Strong Gender Contract, Weak Institutions: Gender Pay Gap in Slovenia and Czechia

Abstract

This article investigates the dynamics of the gender pay gap (GPG) in two post-socialist economies, Czechia and Slovenia, during their capitalist transition and EU membership. Using regression analysis on large samples of linked employer–employee data to estimate the total GPG and within-job GPG and institutional analysis, we find that in both countries women earn significantly less than men for the same work. We interpret this as due to institutions, policies, and practices, including gender inequalities in the division of unpaid work and domestic care, which result in different but strong gender contracts in the two countries. The segregation of women and men into differently paid occupations, jobs, or workplaces does not explain much of the total GPG in Slovenia. In contrast, it explains about half of the total GPG in Czechia. We argue that this is due to weak labor market institutions in Czechia, recently weakening in Slovenia.

Introduction

In spite of legally banning pay differences based on gender many decades ago, recent analyses in Europe and elsewhere still show the existence of considerable gender pay gaps (GPGs) (Penner et al. 2023). Although recent findings highlight the need for further institutional and cultural change for elimination of the GPG (England, Levine, and Mishel 2020) and the significance of labor market institutions and wage-setting processes within and between workplaces (Tomaskovic-Devey et al. 2020), analyses of workplace and job-level GPGs remain scarce. This article contributes to scholarship on sources of GPGs by examining the development of labor market institutions and the gender contract as sources of the GPG and its components in two Central Eastern European (CEE) countries—Czechia and Slovenia. Based on the presumption that the GPG is an outcome of processes embedded in the institutional context of the post-state-socialist transition in Czechia and Slovenia we analyze GPG trends within the analytical framework of the development of the gender contract and labor market institutions. The gender contract represents varied, complex, and changing gender arrangements that underpin the organization of reproduction, labor divisions, and cultural representations (Pateman 1988). Gender equality legislation, high labor market participation, and high educational levels among women, conditions necessary for gender pay equality (Rubery and Grimshaw 2015), were achieved long ago in both Czechia and Slovenia. Legislation on equal wages for equal work exists in both countries, but despite this, the GPG was not eliminated, neither under state socialism nor after the transition to capitalism. Moreover, in both Czechia and Slovenia the GPGs and their institutional causes have been relatively neglected both scientifically and politically. Politically, in both countries solutions to gender pay equality are considered purely individual concerns, with women able to make “free choices” between paid/unpaid work and care.

We pose two main research questions:

1. Did women and men in Czechia and Slovenia earn equal wages for equal work over the 2002–2015 period?

2. What features of the institutional contexts—labor market institutions and/or gender contract—can help explain differences in the extent, trends, and components of the GPG in the two countries?

Czechia and Slovenia are interesting cases for comparative analysis for two main reasons. First, there are differences in both the extent of and trends in the GPG. Based on unadjusted GPG measurements, Czechia and Slovenia have long stood at opposite ends of the GPG continuum within the European Union (O’Reilly et al. 2015). The GPG in Czechia is persistently among the highest (21 percent in 2010; 18 percent in 2022) in the European Union, whereas the GPG in Slovenia has been one of the EU’s lowest but with a significant increase in the last decade (1 percent in 2010; 8 percent in 2022). Second, although both are post-socialist countries, there are many important formal and informal institutional differences in the gender contract and labor market institutions in these two countries. These differences allow us to understand not only the complex post-socialist gendered transition but are also likely important for understanding GPG formation more broadly.

We first briefly present previous research on GPGs in CEE and our theoretical framework, combining labor market and gender contract institutions and the situation in the two countries, followed by hypotheses. We then estimate the contribution of occupation, workplace, and job segregation to the overall GPG and the within-job GPG using linked employer–employee data (LEED) for every year from 2002 to 2015 to identify the trends and sources of GPGs. In our discussion of results, we provide an in-depth interpretation for understanding the GPG developments in the two countries through the use of the concepts of labor market institutions and gender contract.

Previous research taking a workplace perspective on GPGs has typically focused on Western and developed capitalist economies, such as the United States, Japan (Avent-Holt and Tomaskovic-Devey 2012), Germany (Abendroth et al. 2017), Norway (Petersen, Penner, and Høgsnes 2010), and Portugal (Card, Cardoso, and Kline 2016). This earlier research established that within-job GPG is typically small and most of the GPG is attributable to segregation (Ludsteck 2014). Moreover, it underscores how in the face of legal prohibitions against direct wage discrimination, gender segregation becomes more important for the GPG than within-job pay differences (Petersen and Saporta 2004). However, recent comparative research from the OECD found that, compared to Northern and Western European countries where child penalty and gender segregation is the main source of the GPG, in CEE countries social norms and therefore within-job GPG discrimination appear to be more important (Ciminelli, Schwellnus, and Stadler 2021, 27). This research therefore contributes to GPG research with a focus on two CEE countries and seeks to understand similarities and differences in their GPG trends with use of more than two decades of research on gendered labor market institutions and gender contract in Czechia and Slovenia. These two cases of post-state-socialist countries can help understand the persistence of gender contract for the within-job GPG and the important role of labor and wage protections for the sorting of women and men into differently paid jobs, occupations, and workplaces, which we interpret as important sources of the GPG in Czechia but also in Slovenia with the weakening of these institutions.

Previous research on the GPG in Czechia and Slovenia

Research on the GPG in post-state-socialist CEE countries is very scarce and mostly uses official aggregate statistics or surveys. It shows that despite strong overall wage equalization typical of the state-socialist regimes, wide GPGs existed under state socialism because women were placed in sectors and occupations not favored by the system (Brainerd 2000). With the start of transition to capitalism, Orazem and Vodopivec (2000), using a random sample of Slovenian pension and invalid funds, found that the GPG did not widen in Slovenia during the 1987–1992 period. Newell and Reilly (2001) found that by the middle of the 1990s, GPGs decreased in Czechia, Slovakia, Hungary, and Poland. Mysíková (2012) compared GPGs in these countries in 2008 with 2008 EU-SILC data and found the highest GPG in Czechia and that GPG is mostly explained by the fact that men are employed in better-paying occupations and workplaces than women. Brainerd (2000), analyzing GPGs in Eastern Europe (including Czechia) before and at the outset of the transition by means of Household surveys from 1998 to 1992, hypothesized that the breakdown of state control over enterprises might enable employers to discriminate against women more directly. By examining survey data, previous research also demonstrated the existence of a significant motherhood wage penalty in Czechia (Pospíšilová and Vohlídalová 2022; Pytliková 2015; Zajíčková and Zajíček 2020).

Analyses of data from the Information System on Average Earnings (ISAE) used here for the research on Czechia—from 1998 to 2002 by Jurajda (2003, 2005) and from 1998 to 2004 by Křížková, Penner, and Petersen (2010)—concluded that gender segregation represents the major source of the GPG and that the within-job GPG is also significant. In addition, Balcar and Gottvald (2016) found an increase of returns to white-collar occupations between 2008 and 2014, and Eriksson, Pytliková, and Warzynski (2013) and Tomaskovic-Devey et al. (2020) identified an increase in within- and especially between-firm inequality in Czechia caused by weakening labor market institutions, especially decentralized wage bargaining.

For Slovenia, Penner et al. (2012), Adams et al. (2016), and King et al. (2017) have used LEED from the years 1993–2007. Penner et al. (2012) found for the 1993–2007 period that there was a significant increase in the GPG and that the basis of gender inequality in wages had changed. Within-job inequality has been found to be the dominant source of GPGs, but the significance of sorting between occupations and establishments was increasing.

Gender contract and labor market institutions in CEE

Research suggests that discriminatory practices as well as the legacy of gendered cultures have a significant impact on GPG. Another stream of research shows that labor market institutions play an important role in shaping gender pay equality. To explain the GPG and its components at country levels, as well as differences between Slovenia and Czechia and their trends, our theoretical framework is based on the historically changing concept of the gender contract as well as examination of the development of labor market institutions in the two countries.

Gender contract

Anti-discrimination and equal pay legislation has been recognized as a necessary yet insufficient condition for the elimination of GPGs (King et al. 2017; Veldman and Timmer 2017). A recent analysis has shown that although considerable progress has been achieved, especially in the employment rates and educational levels of women, further advances towards economic gender equality would require substantial institutional and cultural change (England, Levine, and Mishel 2020, 6990). We argue that it is important to understand the varieties of formal and informal institutional frameworks for gender equality as well as the functioning of the gender contract. The concept of the gender contract is useful because it can place gender into context and allow comparative analysis. The gender contract concept corresponds to classifications of social policy regimes, which still lack focus on CEE countries, categorizing the gender contract into strong, moderate, and weak breadwinner systems (Lewis 1992), and to the concept of the gender order of society (Connell 2002). Researchers agree on the definition of the role of the state as a central entity institutionalizing gender power and therefore constructing the gender contract (Gottfried 2000; Connolly 2003).

Duncan’s (1995, 265) useful definition stresses that the gender contract includes formal as well as informal and less visible processes around social agreements on what men and women are, what they think and expect, and what they do. Therefore, the gender contract includes (1) legislative, institutional and policy regulations, (2) norms, values, and cultural attitudes, as well as (3) individual and family practices concerning the division of work and care (Ellingsæter and Jensen 2019). We follow this operationalization of the gender contract, which helps us unpack these particular elements and their differences in our analysis of the gender contract in Slovenia and Czechia. Consistency across these levels cannot be assumed a priori (Ellingsæter and Jensen 2019). Uneven development in one or more aspects of the gender contract, for example a policy change, can signify coming change in the gender contract as such.

Our analysis is informed by institutional and feminist literature and research that emphasizes a strong connection between cultural attitudes and individual/employer decisions around considerations such as motherhood (Glass and Fodor 2011) and welfare state design (Budig, Misra, and Boeckmann 2012). Attitudes to the division of paid/unpaid work and care within households influence the aspirations of women and men for paid work. The organizational processes, including hiring, promotions, and wage setting, are gendered (Acker 1990; Akchurin and Lee 2013), and gendered assumptions about care and paid/unpaid work are strongly connected to the undervaluation and underpayment of women’s work (Grimshaw and Rubery 2007). The disproportionate burden of care work performed by women, the root cause of GPGs (EIGE 2021), can be traced to cultural attitudes, and gendered assumptions and expectations. This influences the division of paid/unpaid work and care (Folbre 2006) within households but also welfare state provisions (Budig, Misra, and Boeckmann 2012) and organizational hiring, promotion, training, and wage distribution processes. Feminist sociologists recognize discriminatory practices that arise from company-level personnel management practices and cultural norms and bias about “appropriate” gender roles (Fodor, Glass, and Nagy 2019) which are interconnected to regulatory institutional settings.

Work–family policies are supposed to help women maintain employment during childbearing years and reduce the GPG (Budig, Misra, and Boeckmann 2016), yet broader cultural norms regarding gender, work, and family play a significant role in the formulation, institutionalization, and efficacy of these policies (Budig, Misra, and Boeckmann 2012, 164). Although there is significant literature combining sets of policies into indicators (Mandel and Semyonov 2005), a consensus is growing that different work–family policies are likely to have “divergent impacts” on the employment of women and mothers (Budig, Misra, and Boeckmann 2012, 2016; Cukrowska-Torzewska, 2017; Korpi, Ferrarini, and Englund 2013). The welfare state paradox that gender equality and work–life balance policies have adverse labor market consequences for women is not universally true because work–life balance policies also support gender equality (Barbieri and Cutuli 2016; Brady, Blome, and Kmec 2020). The specific setting and mix of policies is important. Budig, Misra, and Boeckmann (2016) find policies that maintain maternal labor market attachment through moderate-length maternity/parental leave, publicly funded childcare, the existence of paternity leave (as is present in Slovenia), and lower marginal tax rates on secondary earners are correlated with smaller motherhood wage penalties. Although the gender contract is a complex system (re)producing gender inequality at various levels (including segregation of women and men into different occupations, workplaces and jobs), we assume that it has an impact mainly on within-job GPGs.

Both Czechia and Slovenia experienced state socialism after the Second World War and the transition to market capitalism since the 1990s. However, their cultural, regulatory, and overall gender regime paths through these historical changes differed significantly. In Czechoslovakia, under strong Soviet dominance, the socialist (communist) project was imposed from above onto a previously well-developed industrial economy with a strong women’s movement, which was suppressed by the communist regime. By contrast, the Yugoslavian version of “self-management” socialism was accepted by many as an important part of the post-war economic development for this previously primarily agrarian country. The state-led emancipation of women through participation in paid work was experienced as an external pressure at least until the 1960s in Czechoslovakia (Wagnerová 2017). By 1968, women were fully integrated in paid work, constituting 46 percent of the workforce in Czechoslovakia and 41 percent in Yugoslavia (see Supplementary Table 1). A dissident movement and widely shared opposition to the lack of personal, political, and economic freedoms developed in both countries, yet public reflection of the persistent gender inequality and the women’s movement was missing under state socialism (Fodor 2002). While in both countries “state feminism” enabled a sharp increase in women’s participation in paid work after 1948 (Jogan 2004), it did not emancipate women from the burden of unpaid work (Havelková 1993, 2010) or give women full and equal participation in the public sphere (Jalušič 1999). The state-socialist emancipation of women was also experienced and dealt with differently in these two countries during the transitions to democracy and market capitalism. In Slovenia, both women’s emancipation through paid work and gender equality within the public agenda have been stressed (Jogan 2004). Moreover, the previously achieved gender equality was not dismissed in Slovenia during the transition period. In Czechia, however, the state-led emancipation and equalization project has been considered an integral part of the state-socialist oppression and therefore has been opposed as part of the transition to capitalism of the 1990s (Šiklová 1993; Wagnerová 2017). Therefore, in both countries formal gender equality developed in employment and the public sphere; however, while in Slovenia it has been widely accepted as an important modernization project, in Czechia, it has been neglected and/or opposed.

As a consequence of these developments, Czechia and Slovenia significantly diverge in terms of family and work–life balance policies and practices, with explicit familialism in Czechia and supported defamilialism in Slovenia (Javornik 2014). The gender contracts of these two countries, mainly the work–life balance institutions and the practice of division of work and care at home are key explanations for the different GPGs in the two countries. After a period of “emancipation from above” with a focus on participation of women in the labor market, the refamilization policy turn at the end of the 1960s in Czechoslovakia prolonged parental leave and limited the availability of childcare services for children under three years of age (Hašková and Klenner 2010). Contrary to Czechoslovakia, the Yugoslavian state provided the means for easier work–life balance and therefore persistently retained women in the labor market (Kanjuo-Mrčela and Sadar, 2011). In Czechia, this refamilization trend continued into the 1990s during the transition to capitalism (Hašková and Saxonberg 2016). At that time, parental leave was prolonged to three years, and, by 2012, all state-provided childcare for children under three was fully abolished. As a result, only 6 percent of children under three and 81 percent of those three–six years of age use formal (public and private) childcare services in Czechia (Eurostat 2019; for details, see Supplementary Table 1). Despite the discourse of free choice for women, these policies effectively deprive women in Czechia of childcare options and push them to stay on parental leave for longer periods.

In contrast to the striking limitation of childcare services availability in Czechia, the work–life balance legislation in Slovenia is generous when compared to other EU countries (Helfferich and Franklin 2019). The one-year fully paid parental leave and publicly subsidized preschool childcare services allow women to stay connected to their jobs and to perform paid work even when they have young children. While Czechia does not provide designated paternity leave (it only provides two weeks for fathers together with mothers on maternity leave during the first six postnatal weeks), in Slovenia 80 percent of fathers use the paid paternity leave and 20 percent use the unpaid leave. The share of children in Slovenia’s kindergartens between the age of three and mandatory school age is 91 percent—among the highest in Europe (Eurostat 2019).

The weak work–life balance support in Czechia, including long parental leave (with no motivation for fathers to use it), low childcare service availability, and low flexible working arrangement availability limits the choices of women to return to paid work earlier, and most women spend three years on parental leave (Hašková and Saxonberg 2016). In addition, women are viewed primarily as carers and men are considered breadwinners by about a half of the population in Czechia but only by a minority of population in Slovenia (ISSP 2012; see Supplementary Table 1). As women’s careers are interrupted by long care breaks, women predominate in the secondary Czech labor market and in precarious jobs (Hašková and Dudová 2017). Women in Slovenia return to the labor market much earlier after maternity leave, and their careers are not as strongly interrupted. While the Czech gender contract has not changed recently, part-time employment in Slovenia (14 percent women, 7 percent men in 2015; Eurostat) has increased in recent years, which might result in an increased GPG. Despite high, mostly full-time female employment,¹ women continue to bear the double burden of paid/unpaid work and care in both Slovenia and Czechia. Women spend almost double the time (31 hours a week) compared to men (17 hours a week) on unpaid work and care in Czechia and more than double (34 hours a week for women compared to 16 hours for men) in Slovenia (ISSP 2012; see Supplementary Table 1). The male breadwinner/female carer gender contract is strong in both countries and might impact the image of women in paid work leading to gender segregation and unequal wages.

Labor market institutions

Furthermore, we build upon analyses that show how labor market institutions importantly influence gender pay inequality. Besides centralized wage systems, which reduce overall wage inequality and the GPG, minimum wages also help protect workers at the bottom of the wage distribution, who are mostly women (Rubery and Grimshaw 2009). These studies provide orientation for more in-depth research on the role of occupations, workplaces, and institutions. Labor market deregulation increases gender inequality, and in particular reinforces the glass-ceiling effect (Ciminelli, Schwellnus, and Stadler 2021). Higher union density and wage coordination reduce the pay gap, with stronger equalizing effects again in the better-paid jobs (Perugini and Selezneva 2015). Research in developed market economies shows that a reduction of the GPG occurs in work environments that are more regulated either by the state or from the position of other actors in wage negotiations, such as unions and their stronger role in specific sectors, but also in workplaces where organizational processes are more formalized, such as large firms (Anderson and Tomaskovic-Devey 1995; Fuller and Cooke 2018; Triventi 2013).

Studies of the gender pay inequality/institutions covering CEE countries are very limited. Among them, Simon (2012) finds that “the gender wage gap is not significantly correlated across countries either with the minimum wage or with the collective bargaining coverage rate.” Christofides, Polycarpou, and Vrachimis (2010), using quantile regression approaches on an EU-27 sample for 2007, conclude that “unionism appears to be associated with reductions in the wage gap at the center of wage distribution.” Perugini and Selezneva (2015) show that deregulation of labor markets, in the form of less-regulated temporary jobs and weakened minimum-wage provisions, increases the gender earnings gap. On the contrary, a higher presence of unions and stronger bargaining coordination reduce the gap.

Labor market institutions that impact the employment of women and men and GPGs, as well as their development, significantly differ in the two countries. Differences can be found in the development in trust in these institutions. While trust in the justice system decreased significantly since the beginning of the 1990s in both countries (from 44 percent in 1990 to 37 percent in 2017 in Czechia and from 47 to 24 percent in Slovenia, EVS), trust in trade unions slightly increased in Czechia (from 25 to 29 percent) but decreased in Slovenia (from 27 to 21 percent), pointing to some decrease in the strong labor protection in this country (see Supplementary Table 1).

Although there is rather low level of trust in trade unions, their institutional role embedded in the Slovenian wage-bargaining system is still strong and rather positive for gender equality. The wage-setting regulation and the cognitive value of equality and their development during and after state socialism are the key elements of the labor market institutions with potential to explain the GPGs and their differences in the two countries. During the rather turbulent rush to capitalism in Czechia, emancipation, equality, and employment protections, including the role of trade unions, were rejected as a legacy of state socialism (Křížková, Penner, and Petersen 2010). Slovenia did not experience such strong and fast opposition to regulations and values connected to the state-socialist legacy and thus kept significant social and employment protections. Wage setting is highly decentralized in Czechia. Pay tables, classes, and corresponding wage tariffs regulate remuneration only within the public sector. In the private sector, wage bargaining is mostly at the firm or job level (Myant 2019). In Slovenia, not only is collective bargaining in the public sector highly centralized (the entire public workforce is covered by the single payment system in addition to collective agreements), but private-sector wage negotiations take place at the sectoral level. In spite of a recent deunionization trend in Slovenia (Stanojević, Poje, and Broder 2022), 79 percent of the workforce is covered by collective agreements (Stanojevič and Poje 2019), compared with only about 50 percent in Czechia (Myant 2019).

The public sectors in both countries cover the civil service, the military, the police, and almost all of the education and healthcare systems. The public-sector wages of both countries are highly coordinated, providing payment scales for all public-sector employees. However, due to much more extensive privatization in Czechia compared to Slovenia, the public sector is much smaller.

In the Czech private sector, enterprise- and individual-level wage bargaining plays important role, especially in professional occupations. There are no higher-level collective agreements in the Czech private sector (Myant 2019). This strongly decentralized collective bargaining has caused a deterioration of employment protections, and different wage levels between and within occupations and establishments (Tomaskovic-Devey et al. 2020), creating increased opportunity for women to be sorted into workplaces and jobs with lower wages. Collective bargaining in the private sector has also become increasingly decentralized in Slovenia, where there has been no general collective agreement since 2006, and the power of unions has declined. In sectors not covered by the collective agreements, bargaining on the individual level has been gaining in importance (Stanojević, Poje, and Broder 2022), most likely contributing to higher GPGs.

The minimum wage, which always has a significant impact, especially on women’s wages (Rubery and Grimshaw 2009), is higher in Slovenia (53 percent of average wage compared to 40 percent in Czechia; Eurostat 2019), even with recent increases in Czechia. Trade union coverage and density decreased substantially throughout the transition in both countries, but much faster in Czechia than in Slovenia (see Supplementary Table 1).

While both the Czech and Slovenian economies are still characterized by relatively low income inequality (Filauro and Fischer 2021), the different history and experience with state socialism and the ensuing transition is also mirrored in the two countries’ different attitudes toward equality. The European Social Survey shows that equality, in terms of wealth distribution and the refusal to accept inequality, and fairness, in terms of taking care of the poor and needy, are values that resonate much more strongly in Slovenia compared to the much more individualist Czechia (see Supplementary Table 1). This difference in cognitive values of equality and fairness might allow larger GPGs in Czechia than in Slovenia.

Data

LEED from population registers and other administrative sources have significant advantages for GPG research relative to official statistics and survey data.² First, they cover a majority of the population and (almost) all employees within workplaces. This allows for comparisons between the wages of women and men doing the same work at the same workplace, and by extension, the level of the GPG that arises due to sorting into different workplaces and jobs. Second, wages are not dependent on memory and personal estimates but are reported by employers or state authorities. Third, occupational information is rich and based on three-, four-, or five-digit international standard classification of occupations (ISCO) classifications usually containing more than 500 categories.

The data in both countries include more than 80 percent of all employees in organizations with ten or more employees and are administered by the Slovenian Statistical Office and by the Ministry of Labor and Social Affairs in Czechia. Our analysis covers prime full-time worker age (30–55 years).³ In 2015, the Czech sample of 1,511,629 individuals included 1,000,230 individuals employed in sex-integrated jobs. For Slovenia, the sample comprised 519,746 individuals, including 283,640 individuals in sex-integrated jobs.

For Czechia, our analysis uses data from ISAE. For the private sector, the survey takes place every quarter and is compulsory for selected firms with more than ten employees. The ISAE data cover about 51 percent of employees in Czechia and representatively cover 82 percent of Czech employees. For Slovenia, our analysis uses data from the register-based Labor Market Statistics and Statistical Register of Employment. The data are collected by the Slovenian Statistical Office and cover the entire Slovenian labor force as well as registered companies in both the private and public sector across all industries. Firm, establishment, and person identifiers were used to link the databases. Occupation is based on the Slovenian Statistical Office’s national standard classification of occupations, which changes over time: in 1999, the four-digit version of ISCO-88 was used; from 2000 to 2010 the six-digit ISCO-88 was used; and from 2011 to 2015 the four-digit ISCO-08 was used. We use crosswalks to harmonize the data to the current four-digit ISCO-08 categories.

Methodology

This analysis uses LEED, allowing us to match individual employees with their employers in Czechia and Slovenia, and follows Petersen and Morgan’s (1995) methodological approach to understand whether the GPG arises within jobs or because of sorting into different jobs (King et al. 2017; Křížková, Penner, and Petersen 2010; Penner et al. 2012). Our analysis is based on four sets of ordinary least squares (OLS) regression models. The first model adjusts only for age, age squared, and education and provides our baseline estimate of the overall GPG in each country (Model 1). In subsequent models, we compare only women and men who work in the same establishment (Model 2), women and men who work in the same occupation (Model 3), and women and men who work the same job (i.e. occupation-establishment unit; Model 4). We estimate these models separately by year to examine trends in the 2002–2015 period. Comparing results from Models 1–4 allows us to examine the degree to which the GPG arises from women and men receiving different wages for doing the same work for the same employer or from sorting women and men into different occupations, establishments, and jobs.

The equations estimated for our models follow the same general form, using four different specifications:

where the subscripts represent i for individuals, o for occupations, e for establishments, and t for years. The dependent variable is the logarithm of earnings (ln earnings_it) for individual i in year t, and the independent variables are collected in the vector x_it, which includes a constant, the gender, age, and age-squared of individual i and a series of indicator variables for the education of individual i. Models (2), (3), and (4) include fixed effects for establishments, occupations, and occupation-establishment units (represented respectively as η_et, η_ot, and η_oet) and the analytic sample for each of these models is restricted to gender-integrated fixed effect units. The subscripts to the ϑ parameters indicate that these are different coefficients, pertaining to different levels: basic adjustments (B), establishment (E), occupation (O), and occupation-establishment (OE).

We use the natural logarithm of yearly earnings as our dependent variable because, in Slovenia, we only have the sum of all taxed employment income, including wages, wage compensation for periods of absence, overtime, bonuses, and other compensation. In Czechia, we can distinguish contractual hourly wages and various compensation but, for comparative reasons, we also use earnings. Following standard conventions, the resulting coefficients are interpreted as the relative difference between the average male and female earnings; however, more formally, our estimates refer to the difference in relative geometric means for unlogged earnings—the absolute difference in the arithmetic means of logged earnings. We discuss the results mostly in reference to the GPG as a percentage value.⁴

Results

Below we present the key estimates of the GPG, and note the main similarities and differences between the two countries. In our discussion, we analyze our GPG estimates at the four levels of segregation within the institutional context of Czechia and Slovenia to identify particular and interconnected sources of GPGs.

Trends in GPG: GPGs did not decrease during the 2002–2015 period in either country

Table 1 reports the GPG in Czechia and Slovenia over the 2002–2015 period at the four levels corresponding to Models 1–4. We find that GPG adjusted only for education and age is significantly high and increasing in both countries and, in Czechia, is almost double that of Slovenia throughout the 2002–2015 observation period. Within-job (O-E) wage gaps are very high and similar in both countries. Gender segregation into occupations and workplaces plays a significant and growing role in Czechia. Although segregation contributes less to overall GPG in Slovenia, its impact has also grown over time. The remaining within-job GPG represents a significant proportion of the overall GPG.

Figure 1 reports these results graphically as percentages. GPG trends at the four levels are much more compressed in Slovenia than in Czechia, so that every level of segregation explains a significant part of the GPG in Czechia, whereas in Slovenia the within-job GPG is not very different from the base GPG, and segregation into occupations and establishments does not explain a big part of the country’s GPG. Therefore, in Slovenia the GPG is mainly caused by the job-level interactional gender contract, while in Czechia not only the job-level valuation and gender contract attitudes and institutions but also labor market institutions probably play a significant role.

In both countries, the previously described recent trends in the post-socialist transition could explain the general negative trend in GPG change, but some specificities are observable during the analyzed period. Over the 2002–2015 period, the base GPG adjusted by age and education increased in Czechia from −0.284 to −0.324, that is, from 25 to 28 percent. There were increases followed by slight decreases in 2005 and 2012 and an increase again in 2015. The base GPG increase after 2012, despite several increases of the minimum wage in the same period, is particularly concerning. In Slovenia, the base GPG increased from −0.140 to −0.190 (from 13 to 17 percent), with a sharp increase in 2008 followed by a decrease and an increase again to 17 percent in 2015. The decrease in Slovenia’s GPG in 2009 and 2010 could be the result of different factors. Due to extremely high inflation in 2007, social partners agreed upon a general collective agreement in 2008, covering an additional wage adjustment in the Slovenian private sector in 2007, 2008, and 2009. In 2009, due to the global economic crisis, many low-paid workers lost their jobs, many of whom were women, and, in 2010, the minimum wage was increased by 23 percent, increasing wages among many female workers with the lowest wage levels (Poje, 2019), reducing the GPG.

Within-job GPG is very high in both countries: women are paid significantly less for the same work compared with men

In response to our first research question, Czech women earn on average 14 percent less than Czech men who have the same detailed occupational category and who work in the same establishment (i.e. who have the same job). This large gap is very stable throughout the period, with only slight decreases in 2008 and 2010. The within-job GPG is similar in Slovenia, where it increased from 12 percent in 2002 to 15 percent in 2008 and decreased slightly again to 13 percent in 2015.

Institutions in developed Western countries that do not allow for the payment of different wages for the same work (Petersen and Morgan 1995) do not work the same way in Czechia and Slovenia. Both the gender culture and legal cases, where discrimination has been proven in countries with low within-job GPGs, do not allow for open discrimination (Ludsteck 2014). Although legislative regulation for equal wages exists in Czechia and Slovenia, its enforcement is weak due to the combination of a lack of wage transparency measures (see Supplementary Table 1) and a lack of legal cases dealing with wage discrimination that is probably based on low trust in the judiciary system as we documented above.

We understand the similarity of within-job GPG for the two countries as a result of a strong gender contract and norms that consider men the primary breadwinners and women the primary caregivers—the basis of employer bias and worker acceptance of the pay differences. The two countries are very similar in the high time investments of women in paid work (the employment rate of women is 73 percent in Czechia and Slovenia, with women working on average 42 hours per week in Czechia, and 40 hours in Slovenia; see Supplementary Table 1), as well as in the gendered asymmetry in time spent on household work and care. As explained above, women in both Czechia and Slovenia spend, on average, roughly double the time on household work and care compared with men. Despite significantly more traditional attitudes in Czechia than in Slovenia, the traditional gender contract regarding gendered breadwinning and caring is, in practice, strong in both countries.

About half of the base GPG in Czechia, and about three-quarters of the gap in Slovenia, can be attributed to within-job differences. This is in line with previous studies examining the role of occupational sorting in these two countries (see fig. 2; for percentage values, see Supplementary Table 2). Křížková, Penner, and Petersen (2010)^,5 found that occupational sorting explains only about half the wage gap in Czechia, and Penner et al. (2012) found that in Slovenia occupational sorting does not reduce women’s incomes and is sometimes even of benefit to women (in the 1993–2007 period). This underlines that the GPG in Slovenia is driven more by the devaluation of women’s work relative that of men (i.e. within-job inequality), whereas in Czechia the GPG could also be understood in terms of labor market institutions because women do not have access to the higher-paid jobs that men do.

As noted above, both attitudes about the division of responsibilities within households as well as family policies in the two countries differ significantly. While in Czechia the attitudes as well as the family policy setting is favorable to the familialist male-breadwinner model, in Slovenia they are rather in support of defamilization. Attitudes and policies influence women’s choices to use the long parental leave in Czechia in full, although due to the low availability of affordable childcare services and family taxation, it is a constrained choice. In addition, the sociohistorical analysis of women’s employment in Czechia shows that women are firmly integrated in the labor market, and, when they can, they return to paid jobs. The same is true for women in Slovenia, except that the institutional framework makes this return easier for them.

Occupation, workplace, and job sorting is very important in Czechia and plays an increasingly significant role in Slovenia

The main difference between the two countries in our results is the impact of occupation, workplace, and job sorting on the GPG (see fig. 2). The understanding of this feature is based on the same explanation as that describing the higher explanatory value of within-job GPGs in Slovenia. Due to the familialist institutional setting, women in Czechia are systematically relegated to lower-paid occupations, establishments, and jobs, which results in sorting playing a larger role in the GPG. In Slovenia, women were even employed in better-paid occupations and establishments than men at the beginning of the period. However, while the within-job GPG has increased only slightly in Slovenia over this period, gender-based sorting into different jobs is increasingly contributing to Slovenia’s GPG.

In 2002, the establishment-level GPG was higher than the base GPG, meaning that women were on average working in establishments with higher wages, but were paid less than men were in these establishments. In Slovenia, occupational and establishment segregation play smaller roles in the base GPG than in Czechia, likely because of the stronger Slovenian sectoral and centralized occupational wage bargaining, and regulation of firm and occupation pay differences. Decentralized wage bargaining in Czechia allows for significantly higher wage differences between occupations and establishments, with women often employed in those with lower wages.

Figure 2 shows trends in how each level of segregation contributes to the base earnings gap in the base GPG. Sorting contributes substantially less to the base GPG in Slovenia than in Czechia. At the beginning of the period, segregation to establishments and occupations lowered the Slovenian GPG, meaning that were women and men to have worked in the same establishments or occupations, the GPG would have been even higher. The contribution of these two types of segregation had sharply increased by 2008, which marks the beginning of the economic crisis, followed by a sharp decrease. Since 2009, however, the contributions of these segregations have been increasing again. In 2015, sorting into establishments contributed 11 percent to the base GPG, and occupational sorting contributed 17 percent. In Czechia, the contribution of sorting into establishments is very similar to occupational sorting, accounting for between 20 and 30 percent of the base GPG over the whole period. Segregation to occupations contributes slightly more, similar to Slovenia, increasing from 22 percent in 2002 to 30 percent of base GPG by 2015.

Conclusions

We used comparable, high-quality linked employer–employee data from 2002 to 2015 in Czechia and Slovenia to extend previous analyses and establish trends in GPGs, highlighting the contributions of occupational-, establishment-, and job-level sorting, as well as within-job GPGs. Tracing GPGs in Czechia and Slovenia over that period allowed us to follow the post-transition development and the impact of institutional settings. Understanding the differences and similarities between these two CEE countries could contribute to our understanding of GPG formation in general.

Our data contain detailed information on establishments and occupations, with a sufficient density of individuals in the same establishments to allow us to compute GPGs within these units. However, our data do not include information on the different channels through which gender influences earnings; therefore, we are not drawing causal conclusions on wage determination but use an institutional framework to interpret our results.

In response to our first research question, we find that women and men did not earn equal pay for equal work between 2002 and 2015. Within-job earnings gaps were similarly high in both countries. Women in both Czechia and Slovenia earned on average 13–14 percent less than their male co-workers.

The base GPG adjusted only for education and age is significantly high in both countries, and is not decreasing over the period. The Czech GPG is almost double the Slovenian GPG. Whereas the within-job GPG is similar, the role of occupational and establishment sorting in the GPG is very different in the two countries. While we cannot make causal claims, we demonstrated that institutions could explain the patterns of GPG we observe. Specifically, we have two main findings responding to our second research question.

First, the better institutional framework regarding labor market institutions—wage setting and gender equality support in Slovenia can explain why the Czech GPG is almost double the Slovenian GPG. The main difference between Czechia and Slovenia in our GPG estimates is in the importance of occupational, workplace, and job sorting for the GPG. Slovenia’s higher minimum wage, higher trade union coverage and density, centralized wage bargaining, and stronger employment protection mechanisms help minimize between-job inequality in Slovenia (Tomaskovic-Devey et al. 2020) and likely decrease the GPG. Second, the similar job-level GPG underscores how difficult it is to change the gender contract as a principle of social structuring. The fact that women are paid significantly lower wages than their male colleagues doing the same work in both countries, even though Slovenia had a much better formal institutional support for work–life balance for decades (before and after transition) shows the persistent strength of the informal institutional framework maintaining gender inequality. The rising similarities (in the extent of the GPG and the increasingly similar role of sorting) between Slovenia and Czechia could be explained by the decreasing differences in institutional frameworks in the two countries.

Our analysis showed how policies are strongly path-dependent and dependent on the overall interpretation of the state-socialist past and the transition to capitalism. The familialization trend that had already started in Czechia under state socialism continued during the transition to capitalism because it was in line with opposition to a paid work obligation, an ideology of emancipation and equality previously imposed by the state-socialist regime. Slovenian policies continued to support the attachment of women to the labor market and retained policies supporting a work–life balance as well as the importance of equality in society. Despite these different developments and a long history of well-functioning institutional frameworks supporting gender equality and a work–life balance in Slovenia, the traditional definition of the gender contract has survived in both countries. The gender contract was less influenced by differences under state socialism than expected, and although state socialism shifted expectations regarding gender equality in paid work, it never supported full equality or shifted expectations regarding unpaid work at home. This gender contract then, together with very low enforcement of equal-pay legislation and the non-existence of wage transparency measures, can be linked to the substantial within-job GPG in both countries. Although women’s paid work is necessary for households in both Czechia and Slovenia, expectations about paid and unpaid work continue to create a context in which gender pay differences are normalized.

The comparison of these post-state-socialist countries is important to the understanding of GPG (re)production mechanisms. Our results challenge some comparative frameworks that tend to categorize all post-socialist countries into a homogeneous set of welfare states (Pascall and Lewis, 2004). The two analyzed countries are places of observable ongoing social and economic experimentation and, due to the very different pace and character of their transition from state socialism to capitalism, are thus extremely interesting as contrasting cases for comparative analysis. Further research could extend the comparative analysis to other CEE countries as well as focus on the implementation of the new EU Directive on Pay Transparency.

Notes

Acknowledgments

The research and writing of this paper have been supported by the European Social Fund and state budget of the Czech Republic project International mobility of researchers of the Institute of Sociology, Czech Academy of Sciences II reg. n. CZ.02.2.69/0.0/0.0/18_053/0016983 and grant number CZ.03.1.51/0.0/0.0/15_009/0003702 and Rozvoj Výzkumné Organizace (RVO: 68378025), NPO “Systemic Risk Institute” LX22NPO5101.

Supplementary material

Supplementary material is available at Social Politics online.

Conflict of interest. None declared.

Data availability

The Ministry of Labour and Social Affairs of the Czech Republic and the Slovenian Statistical Office provided the data underlying this article with specific permission. Data will be shared on request to the corresponding author with permission of the Ministry of Labour and Social Affairs of the Czech Republic and the Slovenian Statistical Office.

References