Gender differences in fairness evaluations of own earnings in 28 European countries Open Access

While gender gaps in labor force participation, working hours, and wages have narrowed, substantial gender differences in labor market involvement persist. Not only do women typically work fewer hours than men, they also specialize in different occupations and industries (Goldin 2014; Blau and Kahn 2017). Moreover, gender differences in pay, disadvantaging women, exist in all European societies (Boll and Lagemann 2018). These objective inequalities exist simultaneously with evidence that women evaluate their jobs and their pay more favorably than their male peers (Buchanan 2005; Davison 2014; Mueller and Kim 2008; Mueller and Wallace 1996; Valet 2018; Pfeifer and Stephan 2019). This counter intuitive finding is referred to as the ‘paradox of the contented female worker’. Its explanation frequently invokes fairness-based arguments, with empirical investigations focusing on job or pay satisfaction, only indirectly assessing the role of distributive justice (Mueller and Kim 2008). Satisfaction, however, is a much vaguer attitude incorporating a variety of factors that extend beyond the work context (e.g. Judge and Larsen 2001; Judge et al. 2017). Fairness, on the other hand, is closely linked to underlying ideas of a fair pay and behavior to address experienced unfairness (Jasso et al. 2016). Building on early contributions from social exchange and equity theories (Adams 1963; 1965; Homans 1961), distributive justice theory (Jasso 1978; Jasso et al. 2016) is concerned with the question of how fairly or unfairly rewards are allocated and how individuals react when they identify an unfair situation. If individuals perceive unfairness, they aim to leave or alter the unfair situation. Accordingly, unfairness is associated with a number of severe consequences ranging from increased turn-over rates and absenteeism to higher wage demands in the future (Colquitt et al. 2001; Pfeifer and Stephan 2019). If women differ systematically from men in their fairness evaluations of earnings this may, in turn, contribute to the persistence of gender pay gaps.

Therefore, we contribute to the literature investigating whether gender differences exist in the fairness evaluation of own earnings. We draw on the 9th round of the European Social Survey (ESS), collected in 2018/2019 across 28 European countries. The ESS asked individuals, who received income from work, to evaluate the fairness of own labor earnings and, if they perceived unfairness, to identify a fair level of earnings.

First, we ask if women in Europe indeed evaluate their own gross labor income as more fair than men, as suggested in the literature (Pfeifer and Stephan 2019; Valet 2018). Beyond a dichotomy of fair/unfair, fairness evaluations obtained in the ESS R9 allow for a closer investigation into the degree of experienced unfairness by distinguishing between small and large deviations from fairness. First, we investigate if (a) women are more or less likely than men to evaluate their earnings as fair, and (b) if the degree of experienced unfairness differs by gender. Second, we investigate if men and women differ in the fair markup they report. We define fair markup as the relative distance between the level of earnings that respondents declared to be fair for themselves to their reported actual earnings.

Overall, we do not find evidence for the paradox of the contented female worker in Europe. On the contrary, while we do not observe gender differences in the likelihood to perceive own earnings as fair in most countries, we find evidence that women experience more intense levels of unfairness in 15 out of 28 countries. Irrespective of more severe perceived unfairness in some countries, women ask for the same or even smaller fair markups in the majority of countries, thus suggesting that even if women are not as ‘contented’ as suggested in the literature, their conceptions of a fair markup suggest low potential for pay fairness evaluations to contribute to reducing gender pay gaps.

Contrary to the observation that women have been and are objectively worse off compared to their male counterparts on the labor market, studies on the evaluation of work and pay find that, overall, women evaluate their jobs and their pay more positively than men (Clark 1997; Buchanan 2005; Davison 2014; Mueller and Kim 2008; Mueller and Wallace 1996; Valet 2018; Pfeifer and Stephan 2019; Williams et al. 2006). This research on gender differences in job and pay evaluations is often subsumed under the umbrella term ‘paradox of the contented female worker’. Most studies focus on vague attitudes of satisfaction – either with regard to job satisfaction (e.g. Clark 1997; Buchanan 2005; Hauret and Williams 2017; Mueller and Kim 2008) or pay satisfaction (e.g. Davison 2014), lacking a clear link to actual wage gaps and subsequent behavior. Although pay fairness is the more specific attitude and closely linked to behavioral responses, only a few single-country studies focus on gender differences in fairness evaluations. Based on German panel data, both Valet (2018) and Pfeifer and Stephan (2019) find that women in Germany evaluate their own pay as more fair than their male counterparts.

Fairness evaluations of own pay refer to the subjective sense of (un)fairness that is awakened when one's own actual pay is compared to the level of pay that individuals would consider ‘fair’ for themselves. If actual and fair pay are matched, pay is evaluated as fair, but if actual and fair pay are in disagreement, then individuals evaluate their own pay as unfair. If actual pay is higher than what an individual would consider fair for themselves, a state of unfair over-reward exists; if inverse, then a state of unfair under-reward is identified (Jasso 1978; Jasso et al. 2016).

A state of unfairness – whether under- or over-rewarded – triggers consequences aimed at reducing the stress associated with receiving less (or more) than is fair by either altering or leaving the unfair reward situation (Adams 1965; Jasso et al. 2016). The consequences of an unfair reward situation are expected to depend on the intensity of unfairness, i.e. more severe deviations from ‘perfect fairness’ are more likely to trigger behavioral responses (Adams 1965). With respect to unfairly low pay, behavioral responses may include asking for higher pay (in an attempt to narrow the gap between actual and fair pay) or leaving the job; these draw a close link from fairness evaluations and the underlying idea of fair pay to future wage development and labor market behavior (e.g. Jones and Skarlicki 2003; Pfeifer and Stephan 2019). As such, systematic gender differences in pay fairness where women evaluate their pay as more fair, may help to explain persistent gender pay differentials as it may make women less prone to ask for higher pay or leave unfairly paid jobs than men. Pfeifer and Stephan (2019), for example, show that women in Germany are more likely to perceive their hourly wage as fair, but both men and women who reported being unfairly paid are more likely to experience wage growth in the subsequent years, albeit with women receiving smaller markups. This is in line with findings that women have lower pay expectations and negotiate less often, while men are more confident and more successful in negotiations (Kolb 2009; Major and Konar 1984; Mazei et al. 2015; Pelham and Hetts 2001; Sauer et al. 2021).

Lower fair pay conceptions as well as lower subsequent wage growth for unfairly paid women is consistent with the existence of gender status beliefs that paint men as more competent and, thusly, more deserving of higher pay (Ridgeway 2011). More specifically, living and working in contexts where earnings inequalities exist to the disadvantage of women translates into higher earnings for men to be seen as deserved or fair. This is underscored by studies showing that both male and female respondents in Germany consider lower earnings for female workers to be fair (Adriaans et al. 2020; Auspurg et al. 2017; Sauer 2020). Beyond implicit gender biases, the proportionality of inputs and rewards is central for pay fairness (Adams 1963; 1965) and comparisons with similar pay referents are crucial in shaping ideas of fair pay for oneself (Jasso et al. 2016; Kim et al. 2019). The ‘differential job inputs’ argument (e.g. Mueller and Kim 2008) suggests that women perceive their own contributions to be lower and, therefore, expect lower pay in return. Men, on the other hand, tend to overestimate their performance and expect higher returns resulting in more positive pay fairness evaluations among men (Major et al. 1984; Pelham and Hetts 2001). Research further shows a preference for same-gender referents (Major and Forcey 1985) and gendered workplaces and occupational structures facilitate that women compare themselves to other women (Valet 2018), both of these contribute to lower fair pay expectations among women and, subsequently, more positive pay fairness evaluations.

In summary, theoretical explanations suggesting more positive pay evaluations among women rest on the idea that implicit biases, same-sex pay referents, and gendered perceptions of job inputs result in lower fair pay conceptions among women. Accordingly, we not only study gender differences in fairness evaluations but also investigate if conceptions of fair earnings differ between men and women in Europe. Specifically, we investigate the fair markups reported, i.e. the relative distance between the earnings they currently receive and the earnings they consider fair. This fair markup on the currently received earnings serves as a proxy measure for earnings demands that may result from the experience of unfairness. Indeed, if women show lower fair markups than men, this may contribute to persisting earnings inequalities. If, however, women show higher markups, this highlights potential for future reductions in gender earnings gaps.

Our empirical analyses rely on the 9th round of the European Social Survey (ESS). The ESS is a biennial multi-national survey collecting data on attitudes and behavior via cross-sectional, probability samples representative of all persons aged 15 and over, resident within private households in each respective country. The 9th round was fielded in 2018/2019 across 29 European countries and includes a novel module collecting information on the fairness evaluation of own earnings (ESS Round 9 2018). Existing research suggests that countries may differ in the prevalence of the contented female worker paradox (Hauret and Williams 2017; Sousa-Poza and Sousa-Poza 2000), but evidence on the fairness of earnings is lacking. Studying gender differences in fairness evaluations in European countries using harmonized data allows us to address this gap. Given our focus on income from work, our analysis sample is restricted to individuals from 28 European countries that are between 18 and 68 years old, currently receiving income from work.¹ Our sample size ranges from 225 in Montenegro to 998 in Germany, averaging 548.5 observations per country.² All analyses use weights provided by the ESS (pspwght).³

We address two specific principle questions: first, whether men and women differ in their fairness evaluations (i.e. the probability of perceiving actual own pay as fair and the intensity of perceived unfairness) and, second, whether gender differences exist in fair markups (i.e. the relative distance between actual and fair own pay). These two questions require different analytical frameworks and econometric strategies that we discuss in the following sections.

Respondents who receive income from work were asked to evaluate their gross labor earnings on a scale that ranges between −4 (extremely unfair, unfairly low earnings) and +4 (extremely unfair, unfairly high earnings) where the midpoint of the scale – 0 – indicates fair earnings. Positive values are associated with unfair over-payment evaluations and negative values with unfair underpayment, while the scale points can be used to indicate the intensity of unfairness, where −4 indicates extreme unfairness of the under-reward type and +4 indicates extreme unfairness of the over-reward type (Jasso 2015). Similar single-item fairness evaluations that allow to distinguish both the direction (under-reward or over-reward) and intensity of unfairness are used in national as well as international surveys and feature prominently in the distributive justice literature (see, for example Adriaans et al. 2020; Auspurg et al. 2017; Jasso 2015; Schieman and Narisada 2021, Schneider and Valet 2017).

Regression coefficients read as the partial effect of each covariate on the probability of perceiving own earnings as fair. For the female dummy, a positive (negative) statistically significant coefficient indicates that women, on average, evaluate their earnings more (less) often as fair than their male peers in the specific country under analysis.

We then expand our model using the full fairness evaluation scale as the dependent variable. Specifically, we want to test whether there are gender differences in the intensity of perceived unfairness. Given the low prevalence of perceived unfair over-payment, we restrict the sample to only those observations where earnings are declared to be either fair or unfairly low. We run OLS regressions using the re-scaled fairness evaluation ranging from 0 (fair earnings) to 4 (extremely unfair, unfairly low earnings) as the dependent variable, controlling for socio-demographic and job-characteristics as described in equation (1). For the female dummy, positive (negative) and statistically significant coefficients indicate that women experience higher (lower) intensity of unfairness versus their male peers. As before, we apply Heckman correction to account for potential selection issues.

Subsequently, we provide an additional model specification, expanding equation (3) to include the declared intensity of unfairness. This extended specification allows us to test if gender differences in fair markup exist once we control for the fact that men and women might report different unfairness intensities.

In the following sections, we report the main results of our analysis, focusing on the effect of gender on the likelihood to evaluate own earnings as fair, the intensity of perceived unfairness, and the fair markup.

We present and discuss findings on hours and wages – which capture crucial endowment differences between men and women – in Online Appendix Figures A.3.1–A.6. In line with previous research, we find that wages play an important role in determining fairness evaluations and fair markups. The main descriptive statistics for all analyzed countries are provided in Online Appendix Table A.1.

The female coefficient in our model specification is not statistically significant in the majority of countries analyzed. The exceptions are Czech Republic, Denmark, Finland, Germany, Iceland, Lithuania, Netherlands, Serbia, Sweden, and Switzerland, where we observe a significant negative effect. This means that, in these countries, women are less likely to evaluate their earnings as fair than men. Effect sizes range between −.07 and −.18, meaning that women are between 7 and 18 percentage points less likely to evaluate their earnings as fair versus their male peers, depending on the country. Most notably, in none of the countries under investigation do we observe a significant positive gender effect that is consistent with the contented female worker paradox.

After accounting for the intensity of unfairness, we confirm statistically significant gender differences in 15 European countries: Croatia, Czech Republic, Denmark, Estonia, Finland, France, Germany, Iceland, Lithuania, Montenegro, Poland, Portugal, Serbia, Sweden, and Switzerland. In these countries, women report more severe unfairness with regard to own earnings than their male peers. Consistent with findings on the likelihood to evaluate own earnings as fair, in none of the 28 countries analyzed do women evaluate their own earnings more positively than men. All together, these results are not consistent with the paradox of the contented female worker described in previous research.

For both fairness evaluation specifications – dummies indicating fair earnings and unfairness intensity – and consistent with existing research on gender differences in job satisfaction, self-selection proves to not qualitatively affect results (Clark 1997; Hauret and Williams 2017).¹⁵

The paradox of the contented female worker is suggested as contributing to persistent gender differences in pay following the notion that if women are more content with their current earnings, they are less likely than their male peers to ask for higher earnings. However, our analysis of 28 European countries finds that women perceive more intense unfairness than men in 15 countries. If women perceive more intense unfairness, this offers potential to reduce the gender pay gap, assuming it translates into demands for higher pay among women. To investigate this potential for higher pay demands among women, we investigate gender differences in fair markups next.

The number of countries where we observe significantly lower markups among women increases from 3 to 8 after controlling for unfairness intensity. This pattern suggests that, in these countries, similar intensities of perceived unfairness are associated with lower fair markups for women than men.¹⁷ Women in these countries report fair markups that are between 4 and 37 percentage points lower than what men report. The observed negative effect of gender on fair markup is consistent with the explanation that women hold stereotyped beliefs attributing lower deserved pay to women, which may help explain why women tend to gain less from wage negotiations (Kolb 2009; Mazei et al. 2015; Sauer et al. 2021).

The higher fair markups observed among women in Croatia, Denmark, Finland, Germany, and Lithuania vanish after including the intensity of perceived unfairness into the model, thus providing supportive evidence that women in these countries – on average – evaluate their pay as more unfair and, accordingly, report a larger markup necessary to compensate their unfair pay situation. Once differences in the intensity of unfairness evaluations are accounted for, men and women in these countries do not seem to differ in their idea of a fair markup. In Spain, the observed higher fair markups among women remain after controlling for the intensity of unfairness.

Overall, in all countries studied, more intense unfairness evaluations translate into higher fair markups, thus supporting the notion of distributive justice theory that larger gaps between fair and actual pay lead to more severe unfairness evaluations. Gender's role is not unequivocal. For women in Croatia, Denmark, Finland, Germany, and Lithuania, more intense unfairness evaluations align with higher fair markups, indicating potential for reducing actual gender pay gaps if women successfully negotiate pay they consider fair for themselves. However, in the majority of countries, women seek the same level of fair markups, or even lower than men, irrespective of the perceived intensity of unfairness (Figure 4), suggesting a pessimistic assessment of the potential to reduce actual gender pay differentials through the channel of higher fair markup conceptions among women. In Online Appendix B, we discuss in detail the effect of hourly wages and weekly working hours on declared fair markups.

Distributive justice theory posits that individuals perceiving that they are unfairly underpaid are motivated to either change or leave this unfair situation (Jasso et al. 2016). Indeed, research shows that workers who perceived themselves as unfairly paid experienced stronger wage growth in the immediate future (Pfeifer and Stephan 2019). Thus, individual unfairness evaluations may drive the reduction of actual labor market inequalities. However, this is only the case if those who are disadvantaged perceive their own earnings as unfair and seek higher wages. Research on gender differences in pay satisfaction and job satisfaction finds that women evaluate their job and pay more positively than men, despite being objectively worse off (Mueller and Kim 2008). This ‘paradox’ may contribute to the persistence of gender pay gaps if, indeed, contented women do not seek higher earnings. However, empirical evidence on the ‘paradox of the contented female worker’ mostly focuses on job satisfaction measures – a rather vague and broad attitude related to a range of work- and non-work-related factors (Judge and Larsen. 2001; Judge et al. 2017); further, studies typically focus on single countries.

Thus, we extend the literature on gender differences in pay and job evaluations with evidence on gender differences in pay fairness based on recent survey data from 28 European countries, allowing us to study a very specific attitude toward individual pay that is closely linked to labor market behavior. Specifically, we investigate three outcomes: (1) the likelihood to evaluate one's own earnings as fair; (2) the intensity of perceived unfairness; and (3) the fair markup on actual earnings. Overall, we only observe gender differences in fairness evaluations of own earnings in a subset of countries. In particular – contrary to suggestions raised by the ‘paradox of the contented female worker’ – women in Europe do not evaluate their earnings more positively than men. Specifically, controlling for job and socio-demographic characteristics, women are not more likely to perceive their own earnings as fair. In 10 out of 28 countries, gender predicts being paid fairly and in all these countries women are actually less likely to think their own earnings are fair. This finding is supported when we consider the intensity of perceived unfairness: in 15 out of 28 countries, women perceive more intense levels of unfairness.

Two main arguments may serve to explain this seemingly contradictory evidence. First, comparative studies already suggest that gender differences in job satisfaction may not be as prevalent cross-nationally as single-country studies suggest (e.g. Sousa-Poza and Sousa-Poza 2000; Hauret and Williams 2017). Second, pay fairness is a much more specific attitude than generic job satisfaction. It could be that women evaluate the fairness of their pay similar to – if not more negatively than – men but are generally more satisfied with their jobs. Questions on job satisfaction in ESS round 10 will help to investigate whether the reported patterns of fairness evaluations are mirrored for job satisfaction in European countries.¹⁸

Beyond fairness evaluations, we find that, controlling for job and socio-demographic characteristics, higher intensity of perceived unfairness translates into higher fair markups in all studied countries. This underlines the role that unfairness evaluations may serve for subsequent wage growth. With respect to gender differences in fair markups, evidence is mixed. In Croatia, Denmark, Finland, Germany, Lithuania, and Spain, women report higher fair markups, highlighting potential for reducing actual pay gaps if women can enforce their ideas of a fair pay in wage negotiations.

Once we control for the intensity of reported unfairness in exploring the gender-fair markup link, we find that in 8 out of 28 countries – Czech Republic, Estonia, Finland, Hungary, Italy, Montenegro, Serbia, and Portugal – women actually report lower fair markups. However, in none of these countries do we find women to evaluate their earnings more favorably than men when asked directly, rejecting the notion that women are ‘contented’ in these countries while they may still hold gender biased beliefs attributing higher fair pay to men (e.g. Auspurg et al. 2017). Moreover, in 19 countries, no gender differences in fair markup are observed. This should also not be interpreted as women being ‘content’ but rather as women requiring the same markups as men to ‘compensate’ the unfairness they experience. Overall, our findings suggest that it is not sufficient for women to realize they are underpaid to address existing inequalities; especially if women who are discontented with their pay still lag behind their male peers in their underlying wage conceptions and resulting wage demands.

Generally, the subset of countries where women report more intense unfairness is heterogeneous, ranging from countries with high unadjusted pay gaps and a large share of part-time workers, like Germany, to countries with much lower actual wage gaps and part-time shares, like Portugal or Croatia. In Italy – with a comparatively small unadjusted gender wag gap – women report lower markups than men, just like as in Hungary – which has large gender wage gaps (European Commission 2018). Generally, there seems to be no clear link between country-level gender inequalities in labor market outcomes and the observation of gender differences in fairness evaluations. Nonetheless, some groups of countries with similar patterns may be identified: If gender differences in fair markups are observed in Eastern European countries, they usually show lower fair markups among women.¹⁹ This is consistent with more traditional gender norms observed in Eastern Europe (European Commission 2018). Similarly, women in Italy and Portugal report lower fair markups and both countries may also be described as showing traditional gender-role regimes (Cuttilo and Centra 2017; Tavora and Rubery 2013).

Overall, there is no clear country pattern with regard to gender differences in fairness evaluations of earnings. However, research suggests that gender differences in fairness evaluations vary systematically within countries – both on the level of occupations (e.g. Auspurg et al. 2017; Valet 2018) as well as between regional units (e.g. Sauer 2020; Lang and Groß 2020). The potential for differentiated gender effects within countries also points to a shortcoming of our analyses. The ESS is a general population survey that monitors political, moral, and social attitudes across Europe. Thus, only a subset of the full sample is currently employed and only basic information related to employment collected. Consequently, the potential for in-depth analysis of within country differences is limited. Further research should expand on our results, for example, by using country-specific data sources that focus on employment and provide a longitudinal perspective. Follow-up research may determine how elements, like country-specific occupational structures, collective bargaining agreements, and pay transparency regulations, affect gender differences in the evaluation of own earnings, thus tackling the question of why women do not seek higher fair earnings. In extending our findings on the evaluation of own earnings, future research should also address potential discrepancies or overlaps with fairness evaluations of other's earnings (e.g. Auspurg et al. 2017; Sauer 2020), which may help to link fairness considerations to preferences for policy responses to address economic unfairness (e.g. Cavaillé and Trump 2015).

Overall, based on recent survey data, we do not find that women in Europe are ‘content’ with their earnings; but we also find that women do not ‘translate’ their perceived unfairness into higher fair markups. Our findings paint a grim picture of heightened unfairness evaluations among women as a driving force to reduce gender pay inequalities. Further, in some countries, we find counteracting gender effects, where women ask for lower fair markups than men. Future research into the role of attitudes in contributing to persistent gender pay gaps could move from asking ‘Why are women content with their pay?’ to ‘Why do discontented women not ask for higher pay?’

We are greatful for comments on earlier versions of this paper from members of the project ‘Perceptions of Inequalities and Justice in Europe’ (PIJE) as well as from Peter Valet, Alexandra Fedorets, and Mattis Beckmannshagen. Additionally, earlier versions of this work were presented at the SASE 2021 Virtual Conference of the Society for the Advancement of Socio-Economics as well as at the 2021 Virtual Annual Meetings of the American Sociological Association; we acknowledge helpful feedback from participants of both conferences as well as from anonymous reviewers and the editor of European Societies. Jule Adriaans received financial and intellectual support for this paper from the Socio-Economic Panel Study (SOEP) at DIW Berlin.

No potential conflict of interest was reported by the author(s).

Matteo Targa is a researcher at the Socio-Economic Panel Study (SOEP) and a doctoral candidate at the Graduate Center at DIW Berlin. His research focuses on labor economics with a special focus on job-related determinants of economic inequality.

Jule Adriaans is a researcher at the Chair of Social Inequality and Social Structure Analysis and a doctoral candidate at the Bielefeld Graduate School in History and Sociology (BGHS) at Bielefeld University. Her research focuses on the perception and evaluation of inequalities and justice with a special focus on a comparative European perspective.