Gender and job satisfaction: A European analysis in the light of gender ideology

Coron, Clotilde

doi:https://doi.org/10.59876/a-948z-3r1r

Many researchers have investigated the relationship between gender[1] and job satisfaction. Since Hodson’s (1989) seminal article, which defined the “gender-job satisfaction paradox”, i.e., the fact that women tend to report the same or higher levels of job satisfaction while facing worse working conditions, various studies have brought contradictory results.

Job satisfaction constitutes a critical element of employees’ relationships with work (Top, Akdere, & Tarcan, 2015) and even individuals’ lives (Linz & Semykina, 2012). It influences their performance, commitment, and intent to stay in the company (Brunetto, Teo, Shacklock, & Farr-Wharton, 2012; Flickinger, Allscher, & Fiedler, 2016). A considerable amount of research on this topic seeks to identify the determinants of job satisfaction, which include managerial support (Sawang, 2010), values and culture (Huang & Gamble, 2015), work-family conflict (Garner & Laroche, 2016), working conditions (Pujol-Cols & Dabos, 2018), and expected rewards (Linz & Semykina, 2012). Previous research has also highlighted several sociodemographic determinants, i.e., education (Bender & Heywood, 2006; Clark, 1997), level of responsibility (Smith, 2009), age (Clark, Oswald, & Warr, 1996), and sex (Clark, 1997; Hodson, 1989; Huang & Gamble, 2015).

Hodson’s (1989) seminal article highlighted the so-called “gender-job satisfaction paradox” (Hauret & Williams, 2017; Perugini & Vladisavljević, 2019), also known as the “paradox of the contented female worker” (Valet, 2018). This paradox suggests that women report a higher level of job satisfaction despite facing more precarious employment conditions than men (Ficapal-Cusí, Díaz-Chao, Sainz-Ibáñez, & Torrent-Sellens, 2018). Even though comparative analyses have shown that the paradox does not exist in every country (Eskildsen, Kristensen, & Westlund, 2004; Hauret & Williams, 2017; Sousa-Poza & Sousa-Poza, 2000), Clark (1997) explains this paradox by hypothesising that women have lower job expectations than men, which could result in higher satisfaction levels among women than men. This theory has been the focus of much work on the relationship between gender and job satisfaction in recent decades (Hauret & Williams, 2017).

Clark’s (1997) expectation theory has been rarely considered in the light of gender role socialization. However, as Grönlund and Oün (2018, p. 543) write, “another approach would be to widen the lens and consider job satisfaction more broadly in the light of gender role socialization and persistent gender inequalities.” The abundant and inconsistent literature about gender and job satisfaction has given little attention to the different levels of gender ideology that characterize national cultures. Gender ideology refers to the beliefs regarding the appropriate behaviour for women and men in work and family. Gender ideology has been shown to influence the importance attached by women and men to their work (Gaunt & Benjamin, 2007), their training (Kosteas, 2013), their careers (Sweet, Sarkisian, Matz-Costa, & Pitt-Catsouphes, 2016), and, more importantly for the purpose of this study, their expectations toward work (Damaske, 2011). Although the literature suggests that gender ideology is an essential piece of the jigsaw, it has yet to address gender ideology’s role in the gender-job satisfaction paradox as well as the national differences.

This paper complements the literature on the relationship between gender and job satisfaction and the expectation theory by studying the role of national gender ideology in this relationship. Our research uses two representative European surveys, namely, the 2015 European Working Conditions Survey (EWCS), which contains information about job satisfaction, and the 2017 European Values Study (EVS), which contains information about gender ideology. National gender ideology is computed with the EVS and is added to the EWCS, allowing the estimation of hierarchical models that combine individual characteristics (from the EWCS) and national gender ideology (from the EVS). The results show that in Europe women report a lower level of job satisfaction than men, although this gap varies between countries. National gender ideology plays an important role, as people living in countries with a more egalitarian gender ideology report a higher level of job satisfaction. National gender ideology also moderates the relationship between sex and job satisfaction in the sense that, in more egalitarian countries, the gender gap in job satisfaction (to the detriment of women) is more important than in countries with a more traditional gender ideology.

This paper contributes to the literature on both job satisfaction and gender, updating expectation theory and placing it in relation to national gender ideology. It also supports studies which expect the gender-job satisfaction paradox to be a transitory phenomenon, disappearing as soon as women are exposed to more equal gender settings. Finally, it argues that more equal gender settings have a paradoxical effect on the relationship between gender and job satisfaction. Although women may have better working conditions and thus higher job satisfaction in more egalitarian countries, the increase in job satisfaction associated with living or working in a country with an egalitarian gender ideology is more important for men than for women. In other words, women seem to benefit less than men from living or working in a country with a more egalitarian gender ideology.

The following section presents the literature review and the formulation of hypotheses, and the third section reports the data and the research design. Then the results are presented and discussed.

Literature review and hypotheses

Gender and job satisfaction: an unsolved debate

In his seminal article, Locke (1969, p. 316) defines job satisfaction as “the pleasurable emotional state resulting from the appraisal of one’s job as achieving or facilitating the achievement of one’s job values”. Locke’s definition has recently been refined by associating it with individuals’ perceptions of their jobs compared to their expectations (Drydakis, 2017). Perception can depend on the different facets of the work occupied (Christen, Iyer, & Soberman, 2006), as well as on personal values (Lange, 2015). Moreover, perception can refer to affective or cognitive elements (Pujol-Cols & Dabos, 2018) and can depend on individual emotional dispositions.

Since the publication of Locke’s article, conceptual subtleties have gradually been introduced to the literature in order to distinguish satisfaction with the different dimensions of work, e.g., remuneration, working conditions, development opportunities, working climate, working hours. Anglo-Saxon research has proposed a distinction between “work satisfaction” (referring to the intrinsic aspects of work, such as autonomy, self-development, and fulfilment) and “job satisfaction” (referring to the extrinsic aspects of work, such as remuneration, working hours, and contract) (Rose, 2003). This paper uses “job satisfaction” to include both intrinsic and extrinsic aspects of job and work, as seen in recent research (García, Gonzales-Miranda, Gallo, & Roman-Calderon, 2019; Hünefeld, Gerstenberg, & Hüffmeier, 2020).

The relationship between gender and job satisfaction remains unclear among scholars. Several articles have shown that women tend to declare higher levels of job satisfaction than men, even though they generally face more demanding working conditions and have less autonomy in their jobs (Clark, 1997; Ficapal-Cusí et al., 2018; Hodson, 1989). In particular, women face precariousness (Young, 2010), involuntary part-time work (Gottfried, 2009), glass ceiling (Coron, 2018; Ng & Sears, 2017), and lower wages (Acker, 1989; Bergmann, Scheele, & Sorger, 2019; Stojmenovska, 2019) in their workplaces more often than men. However, other studies have shown that in other countries, women do not report higher levels of job satisfaction than men. Table 1 gives a non-exhaustive list of studies that report inconsistent relationships between gender and job satisfaction.

The inconsistent results in the table can be partly explained by the countries in which the studies were conducted. Comparative research has shown that the gender differential in job satisfaction depends strongly on the country (Kaiser, 2005; Sousa-Poza & Sousa‑Poza, 2000). This result has been confirmed by a more recent study (Hauret & Williams, 2017), which shows that in some countries, women report higher levels of job satisfaction than men, whereas the inverse situation occurs in other countries. However, since most studies presented in Table 1 show that women report higher levels of job satisfaction, the first hypothesis of this study is formulated accordingly:

H1: In Europe, women report a higher level of job satisfaction than men.

The role of gender ideology

Two major explanations have been proposed by scholars to explain the “gender-job satisfaction paradox”. Clark’s (1997) expectation theory asserts that this paradox could be explained by women’s lower job expectations due to their poorer labour market condition in the past. In contrast, Hakim’s (2006) preference theory postulates that lifestyle preferences are the main explanatory factor of women’s choices and behaviours, as well as of the importance they grant to work and, therefore, their job satisfaction. However, Hakim’s assumption has since been partly questioned by studies showing that the “gender-job satisfaction paradox” cannot be attributed to the lower work centrality of women and their eventual work-family trade-offs (Grönlund & Öun, 2018).

Table 1

**An inconsistent literature (some examples)**

In line with expectation theory, the national variations of the gender-job satisfaction paradox can be partly due to the national variations concerning the role of women and the sexual division of tasks, both of which shape women’s expectations towards their job (Huang & Gamble, 2015; K. S. Lee, 2006) and their job satisfaction (Clark, 1997). These national variations correspond to two dimensions: gender equality, manifested in women’s rights, situation at work and political empowerment; and gender ideology, manifested in social attitudes and values towards women and gender equality (Otterbach, Sousa-Poza, & Zhang, 2021).

Previous research on the gender-job satisfaction paradox has granted more attention to gender equality, and more precisely professional gender equality, than gender ideology. Professional gender equality refers to several aspects, e.g., equality in the access to the labor market (Blossfeld et al., 2015), gender diversity within occupations (K. A. Bender, Donohue, & Heywood, 2005), equality in the access to responsibilities (Acker, 2009), equal pay (A.-F. Bender & Pigeyre, 2016). Previous research has also shown the relationship between professional equality and the sharing of tasks in the domestic sphere (Pedulla & Thébaud, 2015). Professional gender equality has not reached the same level in every European country. For example, Easter and Western European countries have been characterized by the professional equality between women and men, combined with an unequal share of domestic tasks; this combination of professional equality and domestic inequality has been called the “Soviet gender paradox” (Ashwin & Isupova, 2018). These national variations extend from the individual sphere to public policies, e.g., childcare and gender equality in the workplace (Annesley, Engeli, & Gains, 2015). The limited comparative research on the gender-job satisfaction paradox has shown that women who are exposed to more gender-equal settings at the national level have expectations towards work which are closer to their male counterparts’, which results in smaller gender gaps in job satisfaction than for women who are exposed to unequal settings (Perugini & Vladisavljević, 2019). This result supports Clark’s (1997) expectation theory. Previous research has suggested that the decline in women’s happiness, during decades when their concrete economic situation has improved, can be explained by an increase in their expectations. More precisely, Stevenson and Wolfers (2009) suggested that women’s expectations have increased faster than society was able to meet them. The disappearance of the gender-job satisfaction paradox is actually considered in one previous study as an indicator of labor market modernization and progress towards gender equality (Kaiser, 2005). These studies bring an important new piece to the jigsaw by showing the paradoxical effect of the progress towards gender equality, which improves women’s concrete situation but tends to decrease their satisfaction (Perugini & Vladisavljević, 2019). These previous studies investigate the concrete national situations related to gender equality, and not the national gender ideologies.

However, gender ideology can play an important role in the gender differences in job satisfaction, by shaping individual perceptions and preferences (Otterbach et al., 2021). Gender ideology can be measured thanks to representations of adequate behaviors and roles for women and men. Gender ideology refers to “individuals’ beliefs regarding the appropriate behavior for men and women in the context of work and family” (Gaunt & Benjamin, 2007, p. 343), and is constituted by beliefs and norms about the gendered division of paid and unpaid work (Grunow, Begall, & Buchler, 2018). While previous studies have applied notions such as gender egalitarianism (Otterbach et al., 2021), sex roles (Elm, Kennedy, & Lawton, 2001), gender roles (Bagger, Li, & Gutek, 2008), gender ideology (Davis & Greenstein, 2009; Gaunt & Benjamin, 2007), social role theory (Diekman & Eagly, 2000; Eibl, Lang, & Niessen, 2020), gender role orientation (Zhao, Zhang, & Foley, 2019), gender role attitudes (Kosteas, 2013), and gender role differences (Dolan, Bejarano, & Tzafrir, 2011), this study will draw on the notion of “gender ideology”. Indeed, the notion of ideology refers to a “set of beliefs about the social world and how it operates, and what values and ideals are worth striving for” (Alvesson, 1987, p. 13) and therefore links more directly to beliefs and representations. These beliefs, which are internalized through the early socialization process (Powell & Greenhaus, 2010), influence individuals’ behaviors in various contexts in turn, notably their working behaviors and decisions (Damaske, 2011; Dolan et al., 2011; Kosteas, 2013). Vespa (2009) presents gender ideology as a continuum between traditional and egalitarian gender ideology. Traditional gender ideology refers to the idea that men should take on the role of breadwinner and women should be responsible for household labor and childcare (Bagger & Li, 2012). In contrast, egalitarian gender ideology refers to the idea that women and men should have the same roles and responsibilities regarding breadwinning and household labor (Davis & Greenstein, 2009). Gender ideology varies considerably across countries, with some national cultures being more egalitarian and others more traditional (Beblo & Görges, 2018; Grunow et al., 2018).

To our knowledge, there exists no research which directly studies the role of national gender ideology in the gender-job satisfaction paradox. In this study, we try to bridge this gap by considering the role played by gender ideology. In line with expectation theory (Clark, 1997; Perugini & Vladisavljević, 2019), we assume that in countries with more egalitarian gender ideologies, women have higher job expectations and, thus, lower job satisfaction. H2 is formulated accordingly:

H2: Gender ideology moderates the relationship between gender and job satisfaction in the sense that women living in countries with more egalitarian gender ideology report a lower level of job satisfaction than women living in countries with more traditional gender ideology.

Data and methods

Data

Data for this study were derived from the 2015 European Working Conditions Survey (EWCS), which provides a rich cross-sectional dataset. The general survey is cross-sectional and includes responses from 35 European countries. The EWCS contains detailed information on various sociodemographic characteristics, including sex, age, educational attainment, and sector. Critical to the aims of this study, the EWCS also allows access to information on job satisfaction. It is possible to extrapolate the results obtained in this study using weights provided within the dataset. The initial sample is 43,850 individuals, but it was restricted to employees (permanent and temporary contracts), which represents 32,681 individuals.

This dataset was combined with information drawn from the European Values Study (EVS), which was used to measure national gender ideology. The EVS is a large-scale cross-national study that includes responses from 34 European countries. The survey contains eight items often used to measure gender ideology. The average gender ideology was computed in the EVS for each country, and the value (average national gender ideology) was added in the EWCS database: each individual was attributed the average gender ideology of their country. The initial EVS sample is 56,491 individuals and it was not restricted, as gender ideology is measured in an aggregate way at the national level.

The final sample is constituted by the EWCS-restricted dataset, in which only the countries covered by the EVS were kept: Austria, Bulgaria, Croatia, Czech Republic, Denmark, Estonia, Finland, France, Germany, Hungary, Italy, Lithuania, Netherlands, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden, Montenegro, Republic of North Macedonia, Serbia, Norway, Switzerland, and Albania[2]. This represents 23,839 individuals (12,311 women and 11,528 men). Table 2 gives the number of respondents by country.

Measures

Job satisfaction. Job satisfaction was measured based on five items drawn from the EWCS survey: “On the whole, are you very satisfied, satisfied, not very satisfied or not at all satisfied with working conditions in your main paid job?”; “Considering all my efforts and achievements in my job, I feel I get paid appropriately”; “My job offers good prospects for career advancement”; “I receive the recognition I deserve for my work”; “I generally get on well with my work colleagues”. Therefore, those five items correspond to the dimensions usually looked at in research about job satisfaction (working conditions, pay, career, recognition, and relationships) (Grönlund & Öun, 2018; Jones, Jones, Latreille, & Sloane, 2009; Kosteas, 2013; Zou, 2015). In the EWCS survey, they are assessed with 4-point and 5-point Likert scales. The alpha coefficient was 0.74, higher than the suggested reliability level of 0.70. Job satisfaction was computed as the geometric mean of the five items (controlling for the fact that some are on a 4-points Likert scale and others on a 5-points Likert scale).

Sex. The sex of the respondents is reported by the interviewer, with three modalities: “Male”/“Female/”Don’t know”. Within the initial sample, 22,101 individuals are coded as “females”, 21,740 as “males”, and nine as “Don’t know”. For the purpose of the analysis, the nine individuals coded as “Don’t know” had to be deleted from the sample. In the following text, and in compliance with APA guidelines, sex is referred to as “women” and “men”.

National gender ideology. National gender ideology was measured by aggregating the individual responses from the EVS survey at the national level[3]. Seven items from the EVS survey were used to assess gender ideology. These seven items correspond to the degree of agreement with the following statements: “When a mother works for pay, the children suffer”; “A job is all right but what most women truly want is a home and children”; “All in all, family life suffers when the woman has a full-time job”; “A man’s job is to earn money, a woman’s job is to look after the home and family”; “On the whole, men make better political leaders than women do”; “A university education is more important for a boy than for a girl”; “On the whole, men make better business executives than women do.” All answers conform to Likert scales (1: “agree strongly” – 4: “disagree strongly”). Thus, a high score corresponds to an egalitarian gender ideology, and a low score corresponds to a traditional gender ideology (Gaunt & Benjamin, 2007). These or similar items are found in other studies on gender ideology (Otterbach et al., 2021) and, although the unidimensionality of gender ideology has been questioned (Grunow et al., 2018; Kroska, 2000), this measure of gender ideology is the most common among previous research (Davis & Greenstein, 2009; Gaunt & Benjamin, 2007; Grunow et al., 2018; Zhao et al., 2019).

Control variables (individual level). Several variables identified in previous research as potential antecedents of job satisfaction were controlled for. Previous research has shown that part-time work can have inconsistent effects on job satisfaction (Fenton-O’Creevy, 1995). In the EWCS survey, part-time work corresponds to the binary question: “Do you work part-time or full-time?”. The ability to determine one’s own schedule can also influence positively job satisfaction, especially among women (Dikkers, van Engen, & Vinkenburg, 2010; Huang & Gamble, 2015). In the EWCS survey, respondents have to explain how their working time arrangements are set (“They are set by the company with no possibility for changes” / “I can choose between several fixed working schedules determined by the company” / “I can adapt my working hours within certain limits” / “My working hours are entirely determined by myself”). The type of contract also influences job satisfaction, as shown in Wilkin’s (2013) meta-analysis. The sample was restricted to three types of contracts: “Contract of unlimited duration”/“Contract of limited duration”/“A temporary employment agency contract”, and this variable was controlled for. Managerial status can also influence job satisfaction and has been controlled for in previous research (Huang & Gamble, 2015). In the EWCS survey, the question “How many people work under your supervision” was used to assess the managerial status (“Manager”/“Non-Manager”). Regarding demographics, previous research has examined the influence of age (R. Lee & Wilbur, 1985) and has shown a U-shaped relationship between age and job satisfaction (Clark et al., 1996), which justifies including both age and age-squared among the control variables (Zou, 2015). Other studies have shown the non-linear influence of tenure (Tlaiss & Mendelson, 2014). Therefore, in this research, tenure and tenure-squared were controlled for. Since Clark’s (1997) seminal article, previous research has underlined that educational attainment strongly influences job satisfaction, with higher levels of education associated with lower job satisfaction (Perugini & Vladisavljević, 2019). In the present study, educational attainment corresponded to seven modalities (“Primary education”/“Lower secondary education”/“Upper secondary education”/“Post-secondary non-tertiary education”/“Sort-cycle tertiary education”/“Bachelor or equivalent”/“Master or Doctorate”). The EWCS survey also includes information on overskilling, a feeling which has been shown to influence job satisfaction (Rose, 2003): “Which of the following statements would best describe your skills in your own work?”, the modalities being “I need further training to cope well with my duties”/“My present skills correspond well with my duties”/“I have the skills to cope with more demanding duties”. Variables related to family situation were also added, as most gender stereotypes addressed here deal with the separation between the domestic and the professional spheres. Previous research has shown that parents tend to report higher levels of job satisfaction than non-parents (Grönlund & Öun, 2018). In the EWCS survey, the respondents answer a question about the number of children living in their household, and this variable was coded in three modalities (“No child” / “1-2 children” / “3 or more”). The marital status was also added, indicating whether the respondent lives with a partner or spouse, or not. Indeed, living with a partner can reduce work-family conflict, which can be better for job satisfaction (Buonocore & Russo, 2013). Finally, variables related to the company were added, such as sector (“Private sector”/“Public sector”/“Joint private-public organization or company”/“Not-for-profit sector or NGO”/“Other”), and company size (“1”/“2‑9”/“10-249”/“250+”/“Don’t know”).

Control variables (national level). In addition, control variables at the national level were also added in one of the models, in order to check whether countries’ socioeconomic context and situation regarding gender equality have an influence on job satisfaction and on the gender-job satisfaction paradox. Several scholars have suggested to take into account national gender regimes (Orloff, 1993). Besides, the national concrete situations regarding gender equality can differ from the national gender ideologies (Otterbach et al., 2021), which makes it all the more interesting to add a model including concrete gender equality in order to check whether the role of national gender ideology stays the same. Three variables were added to account for the socioeconomic situation: GDP per capita (source: IMF 2015), HDI (Human Develoment Index, source: UNDP 2015), and the employment protection level for regular contracts (source: OECD 2015) (Caroli & Godard, 2016; Muñoz de Bustillo & de Pedraza, 2010). Six variables were added to control for the gender situation, as suggested by Orloff (1993): The existence of quotas to promote women’s political participation (source: OECD 2014), the feminization rate of national parliaments (source: OECD 2014), the GDI (Gender Development Index, source: UNDP 2015), the gender pay gap (source: EIGE 2014), the feminization rate of board presidents (source: EIGE 2015), and the feminization rate of board members (source: EIGE 2015). Some of these variables were not available for every year, so the available years closest to 2015 (date of the EWCS survey) were chosen.

Table 3 gives the main descriptive statistics for the whole sample and the women’s and men’s samples.

Table 4 indicates the average gender ideology for each country. The standard deviations cannot be reported because the measure is aggregated at the national level.

Table 4

**Gender ideology by country (in descending order for gender ideology values)**

Table 4 indicates that the countries with the most egalitarian gender ideology are, in descending order: Norway, Sweden, and Denmark. The countries with the most traditional gender ideology are, in ascending order: Lithuania, Slovakia, and Montenegro.

As almost all the variables are categorical, the correlation matrix reported in Table 5 only includes job satisfaction, age, age-squared, tenure, tenure-squared, and gender ideology.

Table 5 indicates that gender ideology is positively and significantly correlated with job satisfaction, which seems to justify the inclusion of this dimension in the models.

Methods

In order to test hypotheses H1 (“In Europe, women report a higher level of job satisfaction than men”) and H2 (“Gender ideology moderates the relationship between gender and job satisfaction in the sense that women living in countries with more egalitarian gender ideology report a lower level of job satisfaction than women living in countries with more traditional gender ideology”), first, descriptive statistics with significance tests were conducted. Three generalized hierarchical (multilevel) linear models with job satisfaction as the dependent variable were estimated in the second step. The first one includes sex, national gender ideology, and control variables at the individual level, but does not include the mediation of gender ideology. It corresponds to the following specifications:

The second one includes the same variables and adds the interaction effect between sex and national gender ideology. It corresponds to the following specifications:

The third one includes all the variables and interaction effects of Model 2 and also includes variables at the national level. It corresponds to the following specifications:

Multilevel modeling is beneficial when phenomena can be explained at the individual and national levels (Renkema, Meijerink, & Bondarouk, 2016). Specifically, the present study is based on two levels of analysis: individuals are embedded in countries characterized by various gender ideologies. Multilevel modeling allows the researchers to determine the percentage of variation explained at the national level and is the best way to combine two levels of analysis within the same framework (Goldstein, Browne, & Rasbash, 2002; Smyth & Steinmetz, 2008). All the analyses were performed with the SAS software.

Main results

Descriptive statistics

Table 6 gives the average job satisfaction for women and men in each country. GLM tests were conducted in order to test the statistical significance of the sex differences in each country.

Table 6 indicates that women and men have nearly the same level of job satisfaction in almost all European countries. In only seven countries (out of 26), there is a significant difference between women and men, and in all those countries, women report lower job satisfaction than men. In no country do women have higher job satisfaction than men. The national variations in terms of job satisfaction are important, with levels of job satisfaction varying from 3.29 (Slovakia) to 3.95 (Denmark) among women. This invalidates hypothesis H1 (“In Europe, women report a higher level of job satisfaction than men”): in Europe, there are some countries where women and men have the same average level of job satisfaction, whereas others where women have a lower level of job satisfaction, but there is no country where women have a higher level of job satisfaction.

Models

In order to test more precisely hypotheses H1 (“In Europe, women report a higher level of job satisfaction than men”) and H2 (“Gender ideology moderates the relationship between gender and job satisfaction in the sense that women living in countries with more egalitarian gender ideology report a lower level of job satisfaction than women living in countries with more traditional gender ideology”), with control variables related to the working situations, we use multilevel modeling with job satisfaction as the dependent variable. The results of the models are shown in Table 7.

Model 1 shows that, even after controlling for individual and organizational characteristics, women report a lower level of job satisfaction than men. This means that in 2015, the gender-job satisfaction paradox does not exist anymore in Europe. This invalidates hypothesis H1. The multilevel model also allows us to measure the weight of the national level. In this case, the national level accounts for 4.43% of the individual variations in job satisfaction.

Table 6

**Average job satisfaction by sex**

Model 2, which includes the moderation of gender ideology and sex, indicates that, when controlling for the moderating role of gender ideology, sex does not significantly affect job satisfaction anymore. Gender ideology influences significantly and positively job satisfaction, which means that in countries with a more egalitarian gender ideology, people report higher job satisfaction. Finally, the interaction effect between gender ideology and sex is also significant, but negative. Figure 1 shows the interaction effect. This indicates that in countries with a more egalitarian gender ideology, the gender gap of job satisfaction (to the detrimental of women) is more important than in countries with a more traditional gender ideology. This means that hypothesis H2 (“Gender ideology moderates the relationship between gender and job satisfaction in the sense that women living in countries with more egalitarian gender ideology report a lower level of job satisfaction than women living in countries with more traditional gender ideology”) is not validated because it is more complex than what was expected: women living in countries with more egalitarian gender ideology report, as their male counterparts, a higher level of job satisfaction, but they seem to benefit less from this egalitarian gender ideology than men. In other words, the increase in job satisfaction associated with the fact of living or working in a country with an egalitarian gender ideology is more important for men than women.

Figure 1

**Interaction effect between sex and national gender ideology**

In Model 3, which includes control variables at the national level, gender ideology is no more significant[4]. This result could indicate that gender ideology plays as a proxy of national situation (socioeconomic context and situation regarding gender equality). However, the interaction effect between gender ideology and sex has the same effect as in Model 2.

Concerning the control variables, in the three models, part-time tends to reduce job satisfaction, which brings another piece to the inconsistent literature about the effect of part-time on job satisfaction (Fenton-O’Creevy, 1995). The ability to determine one’s own schedule influences positively job satisfaction, as in previous research (Jijena Michel & Jijena Michel, 2015). Contracts of limited duration and agency contracts decrease job satisfaction, as in previous research (Wilkin, 2013). In conformity with previous research, age has a negative influence and age-squared has a positive influence (Zou, 2015), tenure has no effect, whereas tenure-squared also has a positive influence. Contrary to previous research (Clark, 1997), higher educational levels are associated with higher job satisfaction in this study. Being overskilled decreases job satisfaction, as in previous research (Rose, 2003). The number of children does not influence job satisfaction, whereas living without partner decreases job satisfaction. The company’s sector does not influence job satisfaction, and only the company size “2-9” has a significant effect, job satisfaction being higher than for other company sizes. Among the variables at the national level, only the feminization rate of board members plays an effect and affects negatively job satisfaction.

Finally, we have examined the sensitivity of the results to changes in the definition of the dependent variable. Notably, instead of using the arithmetic mean of the five items to measure job satisfaction, we have conducted a principal component analysis in order to compute job satisfaction as a linear combination of the five items. The results of the models stay roughly the same.

Discussion

This study based on a European representative sample of workers shows that in 2015, the gender-job satisfaction paradox has disappeared in Europe, where women tend to report lower levels of job satisfaction than men. National gender ideology plays an important role, as people living in countries with a more egalitarian gender ideology report a higher level of job satisfaction; however, the increase in job satisfaction associated with living or working in a country with an egalitarian gender ideology is more important for men than for women. The results of this study allow us to discuss previous research on three main aspects.

The first aspect relates to the gender-job satisfaction paradox in itself. The results presented here show that in 2015, this paradox no longer exists in Europe. Women report the same level of job satisfaction as men in most countries, and in the seven countries where the gender difference is significant, women report a lower level of job satisfaction. In addition, Model 1 indicates that, when controlling for individual characteristics, women report a lower level of job satisfaction than men. This result contradicts some existing studies conducted during the same period. Notably, Perugini and Vladisavljević’s (2019) study conducted on data from the 2013 EU SILC (Survey on Income and Living Conditions) shows that on average and all other things being equal, women report higher job satisfaction than men. Even more surprisingly, if we focus on one country, France for example, we find that in Perugini and Vladisavljević’s study and the SILC survey, women report higher job satisfaction than men, while the opposite is true in the EWCS survey. This difference in results can be explained by the way job satisfaction is measured. While it is measured as a single question (“How do you evaluate your current job?”) in the SILC, the EWCS provides a perhaps more precise measure by combining both a general question (“On the whole, are you very satisfied, satisfied, not very satisfied or not at all satisfied with working conditions in your main paid job?”) and more specific questions (e.g., “Considering all my efforts and achievements in my job, I feel I get paid appropriately”). The current study also provides results which contradicts Andrade et al.’s (2019) research, based on the 2015 wave of the International Social Survey Program, but this difference can be explained by the geographical perimeter which is not the same. The result of the current study tends to support previous research that asserted that the gender-job satisfaction paradox would be a transitory phenomenon (Clark, 1997; Kaiser, 2005; Perugini & Vladisavljević, 2019). However, our result goes further than that, as the gender-job satisfaction paradox has not only disappeared but has been replaced by a situation in which women tend to report lower levels of job satisfaction than men. This conclusion seems to support Stevenson and Wolfers’ (2009) general assumption that women’s expectations have increased faster than society has been able to meet them, specifically with respect to job satisfaction. The results found here are consistent with those found by Huang and Gamble (2015) in the Chinese retail sector and Webber and Rogers (2018) in the UK academia.

To our knowledge, no other study has shown that people report higher levels of job satisfaction in more egalitarian countries. Earlier studies that examine the relationship between gender, job satisfaction, and national situation related to gender equality only give attention to the gender gap in job satisfaction and not to the level of job satisfaction in itself (Kaiser, 2005; Perugini & Vladisavljević, 2019). Besides, previous comparative research on the gender-job satisfaction paradox has focused on the concrete national situations regarding gender equality, and not on gender ideology (Perugini & Vladisavljević, 2019). One explanation for the higher levels of job satisfaction in more egalitarian countries might involve a connection between gender ideology and other socioeconomic and institutional indicators. A more egalitarian gender ideology would reflect a better socioeconomic and institutional situation more broadly, in which workers’ rights are more respected and guaranteed, and in which working conditions are more favorable. This interpretation is supported by the fact that gender ideology becomes unsignificant when variables about the socioeconomic context and the national situation regarding gender equality are controlled for (Model 3). This latest result suggests that, even though gender equality and gender ideology do not totally overlap (Otterbach et al., 2021), they play a similar role in shaping the gender-job satisfaction paradox. An important aspect of the current study is that gender ideology was measured through direct questions about the individual beliefs regarding women and men’s roles, with questions closed to other research projects as the GLOBE project (Otterbach et al., 2021), and then they were aggregated to produce an average national gender ideology. Research on values and cultures has relied heavily on the aggregation of individual beliefs to produce a national index (Hofstede, 2011; Leung & Morris, 2015). Although this strategy has been criticized because it postulates a form on consensus within nations (Alvesson, 1987), the different indices computed in the current study (rwg, ICC(1), ICC(2)) justify the aggregation of individual perceptions at the national level.

The third important aspect of the study is that the positive influence of a more egalitarian national gender ideology on job satisfaction is stronger for men than for women. This result gives the impression that women benefit less than men from living and working in an egalitarian country, yet creates a new sort of paradox, which to our knowledge, has not been pointed out by previous research. Indeed, it shows that in more egalitarian countries, women are even less happy than men than in more traditional countries, whereas previous research has mostly shown that women are not happier than men in more egalitarian countries (referred to as the disappearance of the usual gender-job satisfaction paradox) (Kaiser, 2005; Perugini & Vladisavljević, 2019). One way to explain this new paradox would be to use an updated version of the expectation theory, relating it to gender ideology and gender role socialization, thus answering the call made by Grönlund and Oün (2018). In countries with a more egalitarian gender ideology, the differences between women and men are supposed to be less important. Therefore, women tend to compare themselves not only to the group of women but to the whole group of workers. Even when inequalities remain at lower levels, women are consequently less satisfied than men. This interpretation is supported by the fact that the positive influence of a more egalitarian national gender ideology on job satisfaction remains stronger for men than for women even after controlling for national socioeconomic contexts and situations regarding gender equality (Model 3). This explanation relies on the notion of reference group, following Clark and Oswald (1996) and Clark (1997), who introduced the idea that relative working conditions are more important determinants of job satisfaction than absolute working conditions. Another explanation could be that in less egalitarian countries, women participate less in the labor market, and thus there exists a selection bias which suggests that only the women most interested in working are employed in a job. This selection bias could explain part of this new paradox.

Conclusion

Overall, this study offers new empirical insights into the gender-job satisfaction paradox. It incorporates more recent data than previous studies and adds a new perspective by relating it to national gender ideology, updating the expectation theory and showing that women do not have lower expectations than men anymore, especially in more egalitarian countries. The study avoids part of the common method bias because the data come from two different databases, national gender ideology being computed with responses from individuals (EVS) other than the ones who answered about their job satisfaction (EWCS survey). There is a slight difference in dates between the two databases (EWCS data: 2015; EVS data: 2017), but it should not jeopardize the validity of our results, as the changes in gender ideology are very slow. Besides, national variables were added and come from other sources of data (e.g., OECD, EIGE).

Despite these advantages, the current study suffers from several limitations. First, other studies have shown that it is important to take into account not only the current national situation regarding gender equality, but more so the earlier situation, e.g., during the respondents’ childhood (Perugini & Vladisavljević, 2019). Further research could extend this strategy to national gender ideology. Second, this study does not include possible gender effects related to part-time work, flexible work schedule, marital status or the number of children. It could be of interest to extend this research by studying the possible interaction effects between gender and those control variables. Third, the current study measures national gender ideology with the EVS survey, whereas other datasets could be used, such as the GLOBE project. Future research could check the eventual differences between the EVS survey and the GLOBE survey and see whether the results stay the same when measuring national gender ideology with the GLOBE survey.

Gender and job satisfaction: A European analysis in the light of gender ideologyGenre et satisfaction au travail : une analyse européenne au prisme des stéréotypes de genreGénero y satisfacción laboral: un análisis europeo a través del prisma de los estereotipos de género

Abstract

Résumé

Resumen

Literature review and hypotheses