/women-still-earn-less-than-men-across-the-board/ , accessed on 11/12/2018.
• 6. “Gender Pay Inequality: Consequences for Women, Families and the Economy” (Joint Economic Committee, 2016). [no author]
• 7. Hartmann H, Hayes J, Clark J (2014) “How Equal Pay for Working Women would Reduce Poverty and Grow the American Economy” (Institute for Women’s Policy Research, 2014).
• 8. OECD (2015) In it together: Why Less Inequality Benefits All (OECD Publishing, Paris) available at http://www.oecd.org/els/soc/OECD2015-In-It-Together-Chapter1-Overview-Inequality.pdf , accessed on 11/12/2018.
• PubMed/NCBI
• 16. Litman L, Robinson J (In Press) Conducting Online Research on Amazon Mechanical Turk and Beyond. Sage Publications.
• 17. Farrell D, Greig F (2016) Paychecks, paydays, and the online platform economy: Big data on income volatility. JP Morgan Chase Institute.
• 18. Kuek SC, Paradi-Guilford C, Fayomi T, Imaizumi S, Ipeirotis P, Pina P, Singh M (2015) The global opportunity in online outsourcing (World Bank Group, 2015) Available at http://documents.worldbank.org/curated/en/138371468000900555/pdf/ACS14228-ESW-white-cover-P149016-Box391478B-PUBLIC-World-Bank-Global-OO-Study-WB-Rpt-FinalS.pdf , accessed on 11/12/2018.
• 23. Bureau of Labor Statistics, U.S. Department of Labor, Labor Force Statistics from the Current Population Survey, Household Data Annual Averages. Employed persons by detailed occupation, sex, race, and Hispanic or Latino ethnicity, on the Internet at https://www.bls.gov/cps/cpsaat11.htm (visited 9/3/18).