Gender pay gap in education and estimation of unexplained part: The case of the Czech Republic

Abstract

Background

The gender pay gap has long been recognized as an important indicator of labor market inequality. This article focuses on wage disparities between men and women in education sector.

Objective

The aim of the paper is to estimate the unexplained component of the gender pay gap in the education sector in the Czech Republic using the Oaxaca–Blinder decomposition, and to examine how this unexplained component evolved over the decade between 2012 and 2022.

Methods

To estimate the unexplained part of gender pay gap, EU-SILC data were used and the Oaxaca–Blinder decomposition was applied.

Results

The raw gender pay gap in education was approximately 31% in 2012 and 26% to the disadvantage of women in 2022. Differences in personal and firm-level characteristics explain only part of the wage gap. The unexplained part reached approximately 14% in 2012 and 10% in 2022.

Conclusions

The gender pay gap in the education sector, as well as its unexplained component, decreased between 2012 and 2022. However, wage inequalities still persist. By quantifying the unexplained part of the pay gap, the paper contributes to a better understanding of inequalities in remuneration and offers an empirical basis for designing fairer wage policies.

Keywords

gender gender pay gap discrimination education Oaxaca–Blinder decomposition

Introduction

The issue of equality and non-discrimination is currently very topical. Equality between men and women and non-discrimination based on gender is enshrined in the legislation of the European Union and is also part of the promotion of corporate social responsibility in the social field. Nevertheless, significant differences can be noted in the treatment of men and women. In the Czech Republic, women represented 44% of employees and make up approximately 60% of university graduates.¹ Even so, there are significant differences in pay between men and women. According to Eurostat data, the Czech Republic ranks among the countries of the European Union with the highest gender pay gap. The wage difference between men and women was 18.4% in 2022, which was significantly above the European average. In the European Union, the gender pay gap reached 12.2% in 2022.² The Czech Republic is one of the European Union countries with the highest wage gap, and that is why considerable attention is still being paid to this issue in the Czech Republic.

The existence of a gender pay gap does not necessarily mean wage discrimination against women in the labor market. The part of the gender pay gap (or in extreme cases all gap) can be explained by different average characteristics of men and women in the labor market. The traditionally included characteristics are age, tenure, education, occupation, region, and firm size.^3,4 Only the part that these factors could not explain is attributed to wage discrimination against women. Empirical studies show that only part of the wage gap can be explained by the different characteristics of men and women. Horizontal and vertical segregation in the labor market plays a significant role here. Part of the wage gap remains unexplained and is often attributed to wage discrimination against women in the labor market, for example, Refs. ^5,6

Empirical data and scholarly studies also demonstrate that gender pay gaps are not uniform and vary across different industries, sectors, and segments of the wage distribution.^7–10 In general, the public sector displays a lower gender pay gap compared to the private sector; however, wage disparities disadvantaging women are still present even in this domain.^11,12 This study specifically focuses on the issue of wage disparities in the public sector, with particular emphasis on the field of education.

Education is characterized by a number of specific features. It is associated with the positive externality of knowledge dissemination, contributes to the upbringing of the new generation, and to some extent shapes its attitudes and opinions. Ethical behavior and a strong sense of fairness are highly emphasized in this field. From the labor market perspective, education is one of the economic sectors with a high proportion of female employees. Women dominate this sector, representing approximately 80% of the workforce.¹ Similar to the healthcare and social care sectors, education is generally perceived by society as a profession aligned with traditional female roles. All of these factors may influence the size of the gender pay gap in this sector and contribute to its specific characteristics.

This study offers several important contributions to the existing literature and practice. First, the study focuses specifically on the education sector in the Czech Republic, a heavily female-dominated area, and sheds light on the gender pay gap in this specific setting. Given the high share of female employment in this sector and its broader societal role, understanding wage inequalities in this context is of great importance. Second, the paper distinguishes between explained and unexplained components of the gender pay gap, allowing for a more nuanced interpretation of wage differentials. By estimating the unexplained gap, the study contributes to the identification of potential discrimination or structural inequalities that cannot be accounted for by observable characteristics such as education, experience, or job position. Third, the analysis of the gender pay gap over time provides valuable insights into whether progress toward wage equality has been made, stagnated, or reversed. This longitudinal perspective enables a better understanding of the dynamics and persistence of gender-based pay disparities in the public education sector. Finally, the findings of the study may serve as an evidence base for policymakers seeking to design targeted interventions aimed at promoting pay equity and addressing gender-based wage discrimination, especially in public sector employment.

Literature review

Gender pay inequality has long been a persistent issue and remains the subject of extensive research across the fields of economics, sociology, and psychology. The gender pay gap (GPG) is commonly defined in the academic literature as the difference in average earnings between men and women.^3,4 Various theoretical approaches have been proposed to explain the existence of gender differences in pay. Economic theories frequently refer to human capital theory, which attributes wage disparities to women’s lower investment in education and work experience due to expected career interruptions related to childbearing.¹³ Occupational segregation theories are also prominent, highlighting that women are disproportionately concentrated in lower-paid sectors and positions.^3,4 Additional influential frameworks include taste-based discrimination models and statistical discrimination models, which suggest that employers may rely on prejudices or group averages rather than individual assessment when making hiring and wage decisions.^14,15 From a sociological perspective, theories emphasize the role of gender norms and stereotypes, which shape occupational choices, access to leadership positions, and performance evaluations.¹⁶

Empirical studies consistently show that a portion of the gender pay gap can be explained by differences in characteristics between men and women, the so-called explained part, while the unexplained part is often interpreted as an indicator of discrimination.^5,17,18 Research across the European Union reveals that the size of the unexplained gap varies significantly by country, sector, and occupation. The public sector tends to exhibit smaller pay gaps than the private sector, though disparities are not entirely absent.⁷

In the European context, comparative evidence confirms substantial cross-country heterogeneity in both the level and the structure of the gender pay gap.¹⁹ Using cross-country European data, Redmond and McGuinness²⁰ show that gender differences in job preferences contribute to the wage gap and that the relative importance of factors varies across countries and across the wage distribution. Building on this evidence, Das and Joubert²¹ document that, beyond contemporaneous differences in characteristics and returns, historical patterns of workforce composition and career progression still shape gender pay gaps across EU economies. Their dynamic decomposition highlights that part of the persistent gap reflects legacy effects, where differential promotion rates and tenure accumulate over time, indicating that the pace of convergence differs markedly between countries depending on institutional frameworks, labor market regulations, and organizational practices. Together, these studies suggest that gender pay gaps in the EU are not only a function of current labor market differences, but also of dynamic structural processes that operate differently across member states.

The existence of a substantial unexplained GPG has also been confirmed in multiple studies focused on the Czech labor market [e.g., Refs. 24,4,36]. Even when controlling for detailed individual characteristics, including cognitive and non-cognitive skills and personal preferences, a significant portion of the wage differential remains unexplained. For instance, Balcar and Hedija,²² using unique firm-level data including employee ability test results, found that differences in productivity and skills could not fully account for the wage gap. The unexplained GPG was estimated at approximately 13 percentage points in their study.

The evidence also suggests that the unexplained gender pay gap is generally lower in the public sector compared to the private sector. Jurajda,²³ based on data from the Average Earnings Information System, estimated the unexplained component at 9% in the public sector, whereas it reached 19% in the private sector. These findings are consistent with those of Gottvald et al.,¹² who evaluated GPG developments between 2006 and 2011. In 2011, they found the unexplained GPG to be 9.8% in the public sector and approximately 17% in the private sector. Notably, the unexplained gap remained relatively stable throughout the examined period.

The gender pay gap is also unevenly distributed across the wage spectrum. Zajíčková and Zajíček,⁹ applying a quantile decomposition approach to data from 2006 to 2017, confirmed that the GPG is largest at the lower and upper ends of the earnings distribution. Their findings indicate that the unexplained component accounts for roughly half of the total GPG, corresponding to 9–10 percentage points on average. A particularly influential factor is parenthood; the wage disparity between women and men is significantly greater among parents than among childless workers, especially among women returning from maternity leave.

Special attention has also been devoted to vertical segregation, in which women are underrepresented in higher-paid and managerial positions. This lack of representation directly impacts their average earnings and access to decision-making roles. National statistics show that while women are prevalent in sectors such as education, healthcare, and social services, they are less likely to occupy senior or leadership positions within these sectors, roles that typically offer better pay and higher status.¹

Wage inequality is further reinforced by disparities in performance evaluation, which directly influence compensation, career progression, and professional development. Studies show that women are often perceived as less competent, less assertive, and less suited for leadership roles during evaluations.^24,25 For example, Mengel et al.²⁶ found that female academics in the Netherlands received lower student evaluations than their male counterparts, despite comparable teaching performance. Similarly, Hospido and Sanz²⁷ demonstrated that papers authored by women were less likely to be accepted at prestigious economic conferences. Adamus and Ballová Mikušková²⁸ highlight the role of gender stereotypes in performance assessments. Their experimental study using Slovak data revealed clear evidence of wage discrimination: women were offered significantly lower salaries than men, despite being rated similarly in competence and job suitability. Interestingly, the study also found evidence of a backlash effect; men with frequent absences were more likely to be selected for dismissal than equally absent women. Other studies show that evaluation disparities depend on the job context. Diederich et al.²⁴ conclude that women tend to be rated more favorably when they hold positions considered less masculine and more congruent with traditional gender roles. Hedija²⁹ arrives at similar conclusions based on an analysis of peer evaluations of university students’ seminar presentations, finding no evidence that male students receive more favorable assessments than their female counterparts.

Relatively great attention is also paid to the separate issue of remuneration in education and its impact on the quality of education. Empirical studies converge on the finding that the absolute level of wage shows only a limited direct relationship with student achievement.^30,31 Nevertheless, salaries play a crucial role through indirect mechanisms, particularly in relation to teacher recruitment, retention, and motivation.

Research indicates that in regions and countries where teachers are remunerated competitively relative to other professions requiring similar levels of education, turnover rates are lower and the likelihood of attracting high-quality candidates into the profession is higher.³² These effects translate into greater stability of the teaching workforce and, consequently, into improved student outcomes. Analyses conducted within the OECD further confirm that the positive impact of salaries on educational outcomes is most evident in systems where teaching is perceived as a prestigious profession and where higher pay is accompanied by additional support measures, such as professional development and mentoring for novice teachers.^31,33

The existence of the gender pay gap, especially its unexplained components, may also indirectly affect the quality of education. Empirical studies indicate that persistent pay disparities between female and male teachers, often linked to lower access for women to supplemental pay or more prestigious and better-compensated positions, can undermine motivation and increase teacher turnover.³⁴ Such inequalities, in turn, affect the stability of teaching staff, which can have a negative impact on the quality of education. Ronfeldt et al.³⁵ prove that teacher turnover has a negative impact on student results; students in grades with higher teacher turnover achieve weaker results, especially in schools with a higher proportion of disadvantaged students. Thus, the gender pay gap in education is not only an issue of labor market equality but also a factor that can indirectly affect educational quality through its impact on teacher motivation and retention.

Methods and data

The study is based on the data from European Union Statistics on Income and Living Conditions (EU-SILC). Cross-sectional EU-SILC data for the year 2013 (reflecting income earned in 2012) and for 2023 (reflecting income earned in 2022) are used to estimate and decompose the gender pay gap for the Czech Republic. A key limitation of the EU-SILC dataset is the absence of information on hourly wages, which are essential for accurately assessing gender-based wage disparities. Consequently, it is necessary to restrict the analytical sample to individuals for whom hourly wages can be derived from the available data. The reference population is limited to employees who, during the reference year, worked full-time for the entire 12-month period, held no additional jobs, and reported earned income. Self-employed individuals are excluded, as the analysis focuses specifically on wage employees and the potential gender-based differences in remuneration attributable to employer behavior. The gross hourly wage is computed by dividing the sum of annual cash and non-cash income from employment by the total number of hours usually worked per year, including any reported overtime. The final sample contains the data of more than 6000 employees in 2013 and 4000 employees in 2023 of which about 450 employees worked in education sector in each year.

The Oaxaca–Blinder decomposition is used to estimate and decompose the wage differences between men and women.^5,6 The application of this decomposition allows the separation of the gender pay gap into that part which can be explained by different characteristics of men and women, and the part that cannot be explained and that is usually labelled as the unexplained part of the gender pay gap or the effect of discrimination. The unexplained component should be interpreted as an upper bound of potential wage discrimination rather than a direct measure thereof. Its magnitude may also reflect differences in unobserved or imperfectly measured characteristics between men and women.

To use the Oaxaca–Blinder decomposition, it is necessary to estimate wage functions for men and women. They are estimated as follows

{\ln (W_{i})}_{m} = β_{m} {(X_{i} ´)}_{m} + {(u_{i})}_{m}, i = 1, \dots, n_{m},

(1)

{\ln (W_{i})}_{f} = β_{f} {(X_{i} ´)}_{f} + {(u_{i})}_{f}, i = 1, \dots, n_{f},

(2)

where

{(W_{i})}_{m}

is the gross hourly wage of the i-th man,

{(X_{i} ´)}_{m}

is a vector of observed characteristics of i-th men,

{(W_{i})}_{f}

is the gross hourly wage rate of i-th woman,

{(X_{i} ´)}_{f}

is a vector of chosen characteristics of i-th women,

β_{m}, β_{f}

are vectors of coefficients of wage functions for men and women and

u_{i}

is random component, n_m is the number of men and n_f is the number of women in the sample.

For decomposition of the gender pay gap, conventional forms of the Oaxaca–Blinder decomposition will be used

\ln ({\bar{W}}_{m}) - \ln ({\bar{W}}_{f}) = ({\bar{X}}_{m} - {\bar{X}}_{f}) β^{*} + ({\hat{β}}_{m} - β^{*}) {\bar{X}}_{m} + (β^{*} - {\hat{β}}_{f}) {\bar{X}}_{f} .

(3)

where

{\bar{W}}_{m}

is the average gross hourly wage of men,

{\bar{W}}_{f}

is the average gross hourly wage of women,

{\bar{X}}_{m}

is a vector of average characteristics of men,

{\bar{X}}_{f}

is a vector of average characteristics of women,

{\hat{β}}_{m}, {\hat{β}}_{f}

are the vectors of the wage functions coefficients,

β^{*}

denotes the vector of coefficients of wage function in the absence of discrimination (equilibrium wage). The term

\ln ({\bar{W}}_{m}) - \ln ({\bar{W}}_{f})

is a raw pay gap expressed as the difference between the logarithms of the average wages of men and women. The term

({\bar{X}}_{m} - {\bar{X}}_{f}) β^{*}

is the part of the pay gap explained by different characteristics of men and women. This is often called the endowment effect. The term

({\hat{β}}_{m} - β^{*}) {\bar{X}}_{m} + (β^{*} - {\hat{β}}_{f}) {\bar{X}}_{f}

is the unexplained part of the pay gap. This is known as the remuneration effect or the effect of discrimination.

The final step in the application of Oaxaca–Blinder decomposition is the choice of equilibrium wage (β*). In professional studies that work with Oaxaca–Blinder decomposition we can find various concepts of equilibrium wage. Blinder⁶ estimated the extent of discrimination using the wage of men as the equilibrium wage. To decompose the wage gap, Oaxaca⁵ used the wage of both men and women as the equilibrium wage. Reimers³⁶ attributed the regression coefficients of wage functions for men and women the same weight (β*= (βm+βf)/2). Cotton¹⁸ used the weighted average wage of men and women, where, as weights he used the proportion of men and women in the total working population. Neumark¹⁷ estimated the equilibrium wage as a vector of coefficients of the wage function for the whole sample (both men and women).

For the decomposition of the gender wage gap, this study employs the Neumark variant of the Oaxaca–Blinder decomposition, which defines the nondiscriminatory wage structure based on coefficients estimated from a pooled regression across both genders. This approach mitigates the index number problem inherent in traditional Oaxaca–Blinder decompositions, as the magnitude of the unexplained component is otherwise sensitive to the choice of the reference group (male or female wage structure). By relying on pooled coefficients, the Neumark decomposition provides a neutral benchmark that does not a priori privilege either gender as the normative standard. This is particularly important for intertemporal comparisons, as it ensures greater consistency when assessing changes in the unexplained gender wage gap between 2012 and 2022, while reducing the influence of group-specific coefficient instability and compositional shifts over time. Consequently, the Neumark variant offers a more robust framework for evaluating the evolution of the unexplained component of the gender wage gap within the specified period. For these reasons, the Neumark variant of the Oaxaca–Blinder decomposition is widely applied in empirical research on gender pay gap [e.g., Refs. 9; 27; 37].

Recent methodological works highlight that mean-based decompositions may mask heterogeneity across the wage distribution and recommend Recentered Influence Function-based decomposition that can provide distribution-sensitive estimates.^37,38 Given the limited sample size for the education sector subsample, a fully fledged distributional decomposition is left for future research.

The logarithm of the hourly gross wage is used as dependent variable. As explanatory variables, the following characteristics of employees are used: age (in years), age squared, work experience (number of year spent in paid work), work experience squared, level of education (category according to highest attained level of education according to ISCED), sickness (dummy variable, temporary inability to work due to sickness in the income reference year), partnership (dummy variable, having a partner in a common household), occupation (category according to the classification ISCO), firm size (dummy variable, more than 10 employees), contract (dummy variable, having a work contract of limited duration), managerial position (dummy variable, having formal responsibility for supervising a group of other employees), finally sector (category according to economic activity using classification NACE Rev. 2) if the gap is estimated for the full sample not only for education sector. Table 1 shows the mean and standard deviation of individual variables for men and women in the sample.

Table 1.

Descriptive statistics—full sample.

Variable	2012		2022
Variable	Men	Women	Men	Women
Ln hourly wage	1.771344 (0.3893588)	1.555108 (0.3944708)	2.364179 (0.3783132)	2.152594 (0.3511746)
Age	43.27853 (11.74107)	44.42597 (10.30648)	46.27929 (11.45998)	48.03386 (10.1342)
Work experience	22.07468 (11.88519)	21.44493 (10.76228)	24.7657 (11.70526)	24.38646 (10.87598)
Education (ISCED)^a
Less than primary, primary, lower secondary	0.0310148 (0.1733837)	0.0485047 (0.2148694)	0.0285848 (0.1666756)	0.0374718 (0.1899577)
Upper secondary, post-secondary non tertiary	0.7807889 (0.4137743)	0.757841 (0.4284682)	0.7033739 (0.4568773)	0.6717833 (0.4696702)
Tertiary education	0.1881963 (0.3909277)	0.1936543 (0.3952332)	0.2680412 (0.443043)	0.2907449 (0.4542086)
Occupation (ISCO)^a
Armed forces	0.0096357 (0.097702)	-	0.0112465 (0.1054761)	0.0013544 (0.0367856)
Managers	0.0529961 (0.2240594)	0.0423049 (0.2013206)	0.0627929 (0.2426469)	0.0379233 (0.1910538)
Professionals	0.1183379 (0.3230564)	0.1557257 (0.3626613)	0.156045 (0.362983)	0.2311512 (0.4216641)
Tech. and associate professionals	0.1836796 (0.387281)	0.2071481 (0.4053365)	0.1907216 (0.3929622)	0.1837472 (0.3873653)
Clerical support workers	0.0382415 (0.1918076)	0.1896426 (0.3920897)	0.0454545 (0.2083477)	0.1702032 (0.375896)
Service and sales workers	0.0888287 (0.2845391)	0.1976659 (0.3983114)	0.0824742 (0.2751503)	0.1828442 (0.3866261)
Skilled agricultural, forestry and fish. workers	0.0102379 (0.1006783)	0.012035 (0.1090619)	0.0079663 (0.0889185)	0.0076749 (0.0872896)
Craft and related trades workers	0.2746161 (0.4463878)	0.0525164 (0.2231067)	0.2141518 (0.4103288)	0.0388262 (0.1932241)
Plant and machine operators and assemblers	0.1915086 (0.3935476)	0.0751276 (0.2636453)	0.1921275 (0.3940651)	0.0726862 (0.2596794)
Elementary occupations	0.0319181 (0.1758085)	0.0678337 (0.2515062)	0.0370197 (0.1888543)	0.0735892 (0.2611601)
Sickness^a	0.0719663 (0.258471)	0.1232677 (0.3288042)	0.099344 (0.2991934)	0.123702 (0.3293157)
Partnership (having partner)^a	0.6895513 (0.462747)	0.6932896 (0.4612122)	0.6682287 (0.4709597)	0.6487585 (0.4774661)
Sector (NACE Rev. 2)^a
Agriculture, forestry and fishing (A)	0.0418549 (0.2002876)	0.0237053 (0.1521572)	0.0337395 (0.1806001)	0.0121896 (0.1097564)
Mining, manufacturing, electricity, water…(B–E)	0.4197531 (0.4935927)	0.2447119 (0.4299947)	0.4137769 (0.4926249)	0.2361174 (0.424791)
Construction (F)	0.1084011 (0.3109331)	0.0215171 (0.145127)	0.0848172 (0.2786748)	0.0167043 (0.12819)
Wholesale and retail trade… (G)	0.0809997 (0.2728757)	0.1415026 (0.3486028)	0.08388 (0.2772728)	0.1110609 (0.3142785)
Transportation and storage (H)	0.0903342 (0.2867032)	0.0477753 (0.2133295)	0.0899719 (0.2862085)	0.0316027 (0.1749794)
Accommodation and food service…(I)	0.0195724 (0.1385464)	0.0466813 (0.2109937)	0.0135895 (0.1158064)	0.0316027 (0.1749794)
Information and communication (J)	0.031617 (0.1750045)	0.0171408 (0.1298195)	0.0435801 (0.2042069)	0.0194131 (0.138003)
Financial and insurance activities (K)	0.0147546 (0.1205872)	0.036105 (0.1865855)	0.0159325 (0.1252439)	0.0288939 (0.1675462)
Real estate, prof. and admin. activities (L–N)	0.0575128 (0.2328549)	0.0579869 (0.2337613)	0.0557638 (0.2295189)	0.0582393 (0.2342482)
Public admin., defence, social security (O)	0.0674496 (0.2508367)	0.0875274 (0.2826578)	0.0923149 (0.2895378)	0.1079007 (0.310325)
Education (P)	0.0307136 (0.1725667)	0.1232677 (0.3288042)	0.0360825 (0.1865391)	0.1738149 (0.3790358)
Human health and social work activities (Q)	0.0204758 (0.1416423)	0.1261853 (0.3321186)	0.0257732 (0.1584951)	0.1489842 (0.3561533)
Arts, other services…(R–U)	0.0165613 (0.1276398)	0.0258935 (0.1588466)	0.0107779 (0.1032798)	0.0234763 (0.1514448)
Firm size (more than 10 employees)^a	0.8461307 (0.3608778)	0.781911 (0.4130235)	0.6761949 (0.4680363)	0.6325056 (0.4822315)
Contract (limited duration)^a	0.0837097 (0.2769937)	0.0897155 (0.2858259)	0.0440487 (0.2052515)	0.0474041 (0.2125496)
Managerial position^a	0.2354712 (0.4243569)	0.1524435 (0.3595158)	0.2253983 (0.4179423)	0.1534989 (0.3605491)
N	3,321	2,742	2,134	2,215

Source: EU-SILC 2013 and EU-SILC 2023, author’s computations.

^aShare in individual groups (mean of dummy variables), standard deviation in parentheses.

Potential multicollinearity was assessed prior to the Oaxaca–Blinder decomposition by estimating separate regression models for each group and examining variance inflation factors (VIF). The VIF diagnostics revealed the problem of multicollinearity. Multicollinearity is a common feature of wage regressions, as many key explanatory variables capture closely related dimensions of individual human capital and job characteristics. Variables such as age and work experience, and their squared terms are inherently correlated, while education, occupation and managerial position are jointly determined along career paths and therefore exhibit substantial overlap. The presence of multicollinearity in wage equations reflects the economic structure of career progression rather than model misspecification. Despite indications of elevated multicollinearity, the full set of explanatory variables was retained in the wage equations used for the Oaxaca–Blinder decomposition. The aim of the wage specification is to comprehensively control for observable productivity-related characteristics rather than achieve parsimony, as omitting relevant covariates would introduce omitted-variable bias and potentially overstate the unexplained gender wage gap. Multicollinearity mainly affects the precision of coefficient estimates, whereas excluding relevant variables poses a more serious threat to the validity of the decomposition results. To minimize the effect of multicollinearity, in the study, results of Oaxaca–Blinder decomposition are interpreted at the level of two broader groups of personal and firm-related characteristics. This group-level perspective reduces the relevance of multicollinearity, as correlations mainly affect individual coefficients while aggregated contributions remain relatively stable.

To address potential heteroskedasticity, robust standard errors were employed in all regression estimates underlying the decomposition, ensuring consistent inference.

Results

The Neumark’s variant of Oaxaca–Blinder decomposition (equation (3)) is applied to decompose the wage differences between men and women and estimate the explained and unexplained part of gender pay gap in the Czech Republic. The results are shown in Table 2.

Table 2.

Oaxaca–Blinder decomposition.

	Full sample		Education (NACE-P)
	2012	2022	2012	2022
Raw GPG	0.2162*** (0.0101)	0.2116*** (0.0111)	0.3061*** (0.0368)	0.2612*** (0.0375)
Explained part	0.0767*** (0.0080)	0.0823*** (0.0088)	0.1669*** (0.0300)	0.1657*** (0.0256)
Personal characteristics^a	0.0131*** (0.0036)	0.0145*** (0.0041)	0.0419*** (0.0170)	0.0559*** (0.0138)
Firm characteristics^a	0.0635*** (0.0067)	0.0679*** (0.0072)	0.1245*** (0.0236)	0.1098*** (0.0192)
Unexplained part	0.1395*** (0.0068)	0.1293*** (0.0073)	0.1392*** (0.0253)	0.0955*** (0.0266)
N	6,063	4,349	440	462

Source: EU-SILC 2013 and EU-SILC 2023, authors computations.

^aPersonal characteristics: age, work experience, level of education, partnership, sickness. Firm characteristics: occupation, sector, company size, contract, managerial position, Standard errors in parentheses, ***p < 0.01, **p < 0.05, *p < 0.1.

The gender pay gap in education reached 0.2612 in 2022. It means that the average wage for women was 26.1% lower than the average wage for men working in education. The unexplained part was 0.0955. The wage difference of 9.6% could not be explained the differences in the characteristics of men and women and it can be interpreted as the upper limit of the effect of discrimination. However, this component may also reflect differences in unobserved or imperfectly measured characteristics, such as job quality, career trajectories, motivation, or past career interruptions. Consequently, the unexplained component should not be interpreted as a direct measure of discrimination but rather as an upper limit that also captures data and model limitations.

The results also indicate that differences in the characteristics of men and women accounted for approximately 16.6% of the gender pay gap (the estimated explained pay gap was 0.1657). Among the key explanatory factors were firm-level characteristics, which in this study include type of employment contract, job position, managerial status, and organization size. Differences between men and women in these firm-level characteristics explained about 11% of the wage gap (with an estimated explained component of 0.1098). Personal characteristics, including age, work experience, level of education, partnership status, and health status, accounted for approximately 5.6% of the gender pay gap (with an estimated explained component of 0.0559).

The conclusions regarding the predominant influence of firm-level characteristics and the significantly lower impact of personal characteristics in explaining wage differentials are consistent with the findings of numerous empirical studies, for example, Refs. 23,39,40. Data from the Czech Statistical Office¹ show that women have long dominated the education sector. However, their representation in teaching positions declines with increasing levels of education. In nursery schools, women make up 99% of teaching staff. At the primary level of basic schools (ISCED level 1), women account for 94% of teachers, while at the lower secondary level, they represent 74%. In secondary schools and conservatoires, the share of women falls to between 50 and 60%. Finally, in higher education, only 37% of academic positions are held by women. Men are also more likely to occupy managerial positions. While women represent 67% of school management personnel in basic schools, their share drops to approximately 47% in secondary schools and conservatoires. The concentration of men in leadership positions and in higher levels of education (such as secondary schools and universities), where wages tend to be higher, is one of the factors contributing to the lower average earnings of women compared to men in the education sector.

Focusing on change over time, the estimated values show that the gender pay gap in the education sector decreased by approximately 4.5 percentage points between 2012 and 2022. While in 2012 the raw gender pay gap reached nearly 30.6% (raw gender pay gap was 0.3061), by 2022 the gap had declined to approximately 26.1% (raw gender pay gap reached 0.2612). The unexplained component of the gender pay gap also decreased, from 0.1392 in 2012 to 0.0955 in 2022, representing a reduction of about 4.4 percentage points. This suggests that the decline in the overall wage differential can be attributed primarily to a decrease in the unexplained component of the gender pay gap, which represents the upper bound of potential wage discrimination. At the same time, part of this decline may reflect changes in unobserved characteristics or career dynamics of men and women over time rather than changes in discriminatory wage-setting alone.

The results for the full sample are also presented in Table 1 for comparison. Here, it can be seen that gender pay gap in education was above the overall gender pay gap; on the other hand, the unexplained part does not differ significantly. The estimated unexplained gender pay gap in the education sector reached a value of 0.0955 in 2022, which is lower compared to the value estimated for the entire Czech economy (based on the full sample).

This is to some extent at odds with the findings of numerous empirical studies that examine differences in pay between men and women in the public and private sectors. These studies generally conclude that the gender pay gap, and especially its unexplained portion, is smaller in the public sector compared to the private sector [e.g., Refs. 24, 18, and 21]. In the field of education in the Czech Republic, the public sector is dominant. Fewer than 10% of students attend private schools.⁴¹

The larger gender gap can be explained by the relatively low representation of men in the field of education and their concentration in better-paid positions. As discussed above, the proportion of men is very low at lower levels of education (nursery schools and basic schools) and increases at higher levels (secondary schools and universities), where higher salaries are achieved. Despite their relatively low overall representation in education, men often hold managerial, and therefore better-paid, positions.

The lower value of the unexplained (discriminatory) component of the gender pay gap in the education sector compared to the average for the Czech economy may be related to specific features of this sector. Education is predominantly part of the public sector, where wages are determined centrally through pay scales that primarily reflect length of service, level of education, and job position. This system significantly limits the scope for individual wage negotiation, and thus reduces the potential for unjustified pay disparities. Additionally, it is possible that work in education culturally aligns more closely with traditional female roles, which may reduce the tendency for women to be systematically disadvantaged in terms of pay within this field. According to Diederich et al.,²⁴ women are evaluated better when they hold less masculine and therefore more role-corresponding functions.

The results confirm that, despite the overall decline of the gender pay gap in education between 2012 and 2022, substantial inequalities persist. These inequalities cannot be viewed in isolation from the quality of education. Persistent pay disparities, especially in a sector dominated by women, may negatively affect teacher motivation, career satisfaction, and the attractiveness of the profession for talented graduates. This can translate into higher turnover rates and difficulties in retaining skilled educators, which can negatively affect the quality of education. Ronfeldt et al.³⁵ provide empirical evidence that elevated levels of teacher turnover exert a detrimental influence on student achievement. Thus, the persistence of the gender pay gap is not only a labor market issue but also a systemic factor with potential consequences for the long-term quality of education.

On the other hand, the persistence of the unexplained gender pay gap does not necessarily indicate ongoing discrimination. It may partly reflect pipeline effects, whereby historical gender differences in promotion and career progression continue to affect current wage outcomes.²¹ As a result, the unexplained component may capture legacy effects of past inequalities rather than purely contemporaneous discriminatory wage-setting. This opens up scope for institution-level analyses of wage-setting practices within individual schools.

Conclusions

The paper aims estimate the unexplained component of the gender pay gap in the education sector in the Czech Republic using the Oaxaca–Blinder decomposition, and to examine how this unexplained component evolved over the decade between 2012 and 2022. To estimate the unexplained part of wage differences between men and women, we used European Union Statistics on Income and Living Conditions (EU-SILC) data and applied Neumark´s variant of the Oaxaca–Blinder decomposition.

The results show that the raw gender pay gap in education was approximately 26.2% to the disadvantage of women in 2022. The different characteristics of men and women explained approximately 16.6% pay gap, the unexplained part reaches 9.6%. The results thus show that differences in known characteristics of men and women (age, work experience, level of education, partnership, sickness, occupation, company size, type of contract and managerial position) explained only part of the wage gap. The part that remained unexplained can be interpreted as the upper limit of the discrimination effect. This may indicate the persistent wage discrimination against women in the education sector. The results also showed that the gender pay gap in the education sector, as well as its unexplained part, decreased between 2012 and 2022. Although the decrease was not large (by 4.5 percentage points), it nevertheless indicates a positive trend in this area. Especially because the decrease occurs mainly in the unexplained component (a decrease of 4.4 percentage points) of the pay gap between men and women, which can be considered the upper limit of wage discrimination.

However, the persistence of the unexplained component of the pay gap highlights that gender-based inequalities have not been fully addressed, even in a sector largely dominated by female employees and regulated through standardized pay scales. These findings suggest that even centrally governed remuneration systems may not be entirely immune to gender bias, particularly when men are disproportionately represented in better-paid and managerial roles.

From a policy perspective, the results point to the need for continuous monitoring of gender pay disparities even in the public sector and for measures that promote equal opportunities for career advancement and access to leadership positions across genders. Policies aimed at improving gender balance in top-level educational roles, encouraging transparent criteria for promotions, and addressing implicit biases in organizational culture could help reduce the remaining pay gap.

Existing gender wage inequalities in education are not just an ethical problem, it may also indirectly affect the quality of education. Wage disparities that disadvantage women may lead to higher staff turnover, reduced morale, and limited professional development opportunities for teachers. These mechanisms can weaken the stability and effectiveness of teaching teams and, consequently, negatively influence student learning outcomes. Ensuring fair remuneration across genders is therefore not only a matter of equity but also a key prerequisite for sustaining and improving the quality of education. Addressing the gender pay gap in the education sector can thus be viewed as a dual investment, both in equality and in the long-term performance of the educational system.

This study builds upon the established literature on gender pay inequality using the Oaxaca–Blinder decomposition and extends previous research for the Czech Republic that documents a persistent unexplained gender pay gap at the aggregate level [e.g., Refs. 18; 24; and 36]. Earlier studies generally conclude that gender pay inequalities are lower in the public sector than in the private sector^12,23; this article confirms that gender-based wage disparities persist even in highly regulated public sector environments. The main added value of the study lies in its sector-specific and intertemporal perspective. By focusing explicitly on education, a highly female-dominated and centrally regulated sector, and by comparing two time points a decade apart, the paper shows that formalized and standardized wage-setting mechanisms reduce but do not eliminate gender pay disparities. The observed decline in the overall gender pay gap is shown to be driven by a reduction in its unexplained component, thereby extending predominantly cross-sectional evidence with a dynamic perspective. Beyond its empirical findings, the study contributes to theory by emphasizing the importance of institutional factors in shaping wage outcomes. By decomposing the gender pay gap into personal and firm-related characteristics, the findings confirm that the gender pay gap is influenced and shaped by a range of factors. While differences in personal characteristics play a less significant role, the firm-level factors, such as job allocation, and access to managerial positions, play an important role in explaining gender wage differences, which is consistent with organizational and institutional approaches to gender inequality. A large part of the wage gap remained unexplained, which is consistent with discrimination theory.^14,15 Overall, the article contributes to managerial theory and practice by demonstrating that gender pay gaps persist even in industries characterized by high female representation and centralized regulation, thereby calling for management approaches that go beyond formal equality and address the underlying structural and organizational mechanisms.

Despite its contributions, this study is subject to several limitations that should be acknowledged. The analysis relies on EU-SILC data, which, although harmonized and representative, provide only a limited set of observable characteristics. Due to data constraints, the wage equations do not capture all factors that may affect individual productivity and remuneration, such as detailed measures of job performance, teaching quality, workload intensity, task complexity, non-cognitive skills, or institutional-level characteristics (e.g., school type, governance structure, or internal promotion rules). As a result, the estimated unexplained component of the gender pay gap may be overstated and should be interpreted as an upper bound of potential discrimination rather than a direct measure of discriminatory behavior. Furthermore, the empirical analysis is limited to a single country. While this allows for a detailed examination of gender pay inequality within a specific institutional and cultural context, the findings cannot be fully generalized to other countries with different educational systems, wage-setting mechanisms, or gender norms. Cross-country comparative analyses could help distinguish between country-specific institutional effects and more general structural drivers of gender pay inequality in education.

Future research could extend this study in several directions. First, the use of richer administrative or employer-level data would allow for a more precise estimation of productivity-related characteristics and a more accurate decomposition of wage differentials. Second, applying alternative decomposition techniques, such as quantile decompositions, could shed light on how the unexplained gender pay gap varies across the wage distribution within the education sector. Finally, comparative studies across countries or across different segments of the public sector would provide valuable insights into the role of institutional wage-setting mechanisms in shaping gender pay inequalities.

Footnotes

Veronika Hedija

Blanka Plasová

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

Czech Statistical Office . Focused on women and men 2023. https://csu.gov.cz/produkty/zaostreno-na-zeny-a-muze-2023 (2024), (Accessed 9 July 2025).

Eurostat . Gender pay gap in unadjusted form. https://ec.europa.eu/eurostat/databrowser/view/sdg_05_20/default/table?lang=en&category=t_labour.t_earn (2025), (Accessed 9 July 2025).

Jacobsen

. The economics of gender. 3rd ed. Blackwell Publishing, 2007.

Ehrenberg

Smith

. Modern labour economics: theory and public policy. 11th ed. Pearson Education, 2012.

Oaxaca

. Male-female wage differentials in urban labour markets. Int Econ Rev 1973; 14: 693–709.

Blinder

. Wage discrimination: reduced form and structural estimates. J Hum Resour 1973; 8(4): 436–455.

Christofides

Polycarpou

Vrachimis

. Gender wage gaps, ‘sticky floors’ and ‘glass ceilings’ in Europe. Labour Econ 2013; 21: 86–102.

Hedija

. Sector specific gender pay gap: evidence from European Union countries. Econ Res Ekon Istraživanja 2017; 30(1): 1804–1819.

Zajíčková

Zajíček

. Gender pay gap in the Czech Republic – its evolution and main drivers. Prague Econ Pap 2021; 30(6): 675–723.

10.

Finardi

. Determinants of the gender pay gap in the Czech Republic and selected European countries. Eur Financ Account J 2022; 17(2): 33–47.

11.

Stupnytskyy

. Struktura rozdílů ve výdělcích mužů a žen v managementu. VÚPSV, 2007.

12.

Gottvald

Vašková

Janíková

. Determinants of individual wages in the Czech Republic. In: Pytliková

(ed). Gender wage gap and discrimination in the Czech Republic. Ostrava: VŠB-TU Ostrava, 2012, pp. 35–56.

13.

Becker

. Human capital: a theoretical and empirical analysis, with special reference to education. University of Chicago Press, 1964.

14.

Becker

. The economics of discrimination. University of Chicago Press, 1957.

15.

Arrow

. The theory of discrimination. In: Ashenfelter

Rees

(eds). Discrimination in labor markets. Princeton University Press, 1973, pp. 3–33.

16.

Oakley

. Sex, gender and society. Temple Smith, 1972.

17.

Neumark

. Employers' discriminatory behavior and the estimation of wage discrimination. J Hum Resour 1988; 23(3): 279–295.

18.

Cotton

. On the decomposition of wage differentials. Rev Econ Stat 1988; 70(4): 236–243.

19.

Oaxaca

Ransom

. On discrimination and the decomposition of wage differentials. J Econom 1994; 61(1): 5–21.

20.

Redmond

McGuinness

. The gender wage gap in Europe: job preferences, gender convergence and distributional effects. Oxf Bull Econ Stat 2019; 81(3): 564–587.

21.

Das

Joubert

. Women in the pipeline: a dynamic decomposition of firm pay gaps. Labour Econ 2023; 80: 102282.

22.

Balcar

Hedija

. Influence of female managers on gender wage gap and returns to cognitive and non-cognitive skills. Gend Work Organ 2019; 26(5): 722–737.

23.

Jurajda

. Gender wage gap and segregation in enterprises and the public sector in late transition countries. J Comp Econ 2003; 31(2): 199–222.

24.

Diederich

Iseke

Pull

, et al. Role (In-)congruity and the catch 22 for female executives: how stereotyping contributes to the gender pay gap at the top executive level. Int J Hum Resour Manag 2023; 35(7): 1283–1311.

25.

Boring

. Gender biases in student evaluations of teaching. J Publ Econ 2017; 145: 27–41.

26.

Mengel

Sauermann

Zölitz

. Gender bias in teaching evaluations. J Eur Econ Assoc 2019; 17(2): 535–566. https://doi.org/10.1093/jeea/jvx057

27.

Hospido

Sanz

. Gender gaps in the evaluation of research: evidence from submissions to economics conferences. Am Econ Rev Insights 2021; 3(2): 191–204.

28.

Adamus

Ballova Mikuskova

. Gender discrimination and the backlash effect in recruitment and dismissal processes. Gend Manag 2024; 39(1): 107–129.

29.

Hedija

. Gender gap in performance evaluation. Are women stricter? Ekon Cas 2025; 73(3–4): 145–160.

30.

Loeb

Page

. Examining the link between teacher wages and student outcomes: the importance of alternative labor market opportunities and non-pecuniary variation. Rev Econ Stat 2000; 82(3): 393–408.

31.

Diagne

. Is there a link between teacher salary and educational achievement? An analysis in OECD countries. Athens J Educ 2024; 11(1): 37–50.

32.

Dolton

Marcenaro-Gutierrez

. If you pay peanuts do you get monkeys? A cross-country analysis of teacher pay and pupil performance. Econ Policy 2011; 26(65): 5–55.

33.

OECD . Education at a glance 2023: OECD indicators. OECD Publishing, 2023.

34.

Quintero

Hansen

Zerbino

. Uncovering the sources of gender wage gaps among teachers. Econ Educ Rev 2024; 103: 102559.

35.

Ronfeldt

Loeb

Wyckoff

. How teacher turnover harms student achievement. Am Educ Res J 2013; 50(1): 4–36.

36.

Reimers

. Labour market discrimination against hispanics and black men. Rev Econ Stat 1983; 65(4): 570–579.

37.

Rios-Avila

. Recentered influence functions (RIFs) in stata: RIF regression and RIF decomposition. STATA J 2020; 20(1): 51–94.

38.

Firpo

Fortin

Lemieux

. Decomposing wage distributions using recentered influence function regressions. Econometrics 2018; 6(2): 28.

39.

Hedija

Musil

. Gender wage differences in the selected Czech public sector company. Acta Univ Agric Silvic Mendelianae Brunensis 2012; 60(7): 81–88.

40.

Balcar

Gottvald

. Wage determinants and economic crisis 2008–2014: evidence from the Czech Republic. Ekon Čas (J Econ) 2016; 64(1): 3–21.

41.

The Ministry of Education, Youth and Sports . Základní údaje o školách v regionálním školství za období 1989/90 až 2021/22. https://msmt.gov.cz/vzdelavani/skolstvi-v-cr/statistika-skolstvi/zakladni-udaje-o-skolach-v-regionalnim-skolstvi-za-obdobi?lang=1 (2025), (Accessed 9 July 2025).