1.    INTRODUCTION    

Purpose and scope of the evaluation    

2.    BACKGROUND TO THE INITIATIVE    

2.1.    Description of the initiative    

2.2.    Baseline    

3.    PROCESS, METHODOLOGY AND LIMITATIONS    

4.    IMPLEMENTATION AND STATE OF PLAY    

5.    ANALYSIS AND ANSWERS TO THE EVALUATION QUESTIONS    

5.1.    Effectiveness    

5.1.1. Implementation of the concepts of ‘equal pay’, ‘same work’ and ‘work of equal value’ in national legislation and practice    

5.1.2. Effectiveness of EU action in terms of for access to justice    

5.1.3. Effectiveness of actions in terms of monitoring and enforcement    

5.1.4. Effectiveness of actions to tackle occupational segregation    

5.1.5. Effectiveness of the Recommendation    

5.1.6. Effectiveness in terms of the results and impacts achieved    

5.2.    Efficiency    

5.3. Relevance    

5.3.1 Relevance to original needs    

5.3.2. Relevance to current needs    

5.4. Coherence    

5.4.1 Internal coherence    

5.4.2. Coherence with other EU interventions related to the objective of reducing gender discrimination and contributing to gender equality    

5.4.3 External coherence with wider EU policies    

5.4.4. External coherence with international obligations and standards    

5.5. EU Added Value    

6.    CONCLUSIONS    

ANNEXES    

Annex 1 - Procedural information concerning the process to prepare the evaluation    

Annex 2 - Synthesis of stakeholders consultation activities    

Annex 3 – Methodological issues on the gender pay gap indicator and its decomposition using the Structure of Earnings Survey microdata    

Annex 4 - Intervention logic    

Annex 6 – Provisions within the scope of the evaluation    

Annex 7 – Relevant CJEU case-law    

Annex 8 – Concept and measurement of gender pay discrimination    

ACRONYMS AND ABBREVIATIONS

1.INTRODUCTION

Purpose and scope of the evaluation

Equal pay for equal work or work of equal value is one of the European Union’s founding principles 1 . The principle of equal pay between women and men has been enshrined in the Treaties since 1957 (Article 157(1-2)). Article 4 of Directive 2006/54/EC implements the principle of equal pay by providing that, for the same work or for work of equal value, direct and indirect discrimination on the grounds of sex is prohibited in all aspects and conditions of remuneration. This principle, above all, reflects a human rights issue that recognises that male and female workers have equal individual dignity, but it is also an economic issue that benefits overall potential for production through more effective use of skills. However, the effective implementation and enforcement of this principle remains a big challenge 2 , partially reflected in the persistence and magnitude of the gender pay gap 3 (GPG) of 15.7% for the EU28 in 2018.

One of the five main priorities of the 2016-2019 strategic engagement for gender equality 4 is to reduce pay, earnings and pension gaps between women and men, thereby addressing higher risk of poverty among women 5 . The gender pay gap in unadjusted form is also one of the three indicators for gender equality included in the social scoreboard 6 that the Commission uses to monitor the implementation of the European Pillar of Social Rights 7 . Gender equality has also been a prominent topic within the European Semester 8 with the gender pay gap addressed in the Country Specific Recommendations for a number of countries as recitals and as specific recommendation in Estonia.

This staff working document (SWD) provides the results of the REFIT 9 evaluation of relevant provisions (see Annex 6) in EU law implementing the Treaty principle on ‘equal pay for equal work or work of equal value’ (TFEU Article 157 (1-2)). These are Directive 2006/54/EC 10 , complemented by the Pay Transparency Recommendation 11 . In the remainder of this document, Directive 2006/54/EC is referred to as ‘the Directive’ while the Pay Transparency Recommendation is referred to as ‘the Recommendation’. The two elements are jointly referred to as ‘the EU action’.

The evaluation’s scope is limited to analysing the relevant provisions of Directive 2006/54/EC with a view to improving implementation of the equal pay principle, its enforcement and protection of victims of discrimination, taking into account other existing legal provisions (Art. 157 TFEU, having a direct horizontal effect) and the non-binding 2014 Pay Transparency Recommendation and their impact. This scope allows the evaluation to focus on the implementation of the legal provisions related to protecting the victims of pay discrimination and the related pay transparency measures. Given the lack of robust quantitative indicators of pay discrimination, this evaluation looks at the changes to the gender pay gap, as a broad context indicator on pay differences and relies on qualitative sources to illustrate the impact of the relevant legislative measures.

The Commission’s action plan for 2017-2019 12 provides that the Commission will assess the opportunity for targeted amendments to the Directive with the aim, in particular, of improving pay transparency. Gender-focused commentary and the discourse on gender rights following the #MeToo movement have indirectly fuelled demands for better pay transparency. Several countries, such as Germany, France, Portugal and Ireland, have recently introduced new measures on pay transparency. Gender equality issues became more prominent internationally and the attention given to women’s rights in the public sphere clearly increased policy action to tackle gender equality issues.

A 2013 report on implementing the Directive 13 and a 2017 report on implementing the Recommendation 14 highlight persistent problems with the practical enforcement and legal interpretation of the principle of equal pay for equal work or work of equal value. These problems are confirmed by cases brought before the Court of Justice of the European Union (CJEU) 15 and by cases reported by the equality bodies 16 responsible for enforcing the Directive.

In line with the Better Regulation Guidelines 17 , this evaluation assesses the effectiveness, efficiency, relevance, coherence and EU added value of the EU’s legal provisions on equal pay 18 . In particular, this evaluation analyses:

·the relevant provisions of the Treaty (Article 157 (1-2)), which has direct effect 19 ;

·the Directive and its enforcement;

·the relevant provisions of the Recommendation and its follow-up in Member States.

The evaluation also assesses:

·the ways in which the existing legal provisions on equal pay have worked in practice;

·the approaches implemented in the Member States;

·the extent to which their initial goals have been reached;

·how effectively they have been enforced; and

·their overall impact against the overall trend of the gender pay gap.

The evaluation examines different aspects of the impact of the EU action on national legislation and its consequences for different stakeholders (e.g. equality bodies, employers, social partners, female and male employees and victims of sex-based pay discrimination). The evaluation also assesses any potential unnecessary or disproportionate burden or complexity.

The evaluation covers the period from the adoption of the Directive in 2006 to December 2017, with some references to the more recent years when relevant and possible. It focuses in particular on how the Directive has been applied and enforced since 2014, when the Recommendation was adopted. This SWD also tries to take account of developments that have taken place in Member States between the end of the evaluation period (December 2017) and the time at which the support study was prepared (May 2019). The geographic focus was the 28 EU Member States 20 .

2.BACKGROUND TO THE INITIATIVE

2.1.Description of the initiative

The Directive merged four directives 21 on equal opportunities and equal treatment of women and men, taking into account case-law from the CJEU 22 , into a single more readable and accessible piece of legislation that consolidated, updated and modernised EU legislation on gender equality in matters of employment and occupation 23 . The provisions of the Directive (see Annex 6) being evaluated were designed to make the implementation of the equal pay principle more effective, through preventing and combatting pay discrimination on the basis of sex.

To this end, the Directive pursues the following specific objectives:

·establish legal clarity on the existing provisions applying the principle of equal opportunities and equal treatment, including the principle of equal pay for equal work or work of equal value, in particular by clarifying the concepts of ‘pay’ (art. 2(1)(a), ‘equal/same work’ 24 (art. 2(1)(a), and ‘work of equal value’ (recital 9);

·facilitate respect for EU rights on equal pay by putting in place adequate judicial and administrative procedures to enforce the obligations set out in the Directive;

·increase legal protection for a complainant by ensuring that the burden of proof shifts to the respondent where there appears to be a case of discrimination (a prima facie case) by extending the rules on the burden of proof to the area of occupational social security schemes.

The Directive set out actions to achieve these objectives: social partners were encouraged to take part in negotiations on equal pay and to promote gender-neutral systems for evaluating and classifying jobs; courts were expected to implement the shift of the burden of proof to respondent; and Member States were required to introduce real and effective compensation or reparation and dissuasive sanctions, and ensure monitoring and enforcement of the equal pay principle.

Following the report on the implementation of the Directive in 2013 and given the incomplete application of the principle of equal pay for equal work and work of equal value and related implementation challenges, a Recommendation 25 was made to complement the Directive, covering the need for greater pay transparency. Member States were encouraged to adopt at least one of four core measures set out in the Recommendation and employers were expected to comply by responding to employee requests for pay information and/or by carrying out pay reports or audits. The Recommendation sought to encourage companies to analyse and review their pay structures and to improve social partners’ involvement in fighting conscious and unconscious gender bias in setting pay as part of the collective bargaining process.

The four main 26 measures promoted in the Recommendation are:

·Right to information on pay levels: this right allows employees to request information on pay levels (including elements beyond fixed basic salary, such as bonuses), broken down by gender, for categories of employees doing the same work or work of equal value.

·Pay reporting: employers with at least 50 employees should regularly report to employees, workers’ representatives and social partners on the average remuneration by category of employee or position, broken down by gender. This measure is aimed at smaller companies and requires simple reporting of information that is usually already collected by the companies’ human resources systems. 

·Pay audit: this transparency measure includes an analysis of the proportion of women and men in each category of employee or position, an analysis of the job evaluation and classification system used, and detailed information on pay and pay differentials on grounds of gender. The Recommendation suggests applying these measures only to companies with at least 250 employees, as audits may involve additional costs. The results of the audit should be made available to workers’ representatives and social partners on request.

·Collective bargaining on equal pay: equal pay, including pay audits, should be discussed at the appropriate level of collective bargaining.

The intervention logic of the EU action - a diagram summarising how the EU action was (originally) expected to work (i.e. at the time of adoption / implementation), including identifying the underlying assumptions depicted in Annex 4. The objective of the EU action was to promote the EU principle of and fundamental right to equal pay between women and men by allowing easier access to justice for victims of gender-based pay discrimination and the better enforcement of the principle of equal pay, including by promoting pay transparency measures. The operational measures available to Member States to implement the EU action should have allowed to improve pay transparency and should have created a supportive mechanism for victims of pay discrimination. As a result, greater awareness on the right of equal pay, remedies for gender pay discrimination and knowledge on the pay structure applied at the workplace, job satisfaction as well as better application of the equal pay provisions in national courts should have materialised.

2.2.Baseline 

Prior to 2006, the EU adopted a number of legal instruments to address equal pay between women and men, as part of its broader work on gender equality and the CJEU consolidated its interpretation of EU rules with key case-law (see Annex 7). EU Member States were bound by gender equality obligations under the Treaty and a number of directives implementing the principle of equal treatment for men and women. Beginning with the Equal Pay Directive of 1975 27 , the EU has adopted nine directives implementing the principle of equality between men and women in the areas of employment, work and social security. They contained equality guarantees (for example equal treatment in access to employment or equal pay) and special rights (for example pregnancy protection, maternity leave or parental leave). In addition to the Equal Pay Directive, the Equal Treatment Directive 28 and the Burden of Proof Directive 29 were of particular relevance to the institution of equal pay. International obligations and standards also apply to Member States 30 .

The key EU political strategy in place when the Directive was adopted was the ‘Lisbon strategy’ 31 . In the European employment strategy (EES), gender equality issues featured prominently as one of the main tools to achieve high employment levels in the EU 32 , reflecting the parallel development of the Beijing Platform for Action in the context of the 1995 UN Summit on Women. Following the mid-term review of the EES in December 2001, the gender pay gap was recognised as a key indicator for measuring progress towards gender equality 33 . This is calculates as the difference between average gross hourly earnings of male paid employees and of female paid employees as a percentage of average gross 34 hourly earnings of male paid employees 35 . This means that salaries in all sectors, occupations and career patterns are considered together. It summarises all differences between men and women in terms of pay and is therefore a handy tool for communication and policy-making. Nevertheless, interpreting the gender pay gap is not entirely straightforward (See Annex 3). In particular, it is important to stress that it includes issues other than direct gender pay discrimination, i.e. women being paid less for the same work because they are women. It also includes the influence of all possible gender imbalances on the labour market, namely labour segregation, constraints relating to care related career breaks and differences in working hours (part-time vs. full-time) and unobservable variables such as negotiating skills and personal preferences. These factors can be gender biased but are not necessarily related solely to gender pay discrimination.

According to the intervention logic in Annex 4, the expectation of pay discrimination and undervaluation of work influences the extent to which women are active on the labour market. The presented indicators match only some elements of the intervention logic and are limited by data availability. For example, quantitative data across the EU could not be gathered for career progression, reputational impacts on employers and occurrence of gender-based pay discrimination. Impact indicators include gender pay gap, representation on company boards, labour force participation and elderly poverty. Outcome indicators include job satisfaction. Additional relevant indicators (e.g. employment rate and part-time work) are also included in the baseline. To give an overview of the baseline scenario, Table 1 sets out a selection of indicators broadly related to equal pay between men and women for the years 2006 (year of the adoption of the Directive) and 2014 (year of adoption of the Recommendation). 

Table 1. Baseline indicators — 2006 and 2014

Note: Country-level results are available in Annex 3 — EU-25 and EU-28 figures for employment, labour market participation and part-time work are weighted by the number of employees. EU-25 and EU-28 figures for elderly poverty are weighted by population.

a Figures for EU-25 exclude Ireland and Greece due to missing data. Figures for EU-28 also exclude Croatia. b Average values for women (F) and average values for men (M) from 15 to 64 years. c Croatia not included. e Data at end of the year. f Share of respondents who reported being very satisfied or satisfied with their job. Data are available for 2005, 2010 and 2015. Data reported refer to 2006 and 2015 instead of 2006 and 2014.

Overall, the baseline indicators highlight a disadvantage for women in the labour market, which also result in an increased risk of poverty at older age. In 2006, the average hourly wage among women was 18% 36 lower than the average hourly wage of men in the EU. In absolute terms, this corresponds to median hourly earnings, which approximate the ‘typical earnings’ better than the average, of around €2 per hour less for women. A decomposition of the gender pay gap in 2006 using wage data from 21 Member States shows the large proportion of the gender pay gap not accounted for by demographic and occupational characteristics (see Figure 1). This may stem from gender-based discrimination, as well as other factors that cannot be directly measured, such as negotiating capacity.

Figure 1. Decomposition of the gender pay gap (2006)

Note: Analysis carried out by the evaluation team using the Structure of Earnings Survey.

In 2006, women’s participation in the labour market was substantially lower than men’s, in relation to activity rates and employment rates (about 14 percentage points (p.p.) difference for both) and the ratio of part-time work was about four times higher (see Table 1). The impact of these factors on pension entitlements may be one of the factors explaining the higher level of poverty 37 among older women (+4.9 p.p.). The baseline indicators also highlight a clear ‘glass ceiling’, where women are limited in their upwards progression within an organisation, with women making up only an estimated 9% of those serving on the boards of large companies listed on the national stock exchange in 2006. While the intervention logic in Annex 4 suggests that activity and employment rates, part-time work and vertical segregation would not be directly affected by EU action, they are nevertheless closely intertwined with the gender pay gap and must be considered in the evaluation, particularly when considering the impact attributed to the intervention versus other external factors.

In 2006, job satisfaction among employed people was similar for women and men, with slightly higher levels for women. This is consistent with studies that found women report higher job satisfaction despite having jobs with worse conditions. Research shows that the difference in job satisfaction narrows for younger and more highly educated women working in male-dominated fields, which suggests that gender differences in job satisfaction may be due to gender differences in job expectations 38 .

Finally, in 2006, gender-based discrimination was perceived as widespread by an average of 40% of the EU population, with a significant proportion of respondents (an average of 27% across the EU) reporting that gender-based discrimination had increased in the course of the previous 5 years 39 .

3.PROCESS, METHODOLOGY AND LIMITATIONS 

This evaluation complements existing knowledge from the 2013 report on implementing the Directive and the 2017 report on implementing the Recommendation, which identified persistent problems with the practical enforcement and legal interpretation of the principle of ‘equal pay for equal work or work of equal value’, a lack of transparency in pay systems and procedural obstacles.

The analysis in this document is largely based on the outcomes of an external study 40 (support study) carried out by ICF in 2018/2019. This support study includes extensive data on the functioning of equal pay provisions within EU Member States. The data were gathered through desk and national research, including on national and CJEU case law, evidence from equality bodies, a targeted stakeholder consultation survey, interviews with national stakeholders and various national and EU experts, and data analysis including a decomposition of the gender pay gap using Structure of Earnings Survey microdata (see Annex 3). The other source for qualitative data is the outcome of the stakeholders’ consultations carried out by the European Commission (see Annex 2 for more details). These include a public consultation carried out between 11 January and 5 April 2019 which received 386 replies and 16 position papers, and a targeted expert seminar with social partners (7 May 2019, Brussels). Quantitative and qualitative information from this wide range of sources were triangulated and synthesised to generate findings for each evaluation question (see Annex 5).

The evaluation faced some difficulties in producing robust quantitative comparisons. As mentioned (see also Annex 3), there are a number of caveats in the methodology for evaluating equal pay provisions. They mostly relate to data limitations, but the way the legal provisions are framed also plays a role.

First and foremost, it is not possible to precisely identify the percentage of the gender pay gap resulting specifically from pay discrimination. This limitation has obvious implications for the assessment of the scope of the problem.

As mentioned, the persistence of the gender pay gap (at an average level of 15.7% in 2018 41 , ranging from 3 % in Romania to 22.7% in Estonia) suggests gender bias on the labour market which might translate also into gender pay discrimination. Though it is not possible to directly measure this phenomenon, there is a large public perception that such pay discrimination exists 42 . According to a Eurobarometer survey 43 , 69% of EU citizens believe that women earn less than men for each hour of work 44 . This percentage, however, varies substantially across the Member States from 94% in Sweden to 31% in Romania (where 52% believe that women and men earn an equivalent wage). When asked about their perceptions of pay discrimination within their own company, people are less concerned, but many still believe there is a problem: 51% believe that women, on average, are paid the same as men, with a maximum of 56% in Czechia.

From a conceptual point of view, the definition itself of gender pay discrimination can be problematic. Broadly speaking, asymmetries in the opportunities to acquire skills and experience translate into pay differences. It is therefore incorrect to assume that gender pay differences linked to education and working patterns are legitimate, because they might well reflect discrimination and undervaluation of female work and capacities instead of different skills, qualifications and work experience. However, it is worth remembering that the purpose of the provisions analysed here is to enforce the equal pay principle i.e. to redress specific situations where the same skills required to perform a specific job are remunerated differently depending on the sex of the worker. While it is reasonable to expect that correcting these imbalances in the pay structure would ultimately also have an effect on the gender pay gap, the purpose of the provisions is not to correct all gender-motivated pay discrepancies (which could translate into closing the gender pay gap) but to enforce a fundamental right.

The Blinder-Oaxaca decomposition 45 of the gender pay gap identifies a ‘residual’ portion of the gender pay gap that gives a rough quantitative illustration of the possible relevance of pay discrimination to the gender pay gap. However, this portion also accounts for the impact of parenthood on wages and other aspects such as preferences, the relative negotiating advantage linked to scarcity premiums, or individual productivity and so it cannot be used as a proxy. More broadly, it is not possible to provide an estimate of the extent to which pay differentials are gender motivated because there is no aggregate measure that allows measuring (direct and indirect) discrimination per se. It can only be gauged indirectly, looking at structural differences. These limitations were taken into account in the evaluation, which does not consider that the unexplained part of the gender pay gap only result from gender discrimination 46 However, given that the gender pay gap shows a structural / systemic pay difference across the economy that goes beyond the individual characteristics of working men and women, it is considered a good indicator of inequalities in access to work, progression and rewards; and as such relevant in discussing the equal pay issues. Such it is referred to in this evaluation also for lack of more specific indicators.

Secondly, it is not realistic to expect that the gender pay gap can be substantially decreased by better enforcing equal pay laws. The core measures by the EU action address only some of the factors underlying the gender pay gap, with less of an effect on others. Consequently, even with effective enforcement of the existing EU legislation, wider problems affecting pay differences may still persist. This is primarily due to the unequal distribution of unpaid work between women and men. Since women continue to take on more responsibility than men for most domestic and caring tasks within the home 47 , they tend to work fewer paid hours, with consequences for pay and promotion. In this context, the Directive on work-life balance for parents and carers 48 was adopted in June 2019 specifically to support more equal sharing of caring responsibilities between women and men, thereby creating more opportunities for women to participate in the labour market. The Directive is accompanied by non-legislative measures to increase the availability of accessible, affordable and high quality childcare and remove remaining tax disincentives which deter second earners, who are most often women, from engaging in paid work or increasing their working hours.

Thirdly, particularly if they work part-time 49 , women face persisting gender segregation into particular types of occupations, firms or sectors. This kind of segregation limits the applicability and effectiveness of equal pay laws, due to the scarcity of male comparators within the same employing organisation undertaking equivalent work. Since the principle of equal pay is set in EU law as an individual right, litigants must build their complaint for discrimination on the direct comparison of male-female wage from the same employer. This in practice excludes the possibility of comparing ‘equivalent’ jobs (in terms of work demands and/or level of professional skills) across occupations, in particular between male-dominated occupations and female-dominated ones, since this implies different employers. As a result, segregation facilitates the persistent undervaluation of women’s work in certain professions and women often find themselves in occupations with the lowest pay scales. With new and improved rights to request flexible working arrangements in any sector, the 2019 Work-Life Balance Directive could also help reduce this kind of occupational segregation (assuming it is not counter-balanced by a larger increase in female part-time in low paid sectors). These issues might also at least partly explain the scarcity of the number of equal pay cases which has led to a gap in a robust understanding of how EU legislation is implemented by courts. Few claims of gender pay discrimination reach the national courts, as confirmed in this evaluation (see Section 5.1). Therefore, the conclusions drawn are based on the few cases identified and suggest an issue relating to access to justice in general. Finally, the considerable variation in measures and approaches to implementing equal pay provisions across the EU Member States makes it challenging to draw overall comparisons and findings.

4.IMPLEMENTATION AND STATE OF PLAY

The 2006 Directive consolidated EU law on equal treatment between women and men in the area of employment and occupation by bringing together, modernising and simplifying the provisions in previous Directives and incorporating case law from the CJEU. It added a definition of pay, consolidated concepts such as ‘the reversed burden of proof’ and ‘victimisation’ and explicitly extended the application of equal treatment in occupational social security schemes in line with the CJEU’s case law. The transposition and implementation of the Directive by Member States focused primarily on these elements. Article 33 of the Directive required Member States to bring into force the laws, regulations and administrative provisions necessary to comply with this Directive by 15 August 2008 at the latest or to ensure, by that date, that management and labour implement its provisions by way of agreement.

The Commission examined the compliance of national legislation in all Member States. Following detailed discussions with Member States, and following the adoption of additional legislation in some Member States, the Commission considered that most issues of compliance were resolved. Infringement proceedings were, however, launched in 2013 in relation to 3 Member States. One set of proceedings was closed in 2017 following adoption of legislation regarding return to work after paternity and adoption leave in the Member State concerned (Lithuania). At the time of writing, the remaining two cases are still open and being monitored (relating to the definition and scope of compensation in Poland and the scope of protection against victimisation in Belgium). The Commission is in discussions with 2 Member States regarding other questions related to the gender pay gap, namely: the payment of certain categories of workers who are absent from work for pregnancy-related illness, and the effectiveness of the equality body.

The 2013 Directive implementation report

In its 2013 implementation report 50 , the Commission reviewed the transposition of the Directive in all Member States. 14 Member States had adopted new measures or amended existing legislation to transpose the Directive 51 ; 11 Member States already had legislation in place in line with the Directive’s obligations 52 and the 3 remaining Member States (Bulgaria, Romania and Croatia) introduced legislation as part of their accession to the EU. 

The 2013 report assessed in particular Member States’ transposition of the Directive’s novel features and the effectiveness of its application and enforcement 53 . It highlighted that the practical application of the equal pay provisions was one of the Directive’s most problematic areas. The results of the 2013 assessment of the implementation of the relevant equal pay provisions of the Directive are further described in Section 5.1.

The 2014 Pay Transparency Recommendation

The 2013 findings on the application in practice of the equal pay provisions led to the adoption of the Recommendation in 2014. The Recommendation provides guidance to help Member States implement the equal pay principle more effectively and proposes measures to support the identification of unfair wage inequalities within companies and organisations.

Table 2 below lists Member States that have revised existing or adopted pay transparency measures, grouped by the time of adoption or amendment. It also highlights Member States which had measures in place prior to the Recommendation but made amendments to those measures after 2014. The table shows that 7 Member States maintained measures already in place before 2014 and 6 others amended or adopted transparency measures after the adoption of the Recommendation. These measures are, however, not always in line with the Recommendation. For example, in the UK, the pay reporting measure only applies to employers with 250 or more employees, unlike the ‘50 or more’ set out in the Recommendation. In Germany, as another example, the pay audit measure is not mandatory and only applies to companies with at least 500 employees.

Table 2. Pay transparency measures adopted by type and time of adoption

Ireland and the Netherlands have an advanced draft of pay transparency legislation, but in the 13 other Member States not listed above 54 , there are no pay transparency measures.

The 2017 report on the implementation of the Pay Transparency Recommendation

The findings of the 2013 implementation report on the Directive were confirmed in the implementation report on the Recommendation 55 . These reports, as well as a study by the European legal network for gender equality 56 , raise the following key issues which are explored further in Section 5.1:

·Application of equal pay concepts: The concepts of ‘pay’, ‘equal work’ and ‘work of equal value’ are not defined uniformly across national legislation.

·Possibility of justifying differences in pay: in pay discrimination claims, some courts accept employers’ justifications for pay differentials, unduly favouring employers and making access to justice difficult for victims. However, the report also notes that some Member States do not allow any such justifications.

·Implementation of ‘the reversed burden of proof’ varies among Member States and can be unclear. The shift of the burden of proof, leading to a more equitable sharing of the burden of proof between the victim and the employer in cases of discrimination, is not sufficiently actionable, due to a lack of clear rules on the level of evidence required for a presumption of discrimination triggering this shift in burden of proof.

·Limitation periods vary and, in some Member States, they are so short that they may not allow sufficient time for victims to make a claim.

·Lack of deterrent effect of penalties and low compensation levels: compensation awarded by courts varies across countries, tends to be low and rarely accounts for non-material damages. This does not motivate victims to claim their rights.

·Costs of proceedings impede access to justice. The costs of legal proceedings to access justice, both as regards legal representation and court fees, are a key disincentive for many victims, in particular in light of the low compensation levels even if successful.

·Lack of knowledge of gender pay discrimination rules on the part of employees. There is sometimes limited knowledge of equal pay legislation and how to enforce existing rights on the part of employees, trade unions, lawyers and even judges.

·Lack of clarity on and awareness of the powers of national equality bodies. The powers of national gender equality bodies vary across the EU and only some Member States allow equality bodies to represent individuals in equal pay claims.

·Lack of promotion, development and use of gender-neutral job evaluation and classification systems. While in some Member States, legislation or collective agreements explicitly require job evaluation and classification systems to be gender-neutral, in others there is no such requirement, or at least not an explicit requirement.

·Barriers to pay transparency and lack of transparency in pay systems. Accessing pay information is difficult, which restricts awareness of gender pay discrimination and the ability to support pay discrimination claims. Individual pay information is confidential, which undermines pay transparency.

This evaluation
This evaluation investigates the extent to which the issues identified by the 2017 report remain problematic in the Member States. Table 3 below summarises the assessments carried out as part of the support study.

Table 3. State of play — mapping of the issues by relevance

Source: ICF country fiches (2019).

5.ANALYSIS AND ANSWERS TO THE EVALUATION QUESTIONS

The EU law action was evaluated on the basis of five criteria: effectiveness, efficiency, relevance, coherence and EU added value.

5.1.Effectiveness 

The assessment of effectiveness considers how successful EU action has been in achieving or making progressing to achieving its objectives. Performance indicators allow the extent to which that purpose has been achieved to be measured. As summarised in the intervention logic (see Annex 4), the needs that the Directive aimed to address were on three interlinked levels:

1.the guarantee of an EU fundamental right, i.e. non-discrimination on grounds of sex;

2.the economic necessity to facilitate female labour market participation by removing one of many barriers; and

3.the enforcement of the principle of equal pay between women and men enshrined in the Treaty.

Since it is not possible to measure specifically the impact on gender pay discrimination of measures driven by the EU action, because it is not possible to disentangle the portion of the gender pay gap linked to it, the impact of these measures can be only indirectly gauged through the broader impact on gender equality in the labour market and society.

The Directive’s aim, reinforced by the Recommendation, is ‘to ensure the implementation of the principle of equal opportunities and equal treatment of men and women in matters of employment and occupation’ (Art. 1). This broader objective builds on the three specific aspects mentioned in Section 2.1:

1)legal clarity on the application of the principle of equal opportunities and equal treatment, including the principle of equal pay for equal work or work of equal value, especially by clarifying the concepts of ‘pay’, ‘equal work’, and ‘work of equal value’;

2)facilitating respect for EU rights on equal pay by putting in place adequate judicial or administrative procedures for monitoring and enforcement of the Directive’s requirements e.g. the introduction of job classification and evaluation systems based on the same criteria for both women and men; and

3)facilitating access to justice with legal protection for a complainant by shifting the burden of proof to the respondent where there is a prima facie case of discrimination.

The analysis of the effectiveness of EU action in relation these three aspects relies mostly on a qualitative assessment provided by national experts involved in the evaluation. Specific quantitative data to construct relevant indicators for the broader impacts of the measures hypothesised in the intervention logic (see Annex 4) are, for the reasons mentioned above, scarce. Since external economic and social factors (such as economic growth and demographic trends) are more likely to shape the overall trend of the gender pay gap, it is not possible to disentangle the specific effect of the Directive. Member States were starting from different positions in 2006 as this was a ‘recast’ directive integrating previous directives and case-law of the Court of Justice. Therefore, the degree of implementation of the Directive itself does not reflect the level of compliance with the Directive’s content in the Member State concerned. Nevertheless, a quantitative analysis of the trend of the relevant indicators selected for the baseline and an econometric analysis of gender pay gap components attempts to shed light on the effects at macro-level.

5.1.1. Implementation of the concepts of ‘equal pay’, ‘same work’ and ‘work of equal value’ in national legislation and practice

The concepts of ‘pay’, ‘equal pay’, ‘same work’ and ‘work of equal value’ are derived directly from Article 157 TFEU. The Directive’s primary objective, set out in the first recital, was to clarify the provisions of existing directives implementing these Treaty provisions, in particular by incorporating certain developments from the case-law of the CJEU, including regarding the legal concepts of ‘equal pay’, ‘same work’, and ‘work of equal value’.

The concept of ‘pay’

The CJEU interpreted the concept of ‘pay’ to include not only salary, but also additional benefits such as bonuses, overtime compensation, travel facilities, compensation for attending training, payments in case of dismissal, statutory sick pay, statutory required compensation and occupational pensions. In order to clarify the concept of ‘pay’, Article 2(e) of the Directive recalls the Treaty definition of ‘pay’ as ‘the ordinary basic or minimum wage or salary and any other consideration, whether in cash or in kind, which the worker receives directly or indirectly, in respect of his/her employment from his/her employer’. The Directive took account of CJEU case-law clarifying the concept of ‘pay’ 57 . As such, the concept of ‘pay’ in the Directive also includes:

·all forms of occupational pensions 58 (recital 13);

·benefits from pension schemes for public servants if those benefits payable under the scheme are paid to the worker by reason of his/her employment relationship with the public employer, even if such schemes form part of a general statutory scheme. That will be the case if the pension scheme concerns a particular category of workers and its benefits are directly related to the period of service and calculated by reference to the public servant’s final salary (Article 7(2) and recital 14).

In addition, the Directive clarifies that an inequality in employers’ contributions paid under funded defined-benefit schemes which is due to the use of actuarial factors differing according to sex does not come in the scope of Article 157 TFEU (Article 9(1)(h) and recital 15). While actuarial factors differing according to sex cannot be used to calculate male and female workers’ contributions (Article 9(1)(j)), the use of such differing actuarial factors to calculate employers’ contributions is under certain circumstances permitted by the Directive as not contrary to the principle of equal treatment (Article 9(1)(j)). Recital 16 further clarifies to which elements of funded defined-benefit schemes this exemption may apply.

The Directive also draws attention to the content of Protocol 17 concerning Article 141 (now 157) of the TFEU, according to which benefits payable under occupational social security schemes are not to be considered as remuneration insofar as they are attributable to periods of employment prior to 17 May 1990, except in some well-defined cases.

According to Better Regulation standards 59 , the definition of ‘pay’ was considered to be a ‘substantial change’ compared to the pre-existing directives. This required, according to Article 33 of the Directive, explicit transposition in national law.

The Directive’s definition is broad and encompassing, but lacks the necessary complexity and precision that characterises the CJEU case-law. All Member States but six 60 have implemented the concept of pay contained in the Directive 61 . In the UK and Latvia, where no codified definition exists, national case-law has developed a definition in line with EU law 62 . In two Member States Lithuania and Romania) which have implemented the definition, uncertainty in the interpretation of the law remains. This may be due to an incomplete implementation: for instance, indirect payments are not mentioned in the Lithuanian law, thus excluding various benefits or services provided by third parties (including insurance or pension benefits) from the domestic concept of pay 63 . Finally, in the Netherlands and Portugal, experts have noted that even if the definitions of ‘pay’ are not as well-developed as in EU legislation, the meaning is the same in practice.

The concepts of ‘same work’ and ‘work of equal value’

The Directive does not contain a binding definition of ‘same work’ and ‘work of equal value’ (these concepts are only referred to in Recitals 9 and 10 of the Directive). In order to determine whether workers are performing the same work or work of equal value, it should be determined whether they can be considered to be in a comparable situation. In order to make that determination, a range of factors should be taken into account, including the nature of the work and training and working conditions. Furthermore, under certain conditions established by the CJEU 64 , the comparison should not necessarily be limited to situations in which men and women work for the same employer.

The 2014 Pay Transparency Recommendation invited Member States to clarify the concept of ‘work of equal value’ in their national legislation. It indicated that the value of work should be assessed and compared based on objective criteria, such as educational, professional and training requirements, skills, effort and responsibility, work undertaken and the nature of tasks involved.

Even if EU Member States have included the concepts of ‘equal/same work’, ‘work of equal value’ or ‘equivalent work’ in their legislation, and despite the guidelines given by the CJEU that are incorporated in the preamble to the Directive and the 2014 Pay Transparency Recommendation, these concepts still are not always defined in national legislation, and where they are defined, it is not done uniformly in various Member States. This affects the ability of complainants in pay discrimination cases to establish a valid comparator; moreover, it also creates substantially heterogeneous situations in EU Member States. Some countries (BG, FI, IT, LV, MT 65 , PL 66 , SI, EE) have no legal definition at all, and even where a definition exists (CZ, DK, EE, EL, FR, HR, HU, IE, PT, SE, SK, UK) experts report practical problems in clarity and lack of understanding of the concepts.

Experts report difficulties, for instance, in identifying a ‘comparator’ because national legislation is restrictive on what is considered ‘work of equal value’. The identification of a comparator is a crucial parameter in determining whether work may be considered ‘of equal value’. The CJEU has developed jurisprudence 67 to determine whether two different groups of workers perform work of equal value, taking into account the overall nature of the work, the responsibility involved, the training requirements and the working conditions. The CJEU indicated — as far back as 1982 — that Member States should ensure workers can claim before a competent authority, by a binding decision and through effective means, the application of the right to equal pay for work of equal value. That authority, which can be courts, would be empowered to assess which work has the same value in application of the Treaty and the Directive 68 . However, the lack of parameters for the implementation of the concept can lead to difficulties in its application by courts.

National legislation does not always elaborate on the definitions and parameters, as this is not imposed by the Directive and thus is left to the discretion of the courts’ interpretation. Only 10 Member States 69 have laid down specific parameters and criteria for establishing the equal value of the work performed (BG, HR, CZ, FR, HU, IE, PL, PT, SE, the UK). Since the number of cases brought before the courts is low, it is difficult to develop consistent jurisprudence; at the same time, as in a vicious circle, there is little incentive to file complaints.

The CJEU has held that although the comparator does not necessarily need to be employed at the same time as the complainant, but comparisons must be made ‘on the basis of concrete appraisals of the work actually performed by employees of different sex’ 70 . To date, 11 Member States 71 require an actual comparator to be identified, while 11 others 72 allow a hypothetical comparator. In the 6 remaining Member States 73 , case-law and national legislation remain unclear on this point.

The requirements for a real-life comparator can be very narrow. In seven Member States 74 , the comparator must be employed by the same employer, thus preventing comparisons across sectors and companies. According to some authors, this requirement reduces the effectiveness of the concept of ‘work of equal value’ in highly segregated labour markets 75 and limits the probability of finding an individual undertaking similar tasks within the company or similar companies in the same highly segregated sector. However, in some Member States, wages in these sectors (such as education and healthcare) are set by the public sector. While pay may be equal within the sector, pay inequality can transpire when compared to highly segregated male-dominated sectors with similar training requirements and working conditions, where equal value is harder to establish. Those male-dominated sectors are likely to be private sector driven with higher set wages. This contributes to the gender pay gap and arises from indirect discrimination, which is harder to establish.

In the UK, in the ASDA case 76 , the claimants (mostly women working in the respondent’s supermarkets) were entitled to compare themselves, for equal pay purposes, to employees working in the respondent’s distribution operation. Even though they did not work in the same establishment, claimants claimed that ‘common terms of employment’ applied to both retail and distribution workers. The Court agreed that the respondent applied common terms and conditions to both retail and distribution workers ‘wherever they work’. The phrase ‘wherever they work’ allows reference to comparators who work in a different workplace because the nature of their tasks is different. National experts involved in the evaluation emphasise, however, that the use of such flexible comparators remains very rare.

Finding the right comparator may be even more challenging in Member States 77 where secrecy or confidentiality clauses regarding pay are included in employment contracts. In Luxembourg, for instance, courts have denied access to information about colleagues’ pay in several cases on this basis 78 .

Allowing reference to a hypothetical comparator enables more effective correction of the gender pay gap in highly segregated labour markets 79 . In France, Portugal and Spain, a comparison may be made with employees that previously worked in the same job. In France, Hungary and Denmark, claimants can argue that a discriminatory preference exists without referring to any particular employee as a comparison.

Finding the right comparator depends on many factors. It will be more difficult in the absence of established job evaluation/classification systems and in the absence of pay transparency and pay information rights. In Member States where pay inequalities are not reported, it is evidently very difficult to find a real-life comparator with a higher salary.

The lack of clear concepts used to evaluate whether men and women receive equal pay and of clear assessment criteria for comparing different jobs is a major obstacle in addressing pay discrimination, due to the gender-segregated structure of the labour market and the undervaluation of female-dominated professions 80 .

Conclusion on the effectiveness of EU action concerning the clarity of equal pay concepts

To ensure enforcement of the fundamental right to equal pay, the concepts of ‘pay’, ‘same work’ and ‘work of equal value’ must be clarified. While the Directive has improved the consistent interpretation of these concepts, they remain difficult to circumscribe in practice. 15 national legal experts rated the relevance of the problem as ‘moderate’ 81 or ‘significant’ 82 , leaving 12 experts 83   considering it of ‘limited’ 84 or no importance 85 . Over one third of the respondents in the public consultation highlighted the lack of legal clarity as the main obstacle to the effective implementation of the equal pay principle.

Both the European Parliament 86 and the European legal experts on gender equality 87 point out that most Member States simply transposed the provisions of the Directive textually, without any further guideline or indication on how to apply them. This leads to a lack of parameters and indicators permitting implementation at national level.

The Court of Justice has played a very significant role in ensuring enforcement of the right to equal pay between men and women. First of all, the Court has given substantial guidelines on how to interpret the concepts used to guarantee equal pay 88 . In addition, the Court has ensured that workers have the right to effectively claim their rights in practice: in case C-61/81 89 , the CJEU, with regard to the predecessor of the Directive, namely Directive 75/117/EEC, held that Member States must guarantee the right of workers, in the absence of a job classification system, to claim before an appropriate authority that their work is of the same value as other work, and, if that is found to be the case, to have their rights under EU law acknowledged by a binding decision. The case related to workers in the UK, who were only able to claim their right to equal pay on the basis of a job classification system. Such a system was, however, not mandatory for employers. As a result, people who worked for employers that did not have a job classification system were not able to claim their rights. The Court held that UK had not taken the necessary measures to ensure that the principle of equal pay is applied.

5.1.2. Effectiveness of EU action in terms of access to justice

The Directive contains a number of important procedural provisions 90 which are key to ensuring enforcement of the equal pay principle. These provisions include:

·the opportunity to defend victims’ rights;

·the sharing between employers and victims of the responsibility to prove the absence/presence of discrimination where a prima facie case is established;

·the curtailing of retaliation against employees who complain about pay discrimination (such behaviour is known as ‘victimisation’);

·the incentives to seek justice through the payment of compensation or reparation to victims of discrimination.

All of these provisions are considered important by national experts and respondents in the public consultation. Nevertheless, several problems have been highlighted as regards their application in practice. In general, national experts consider the protection against victimisation entirely satisfactory in four Member States. The application of the rules on the shared burden of proof is considered the most problematic issue in 27 Member States, followed by defence rights (23) and compensation (22). A detailed assessment is presented below, broken down by provision.

Defence of rights (Article 17)

Article 17 requires Member States to ensure access to judicial procedures for the enforcement of the rights and obligations granted by the Directive. Such access must be available, after possible recourse to conciliation procedures and even after the end of the employment relationship in which the discrimination is alleged to have occurred. In addition, the Directive provides that Member States must ensure that organisations or other entities with a legitimate interest may act on behalf or in support of discrimination victims in judicial or administrative proceedings. Finally, Article 17(3) provides that the time limits to bring actions remain exclusively governed by national law.

The support of victims by organisations or other entities with a legitimate interest

Article 17(2) is implemented in Member State legislation by enabling collective redress or by allowing victims to be represented in court by certain organisations such as trade unions, NGOs or equality bodies (See Table 4 below).

Table 4. Organisations that can act on behalf or in support of victims

Source: ICF (2019), Country fiches; Equinet website

The degree of involvement of these organisations varies. In 18 Member States, equality bodies have the power to bring cases to court or to represent the claimant before courts or administrative bodies. In Member States where the equality body is not empowered to bring cases to court, it can nevertheless assist and advise victims (e.g. AT) and, in some cases, issue binding (BG, CY, LT, MT, PT, RO) or non-binding decisions (CZ, DE, FR, LU). Collective action is possible in only four Member States (CZ, FR, NL, UK).

Some legal experts (especially from EL, LT, LU, SI and UK) noted that while organisations may be empowered by law to bring cases or represent victims before courts, in practice they do not make use of these opportunities, usually because of lack of resources or expertise. Such organisations report being ill-equipped and insufficiently funded to handle the complex and lengthy litigation that equal pay claims often involve. Some opt to support strategically important equal pay cases only.

Limitation periods

One important problem highlighted by experts and stakeholders concerns the length of limitation periods, which remain exclusively governed by national law (Article 17(3)). In some Member States (DE, LV, EE, SI), the time limit to bring a case to court is simply too short, ranging from 30 days to 3 months. Victims need time to gather the necessary evidence, which can be a complex and difficult exercise, particularly if there is no transparency on pay structures and wages. Victims therefore often lack the time to take action once they become aware of the discrimination. Some victims already move on to another job before considering filing an action against their previous employer. Many victims decide simply not to bother because of the unsurmountable difficulties involved in taking the necessary action within the required time limits.

Effective compensation or reparation (Article 18)

The Directive requires Member States to ensure real, effective, dissuasive and proportionate compensation or reparation for loss and damage sustained as a result of pay discrimination on the grounds of sex. Member States must not impose a prior upper limit on the compensation or reparation, unless the employer can prove that the only damage suffered by the claimant is the refusal to take his/her job application into account.

As presented in Table 5 below, 18 Member States ensure compensation for material damages; non-material damages may be compensated in 14 Member States.

The level of compensation is usually (e.g. in AT, BE, FI, IT) calculated by the courts taking into account the circumstances of the individual case, such as the nature, extent and duration of the discrimination. In 13 Member States, the level of financial compensation may be up to the difference between the victim’s wage and the wage of the comparator. In two countries, fixed compensation amounts have been identified. In Belgium, the victim can opt for either a standard amount, which is 6 months’ salary covering both material and non-material damage, or the amount of the damage actually suffered by the worker. As this system does not fix a prior upper limit, it can be considered in line with the Directive.

In Spain, the compensation must aim to restore, as far as possible, the integrity of the victim to the situation prior to the injury, as well as contribute to preventing harm. In Sweden, the Discrimination Act requires the compensation to fulfil both a reparative and a preventative purpose. Even though the preventative function was introduced (in 2008) to promote higher compensation levels, damages awarded generally remain low.

Table 5. Types of compensation and reparation

Source: ICF (2019), Country fiches

Most national experts involved in the evaluation regard compensation or reparation of victims as an important (12) or moderate issue (10). If compensation amounts are too low, they are not sufficiently dissuasive for defendants; in addition, especially when compared to litigation costs, they discourage victims from taking judicial action against gender pay discrimination. While the issue of dissuasiveness may be offset by criminal sanctions (See 6) in certain countries (BE, CY, FI, LT, MT), legal experts note that civil compensation remains an important avenue of redress and dissuasiveness. In Belgium, victims prefer civil redress because the threshold for evidence is easier to reach than in criminal proceedings. In Cyprus, it is reported that the courts prefer to award civil compensation to address pay discrimination, shying away from imposing fines.

Table 6. Types of penalties

Source: ICF country fiches (2019)

* Administrative fines can be imposed in Estonia, but in practice the labour inspectorate cannot enforce them due to limited resources.

Burden of proof (Article 19)

It is generally very difficult for victims of discrimination, including pay discrimination, to meet the necessary legal standards to prove the existence and extent of the discrimination. The EU addressed this matter already back in 1997, in a directive specifically intended to alleviate the burden of proof in cases of discrimination based on sex 91 . That directive was part of the recast exercise which resulted in the current Directive.

Under the Directive, Member States must ensure that, when a claimant establishes ‘facts from which it may be presumed that there has been direct or indirect discrimination’, it is for the employer to prove that there has been no breach of the principle of equal pay (Article 19).

While the reversal of the burden of proof and protection from victimisation can be considered to confer improved access to justice, these provisions depend heavily on their implementation in practice at national level. This includes how victims can establish the facts to qualify for the presumption of pay discrimination necessary to trigger the shared burden of proof. The burden of proof, in practice, continues to fall heavily on victims (see below).

Pay transparency measures have the potential to support the implementation of the reversed burden of proof. This can assist female workers in determining the average pay for women and men. The Directive does not provide for binding pay transparency measures (such as the right to request pay information) that could facilitate the establishment of presumed pay discrimination. EU action provides for such provision in the form of soft law, as laid down in the Pay Transparency Recommendation. In Danfoss 92 , the Court of Justice already held that where an employer’s pay system lacks transparency, the employer is responsible for proving that the system is not discriminatory. This enables workers to provide evidence from which discrimination can be presumed, thus triggering the reversed burden of proof. 

The provision on burden of proof has been transposed in all Member States’ national legislation, although the specific formulation varies among Member States:

·19 Member States (BE, BG, CY, EE, EL, ES, FI, IE, IT, LU, LV, MT, NL, PL, RO, SE, SI, SK, UK) require victims to establish prima facie discrimination in order to trigger the mechanism regarding the burden of proof.

·In Austria, the claimant must make pay discrimination based on sex ‘credible’ before the courts. Depending on the judge’s knowledge of the legislation and pay discrimination, this task may be more or less challenging.

·In DK, EE and HU also, once the employee presents credible arguments/facts, the employer must prove the lack of discrimination. These Member States refer explicitly to the concept of a ‘shared’ or ‘divided’ burden of proof.

·In France, employees must present evidence ‘that leads to believe that direct or indirect discrimination has taken place’.

·In Croatia, on the other hand, the claimant must present facts which ‘justify the suspicion that discriminatory treatment has occurred.

·In Lithuania, the victim must ‘indicate the fact of discrimination’.

·In Germany, victims need to supply circumstantial evidence that can prove the likelihood or lead to the assumption that a discriminatory act/fact exists, but it is not clear exactly what type of proof should be delivered.

·In Portugal, the claimant must indicate the worker(s) in relation to whom they feel they have been discriminated, and the employer must prove that the difference in treatment is not based on discrimination.

Experts from Member States highlight that in general it is not clear what standard should be met to establish the required presumption of discrimination that would tilt the burden of proof towards the employer, in favour of the victim. The concept used in the Directive, transposed in national law as requiring a ‘probability/credibility/likelihood’ of discrimination, may not correspond to standards of evidence applicable under national procedural law, making it difficult for courts to apply it in a consistent and transparent manner. For example, in Belgium, the Gender Equality Act provides for a presumption of discrimination and the defendant must prove that there has been no discrimination. It further details which elements can contribute to a presumed discrimination, such as the recurrence of the discrimination, or the victim being in a comparable situation to the reference person, as well as statistics in cases of indirect discrimination. In Portugal, the presumption of pay discrimination is established in cases where the worker claims to be discriminated against and the employer does not present a transparent remuneration policy that demonstrates that the alleged differences are based on objective criteria.

Experts from Belgium, Greece, Italy, the Netherlands and the UK report that data or statistics can be used as evidence in court. The use of data and statistics can facilitate the establishment of discrimination, in particular in cases of indirect discrimination or in cases where reference is made to a hypothetical comparator. However, there is a lack of clarity on whether or not statistical data to establish pay discrimination can be used in the courts across the EU. Moreover, even if such data and statistics are accepted as evidence, such data and statistics may not exist and, even if they do exist, workers usually do not have easy access to them. Workers usually also have no access to experts or technical studies on the equality and value of the work concerned that would make comparisons across the board. The support and expertise of NGOs, trade unions and equality bodies in litigation in these cases can be crucial.

One way to establish the presumption of discrimination is by identifying a comparator. However, as mentioned above, this may be very challenging. Issues with providing evidence about a comparator have been reported in 14 Member States (BE, CY, CZ, DK, EE, EL, FI, HR, IE, LT, LU, NL, SI, UK).

In Member States where pay discrimination is a criminal offence (BE, CY, FI, LT, MT), the difficulty in gathering the evidence necessary to establish a prima facie case of discrimination may be mitigated by the fact that the public prosecutor has the power to investigate and gather elements to establish a discrimination case before the criminal courts. Evidence gathered in this way may usually be used by the victim in subsequent/parallel civil proceedings. However, according to stakeholders interviewed by national experts, criminal investigations are rare in cases of alleged pay discrimination, probably because the scarce public resources to investigate criminal offences can only be made available when there are sufficient elements pointing to a criminal offence in the first place, thus creating a vicious circle of lack of evidence.

Overall, experts identify applying the rule on the burden of proof as the most problematic provision relating to access to justice: experts from 11 Member States assess it as ‘important’, while experts from another 16 Member States assess it as ‘moderate’.

Protection against dismissal or other adverse treatment (Article 24)

The Directive requires Member States to ensure that measures are in place to protect employees against dismissal or other adverse treatment by their employer as a reaction to a complaint within the undertaking or to any legal proceedings related to pay discrimination. National experts involved in the evaluation have not detected any issues regarding the application of this legal provision.

Member States have all established a prohibition on dismissal and adverse treatment for workers who file a complaint or action before the courts on equal pay. National legislation provides for sanctions in the form of fines (CY, LT, LV, RO, SI), compensation (AT, BE, DK, EE, EL, FI, IE, LV, MT, NL, SE, UK) or the requirement to reinstate the worker or consider void the dismissal or adverse decision (BE, ES, FR, IE, LU, MT, NL, PL, PT, SK). In some Member States, the victim can only claim this protection before the courts in different proceedings from the ongoing pay discrimination proceedings; in practice this makes the protection far from effective because in order to be protected victims must file a separate, additional claim before a court (BG, CZ, DE, HR, HU, IT). This requires an additional investment of time and costs, which acts as a deterrent for victims to claim their rights.

The study and the consultations show that fear of dismissal or adverse treatment remains a significant barrier to seeking redress 93 . 76% of public consultation respondents 94 rate this matter as ‘rather important’ or ‘very important’ (see Annex 2). The relevance of the issue was confirmed by the legal experts pointing to adverse treatment such as harassment or barriers to career advancement being difficult to prove in court.

Other barriers to access to justice

Further barriers to access to justice mentioned by legal experts and stakeholders include:

·costs of proceedings, with legal aid only available to (very) low income earners: in some countries (e.g. AT, BG 95 , SE) the losing party also bears the costs of the winning party, making the stakes of pursuing litigation very high;

·lack of courts’ expertise on the legal framework for gender pay discrimination 96 ;

·low social awareness among victims of pay discrimination and lack of pay transparency 97 ;

·length and complexity of the proceedings linked to difficulties in proving discrimination and in finding the required comparator.

Conclusion on the effectiveness of EU action in ensuring access to justice for victims of pay discrimination

The legal protection provided for in Articles 17 through 19 and 24 of the Directive is considered very important in ensuring proper enforcement of the equal pay principle. However, in practice, it seems that those seeking redress in cases of alleged pay discrimination continue to face substantial barriers. Around 60% of respondents in the public consultation (Annex 2) rate all factors listed in Table 2 of the Annex 2 as ‘rather important’ or ‘very important’. The most relevant issues mentioned are the fear of professional consequences (76%), the length and costs of litigation (81%), and the problems relating to the burden of proof.

The literature review confirms that very few claims of gender pay discrimination reach the national courts 98 . This is also reflected in the legal research carried out in the support study, which identified only a few national cases in Cyprus, Finland, Germany, Greece, Malta and the UK. The research also points to a lack of expertise on EU anti-discrimination law, including issues of equal pay, among judicial authorities at national and regional level.

In a European Trade Union Confederation (ETUC) survey carried out in 2019, most confederations (20 out of 38) reported that no information on the number of cases was available, and 10 reported that they did not know or did not consider the issue to be relevant. Only eight confederations confirmed the existence of statistics, but considered the number of cases ‘very low’ or ‘somewhat low’ compared to the scale of the issue 99 .

5.1.3. Effectiveness of action in terms of monitoring and enforcement

Effective, proportionate and dissuasive penalties (Article 25)

Under the Directive, Member States must lay down penalties for infringements of national provisions adopted pursuant to the Directive and should take measures necessary to ensure that they are applied. The Directive allows for penalties to comprise the payment of compensation to the victim. In any case, penalties must be effective, proportionate and dissuasive.

Nine Member States (AT, DE, DK, EE 100 , HR, NL, PL, PT, SE) limit penalties to compensation awarded by courts. In these Member States, the burden of enforcing the principle of equal pay therefore rests on victims who need to take action before the courts. Considering the barriers to access to justice mentioned above, it is doubtful that compensation awards alone qualify as an effective form of penalty, especially since compensation and penalties are regulated by two separate articles of the Directives (Article 18 and 25 respectively). It is equally doubtful that compensation awards qualify as dissuasive penalties, considering the reported generally low level of compensation awarded by courts.

19 other Member States have established specific penalties in addition to compensation. The most common penalties are of an administrative nature (BG, CZ, EL, ES, FI, FR, HU, IT, LT, LU, LV, RO, SI, SK, EE) and are usually ordered by the national labour inspectorate or equality body. Other measures are of a criminal nature (BE, CY, FI, LT, MT): they include fines or even imprisonment 101 . Other penalties include the publication of the decision concerning the pay discrimination (HU, IE, RO), the loss of subsidies by the defendant company/employer (ES, IT) and pay audits (IE, UK 102 ).

Criminal penalties may appear more dissuasive and effective than administrative ones, but they require potentially lengthier criminal proceedings. In Belgium, the equality body, Unia, actually suggested that it would be better to reintroduce administrative fines. This happened, for instance, in Italy, where criminal penalties were replaced by administrative penalties in 2016, with higher efficiency in the sanctioning procedure.

Penalties are generally applied through administrative means (labour inspectorates), quasi-judicial means (equality bodies) or judicial means (courts).

Labour inspectorates and equality bodies

In most Member States (BE, CZ, CY, EE, EL, FR, LU, ES, LT, LV, MT, PL, RO, SI, SK, PT), the labour inspectorates have a formal role in enforcing equal pay provisions. Gender-based pay discrimination is, however, not their core business, and labour inspectors are not necessarily trained to identify such discrimination in the first place 103 . Several experts (BE, CZ, EL, HU) pointed out that labour inspectors only investigate following a complaint from workers or a warning from the national equality body.

Country experts from SI, LU and FR also reported that the principal barrier to efficient enforcement of the equal pay principle by labour inspectors stems from the lack of human resources.

As for equality bodies, all Member States have established at least one. Their core purpose is to implement equal treatment legislation generally. Article 20 of the Directive mandates them to provide assistance to victims of discrimination in pursuing their complaints, to conduct independent surveys and to publish independent reports that make recommendations on issues relating to discrimination. The freedom the Directive gives to Member States to decide on equality bodies’ mandate and level of responsibilities has led to a wide diversity in the bodies’ competences and powers, financial support and structure.

In many Member States, equality bodies assist alleged victims in bringing their claims before the courts (BE, BG, DE, ES, FI, HR, IE, IT, MT, PT, SI, SK, UK) or may bring claims on their behalf (BE, IE, IT, LV, PL, SE, SI, SK). Certain Member States entrust equality bodies with quasi-judicial competences. In these cases, the equality body can grant compensation to victims directly (DK), sanction perpetrators through administrative and/or criminal sanctions (BG, HU, RO), or resolve discrimination disputes by taking binding decisions (CY, EE, FI).

Most national equality bodies have also a non-judicial enforcement role 104 . Non-judicial activities include the competence to engage in alternative dispute resolution 105 (‘out-of-court procedures’ such as mediation) and to provide legal non-binding opinions/recommendations 106 . In addition, equality bodies’ mandates include other types of assistance to victims, as well as the promotion and support of gender equality rights.

The number of complaints or cases on equal pay brought before equality bodies varies across the EU and tends to be low (Austria has the highest number of such cases, ranging from 58 to 110 per year). Reasons mentioned for the low number of cases are the fact that victims lack of awareness of the equality body’s competences or even of its existence (LU, DE, EE) and a more general lack of awareness of the rights relating to equal pay (BE).

To ensure the independence and effectiveness of national equality bodies, the European Commission adopted a recommendation on standards for equality bodies in June 2018 107 . This legal act sets minimum standards concerning the mandate of equality bodies, their independence, their effectiveness (including sufficient resources and appropriate powers) and the national institutional architecture for equality.

5.1.4. Effectiveness of action to tackle occupational segregation

Gender-neutral job evaluation and classification systems help tackle occupational segregation by elucidating the true value of different jobs, skills, experiences and responsibilities. They detect indirect pay discrimination related to the undervaluation of jobs (typically done by women) by measuring and comparing jobs whose content is different but of equal value. In this way, they support a transparent pay system and enhance clarity on what is considered work of equal value, thus contributing to reducing the gender pay gap.

Article 4 of the Directive provides that when job classification systems are used to determine pay, they must be based on the same criteria for men and women and must be drawn up to exclude any discrimination on the grounds of sex. The Pay Transparency Recommendation tries to reinforce and extend this provision by inviting Member States to promote and develop such systems, including in their capacity as employers in the public sector, to prevent or identify and tackle possible pay discrimination based on gender-biased pay scales. Member States are specifically invited to encourage employers and social partners to introduce gender-neutral job evaluation and classification systems. In doing so, Member States could draw inspiration from Annex 1 to the Commission staff working document accompanying the report on the application of Directive 2006/54/EC 108 .

The Recommendation introduced the additional term ‘job evaluation systems’ as a new concept in the EU legal framework, besides job classification systems. The concept of job evaluation systems is a broader term 109 and considered more effective 110 . Both job evaluation and job classification systems must be ‘gender neutral’.

Since 2006, 14 Member States (BE, CY, DE, EL, ES, FR, IT, LT, MT, PT, SE, SI, SK, UK) have introduced the obligation that job classification or evaluation systems must be gender neutral. Overall, the use of job evaluations and classifications as part of wage determination is generally limited in the private sector. The network of legal experts in the EU 111 reports the common use of job evaluation or classification systems in only 8 countries (FI, FR, DE, IT, NL, LU, BE, UK). National experts involved in the evaluation underline that these schemes seem more common in larger companies, or in companies under collective agreements (CZ and LU).

Promotion of job classification by social partners and employers

Some Member States (BE, FR, LU) have included an obligation for social partners to negotiate on equal pay, including the requirement to establish job classification/evaluation systems. However, this obligation is not always applied in practice. In Luxembourg, for instance, collective agreements simply mention that equal pay has been discussed 112 . The main obstacle to effective negotiation of job classification systems by social partners lies in the absence of monitoring by the authorities.

A good example of negotiation by social partners is the adoption in Belgium of Collective Agreement No 25 113 , as amended most recently in 2008, which provides that enterprises and joint sector committees must assess whether their job evaluation systems and pay classification schemes are gender neutral and amend them when necessary. The agreement required all companies and sectors to assess and ensure the gender neutrality of job classification and evaluation systems. The collective agreement defines the notion of pay and establishes a special committee competent to advise courts on equal pay. 

To ensure proper implementation, the legislation established a close monitoring system. The Joint Committee in charge of concluding the agreement must have ensured the gender neutrality of the evaluation and classification system before the mandatory submission of collective agreement to the Ministry of Employment. The ministry rates the monitoring system using a 12-point screening tool and can suggest changes or deny approval if the function title is not neutral e.g. if it is expressed with the feminine or masculine form. This was the case for 45 out of 255 job classifications. After resubmission in the 2 years allowed, 12 were still found to be non-gender neutral and are being renegotiated. The initiative as a whole was very successful in bringing awareness of stereotypes in job classifications, which eventually helped to have more analytical classifications and increased the status and value of sectors in which a high proportion of workers are women. In 2010, the Institute for the Equality of Women and Men (part of the national government) developed and published a checklist to help design gender-neutral job evaluation and classification systems. Despite this comprehensive framework, the study shows that the implementation of the obligation varies substantially from one sector to the other.

Training programmes to assist employers in implementing gender-neutral job evaluation and classification systems are reported to exist in some Member States (BE, CY, EE, LU, SE). Other Member States (AT, BG, BE, EE, FR, NL, PT, SE, LU and UK) have drawn up guides and checklists for job evaluation and classification that make it possible to avoid gender bias. For instance, in France, the equality body (le Défenseur des droits) published a guide on gender-neutral job classification in 2013 to promote effective implementation in collective bargaining procedures at sector and company levels 114 . The Higher Council for Employment Equality and social partners then developed a guide listing the parameters that ensure job classifications are gender neutral 115 . The guide is now the point of reference for collective bargaining procedures at the sectoral and company levels, and trade unions have referenced it in publications about reducing the gender pay gap 116 .

National experts report a lack of promotion and monitoring of gender-neutral job evaluation/classification systems in a number of Member States (AT, BG, CZ, DK, EE, EL, IT, MT, LT, LV). In some Member States (AT, BG, DK, LT and LV), experts have not found any evidence that private sector employers currently use a gender-neutral job evaluation or classification system. Even where systems exist, they are often based on traditional, non-transparent criteria, resulting in the under-classification of predominantly female categories and making indirect discrimination very probable (this is the case in countries such as EL and HR).

The national research identified several reasons underlying the challenges faced in implementing gender-neutral job classification/evaluation systems. For example, trade union delegates do not always have the knowledge and expertise to negotiate on gender-neutral job classification systems (this was identified as an issue in France and Bulgaria in particular). In some countries, there is a lack of awareness of the systems’ purpose and added value 117 . A way to tackle this issue is through training (BE, EE, CY, LU and SE) and guidance. Even in cases where gender-neutral job classification systems must be negotiated by social partners, the absence of monitoring affects the systems’ implementation, as was noted particularly for Luxembourg. In Estonia, the development of gender-neutral job classification/evaluation systems is not on a priority for trade unions. In several Member States, a review of existing systems is needed to remove traditional, non-transparent criteria which are not gender neutral and thus reinforce segregation and inequality in pay.

5.1.5. Effectiveness of the Pay Transparency Recommendation

The lack of pay transparency contributes to scarce awareness of actual and existing pay discrimination and represents an obstacle for victims to claim equal pay. As outlined in Section 2.1, the 2014 Recommendation proposed four main measures for employees, employers and public authorities to uncover unjustified wage inequalities within organisations.

The pay reporting and pay audit measures proposed in the Recommendation depend on the size of the company: over 50 employees for pay reporting and over 250 for pay audits. The threshold is important because a lower employee threshold leads to greater coverage of the workforce 118 and, consequently, more potential to identify opportunities to eliminate gender pay discrimination. For example, the threshold of 250 employees in the UK implies that less than half of UK workers are covered by the legislation 119 . The employee threshold applied varies across the 13 Member States 120 that had pay reporting or pay audit measures in place by December 2018, ranging from less than 50 (DK, FI, LT, PT, SE) to 250 and more (UK, DE). As mentioned, there was limited follow-up by Member States (See Table 2 ) to the four main measures that are the focus of this evaluation in addition to the Directive.

Enforcement mechanisms and compliance of pay transparency measures in Member States

Strong enforcement mechanisms should lead to higher levels of compliance and more widespread application of the legislation. In 11 Member States (AT, BE, DK, DE 121 , ES, FI, FR, IT, PT, SE, UK), pay transparency measures are mandatory. Enforcement mechanism and/or sanctions in place are fines (BE, DK, FI, FR, IT, LT, PT, SE and UK) and a ban on public benefits (PT, IT).

Information about the actual level of enforcement is, however, scarce. In Sweden, three employers were fined for non-compliance with the pay audit requirements, with the penalties ranging from €9,400 to €188,000 122 . In the UK, the Gender Pay Gap Information Regulations mention a fine for companies not complying with the legislation, but do not provide a figure 123 . 

The level of compliance also varies substantially. It appears to be higher in Sweden, the UK and Finland. For the UK, almost all firms to which the regulations apply (i.e. those with more than 250 employees) comply with the legislation 124 . In Sweden, the compliance rate has increased over time and reached high levels in recent years. In 2008, only 48% of employers complied with the legislation 125 . By 2016, an estimated 7 out of 10 companies applied the pay transparency regulation 126 and about a quarter of employers did not comply with the pay audit measure 127 . In Finland, an estimated 60% of companies complied with the measure in 2008-2009, increasing to 50-66% 128 .

In other Member States, compliance with pay transparency measures is quite low. In Germany, low compliance seems to be related to low awareness of the legislation among employers. A survey of 2,300 German companies revealed that 74% of companies with 500 employees were not aware of the existence of the Pay Transparency Act and only a minority of companies (12%) are reviewing their pay structures to make them more transparent. Another survey of companies reports that 16% of employers had employees asking for pay information 129 . In France, figures for 2019 show that only 50% of employers legally bound to pay transparency measures had actually published their gender equality index on time 130 . Belgium does not appear to have a record of the number of companies that actually submit pay reports 131 . In Austria, the level of compliance in preparing income reports was higher among larger companies. It was also found to increase over time 132 .

Disclosure of pay reports and pay audits

The Recommendation suggests that employers should regularly inform employees, workers’ representatives and social partners of pay reports. In addition, the results of pay audits should be made available to workers’ representatives and social partners upon request.

The results of the first empirical study 133  on the impact of mandatory wage transparency suggest that disclosing gender disparities in pay can have some effect in narrowing the gender wage gap at company level. Moreover, disclosure seems to have a positive impact on the number of women hired and their career progression. At the same time, the overall wage cost for companies has not increased as the growth of men’s wages slowed down. The study analysed companies’ wage statistics before and after the introduction of the 2006 Act on gender-specific pay statistics in Denmark. It compared companies with 35-50 employees (who had to report their wage gaps) to data of a group of similar-sized firms with 25-34 employees. It appears that from 2003 to 2008, the gender pay gap in the first group of companies declined by 7% relative to the pre-treatment mean, from 18.9% to 17.5% against a steady 18.9% in the other group. A 7% reduction in the pay-gap should be considered as a significant effect. Only a limited number of firms in Denmark are governed by this legislation and the study also proves the effect amongst firms, that were not required to provide gender segregated pay-statistics.

These findings suggest that the obligation to provide information on the gender pay gap can be an effective step towards addressing gender wage disparities. The salary growth for men in the companies bound by the obligation increased less than in the companies not required to report the data and the overall wage bill at the end of the observation period was therefore 2.8% lower. These lower wage costs for companies might compensate for the potential loss related to an observed decline in productivity of 2.5%, suggesting overall that fears from companies on the costs of these measures are not justified (see also Section 5.3).

It is not possible to generalise these results, which are specific to Danish experience. However, since Denmark already had a strong record of supporting women in the workforce, it is possible that a mandatory wage transparency reform would have a even larger effect on the gender wage gap in less egalitarian countries.

In practice, the disclosure of pay reports and pay audits varies by Member State 134 . In Austria, an estimated one third of works councils were reported to have held company meetings, typically at their own request. However, awareness of pay reports among employees was still low. In Denmark, less than half of employee representatives (42%) reported receiving pay reports from companies. In Finland, an estimated 35-40% of employees reported that the pay audit results were available online 135 . In Belgium, pay reports from employers are only submitted to the works council. No information is available on the extent to which this obligation is respected, as for confidentiality reasons neither the equality body nor any other government body can monitor it. It is unclear if employers discuss the reports in works councils and adopt corrective measures as a result 136 .

By contrast, the UK and France publish the results more widely, making the threat of reputational damage more credible and therefore more effective 137 . In the UK, where the pay report is published on the employer’s and the government’s website, there was 100% compliance in the first year of reporting 138 . Though the level of compliance was lower in France (about 50%), an estimated 94% of employers changed their internal company regulations to align them with the equal pay measures 139 .

Although it is perhaps too soon to assess the measures’ impact at a higher level, the experiences of the UK and France suggest that the threat to businesses’ reputational capital may be more effective at preventing pay discrimination and ensuring compliance and follow-up than ex post fines or sanctions.

Effectiveness in terms of bringing about change in Member States’ legal systems

As mentioned in the 2017 report on the implementation of the 2014 Pay Transparency Recommendation, the Recommendation’s impact has remained limited. In 13 Member States (BG, CY, CZ, EE, EL, HR, HU, LV, MT, PL, RO, SI, SK), pay transparency measures are still entirely absent, while two are considering introducing such legislation (IE and NL).

The report noted how secrecy around pay levels makes it difficult to detect discrimination cases. It also found that the limited follow-up to the Recommendation suggested a possible need for further targeted measures at EU level. Measures implemented in Member States vary substantially in their design and implementation. The specific features of the adopted measures have a significant impact on their overall effectiveness, either promoting or hindering their potential to generate benefits.

5.1.6. Effectiveness in terms of the results and impacts achieved

As mentioned in Section 2.1, the EU action evaluated builds on already existing legislation and related CJEU case-law. This limits the possibility to attribute results and impacts specifically to the Directive. The specific additional measures brought about by the Recommendation have more potential for being identified as having effects, but these measures can only have a long-term effect on the gender pay gap in combination with other measures. As a consequence, outcomes and impacts are more evident in countries where national transparency measures have been in place for several years and in countries where the level of compliance is relatively high.

This section reviews progress made in the EU over the evaluation period, first by drawing on the set of key indicators defined for the baseline (see Section 2.2) and then by investigating the extent to which progress observed can be attributed to EU action. Several studies 140 show evidence that transparency measures are effective. In addition, there is evidence that collective bargaining also has an impact on the gender pay gap, possibly by reducing the risk of discrimination through increased pay transparency. All this shows that EU legislative action in this area has the potential to yield results in the future. Stakeholders also generally agree with the potential benefits of future EU action.

Progress during the evaluation period

Table 7 below presents an overview of progress during the evaluation period on the key indicators selected for the baseline. A traffic light assessment is made for each indicator to reflect the trend over the 2006-2018 period (EU-25) and the 2014-2018 period (EU-28). In most cases, a positive trend was observed. For example, labour force participation and employment rates increased, especially among women, and the gender pay gap decreased. The representation of women on company boards also rose markedly. Together, these trends suggest a more efficient allocation of resources in the economy, increased productivity and greater benefits generated for the EU. Nevertheless, gender gaps persist in labour market participation, in employment rates and in the quality of employment and work-life balance, particularly for women with lower educational attainment and/or with a disability 141 .

A number of factors may have shaped these trends. As highlighted by the intervention logic (see Annex 4), these factors may include demographic and economic trends, other policies introduced in the EU and the Member States, rising social awareness of gender inequalities, as well as the economic crisis. The increased involvement of women in the labour market is part of a broader trend that has been evident since at least the early 1990s. This trend is driven to some extent by demographics (e.g. increasing number of working-age women and lower fertility rates), as well as legislative initiatives in European countries 142 . The sharp increase in the share of women represented on the boards of large companies has been driven largely by EU 143 and national policies.

Table 7. Changes during the evaluation period (baseline to endline)

Notes: Country-level results available in Annex 3. EU-25 and EU-28 samples do not include Ireland, Greece and Croatia. Job satisfaction data are not available post-2015 to assess changes that may be due to the 2014 Recommendation. EU-25 values are weighted for the number of employees in 2014 (2018 data not available). a All endline estimates are for 2018 except for the gender pay gap and elderly poverty, which are from 2017; b Endline estimates are from 2017; c Average values for women (F) and average values for men (M).

It is possible to investigate further the evolution of the gender pay gap by looking separately at ‘unadjusted’ gender pay gap as such, as well as the trend for the ‘explained part’ and the ‘unexplained part’ of the gap. The ‘explained part’ of the gap can be linked to differences in some observable variables such as education, sector, age, etc. The ‘unexplained part 144 ’ is the trend for the residual part, which includes all possible but not observable or not available variables such as negotiating skills, career breaks, personal preferences, etc.

A decomposition analysis to break down the data was carried out for this purpose (See Annex 3, table 3) using data for three time points (2006, 2010 and 2014) available for 16 countries 145 . This analysis did not allow to identify relevant trends, correlations or cluster of countries.

Shifts in women’s labour force participation over the period may explain changes in the unexplained component to some extent. This is because a larger participation of women on the labour market might by itself push towards a shift in the gender pay gap. The unadjusted gender pay gap declined in most countries. Differences in the evolution of the gender pay gap over the 2006 to 2014 period were not evident between countries where women’s labour participation increased substantially and where it did not.

Attribution of progress over time to EU action

Assessing the extent to which EU action contributed to results at a higher level i.e. the outcomes and impacts identified in the intervention logic (Annex 4) is challenging for several reasons. First, progress over time may be driven by external factors in addition to or besides EU action. Secondly, EU action encompasses a wide range of measures, which, together with their degree of implementation, vary across the Member States. Establishing a robust approach to assessing such varying action is difficult, with the desk research identifying very few evaluations of specific measures. The findings from the few robust studies identified were triangulated with evidence gathered from stakeholders to assess the overall attribution of EU action to the higher-level results through a qualitative ‘strength-of-evidence’ assessment.

According to the intervention logic (see Annex 4), the first expected direct impact of EU action is a reduction in the gender pay gap. A study from Denmark 146 found that the pay transparency measures introduced there in 2006 (gender pay gap reporting duties for employers with more than 35 employees) caused a 7% reduction in the gender wage gap in the 3 years after the legislation was introduced. The study found that this effect was driven by men’s wages growing less after the introduction of the legislation, with no simultaneous change in women’s wage growth. Studies investigating the same matter for non-EU countries found similar results and actually also found that higher female salaries contribute to narrowing the gender gap  147 . The key message from these studies is that pay transparency measures help to narrow the gender pay gap, but may also lead to lower wage growth on average.

The potential impacts of EU action at workplace level relate to workforce retention, worker and firm productivity and business profitability. As outlined in the intervention logic, the application of the equal pay principle is expected to contribute to generating fairer and more diverse work environments. This in turn improves the retention of female workers, attracts more talented female employees and improves women’s career advancement. The relevant literature provides robust findings in support of the above hypothesis. For instance, the study on Denmark (mentioned above), though it cannot be generalised to other EU countries given Danish specificities, found that in the 3 years after the introduction of pay transparency legislation, firms subject to the regulation hired around 4.4 pp. more women in intermediate positions than comparable firms not subject to the legislation. In addition, the share of women promoted to more senior positions was 1.2 pp. higher in firms applying pay transparency measures than in other comparable firms. These findings are in accordance with other descriptive evidence collected. For instance, Dromey and Rankin (2018) report the case of Fujitsu UK, which is subject to pay transparency regulation, where several actions were recently taken to facilitate women’s access to senior positions and to recruit talented women, including flexible working time and mentoring programmes for female employees.

Findings from the desk research indicate that EU action can have an impact on worker productivity, job satisfaction and morale. Whether these impacts are positive or negative seems to depend on the specific measure considered and, on firms’ initial compensation practices. Pay transparency is likely to have a positive impact on workers’ motivation if the underlying pay distribution is perceived as fair; but it may have a negative impact if the underlying pay distribution is perceived as unfair 148 (which might also depend on whether employees compare themselves to a group or to specific co-workers). Therefore, obliging employers to be transparent about pay can be a driver for change towards higher fairness and positive impacts are more likely to dominate over the longer term. It may as well have a positive impact on levelling the playing field for companies with fair compensation structures.

As for productivity, some studies find a positive effect 149 . The results of the public consultation,Public consultation as well as anecdotal evidence reported in the popular media 150 , also indicate that initiatives to decrease gender-based pay discrimination can have a positive impact on workforce morale and motivation. On the other hand, research has shown that disclosing pay data may also have a negative impact on workers’ productivity if workers find out that they are paid less than comparable colleagues. In such cases, workers may put less effort into their work as they feel they are being treated unfairly. There is no evidence on whether this ‘disgruntlement effect’ is a short-term or long-term outcome of regulation. However, in countries where employers are obliged to take action once pay discrimination is uncovered, disgruntlement can be expected to have less negative effects or to last for a shorter time. This shows that for pay transparency measures to have positive effects on workers’ productivity, they should aim not only to create transparency but also to ensure appropriate follow-up to the discovery of any pay inequalities. Even more the case, considering the expanding availability of pay information online that reduces the effectiveness of pay secrecy tools.

The potential impact of EU action on companies’ profitability is highly linked to variation in productivity and wages. For instance, the study on Denmark found that profits did not decrease after the legislation was introduced, although productivity decreased. This happened because the negative effect on productivity was offset by the lower wage growth of male employees. Cullen and Pakzad-Hurson (2018) found that firms’ profits increased after the implementation of pay transparency, due mainly to the fall in average wages.

Finally, there is only weak evidence of the impact of EU action on the prevention of gender-based pay discrimination. A country like Sweden is known to have made large wage adjustments to narrow wage differences between female and male employees. However, the extent to which this is attributable to EU action is not known since most of the measures date from before EU action.

Regarding increasing access to justice, Dromey and Rankin (2018) report that equal pay cases made up 11% of all labour court cases in the UK in 2017. Even if equal pay cases seem to make up a significant part of labour litigation, ‘many more cases on unequal pay go unchallenged’. According to most trade union representatives taking part in the European Trade Union Confederation (ETUC) survey in 2019, the Directive brought only minimal changes in terms of access to justice 151 .

There seems to be evidence that stakeholders still value EU action, as it puts pressure on several Member States to act in this area and provides a legal toolbox on which victims can rely in litigation. Trade union confederations tended to hold the view that the Directive did not have any major impact on preventing pay discrimination, although most of them considered that the situation would be worse without the Directive 152 .

In conclusion, though it is not possible to identify the specific contribution of EU action, there was an overall positive effect on workforce retention, worker and firm productivity and business profitability.

The overall assessment of the effectiveness of EU action is cautiously positive. The measures introduced at EU level seem to have had a positive impact on reducing the gender pay gap and therefore possibly reducing pay discrimination. At least to some extent, this impact may be attributed to EU action.

Critical issues remain regarding the uniform application of the concepts of ‘pay’, ‘equal pay for the same work’ and ‘work of equal value’ and regarding access to justice, particularly in relation to the reversed burden of proof.

Following the Recommendation, some evidence was identified that the EU action had a positive impact on workforce participation and retention, had enhanced career progression for women and had resulted in the issue of equal pay achieving greater visibility in the public consciousness and the political agenda.

 Table 8 below summarises the findings of the qualitative review.

Table 8. Overall assessment of the Directive in the Member States — effectiveness

Finally, the persistence of the gender pay gap does not imply that equal pay legislation has been ineffective: rather, it demonstrates that measures to eradicate differences in wages between men and women require significant changes in labour market behaviour. Since the gender pay gap captures multifaceted differences in gendered participation patterns, its elimination requires a multi-pronged approach.

5.2.Efficiency

The analysis of efficiency assesses the relationship between the resources used by an intervention and the changes generated by it.

The widespread variation in the implementation of specific provisions and measures has a significant impact on the efficiency assessment. The efficiency assessment therefore focuses on Member States which saw a relatively high level of action linked to implementing the Directive and the Recommendation and where robust evidence could be gathered. For a selection of countries, this section reviews the financial and human resources that different stakeholders have invested in putting measures in place to comply with new obligations. The efficiency assessment presented in this section remains to a large extent theoretical, even given this evaluation’s ex post character. The reason is that the Recommendation dates from 2014 (the last year for which data at EU level are available) and several Member States are still in the very early stages of implementing follow-up measures.

This efficiency assessment focuses in particular on administrative and compliance costs. This is because the key stakeholders broadly fell into two opposing camps on this issue. A substantial majority of employers (84%) reported in the public consultation that pay transparency measures would increase administrative costs to a rather large or very large extent. In contrast, that view was shared by only 25% of individuals and 12% of trade unions/NGOs. Examining evidence for any administrative cost placed on businesses therefore plays a prominent part in the present evaluation.

Compliance costs and administrative burden

The Directive’s compliance and administrative costs have been assessed as limited, and were primarily felt by the Member States. In contrast, compliance and administrative costs stemming from the Recommendation were experienced more strongly by employers. They are summarised in Table 9 below.

The costs to Member States consisted of the financial resources needed to set up and operate monitoring and enforcement mechanisms. These costs have proved to be low at most, also because resources dedicated by Member States were often none or low, and not necessarily sufficient to ensure the legislation’s effective implementation. Would the effort to monitor and ensure enforcement of the rights be more solemn, some additional resources would likely need to follow. In particular, pay discrimination enforcement has been rarely a focus of labour inspections given the lack of resources allocated to that task. The failure to ensure strong enforcement by labour inspectorates and/or equality bodies has pushed the enforcement costs onto victims, especially costs related to the burden of proof (see section on effectiveness). Several Member States did not devote any resources at all to implementing the Directive. In some Member States, implementation was organised through reallocation of existing financial resources (e.g. ES, SE, SK). Some Member States offered support to companies to implement equal pay measures. The type of support offered involved for instance the development of tools to apply gender-neutral job classification systems and the financing of training to apply those tools (e.g. Luxembourg).

Pay transparency measures introduced following the Recommendation did not impose substantial new costs on public administrations in most Member States. In Germany, the impact assessment of the national pay transparency law estimated that the total additional yearly costs for public administration would be about €27,000 at federal level and around €83,000 at regional (Land) level. Some Member States (e.g. AT, DE, PL, UK) have developed wage calculators to support employers in complying with pay transparency legislation or in developing web calculators. Luxembourg financed the development of software (Lobig-Lux) that supports employers in producing gender pay gap reports. The exact cost of these measures is difficult to estimate, but national experts involved in the evaluation judged them to be low.

Table 9. Overview of actual costs by stakeholder type

Despite the concerns expressed by businesses, the evaluation found low to medium compliance costs for employers, including low administrative burden costs related to the implementation of pay transparency measures. It should be noted though, that choices made by analysed Member States, in particular concerning the company threshold for pay reporting, are sometimes laxer than those suggested by the 2014 Recommendation. Part of the administrative cost for employers is linked to one-off adjustments for companies’ HR and IT departments and/or payroll service providers (understand pay information duties; set up (if not in place yet) or make adjustments to the pay database system and related software; provide staff with the necessary training).

The evaluation demonstrated that costs for business largely depend on the complexity and level of implementation requirements Member States put in place. Such requirements focus on the types of action that employers are required to implement, and the specific reporting requirements imposed by the national legislation. In the case of reporting, employers may need to report annually, every 2 years or every few years, and may need to report mean or median average pay and/or bonus figures. The administrative burden cost was found to be highest for pay audit measures, although this cost was limited too and the measures were also found to have the advantage of detecting pay discrimination and enabling earlier intervention.

Table 10 presents an overview of the quantitative estimates of the compliance costs and administrative burden costs faced by businesses in implementing pay report and audit measures. These estimates are simplified and based on anecdotal information gathered during the study. They are likely to vary between different companies in the same country, but provide useful insights into the extent of costs faced by these stakeholders.

Table 10. Estimated costs of pay reporting and pay audit measures per business in a selection of Member States

Note: One-off costs may include staff training and costs of developing a pay database.

a From information collected, the upper bound was estimated at 80 hours multiplied by 34.6 (i.e. the labour cost of an average employee in Germany).
b We assume three persons from the human resources department were trained over a three-day period and earned the average hourly wage and €10,000 as the average cost of engaging a consultant.
c The total internal cost of pay audits in 2006 (€23,734,100) was divided by the estimated number of employers in Sweden in 2006 (553,918), multiplied by an estimated compliance rate of 48% (Nyberg, 2019 154 ). This gave an internal cost of pay audit of €89 per firm. The average cost of an external auditor was estimated assuming that it takes between 1 and a half and 3 months to conduct a pay audit and applying the average unit labour cost in Sweden in 2006 (€31.6). This leads to a lower bound estimate of €7,584 and an upper bound estimate of €15,168, to which the value for internal costs of pay audits can be added to obtain the figures in the table.
d The estimated €2 million was divided by the number of employers subject to the legislation. Dromey and Rankin (2018) 155 note that 10,528 employers in the UK complied with the legislation.

Based on the information available, the rough estimate of the one-off cost for the HR and IT department mentioned above could not exceed 10-30 person-days per company. In the UK, total one-off costs were estimated at €2 million (of which an estimated 75% was staff training costs). As for regular reporting, estimates varied from 10 person-days per report in a large German bank, to estimated annual costs of €287 per employer in the UK, €146-192 156 in Sweden and €1,470 157 in Belgium.

The limited impact of equal pay measures on the costs for companies is confirmed by the outcomes of the mutual learning programme on equal pay 158 organised by the European Commission in May 2019. Experts consider that the implementation of the most progressive measure to promote equal pay, equal pay certification (which does not exist in the EU, but which does exist in Iceland), may reasonably be considered as giving rise to an upper boundary of costs borne by business. Available evidence from the first phase of implementation of equal pay standards in Iceland shows an average total cost of certification estimated at approximately €21,700 over an average of 6 months to 2 years of implementation, accounting for training, consultancy and fees to the accreditation institution. The costs of such certification system does not seem to have been a major barrier to implementation according to Icelandic companies that implemented the standard, though it was judged expensive by smaller companies.

Costs incurred in the short term are likely to smooth out over time and to be absorbed into standard business practices; these should be less of an issue for companies with well-established HR policies. On a recurrent basis, the administrative burden cost for companies is related to regular pay reporting or replying to requests for pay information.

The evaluation found no evidence that compliance and administrative costs would be higher for small and medium-sized enterprises, even though some countries had in place more stringent thresholds for pay transparency measures than those imposed by the Recommendation (e.g. Denmark). One explanation could be that Member States with a low employee threshold for reporting face low costs in general because of well-developed pay and monitoring systems that pre-date the legislation. These countries are typically located in Scandinavia (e.g. Denmark). By contrast, Member States with higher employee thresholds (e.g. Germany) may face higher costs due to the need to update less-developed pay systems in SMEs.

The evaluation found particularly high access to justice costs for citizens. These relate to the direct costs involving in making a request for pay information, filing complaints and bringing equal pay cases to court. The main factor making costs disproportionate for individuals was the poor application in most Member States of the burden of proof principle.

In most Member States, organisations such as trade unions and equality bodies can help potential victims to file pay discrimination complaints. Despite this, the problems with the burden of proof and the complexity of building a legal claim mean that victims would most likely have to seek legal representation. Costs relating to legal representation are the highest: in most Member States, lawyers’ fees are (on average) between €80-100 an hour, and it cannot be assumed that legal aid is likely to be available to a victim, although ‘no-win, no-fee’ arrangements may enable victims to pursue claims (e.g. UK). In some countries, potential pay discrimination victims can obtain legal representation from equality bodies (e.g. DK, SK, SE), trade unions (e.g. BE, BG, EE, DK, DE, FI, FR and SE) and other organisations dealing with equal pay issues (e.g. NGOs). However, even in such cases, availability of resources determines how far such bodies can get involved in bringing claims in practice (see Section 5.1).

Lengthy procedures contribute to disproportionate judicial costs for victims. Limited information was available on the average time it takes for the typical pay discrimination case to be resolved, though the national research found that such cases may be quite long, taking even up to 10 years to be resolved (UK, CZ).

Cost of no action and longer-term effects

While most of the costs of EU equal pay legislation are generally incurred by stakeholders in the short term, most of the benefits are expected in the longer term. In the selection of countries for which cost estimates for companies were available (see Table 10), the efficiency assessment had a closer look at the level of potential benefits that can be generated by those pay transparency measures — specifically, pay reporting and pay auditing — that could potentially, depending on the circumstances (as mentioned above) require more costs than the others.

The key benefit in the long term has been assumed to be the reduction in the gender pay gap (see Table 1 in Annex 3). This follows from the assumption that making pay, and any differences in pay for similar work, more transparent is a step towards equal pay for the same work or work of equal value. This assumption is confirmed by studies showing that when respondents agree that pay is transparent in their organization, the gender wage gap vanishes for women doing similar jobs to men when all compensable factors are adjusted 159 . While it is not possible to calculate precisely how far the gender pay gap would have decreased in the absence of pay reporting measures, it is reasonable to assume that at least a share of the observed decline in the gender pay gap is a result of the legislation 160 .

The extent to which legislation can have an impact on the gender pay gap depends partly on the level of compliance with the legislation, as well as on the extent to which firms take follow-up measures as a result of the information obtained through compliance with the legislation. This differs significantly across Member States. For the UK, recent estimates suggest that compliance is close to 100% 161 , while for Sweden the evidence suggests compliance has increased over time, from 48% in 2008 to 70% in recent years 162 . Factors affecting compliance can include whether fines can be imposed on companies which fail to comply, as well as whether Member States invest in attracting media interest to the topic, pushing for naming and shaming, thus creating an incentive for companies to comply to avoid reputational damage.

Finally, among the long-term benefits, it is worth mentioning the reputational benefits for employers, as adopting equal pay measures signals to customers and workers that they run a fair business. This can have a positive effect on attracting and retaining talent and on the competitiveness of EU companies. Workers are generally more motivated and satisfied if working for a fair employer, which in the long run can improve firm productivity. In addition, another important long-term benefit is the increased awareness of gender equality matters throughout society. This itself can improve the effective implementation of the legislation and generate more widespread benefits for all stakeholders.

Table 11. Overview of benefits by stakeholder type

5.3. Relevance

‘Relevance’ here means how far the provisions of the Directive and the Recommendation relating to implementation of the equal pay principle were and remain relevant to the need to ensure equal pay between men and women in the EU and prevent gender-based pay discrimination. The evaluation assessed the relevance of the EU action in relation to the original and current problems and needs, analysed the key factors supporting the relevance of the EU action, and identified factors limiting the relevance of the EU action.

5.3.1 Relevance to original needs

As presented in the intervention logic (see Annex 4), the Directive aims to ensure that the implementation of the principle of equal treatment, including equal pay, ‘is made more effective by the establishment of appropriate procedures’ (Article 1) and prevents gender pay discrimination in the EU. The Recommendation was introduced in 2014 to tackle one of the issues identified as a barrier to implementing the principle of equal pay, namely the lack of transparency in pay systems.

The principle of equal pay was originally established under the Treaty of Rome in 1957 for economic reasons at the demand of a Member State, over concerns about a competitive disadvantage for those Member States applying the principle of equal pay. The EU was tasked with guaranteeing the principle of equal pay for the same work and work of equal value between women and men in its role of safeguarding a level playing field across the EU. The CJEU later elevated equal pay to a principle of both economic and social aims with the stature of a fundamental right directly applicable across the EU, which individuals can invoke before national jurisdictions. This initial need for a level playing field remains relevant and the role of the EU in fulfilling this need is essential.

As highlighted in the baseline assessment, the gender pay gap in the EU-25 was 18.0% at the time the Directive was introduced. It was particularly high (around or above 20%) in Estonia, Austria, Cyprus, Slovakia and Czechia, but below 10% in Poland, Malta, Romania Slovenia and Italy. As mentioned earlier in this staff working document, insight into the specific national circumstances may be needed in order to interpret these percentages.

A statistical working paper from Eurostat 163 analyses more in detail the gender pay gap data 164 .. It finds that economic activity and working time are the main drivers for the explained share of the gender pay gap. These factors could explain 32% and 13% respectively of the gender pay gap at EU level in 2014. The difference between employers in the public and private sectors was not found to be significant overall, although it was notable in some Member States (BG, DK, HU, NL, SE). In the case of the Netherlands, for instance, this difference plays a role in explaining the gender pay gap, due to the higher likelihood of women working in the public sector, where wages are lower on average.

5.3.2. Relevance to current needs

The #MeToo global movement brought awareness of violence and discrimination against women on a global scale, including in the EU, with different variants of the famous hashtag being used across the Member States. While it is too early to assess the impact of the movement on equal pay claims, some stakeholders interviewed as part of the data collection for this evaluation mentioned the link between the #MeToo and higher level of complaints to equality bodies.

The gender pay gap slowly declined between 2006 and 2017, but it remains substantial: 16.5 % in the EU-25 and 16.1 % in the EU-28 (see Annex 1 for national data). Differences are evident across the Member States. The gender pay gap has decreased to varying degrees in 18 Member States, while 7 others have either shown no progress at all (FR, SI) or even a significant increase in the gender pay gap (BG, MT, PT, LV, IT 165 ).

National studies 166 carried out in Member States having pay reporting and/or auditing in place demonstrated the capacity of those measures to detect a company-level gender pay differences. In Finland, a study showed that the gap could not be explained in 11 % of the cases detecting pay difference. In Sweden, 3 out of 5 companies which conducted pay audits discovered pay differences which needed adjustment or action to be taken. In Austria 35 % of respondents claimed their reports showed that women earned less than men in the same pay category.

Other sources of evidence, including European public opinion polls, the public consultation and stakeholder interviews, highlight the continued prevalence of gender pay discrimination. The relevance of EU action to current needs is also underlined by citizens’ perceptions of the existing gender pay gap and the need to ensure equal pay for work of equal value. Eurobarometer surveys 167 show that almost 7 in 10 (69%) European citizens believe that women are paid less than men per hour of work, with less than one quarter (23%) believing that women are paid the same (see Figure 2). Women themselves are more likely than men to think that women are paid less (74% vs 65%).

The social perceptions of the gender pay gap differ widely across Member States. In 24 Member States, the majority of respondents believe that women are paid less than men per hour of work, including 94% in Sweden, 88% in the Netherlands and 84% in Germany. At the other end of the scale, only 31% of respondents in Romania, 34% in Bulgaria, 39% in Malta and 43% in Greece believe this to be the case 168 .

Citizens’ perceptions of gender pay discrimination at their own place of work are generally lower. In most (19) countries, the majority of respondents thinks that women and men in equivalent positions are paid the same in their company, with respondents in the Netherlands (68%), Finland, the UK and Romania (66% each) most likely to believe this 169 . At the other end of the scale, in Czechia an absolute majority of respondents believe that women in equivalent positions are paid less than men (56%). This view is held by a relative majority in Slovakia (49%), Germany (47%), Austria (43%) and Hungary (42% 170 ).

 At the same time, there is broad social support across the EU for the principle of equal pay for equal work.

The Eurobarometer survey 171 further shows that 90% of Europeans consider it unacceptable for a woman to be paid less than a man for doing the same job in any circumstances.

The results from the 2017 Eurobarometer are echoed in the results of the public consultation (which is not, however, based on a representative sample). Overall, 44% of respondents felt that men and women are not paid equally for the same work or work of equal value in their countries (responding either ‘yes’ or ‘partially’). By contrast, 23% of respondents believed that men and women are paid equally, while 5% had no opinion or did not answer the question. Citizens and trade unions/NGOs were substantially more likely to believe that women and men are not paid equally for the same work or work of equal value (55% and 57%, respectively) than companies and businesses (3%).

The ongoing need to tackle gender pay discrimination — and the continuing relevance of EU action — was also emphasised in the EU-level interviews carried out for the evaluation. The stakeholders interviewed pointed to the need to have provisions in place to draw attention to the persistent gender pay gap. They highlighted that EU-level legislation remained relevant to addressing the needs of the current situation, but more effort is required to ensure its better implementation on the ground, including more awareness and support for cultural and societal attitude change. From a policy perspective, EU action remains or is even increasingly relevant, as reflected in the Commission priorities relating to the 2014-2019 Gender Equality Strategic Engagement 172 and the European Pillar of Social Rights 173 .

5.3.3. Key factors promoting the relevance of EU action

The evaluation highlighted a range of factors, both internal and external to EU action, which contributed to this action’s increased relevance.

The Recommendation’s introduction in 2014 increased the relevance of EU action to current needs. This is because lack of transparency in pay systems was identified as a key issue in the implementation of the Directive, where measures to increase pay transparency were noted as the key missing factor 174 . The lack of transparency in pay systems is directly linked to the lack of awareness of gender bias and discrimination in pay, and constitutes a significant obstacle for victims to claim for equal pay. The Recommendation aimed to address this need for further transparency in pay structure. Adopting the Recommendation thus contributed to increasing the relevance of measures adopted at EU level and contributed to increasing transparency, in terms both of: (i) knowledge of actual unjustified pay differences between women and men for the same work and work of equal value; and (ii) access to information. The adoption of the EU’s 2017-2019 gender pay gap action plan further strengthened the relevance of the EU action by specifying a set of measures to tackle root causes of the gender pay gap in a systematic and integrated way 175 .

Given the increased policy debate and action on pay transparency in national policy and the mainstreaming of pay transparency in policy-making and pay-setting mechanisms, the relevance of EU action is also reflected in stakeholder opinions expressed through the public consultation. Overall, more than 40% of respondents agreed — or somewhat agreed — with six broader impacts of EU legislation (see Figure 2 ). Opinions on the extent to which public debate had increased varied to some extent by stakeholder group, but not substantially.

Figure 2. Shares of answers to the public consultation concerning the broader impact of the Gender Equality Directive and Pay Transparency Recommendation

Source: ICF (2019), Public consultation report (Question 5)

The greatest level of impact was found in the increased public debate about the need to ensure pay transparency (more than 70% of respondents agreed or somewhat agreed), with little variation by stakeholder group (see Figure 3). Some respondents noted that the debate concerning the gender pay gap has intensified over time, but that this may not be fully attributable to EU action.

Figure 3. Shares of ‘agree’ and ‘somewhat agree’ responses about the need to ensure pay transparency and the need for effective measures by stakeholder group

Source: ICF (2019), Overview of replies on question 5 of the public consultation. A breakdown is provided for three stakeholder groups where the sample size was greater than 20 responses.

In the targeted survey, the majority (67%) of stakeholders affirmed that the introduction of pay transparency measures in their country along the lines proposed in the Recommendation had led to pay transparency issues being considered more widely in policy-making and in company policy.

The evaluation also highlights that, in parallel, a set of external factors in the wider political, economic and labour market context have also contributed to the continuing relevance of EU action.

As shown in the Baseline scenario and in the Section 5.2 on the Effectiveness in this document, the long-term trend across the EU has been towards increased labour force participation of women. More women entering the labour market means more women at risk of pay discrimination, which in turn increases the relevance of EU action.

Across European society, there is greater awareness and more political attention paid to overall gender equality and equal pay challenges in particular. This has been supported more recently by the #MeToo movement that brought awareness on issues of violence and discrimination against women, the broad social support for the gender equality agenda (see Annex 2 for an overview of European public opinion), and by the rapid change in European labour markets, which have steadily made employment opportunities more diverse 176 . Lastly, awareness of pay transparency may have been boosted by increased digitalisation and the availability of wage calculators (e.g. the Pay Transparency Monitor in Germany 177 ) and web platforms (e.g. Glassdoor and Payscale), thus complementing EU action.

5.3.4. Key factors requiring strengthened EU action

Evidence collected and analysis of the existing literature indicates that a number of factors have limited the relevance of EU action. These are set out below.

Lack of transparency in pay systems

The lack of pay transparency contributes to the lack of awareness of gender pay discrimination and represents an obstacle preventing victims from claiming equal pay. It also makes it difficult for victims to access data and evidence to demonstrate direct or indirect difference in pay. That is why the Commission adopted the Recommendation in 2014. As illustrated in Section 4, the evaluation of the implementation of the Recommendation revealed limited follow-up by Member States.

National experts involved in the evaluation cited a number of reasons for the absence of pay transparency measures:

·resistance from employers due to costs (BG, CZ, HU, LV, NL);

·the argument that it is against the culture of confidentiality around remuneration (BG, CZ, MT, PL);

·the belief that gender pay inequality is not (crucial) issue in the Member State (CY, HU, LV, SI);

·the belief that the gender pay gap would be solved by work-life balance measures rather than the transparency measure (CY, NL);

·use of the confidentiality clause (CZ, MT);

·a lack of political will (EE, LV, PL, SI);

·existence of other transparency mechanisms such sex-disaggregated data (LU), access to pay information by trade unions or employee representatives (RO) or an obligation to publish basic pay in job advertisements (SK).

Remaining barriers to access to justice

The Directive requires Member States to ensure access to justice for the victims of sex discrimination, including in relation to pay. The evaluation found that despite these provisions, awareness of discrimination rights and redress mechanisms remains low among the EU public. Low awareness of the judicial redress mechanisms for victims of gender pay discrimination is confirmed by public opinion polls. Just over one quarter of European citizens believe that equal pay for equal work is guaranteed by law, with the majority actually believing the opposite (26% and 59% respectively 178 ).

Awareness of the legal provisions guaranteeing equal pay ranges from above 40% in Romania, Luxembourg and Latvia, to lows of 19% in Spain and even 15% in Germany and France. The proportion stating that they do not know is also significant in some other countries: Bulgaria (34%), Estonia (32%), the UK and Malta (22%) and Poland and Italy (21% 179 ).

Such low levels of awareness, combined with the range of barriers to access to justice already presented in Section 5.1 on effectiveness, explain why the number of pay discrimination cases referred to national courts remains very low in most Member States. Where pay discrimination cases are brought, they tend to be lengthy 180 . The scarcity of national case-law on equal pay may relate to difficulties in establishing convincing cases for pay discrimination. Furthermore, the issue of access to justice over pay discrimination in conjunction with broader structural problems relating to citizens’ lack of trust in the judicial system 181 , limited availability of legal aid, the cost of legal proceedings, and the high level of uncertainty about the outcome involved.

Legal experts have confirmed the continuing relevance of the provisions on ensuring effective access to justice 182 . Nevertheless, legal experts suggest that a number of legal concepts and their application could be further clarified, particularly those relating to the burden of proof and the concept of equal pay for work of equal value.

Besides the difficulties in applying the legal concepts, the evaluation (see Section 5.1) also highlights the difficulty in defining a meaningful comparator to establish what constitutes ‘work of equal value’, especially in female-dominated sectors and professions. The deep gender segregation in Member State labour markets is a core problem that impedes how far EU action can ensure equal pay. The problem is further compounded where strong sectoral social dialogue is the primary determinant of wage-setting mechanisms in individual sectors, with limited space for cross-sectoral collective agreements. In such situations, the concepts of equal pay for the same work and for work of equal value are only applied within the same type of occupation or sector and not across sectors. This has a negative impact on the gender pay gap. As labour markets are gender-segregated, a key structural problem for the gender pay gap in many Member States is the lack of comparison between the value of work in female-dominated sectors (e.g. care sectors or education) and the value of work in male-dominated sectors (e.g. construction or transport). The evaluation indicates that here the relevance of EU action could be further increased by further improving pay transparency in Member States and by ensuring better legal clarity in application of the key legal concepts.

Limited mandate of national bodies to enforce equal pay measures

Under the Directive, national bodies have a role in monitoring, enforcing and implementing equal pay measures. The evaluation found that in practice, the relevance of this role has been hampered by the different mandates and roles those national bodies across the EU have. This applies both to national equality bodies and labour inspectorates.

The relevance differs depending on the specific mandate and context for the work of the national equality body in the Member State (see Section 5.1.3). The analysis by European legal experts shows that there is considerable variation in equality bodies’ competences, scope of action, resources and degree of specialisation 183 .

A common challenge across the Member States is the need to ensure that equality bodies have a sufficient mandate, have sufficient human and financial resources to support victims of discrimination, and perform their enforcement duties effectively and efficiently. The Commission addressed this issue with a 2018 recommendation on standards for equality bodies 184 , which set out measures that Member States are invited to apply to help improve the equality bodies’ independence and effectiveness. These measures relate in particular to the bodies’ capacity to ensure that individuals and groups that are discriminated against can fully enjoy their rights.

The relevance of EU action to labour inspectorates’ activities is more limited, as in a number of countries 185 the inspectorates do not have a mandate to intervene in matters related to equal pay (see Section 5.1). Where the labour inspectorates do have a formal role, challenges have been identified over adequate resourcing, the inspectorates’ capacity to take a proactive approach, and awareness of hidden discrimination among inspectorate staff i.e. gendered preconceptions and biased expectations regarding the job in question. In this sense, the Directive’s relevance for the work of the labour inspectorates varies, as they do not hold a uniform formal role in enforcing equal pay provisions.

5.4. Coherence

The assessment of coherence considers:

-the extent to which the original objectives of EU action mutually reinforce each other or whether they include contradictory elements (‘internal coherence’); and

-how they work together with other actions at national, EU and international level (‘external coherence’).

The assessment looks at coherence in the wider EU action, including existing legislation in the area of gender equality and beyond (e.g. EU data protection legislation) and proposals for new legislation.

5.4.1 Internal coherence

The evaluation assessed the internal coherence of the EU action by looking at how its various components operate together to ensure effective implementation of the principle of equal pay. It also tried to identify any elements that contradict one another or pose problems in practice. The evaluation considers that lack of clarity of key concepts in the Directive (‘pay’, ‘work of equal value’, etc.) is a source of partial incoherence in fully achieving the practical implementation of this complex principle at national level and ensuring its legal clarity. In addition, the evaluation found that a derogation in the Directive potentially accentuates inequalities in the area of occupational pensions, which generally fall within the scope of Article 157 TFEU. The derogation, laid down in Article 9(1)(h), (j) and (k), relates to the application of equal treatment by laying down the possibility for Member States to opt to allow the use of sex as an actuarial factor in the calculation of defined-contribution schemes.

The provisions, aims and objectives of the Directive itself were also assessed as regards their coherence with those of the Pay Transparency Recommendation. The Recommendation, to reiterate, (i) reaffirms the principle of equal pay for equal work or work of equal value; (ii) addresses the lack of wage transparency, which prevents awareness of pay discrimination and the gathering of sufficient evidence for equal pay claims; and (iii) highlights the need to promote gender neutral job evaluation and classification systems. The Recommendation aims to reinforce the Directive and to provide a toolbox of pay transparency measures to help national authorities step up their efforts to eliminate sources of gender-based pay discrimination; it also invites Member States to ensure consistent monitoring and enforcement of the implementation of the equal pay principle. Pay transparency measures are shown in this context to have the potential to make employers, who may not be aware of bias in the pay structure, and employees more aware of pay discrimination. In addition, pay transparency measures can support the implementation of the shared burden of proof in assisting workers to determine the average pay for women and men. However, as the Recommendation is not binding, the provisions on monitoring and enforcement fall short of ensuring that concrete measures are taken. Nor can it impose appropriate sanctions where the transparency measures are not respected.

The evaluation considers that the non-binding nature of the Pay Transparency Recommendation limits the impact and ability of its measures to reinforce the Directive.

5.4.2. Coherence with other EU action contributing to gender equality and to a reduction in gender discrimination

This section assesses how far the Directive’s goals and provisions, as complemented by the Recommendation, are consistent with other EU legislation, including other directives and several proposals for directives.

Directive 79/7/EEC on the progressive implementation of the principle of equal treatment for men and women in matters of social security 186  

This Directive lays down provisions ensuring implementation of the principle of equal treatment between men and women in the field of statutory social security schemes. Overall, three EU instruments aim to ensure equal treatment in matters of social security in general (known as ‘pillars’): Directive 79/7/EEC covers statutory social security (first pillar), while the Directive (i.e. Directive 2006/54/EC) covers occupational social security (second pillar), and Directive 2004/113/EC covers private insurance (third pillar). Directive 79/7/EEC and the Directive regulate different areas of social security and complement one another in terms of implementing the principle of equal treatment in social security in general.

The evaluation highlights one point of inconsistency, which concerns equal treatment in relation to defining old age in retirement pensions. Directive 79/7/EEC explicitly allows Member States to exclude the determination of pensionable age from its scope, at least temporarily. The Directive does not allow such an exemption.

While Directive 79/7/EEC made the derogation on determination of pensionable age temporary and obliges Member States to periodically examine and justify keeping any such derogations, it does not contain any time frame for ending the derogation and the Directive has so far never been amended 187 .

Moreover, and this is very important, the CJEU held in Case C-318/13 (X.) that the use of life expectancy as an actuarial factor to calculate a statutory social security benefit is in conflict with the principle of equal treatment.

In that case, only statutory social security was at stake. However, the Court had already made a similar decision with regard to private insurance schemes 188 . There is little doubt that the Court would reach a similar conclusion with regard to the remaining pillar of social security, i.e. the occupational social security pillar covered by the Directive. The coherence of the EU legal system would therefore seem to require that the Directive be brought in line on this point with the legal situation under Directive 79/7/EEC. This alignment should be realised as soon as possible given the significant impact of the Court’s decisions on the pension schemes, and thus on the pay, of millions of citizens, not to mention the increasing relevance of this matter in Europe’s rapidly ageing society.

Council Directive 92/85/EEC concerning safety and health at work of pregnant workers and workers who have recently given birth or are breastfeeding 189

Directive 92/85/EEC aims to ensure the protection of pregnant workers and young working mothers following the birth of their child or breastfeeding against exposure to health risks and dangerous working conditions. Directive 92/85/EEC and the Directive mutually reinforce each other in the objective of ensuring equal treatment between women and men: the Directive clearly acknowledges (in recitals 23-25), in line with CJEU case-law, that any unfavourable treatment of women related to pregnancy or maternity constitutes a direct ground for discrimination on the basis of sex. The Directive aims to ensure that maternity leave (Article 15) and parental leave (Article 16) do not have any unfavourable impact on women’s pay or conditions. Both Directive 92/85/EEC and the Directive lay down minimum requirements, leaving Member States room to improve on these standards. Furthermore, Directive 92/85/EEC prohibits dismissal on the grounds of pregnancy or maternity leave (Article 10). Both Directives are thus consistent with each other.

Directive 2004/113/EC on the principle of equal treatment between men and women in access to and supply of goods and service 190

Directive 2004/113/EC prohibits discrimination based on sex in accessing and supplying goods and services. Directive 2004/113/EC has an impact on the economic situation of women, in particular as regards access to and supply of goods and services relating to women’s health, care services, insurance and pensions, as well as other financial services. Directive 2004/113/EC and the Directive are mutually reinforcing in that they aim to protect against discrimination based on sex, with one focused on employment and the other going beyond employment.

However, similar to the inconsistency highlighted in regard to Directive 79/7/EEC, there is a discrepancy between the Directive and Directive 2004/113/EC over the calculation of premiums and benefits for the purpose of insurance and related financial services. Directive 2004/113/EC covers private (third pillar) pension insurance and schemes, thus complementing Directive 79/7/EEC (statutory (first pillar) schemes) and the Directive (occupational (second pillar) schemes). As a high proportion of women work part-time, they may experience problems in accessing private insurance and pension schemes. Therefore, Article 5(1) of Directive 2004/113/EC precludes the use of sex as an actuarial factor in the calculation of premiums and benefits for the purpose of insurance and related financial services. As in its ‘sister’ directives, however, Article 5(2) of Directive 2004/113 allows Member States to permit exceptions to the rule if based on relevant and accurate actuarial and statistical data showing that sex is a determining factor in the assessment of risk. In the Test-Achats case 191 , the CJEU eliminated the use of such exceptions by Member States in third pillar schemes, ruling that such exceptions are contrary to the principle of equal treatment. As a result, while the illegality of using different actuarial factors based on sex has been established for first and third pillar pension schemes, no such finding has been made regarding second pillar schemes, which are covered by the Directive. Therefore, ensuring consistency between the three EU instruments on equal treatment in relation to pension insurance and schemes requires the alignment of the relevant provisions of the Directive (see also below).

Directive 2010/18/EU on the Framework Agreement on parental leave 192

Directive 2010/18 aims to ensure that workers are entitled to parental leave on the birth or adoption of a child. It applies equally to all workers, men and women, irrespective of their type of employment contract (open-ended, fixed-term, part-time or temporary) and requires that 4 months from parental leave should not be transferable from one parent to the other. In this way, it aims to achieve more equal uptake of parental leave by both parents.

Generally, both directives mutually reinforce each other. Directive 2010/18/EU addresses the (structural) need to better reconcile work and family responsibilities to improve equality between women and men, as well as women’s access to the labour market. It thus seeks to combat gender stereotypes by setting out parental leave rights that encourage fathers to take up their leave entitlement. The results of the 2018 Eurobarometer survey on work-life balance 193 show that only 4 in 10 European men take up paternity leave, while 32% take up parental leave. Ensuring take-up among fathers is key to relieving women of the burden of care. The main reasons for men not availing themselves of their parental leave entitlements are financial reasons and the use of the entitlement by the spouse. The Directive addresses the latter but not the former. Directive 2010/18 remains in force until 30 July 2022, when Directive 2019/1158 will replace it.

Directive 2019/1158/EU on work-life balance for parents and carers and repealing Directive 2010/18/EU on parental leave 194

This Directive aims to modernise the legal framework on family leaves and flexible working arrangements to improve gender equality by providing for new measures to encourage men’s uptake of such leave/flexible working arrangements (and thus their engagement in family responsibilities). It tackles the gender pay gap beyond pay discrimination by addressing one of the gap’s key drivers, namely the motherhood penalty and the lack of shared unpaid care work between women and men. It also provides a right to paternity leave of 10 days, compensated at the level of sick pay in a Member State, and guarantees an individual right for each worker to 4 months parental leave, of which 2 months are not transferable and paid at the level determined by a Member State.

The general considerations of Directives 2019/1158/EU and 2006/54/EC are mutually reinforcing. The new Directive refers to the need for work-life balance policies to contribute to achieving gender equality and closing the gender pay gap.

Directive 2010/41/EC on equal treatment between men and women in an activity in a self-employed capacity 195  

Directive 2010/41/EC guarantees the principle of equal treatment between men and women engaged in self-employment for aspects not covered by Directives 2006/54/EC and 79/7/EEC. The two directives are consistent in that both complement each other in guaranteeing the right to equal treatment based on sex in self-employment with respect to: (i) social security schemes generally and (ii) occupational social security schemes.

Considering the scope of the EU action and the entire EU legal framework on gender equality mentioned above, one can see a significant gap in the protection of workers as compared to the self-employed. Several pieces of EU legislation, especially Directive 2010/41/EC and Directive 2019/1158/EU, do not apply to self-employed workers. In light of the importance of work-life balance policies in reducing the gender pay gap, this different treatment may be considered as inconsistency in the overall EU legal framework. Directive 2019/1158/EU recognises this problem and requires the Commission to present a study on the rights to family-related leave that are granted to the self-employed (see Article 18(2)(b)).

Directive 97/81/EC concerning the Framework Agreement on part-time work 196  

Directive 97/81/EC aims to implement the Framework Agreement on part-time work concluded by social partners. The Agreement provides for the removal of discrimination against part-time workers, and aims to improve the quality of part-time work, facilitating the development of part-time work on a voluntary basis, and contributing to the flexible organisation of working time in a manner which takes into account the needs of employers and workers. As such, it focuses on access to employment and employment creation rather than on combating sex discrimination in employment or part-time work, though equal treatment in terms of employment conditions for part-time workers (including for female workers) and non-discrimination (Clause 4) is the main provision of the Directive.. Directive 97/81/EC could potentially play an important role in establishing cases of indirect sex discrimination, as more women generally work part-time than men. Nevertheless, it appears that in practice, cases of discrimination against female part-time workers are handled as cases of sex discrimination rather than cases of discrimination against part-time workers. This shows the importance of EU gender equality regulation in making sure that gender equality is incorporated across the board.

Two main instances of incoherence can be highlighted: one in relation to comparators and a second related to the personal scope. To establish discrimination between part-time and full-time work, Directive 97/81/EC requires a comparator in the same establishment (Clause 4 of the Agreement 197 ), while the Directive on the principle of equal treatment refers to a comparator in a comparable situation (Article 2(1)(a)). The CJEU has interpreted ‘comparable situation’ as referring to ‘the same establishment or same service 198 ’. The Court has held that ‘nothing in the wording of [Article 157 TFEU] suggest that the applicability of that provision is limited to situations in which men and women work for the same employer’. However, the inequality must be attributed to a source which can restore equal treatment and cannot be objectively justified 199 . The Court specified that the comparisons must be made ‘on the basis of concrete appraisals of the work actually performed by employees of different sex’ 200 ’ and must take into account a number of factors: the nature of the work, the training requirements and the working conditions 201 .

The personal and material scope of the Directive is broader than that of Directive 97/81/EC as it applies to all workers, including the self-employed, and does not allow for the exclusion of categories of workers. The Court affirmed in the Jenkins case 202 that a difference in pay between full-time workers and part-time workers is not de facto pay discrimination unless it is an indirect way of reducing the pay of part-time workers on the grounds that those workers are predominantly women. Nevertheless, the CJEU has also held that financial conditions, including remuneration and pensions, fall within the scope of Directive 97/81/EC 203 .

Finally, while Directive 97/81/EC aims to combat structural causes of inequality, it does not take into account the fact that part-time work can itself be a trap and even a source of inequality for women, although the Directive encourages voluntary part-time work and ensures non-discriminatory working conditions in such cases. For example, it does not provide for firm obligations on employers to offer part-time workers equal career opportunities or the option to move to full-time employment. As such, Directive 97/81/EC could have an adverse impact on gender equality if no accompanying measures are taken to prevent gender-based discrimination. This is to some extent remedied by the new Directive 2019/1158/EU on work-life balance for parents and carers, which provides for broader rights to request flexible working arrangements. Moreover, another new Directive on transparent and predictable working conditions 204 introduces a new right for workers to request a more predictable and secure form of employment and to receive a reasoned written reply from the employer.

Proposed directive on improving the gender balance among non-executive directors of companies listed on stock exchanges and related measures 205  

Both the Directive and the proposed directive on women on boards mutually reinforce each other and contribute to achieving equal opportunities for women and men in employment and access to promotion. The proposed directive aims to address a phenomenon known as ‘glass ceiling’, where women are limited in their upwards progression within an organisation, one of the causes of the gender pay gap, as highlighted in the 2017-2019 action plan on the gender pay gap. The proposal thus complements the Directive, aiming for a higher number of women in decision-making positions. While the Directive prohibits discrimination in employment and occupation, and encourages Member States to adopt positive actions, the proposed directive requires Member States to take specific positive action to ensure that selection procedures for corporate boards are transparent and based on clear criteria such as merit and qualification. A higher number of women in corporate management positions would have a significant impact on reducing the gender pay gap.

The reporting obligations set out in Article 5(2) of the proposed Directive require companies to report on gender representation on their boards, distinguishing between executive and non-executive directors; these reports should be published in an accessible manner on their website and include information on measures taken to achieve the set objectives. The proposed directive should have a more general positive impact on women’s promotional prospects and on vertical segregation, and will reduce possibilities for discrimination based on sex, and consequently the gender pay gap itself.

Coherence in relation to the use of sex-based actuarial factors following the CJEU case-law (Test-Achats 206  and C-318/13 (X 207 ) cases)

The evaluation examined the coherence of the Directive with these two rulings of the Court interpreting Directives 2004/113/EC and 79/7/EEC. The relevant provisions of the Directive, Article 9(h) and (j), allow the use in certain circumstances of actuarial calculation factors which differ according to sex and the setting of different levels for employers’ contributions. A consistent approach would require the use of gender-neutral actuarial factors across all gender equality directives covering different types of social security, including in the calculation of contributions to occupational pension schemes that form part of what is considered to be ‘pay’. In the Test-Achats and X cases, the Court held that the use of sex as a determining factor in the assessment of risk for insurance premiums and benefits in other financial services, in regard to private and statutory schemes respectively, is contrary to the principle of equal treatment between women and men. However, matters related to occupational social security schemes, i.e. those contracted by the employer, are not covered by Directives 2004/113/EC and 79/7/EEC and are therefore not directly addressed by these two Court rulings.

The evaluation found that based on these derogations allowed by the Directive, the use of gender-based actuarial factors is still allowed in several Member States’ legislation on occupational social security schemes 208 . Based on the analysis of these two Court rulings, the provisions of the Directive concerned and the impact of the rulings on equal pay, it is clear that the use of gender-based actuarial factors would in some cases lead to higher occupational pension benefits for either men or women (Article 9(1)(h) of the Directive) or to higher contributions by employers for men or women (Article 9(1)(j)). In such cases, the use of such factors can have a negative impact on pay or lead to a disadvantage regarding employment costs for either women or men. On that basis, the Directive’s consistency with the Court case-law is questionable, because the Directive allows for the use of gender-segregated actuarial data in setting different levels of benefits or employer’s contributions. It is still possible, according to the text of the Directive, to apply and use Article 9(1)(h) or (j), allowing the use of sex as grounds for differential treatment. However, if such application or use were to be challenged before the CJEU, these provisions would most likely be considered contrary to the equal treatment principle. This situation creates legal uncertainty, while having a significant impact on the occupational pension schemes, and thus on the pay, of millions of citizens. The evaluation suggests that given this issue’s increasing relevance in Europe’s rapidly ageing society, the provisions of the Directive at stake should be revised.

5.4.3 External coherence with wider EU policies 

GDPR
In terms of wider EU policy in the context of the evaluation, a key area is data protection legislation. Wages and related information are considered to be personal data under EU law and data protection legislation is sometimes flagged as being potentially problematic or even incompatible with pay transparency measures. The main instrument considered is the General Data Protection Regulation (GDPR 209 ).

The GDPR aims to protect fundamental rights, in particular, the right to protect personal data. EU equal pay provisions (particularly in the Recommendation) aim to increase the transparency of pay information to identify potential inequalities and guarantee equal pay for the same work or work of equal value.

The Directive and the Recommendation are assessed as being consistent with the GDPR. The Recommendation explicitly requires Member States to comply with data protection requirements. The measures proposed in the Recommendation are sufficiently broad, allowing Member States to design national pay transparency measures in compliance with data protection safeguards.

When it comes to the practical application of pay transparency measures, it is important to bear in mind the following: information about an individual’s pay is ‘personal data’ as it is information relating to an identified or identifiable person (‘data subject’). A key aspect of personal data is that they can be linked to an individual, either directly or indirectly (e.g. through the use of additional information). Data protection rules in the GDPR only apply to personal data. Anonymised data, namely information which cannot be traced to an identified or identifiable person, are not personal data. Data protection rules do not apply to anonymous information or personal data rendered anonymous. On the other hand, if data cannot be directly attributed to an individual, but can be traced back to an individual by means of using a key or additional information, the data is considered pseudonymised and is personal data. Therefore, pseudonymised data must be processed in accordance with the data protection rules.

Pay transparency measures that require fully anonymised data to be provided in the gender pay gap report or pay audit do not fall within the scope of EU law on protection of personal data. In addition, pay transparency measures should put in place safeguards to ensure the data published do not allow the identification of a person, directly or indirectly (e.g. not publishing pay information where a job category has fewer than three people).

Pay transparency measures that require personal data or data related to an identifiable individual to be provided fall within the scope of the GDPR. In such cases, pay transparency measures must comply with the GDPR, including the principles laid down in Article 5, namely that data is:

·processed lawfully, fairly and in a transparent manner in relation to the data subject;

·collected for specified, explicit and legitimate purposes and not processed in a manner incompatible with these purposes;

·adequate, relevant and limited to what is needed in relation to the purposes for which they are processed;

·accurate and, where necessary, kept up to date;

·kept in a form which permits identification of data subjects for no longer than is necessary;

·processed in a manner that ensures appropriate security of personal data.

The GDPR requires the processing of personal data is based on one of the legal grounds set out in its Article 6(1) including i) where it is necessary for compliance with a legal obligation which employers are subject to or ii) where it is necessary for the performance of a task carried out in the public interest or in the exercise of official authority vested in the controller. The basis for the processing on these legal grounds must be laid down by EU or national law. The CJEU 210  has confirmed that EU data protection rules allow for national legislation that requires an employer to make employees’ personal data available to the national authority responsible for monitoring legislation. This is to allow the authority responsible to be immediately consulted, provided that this is necessary for the authority to monitor how legislation is being applied. In such cases, the processing must be proportionate to the legitimate aim pursued. Equal pay, as a fundamental EU right, is an area of public interest. The evaluation found that in some Member States tackling the gender pay gap has been recognised as a justifiable reason for a proportionate interference in the right to privacy 211 .

Furthermore, it is possible to comply with the Directive and the Recommendation in accordance with data protection legislation. All instruments concerned are therefore coherent.

5.4.4. External coherence with international obligations and standards

The coherence assessment aimed to reveal any major incoherence that could stem from transposing the international obligations into national law in light of Member State obligations resulting from the EU legal framework on equal pay. In this regard, it is important to emphasise that Member States have individually adhered to these international obligations and standards. They regularly submit implementation reports, which are evaluated by various international bodies.

A coherence assessment was carried out for the following international obligations and standards. The result of the analysis of the external coherence with international obligations and standards assessment shows that the EU legal framework could be rendered more aligned with the international legal framework on some points:

International Covenant on Civil and Political Rights (ICCPR) (1966)

The evaluation raises an issue that affects coherence regarding the right to equality before the law, which prohibits any form of discrimination, thus providing for effective protection. Under EU law, a set of directives address the various grounds of discrimination. This may undermine effective protection in cases of discrimination on multiple grounds. Women from ethnic minorities, migrant backgrounds or with disabilities typically face more discrimination, including inequalities with regard to pay. CJEU case-law has restricted the use of positive action where it grants automatic preference to the under-represented gender and where it is not proportionate to the policy aim to be achieved. This may hinder claims on the basis of intersectional discrimination on multiple grounds – in situations when different types of discrimination intersect and interact, e.g. gender and ethnic discrimination’.

Convention on the Elimination of Discrimination against Women (CEDAW) (1979) and CEDAW Committee General Recommendation No 13 on Equal Remuneration for work of equal value

The Convention and the Directive both protect the right to equal remuneration and to equal treatment as regards work of equal value. However, CEDAW sets out rights specifically for women, while the Directive prohibits discrimination on the ground of sex. This is a fundamental difference, as CEDAW recognises women as the disadvantaged sex that needs additional protection. CEDAW aims to achieve specific results by obliging positive action and tackling gender stereotypes in relation to family responsibilities. A key inconsistency identified by the evaluation is that under the Directive, positive action is set out as the exception and restricted by the CJEU case-law 212 , under the lenses of proportionality and objective justification. This may limit Member States in taking positive action.

CEDAW’s Recommendation No 13 on equal remuneration for work of equal value recommends job evaluation systems based on gender-neutral criteria be developed and adopted in order to compare the value of different types of jobs in which women predominate. This recommendation has not been taken up by the Directive, which only specifies that job classification systems should not be discriminatory. Nor does the Recommendation specify or promote the development of job classification systems that can be applied to jobs of a different nature.

ILO Convention No 100 concerning Equal Remuneration (1951, ILO) and ILO Recommendation, 1951 (No 90) Report III (Part 6B) International Labour Conference (72nd Session, Geneva, 1986)

The Convention’s intention via the ‘work of equal value’ principle is to compare occupations in different sectors. However, under EU law, this is does not seem to be strictly the case. There is no clear legal basis for such comparisons. While the Lawrence case (C-320/00) stipulated that the comparator can be found beyond the same establishment, without a job classification system that provides a legal basis for such types of comparison, beyond the sector it may not be possible in practice. The Directive does not further define the concept of ‘work of equal value’ or codify earlier CJEU case-law interpreting this concept. It merely reflects in its recital 9 the factors (nature of the work, training and working conditions) set out by the CJEU case-law in order to determine what constitutes work of equal value. While the CJEU case-law is binding on Member States, the Directive would be clearer if these factors were specifically included in the defined concepts of the instrument.

The aim of ILO Recommendation No 90 is to ensure that the principle of equal pay for work of equal value is applied. It refers to the public sector’s role in setting an example for the private sector. Furthermore, the ILO Recommendation highlights the use of job classification and evaluation systems and the role of social partners in applying the principle of equal pay for work of equal value. This ILO Recommendation has not been taken up by either the Directive or the EU Recommendation. The EU Recommendation aims to close this gap by making a similar recommendation to Member States, along the lines of that of the ILO.

Beijing Platform for Action (1995, UN). F. Women and the Economy, Strategic objective F.5 ‘Eliminate occupational segregation and all forms of employment discrimination’

The Beijing Platform for Action and the EU action complement each other in their mutually reinforcing objectives of eliminating discrimination in employment based on sex.

Measures to tackle segregation across different occupations or sectors (horizontal segregation) were not directly included in the Directive or addressed by the Recommendation due to the restricted scope of the provision dealing with work of equal value, which does not allow for cross-sector comparison under the current EU legislation on equal pay. Job classification and evaluation systems are not perceived as a tool to tackle gender occupational segregation. This should be emphasised if Member States are to consider effective use of these tools. In terms of vertical segregation, pay discrimination provisions under the Recast Directive do not directly address or aim to address this issue.

Sustainable Development Goals (2015, UNDP), Goal 5: Gender Equality (SDG 5)

Among others, SDG 5 sets a goal to end all forms of discrimination against women and girls everywhere. The EU action is considered to be aligned with this objective providing a different range of legal provisions aiming to ensure equal treatment between women and men, including the prohibition of pay discrimination. To make the existing provisions of the EU action more effective in order to fulfil the objective set by SDG 5, they need to be complemented by other measures already discussed in this section — such as a work-life balance directive or the proposed directive on gender balance in boards of largest listed companies.

5.5. EU added value

Equal pay between men and women is one of the EU’s founding values. The principle that men and women should receive equal pay for equal work or work of equal value has been enshrined in the European Treaties since 1957 213 . Although the principle of direct effect enabled women to challenge discrimination before national courts, the Treaty itself gave no specific legal means at national level to fight wage discrimination or to tackle the gender pay gap 214 . The analysis of EU added value looks at changes that have been triggered by EU secondary legislation over and above what could have reasonably been expected from the Member States and stakeholders alone in the implementation or the lack of implementation of the Treaty principle itself.

The evaluation points to the EU action’s positive added value, especially in light of the overall progress, even if slow, in reducing the gender pay gap. The effectiveness assessment found that the unexplained gender pay gap tended to decline in Member States that introduced legislation over 2006-2014 more than in those that did not introduce legislation. Member States acting alone cannot more effectively improve overall EU legislation in this area.

In 1957, equal pay was already considered as a key element required to harmonise social systems and ensure that the internal market functioned properly, where companies could compete on a fair and equal basis. On their own, Member States cannot achieve this level of harmonisation nor can they create a level playing field for companies in the internal market. Member States alone cannot ensure that the Treaty concepts such as ‘pay’ and ‘work of equal value’ are uniformly applied. In addition, a well-functioning labour market free of discrimination, including on grounds of sex, is important for promoting labour mobility within the internal market. A coherent legal framework of equal treatment legislation, ensuring uniform rights to equal pay for the same work or work of equal value and with comparable enforcement mechanisms, avoids disparities between Member States which would have an impact on the free movement of workers. This evaluation finds that the EU action has triggered action in Member States, which otherwise may not have happened (see Section 5.1 on effectiveness). However, the evaluation also suggests that the EU action was not enough in many respects: even with the Directive and the Recommendation in place, a wide variation in implementation and compliance persists across the Member States.

The limited resources dedicated by the Member States to implementing the Directive, and to further following up on the Recommendation, contribute to persistent problems with enforcement. In particular, the Recommendation’s value added has been limited, because of its non-binding nature, despite triggering change in some countries and bringing the matter of equal pay and pay transparency more specifically to the political scene, strengthening policy guidance on the implementation of equal pay measures. All this does not put in question the EU added value, but poor and varied implementation in Member States has failed to give real ‘teeth’ to equal pay as a legal provision. Not surprisingly only just over a quarter of European citizens (26%) believe that equal pay for equal work is guaranteed by law (see Eurobarometer’s Figure 8) 215 .

The value added of the EU action also lies in providing a clearer, more readable and more accessible legal text, providing for legal certainty and coherence. This happened to a lesser extent than expected. The definitions of ‘pay’, ‘the same work’ and ‘work of equal value’ — critical to implementing equal pay legislation and measures effectively — have not been sufficiently clarified in the legislation, even if guidance from the CJEU exists. However, the EU action did clarify roles and responsibilities for key players, particularly with respect to promoting access to justice. For example, the Directive extended the rules on the ‘burden of proof’ to the area of occupational social security schemes and expanded the mandate for equality bodies. Yet challenges remain as regards implementing legal provisions, in particular when it concerns the burden of proof, calling for further EU action in that field.

With 15 Member States not yet adopting pay transparency measures and 13 Member States adopting a range of measures targeting different companies and establishing different legal requirements, there are clear discrepancies in regulatory standards across the EU which can negatively affect the cross-border mobility of companies and prevent a level playing field. Divergences in legal requirements in terms of pay transparency measures across the EU can reduce the attractiveness for companies to move within the EU and set up branches in other EU countries as they would need to get acquainted with different regulatory standards and reporting mechanisms. It also means that respect for one of the EU’s fundamental rights is not uniformly ensured across the EU and EU citizens do not have equal access to justice.

Table 12. Summary of the EU added value assessment

6.CONCLUSIONS

This evaluation assesses the EU’s legal framework on equal pay, in particular the progress made in improving the implementation of the ‘equal pay principle’ enshrined in the Treaties since 1957, including:

·its enforcement;

·the protection of victims of gender-based pay discrimination; and

·its contribution to ensuring men and women are treated equally.

It has been carried out in accordance with the Better Regulation Guidelines and its five evaluation criteria. A wide range of findings have been generated from primary and secondary data collection and analysis. The key sources of data and information include an extensive desk review of EU and Member State sources, interviews with relevant stakeholders, a public consultation and an original gender pay gap decomposition analysis of the Structure of Earnings Survey. The evaluation confirms that while the legal framework has triggered significant change at national level, its effectiveness has been hampered by several factors which result in Member States continuing to face challenges in implementing the principle of equal pay effectively.

Regarding effectiveness, the evaluation has found that some progress has been made in achieving the EU action’s objectives to prevent pay discrimination and achieve equality in pay between women and men. Progress was also made in other key labour market and social inclusion indicators, i.e. elderly poverty, female labour market participation and employment rates. However, the extent to which these trends can be attributed to the EU action cannot be assessed with certainty given the confounding impact of other social and economic trends as well as other legislative measures.

The elements which have contributed most to the objectives are job classification and evaluation, as well as sufficiently dissuasive compensation and penalties. On the other hand, the effectiveness of the legal framework is hampered by the lack of clarity and awareness of the concept of equal pay for equal work and work of equal value. Progress in access to justice has been hindered by several challenges, the most common being the application of the burden of proof and the establishment of a prima facie case of pay discrimination.

The issuance of the Pay Transparency Recommendation in 2014 triggered some Member States to adopt or amend legislation on pay transparency. While it is too soon to assess their higher-level impacts, the experience of Member States that adopted measures before 2014 as well as the opinions of stakeholders provide significant evidence for a range of impacts including:

·an improved workforce participation and retention;

·strengthened career progression for women; and

·the issue of equal pay being elevated as regards public awareness and the political agenda.

The experiences of the UK and France with pay transparency measures also suggest that the threat to the reputational capital of businesses can be effective in ensuring compliance and follow-up actions, sometimes even more so than penalties.

The assessment of efficiency was based on the analyse of costs and benefits, quantifiable or intangible, triggered by the EU action for three main categories of players: Member States, business and individuals/trade unions in the short-term and long-term and they can be.

As to costs, the efficiency assessment found no evidence of significant administrative burden linked to the EU provisions. It suggests that the better enforcement of equal pay — one of the EU fundamental rights — can be achieved at a reasonable cost. Even though, the costs for companies would differ between Member States, depending on the scope of the implementing measure, they would in principle be limited, even in the case of the most burdensome measures such as pay audits. For individuals, besides judicial costs in the case of litigation (the costs of filing a pay discrimination complaint), other significant potential costs were identified in the case of pay transparency measures. The latter would potentially lead to lower wage growth due to wage compression (i.e. lower and higher wages tending toward one another) and have in turn have negative effects at the system level (e.g. in terms of potentially lower tax revenue). The evaluation’s findings suggest however that these may be negligible and are likely to be overcome by the benefits of the EU action in the longer term.

As to the benefits, the analysis compares the magnitude of short-term costs for business to the potential long-term benefits from closing the gender pay gap. These simple computations show that the benefits in terms of higher female earnings from closing the gender pay gap would outweigh any possible costs incurred by employers and Member States in a relatively short period of time (2 weeks in our rough calculations).

The evaluation further identifies significant intangible costs and benefits associated with the EU action. For individuals, there is the possible cost of being stigmatised when filing a pay discrimination complaint and the risk of retaliation from the employer’s side. For businesses, the intangible costs can result from possible disgruntlement (and hence lower productivity) from the workers’ side in case large wage differences are revealed. On the other hand, intangible benefits are associated with i) improved reputation from ensuring equal pay, ii) potential improvements in talent attraction and retention, and iii) the potentially higher worker motivation (and productivity) stemming from workers who feel that they are working in a fairer environment and that they are being fairly remunerated. Furthermore, costs and benefits for both individuals and employers are related to reduced negotiating power during wage bargaining, as equal pay might affect the possibility for both employers and employees to bargain wages if considered in comparison with other employees carrying out the same work or work of equal value. For Member States, intangible benefits in the long run are related to i) the increased feeling of social justice and equality among the population, ii) the increased awareness of the importance of and rights/obligations related to equal pay, and iii) a stimulated political and social debate on gender equality more generally.

On relevance, the evaluation finds the EU action to be relevant given the lower hourly wage evident among women as compared with men when the EU action was introduced, which, to a large extent, cannot be explained by observable factors. The evidence and opinion of stakeholders confirm that the Directive strengthened key legal provisions that existed before 2006 and added greater clarity in areas where it was needed such as the burden of proof. The adoption of the Recommendation in 2014 has boosted the relevance of measures at EU level by targeting the lack of transparency as regards pay — a key barrier to implementing rules on equal pay. The relevance has also been boosted thanks to external trends such as i) the increasing labour force participation of women, ii) the greater awareness and political attention given to gender equality, and iii) the development of other legislative and policy measures which tackle root causes of the gender pay gap.

The relevance of the EU action has been limited by its limited scope with respect to the ‘problem drivers’ of the gender pay gap as it does not directly address occupational segregation and the constraints faced mostly by women due to care responsibilities (the ‘care penalty’).

The EU action is found to be only partially coherent in the area of equal treatment between men and women. In particular, CJEU case-law is not reflected in the definition of ‘equal pay for work of equal value’ in the Directive. This has resulted in the concept not being applied effectively in the Member States. Furthermore, other sources of gender-based discrimination are not sufficiently accounted for in the EU action (e.g. gender segregation in the labour market and the part-time work trap). Another key shortcoming is that it is possible to use gender-specific actuarial factors in occupational social security schemes. This is not consistent with the recent CJEU approach taken in relation to statutory social security and financial services, including individual supplementary social security schemes — to prohibit reference to gender in actuarial factors for the calculation of contributions. Separately, the EU action was found to be coherent with EU and national policies on data protection as they do not limit the collection and processing of pay information in order to monitor how the legal obligations of the principle of equal pay for work of equal value are applied. The EU action was also found to be consistent with the international obligations and standards. It is also consistent with a number of important areas not extensively covered by EU legislation, especially in relation to i) positive action, ii) the concept of work of equal value and iii) the use of comparators and job evaluation and classification systems to compare the value of different job areas including those in which women predominate. The latter area contributes to a lack of opportunity to increase the mutually reinforcing effect of these instruments. With EU law having stronger enforcement mechanisms than those in international law, this is a missed opportunity for the EU to reinforce those international standards related to positive action and the concept of work of equal value.

As regards implementing the principle of equal pay, the evaluation confirms the Directive’s clear EU added value. The Directive triggered actions to promote equal pay in Member States that would otherwise likely not have occurred. Similarly, EU added value is attributed to the Recommendation. However, the added value of these actions was limited to a large extent by insufficient and varied implementation (especially since the Recommendation is not legally binding) and because the Directive integrated previous directives with new elements. The added value of the Directive has been hampered by the lack of clarification of the key legal concepts that are central to implementing equal pay measures, particularly for employers, employees and the courts. As one of the EU action’s main objectives is improving legal clarity, the EU added value in clarifying the notion of equal pay for the same work or work of equal value has proved to be particularly limited, as the research reveals that the notion still remains problematic in 22 Member States. There is a similarly problematic situation with regard to applying the burden of proof for which the provision is not sufficiently clear, leaving it to the courts to decide the level of required evidence.

Studies from Sweden and Denmark present evidence which suggests that transparency measures (adopted prior to the Recommendation) are effective, highlighting that EU action in that area has the potential to yield results. These findings are bolstered by stakeholder opinions on the potential benefits of the EU action. There is scope for the EU action to have further added value in the area of pay transparency, given the persistent high level of the gender pay gap. In particular, the EU action can clarify key legal definitions, at a minimum reflecting explicitly CJEU case-law and guiding the application of the concepts, and promoting coordination across the Member States with respect to implementing equal pay measures effectively. A social and economic case can be made for implementing pay transparency measures effectively.

Further EU action on other key drivers of the gender pay gap, particularly those related to the workplace, may offer synergies with the Directive and pay transparency measures. One of these actions where such synergies may be generated is the recent adoption of Directive 2019/1158/EU on work-life balance for working parents and carers 216 .

A number of lessons emerged from the evaluation, reflecting further needs for better implementation of the principle of equal pay.

1.The need to clarify how to apply and use some existing legal concepts:

Existing legal concepts such as ‘pay’, ‘equal/same work’ ‘work of equal value’ are not defined uniformly across national legislation and require facts to be assessed which can be complex. It is not always easy to determine what counts as equal work or work of equal value. This deters victims of pay discrimination from bringing claims and makes running a claim more complex for litigants and courts. Therefore, stronger clarity on how to interpret and apply the existing concepts could be helpful.

2.The need to better define a mandate and increase resources of monitoring bodies (e.g. equality bodies and labour inspectorates) in terms of their formal role in enforcing equal pay provisions — to enable them to issue an opinion or to take a decision on a case or to bring a case to court.

3.The need for a minimum level of fines and compensation in cases of gender-based pay discrimination.

4.The need to promote awareness of rights to redress which would increase the incentive for victims to seek access to justice.

5.The need to improve the practical application of the reversed burden of proof. This can be improved by clarifying the level of evidence required to reach the threshold of presumption of discrimination or by strengthening of the Directive’s provisions on the burden of proof; pay transparency measures can also potentially support the implementation of the reversed burden of proof by enabling workers to provide evidence from which discrimination can be presumed.

6.The need to promote the use of gender-neutral job classifications/evaluations in Member States, in addition to requiring equal treatment of women and men where a job classification system is in place.

7.The need to ensure gender neutrality in terms of the actuarial factors used to calculate contributions and benefits in occupational social security schemes.

8.The need to promote training and the exchange of good practices on building gender-based pay discrimination cases; this could promote access to justice, and could build on existing examples such as the handbook developed by the European Network of Equality Bodies (Equinet) that provides concrete insights into how to build a case on equal pay 217 .

9.The need to ensure pay transparency measures are better implemented across the EU, which would spur more widespread action and higher compliance.

10.The need to promote data collection and research on the impacts of policies that promote the application of the equal pay principle, in particular counterfactual impact evaluation studies that assess the higher-level impacts.

Addressing these needs would potentially strengthen the implementation of the ‘principle of equal pay for equal work’ in the EU. Therefore, it is important to recall the opinion adopted on 19 April 2018 by the European Economic and Social Committee, which supported the idea of a proposal to introduce pay transparency and pay audits in order to facilitate the collection of individualised data and to develop appropriate action plans at sector and business level 218 .

ANNEXES

Annex 1 — Procedural information concerning the process to prepare the evaluation

Organisation and timing

The evaluation took place between October 2018 and July 2019, following the commitment established in the EU Action Plan 2017-2019 — Tackling the Gender Pay Gap 219 . In the REFIT Scoreboard, under ‘Equal Pay for women and men’ 220 the Commission also indicated that it will evaluate the Directive.

The evaluation was carried out by Unit D2 (Gender Equality) of DG Justice and Consumers (DG JUST) of the European Commission. This was done in cooperation with other Commission DGs in the context of the Inter-Service Steering Group (ISSG) convened by DG JUST. The following DGs participated in the steering group: DG CNECT, DG EAC, DG ECFIN, DG EMPL, DG ESTAT, DG GROW, DG HOME, DG JUST, DG MARE, DG MOVE, DG RTD, DG SANTE, together with the Secretariat-General and the Legal Service. During the first meeting on 26 July 2018 the ISSG discussed the following items: the aim of the evaluation; the draft Terms of Reference and the draft consultation strategy.

In the later stages ISSG members were communicated to in writing on the following:

·draft public consultation questionnaire sent for comments on 9 November 2018;

·draft inception report sent for comments on 14 November 2018;

·draft interim report circulated for comments on 8 May 2019;

·draft final report circulated for comments on 24 June 2019;

·meeting of the ISSG to discuss draft final report took place on 28 June 2019;

·final Report approved by the ISSG on 5 September 2019;

·draft SWD discussed by ISSG on 5 September 2019;

·draft SWD submitted to Interservice consultation on 27 January 2020.

Evidence used

Study to support the evaluation of the relevant EU equal pay and related provisions

ICF was commissioned to carry out a study in October 2018 (referred to in this SWD as the ‘Support study’). The study aimed at gathering all the available evidence that would help the Commission DGs assess the Directive on the basis of the 5 REFIT criteria. The support study included extensive data gathering on how Member State applied equal pay provisions. The data were gathered through a range of approaches: desk and national research, data analysis, including the decomposition analysis of the Structure of Earning Survey (SES) microdata, and targeted consultation with various national and EU experts and stakeholders.

Structure of Earnings Survey (SES) analysis

As a part of the evaluation conducted by ICF, the analysis was built on a Eurostat statistical working paper (SWP 221 ) by applying a consistent methodology to analyse the unadjusted gender pay gap across time looking at microdata from SES 2006, SES 2010 and SES 2014. The analysis could not include data for SES 2018 because, despite the fieldwork having been completed, microdata for SES 2018 have not yet been released at the time of the evaluation process.

Stakeholder seminar

On 7 May 2019, a stakeholder seminar was organised by DG JUST (conducted by the contractor), which aimed to stimulate and support a dialogue with and among the social partners about the role, costs and benefits of pay transparency measures in reducing the gender pay gap. It aimed to encourage the European Commission and the social partners to explore together possible further steps to reduce the gender pay gap. The seminar was attended by representatives of the social partners from various levels.

Targeted consultation survey

This stakeholder survey (steered by the ICF) took place up until 21 May 2019 and aimed to support the evaluation and to gather national practices in implementing the equal pay provisions.

Mutual learning seminar

On 27-28 May 2019 a mutual learning seminar took place in Iceland. It was attended by national representatives and experts from 15 EU Member States, and discussed the Icelandic pay certification model. This model requires formal written procedures so that all salary decisions are justified. It involves a job classification and ranking system, with each cluster of jobs of equal or same value given the same monetary value. There was an in-depth discussion about how other countries are addressing the gender pay gap from a structural perspective and the extent to which the Icelandic pay certification model could be transferable.

Online discussion

An online discussion was held on 11 June 2019 to review the draft recommendations stemming from the support study. In total, 35 legal and policy experts in the field of gender equality were invited to join. Of these, 10 accepted and joined the discussion, which lasted 2 hours. The table below lists the participants and their affiliation.

The online discussion was hosted and moderated by ICF via the GetResponse platform 222 . The participants underscored the need to review the structure and logic of recommendations. The participants also raised the issue of what pay transparency actually means in practice and how this can counteract the drivers of the gender pay gap. Specifically, they commented that actual pay transparency is low, even in countries with pay transparency legislation (and measures) in place.

Public consultation

A public consultation on the EU legal provisions implementing the principle of equal pay was undertaken on the EU Survey website from 11 January 2019 until 5 April 2019. All interested stakeholders, experts, academics as well as citizens were invited to contribute. The online questionnaire covered both the evaluation on how the current Directive is being implemented and the problems, objectives and possible options for the future (see Annex_2 for more details). With several targeted actions, the Commission made stakeholders aware of the public online consultation. In accordance with the Better Regulation Guidelines, the questionnaire was available in all EU languages and attracted 386 responses and additional 16 position papers.

An initial summary report of the findings was published on 7 August 2019, and the full report of all stakeholder consultations undertaken for the evaluation can be found in Annex 5.

Literature review

Apart from all the available academic literature on the subject of the evaluation, a series of documents that the Commission previously produced or commissioned were important sources of information and insights of this evaluation.

Annex 2 — Synthesis of stakeholders consultation activities

This annex offers a brief overview of the data-gathering process, elaborating on the different types of measures used to consult stakeholders.

Following the Better Regulation Guidelines, opinions from a wide range of stakeholders were gathered through:

·a public consultation;

·a targeted consultation survey and semi-structured interviews with representatives of national equality bodies, relevant ministries, trade unions, employers’ representatives and labour inspectorates; and

·through a survey targeting social partners, NGOs and experts.

At the end of the evaluation, an online discussion was held which provided for discussion with selected experts from different Member States.

Table 1. Overview of stakeholder consultation activities

Each activity had a specific purpose and objective. For example, the scoping interviews with EU-level stakeholders served to kick-start the evaluation by identifying additional literature for the desk research. The public consultation provided the opportunity for all types of stakeholders to give opinions and views that were triangulated with other sources to support the assessments of all five evaluation criteria (relevance, coherence, effectiveness, efficiency and EU added value). A targeted stakeholder survey and semi-structured interviews were carried out with national stakeholders to complement the desk research in preparing the country fiche and to fill in key gaps. Furthermore, an online discussion was organised with a selection of experts on equal pay in order to review the preliminary conclusions and recommendations

Semi-structured interviews

The ICF team carried out a total of 128 interviews with national stakeholders (for each country the expert contacted the equality body, labour inspectorate, relevant ministry, an employers’ representative and a trade union representative), while an additional 10 interviews were organised with EU-level stakeholders. The people interviewed helped to identify additional literature for the desk research. Information gathered through the national interviews was subsequently integrated into the ‘country fiches’.

Targeted consultation survey

Compared to the public consultation, the targeted consultation survey included more detailed questions and more opportunities for open responses. Both were structured according to the five evaluation criteria. It received 72 replies from national stakeholders in all Member States, with more than half coming from employer representatives and trade unions. Similar to the public consultation, equality bodies and ministries provided comparatively fewer responses to the survey, as they were targeted more through the interviews rather than through this consultation.

Public consultation

The European Commission’s action plan 2017-2019: Tackling the gender pay gap 223  included an assessment the possibility of making targeted amendments to Directive 2006/54/EC (Gender Equality Recast Directive), to ensure the principle of equal pay for work of equal value is better enforced in practice.

In this context, the Commission launched an online public consultation to gather the views of European citizens, stakeholders and Member States on how the relevant provisions of Directive 2006/54/EC are applied and enforced. More specifically, the public consultation gathered stakeholder opinions on the relevance, efficiency, effectiveness, coherence and EU added value of the legislation to feed into the overall evaluation. The public consultation featured mostly closed-response questions, with an option to provide further information. The Commission’s minimum standards for public consultations were all met.

The consultation opened on the Commission’s ‘Have Your Say’ portal 224  on 11 January 2019 and ended on 5 April 2019. In total, 386 replies and 16 position papers 225 were received. This report summarises the results of the public consultation on equal pay and gives a qualitative analytical overview of these results.

Overview of respondents

The 386 respondents indicated that they were replying as: an ‘EU citizen’, a ‘business association’, a ‘company/business organisation’, a ‘trade union’, a ‘non-governmental organisation (NGO)’, ‘other’, a ‘public authority’, an ‘academic/research institution’ or a ‘non-EU citizen’.

Most replies were from individual EU citizens (60% of all responses). The consultation also gathered input from business associations (10%), companies/business organisations (8%), trade unions (6%) and NGOs (6%). Despite the Commission’s efforts to promote the public consultation via permanent representations to the EU, the number of public authorities that responded was relatively low (6%). So was the number of responses from academic/research organisations (2%) and ‘other’ (2%).

Overall, responses were received from across all Member States and from 11 non-EU countries (4%). The largest proportion came from Germany (23%) followed by Spain (12%) and Italy (10%). Other countries for which the number of replies represented more than 5% of the total included Belgium (7%), France (5%) and Hungary (5%). The level of response was low from several Member States most notably Bulgaria, Croatia, Czechia, Estonia, Latvia, Slovakia and the United Kingdom 226 .

Methodology and limitations

The public consultation questionnaire was composed of 16 questions, divided into a general part (considering different sectors, the implementation of the Gender Equality Recast Directive and the EU Pay Transparency Recommendation, national measures, progress and challenges), a part concerning pay transparency (benefits, drawbacks and costs of national pay transparency measures), and a third part about enforcement measures and protection of victims of gender pay discrimination.

Respondents were invited to provide their opinion on:

·the different aspects of the implementation of the EU legal framework on equal pay, including the Gender Equality Recast Directive and the Commission Pay Transparency Recommendation;

·the effectiveness of national measures;

·progress made over the past 10 years (since the Directive was implemented by Member States); and

·the remaining obstacles.

In addition, they could provide open responses to any of the questions.

The evaluation team carried out an overall assessment of each question and a review of differences by stakeholder groups such as EU citizens, organisations representing employers 227 and organisations representing individuals 228 . The sample size for research institutions and public authorities were not sufficiently large to be analysed separately.

Responses were screened for duplicates. Though no multiple replies from the same email address were received, the screening showed that 20 German respondents had identical replies to several, though not all, of the survey questions, suggesting a coordination campaign among business associations in Germany. Given their low number relative to the overall sample of responses (20 out of 386, i.e. 5%) these replies did not skew the overall results, but it is important to highlight that they certainly influenced the results related to business associations (20 out of 38, i.e. 50%).

Outcome of replies

While 44% of all respondents (171 out of 386) feel that men and women are not paid equally for the same work or work of equal value in their countries, significant differences emerge when analysing the results by stakeholder groups, which are presented in the following part. The structure of this section is ordered by stakeholder position and follows roughly the structure of the survey.

Companies and businesses

Companies and businesses were substantially less likely to consider equal pay an issue compared to EU citizens and trade unions/NGOs. Only 3% of companies and businesses feel that men and women are not paid equally for the same work or work of equal value in their countries. This divergence extends to opinions on the benefits of pay transparency, views on the effectiveness of measures that could encourage victims of gender-related pay discrimination to seek redress and the effectiveness of measures in the private sector.

General part

On the effectiveness of the implementation of the equal pay principle, which overall is viewed as especially low in the private sector (30% of respondents responded with ‘very ineffective’), it turns out that a majority of companies instead view the implementation as effective in the private sector.

Accordingly, if one compares the opinion of EU citizens with those of companies and businesses on the obstacles to effective implementation of the equal pay principle, the latter were far more likely to disagree with the obstacles listed in the survey. The most obvious difference between stakeholder positions concerns the lack of dissuasive penalties for employers, a point on which EU citizens were twice as likely to agree as companies and businesses.

Looking more closely at the evaluation of the effectiveness of the individual measures, with respect to access to information on pay levels, companies and businesses were more likely to report that private sector measures were effective. The same result could be observed for the promotion of social partner dialogue and enforceable rules on equal pay, which were also twice as often considered as effective by companies and businesses compared to EU citizens and trade unions and NGOs.

On the effectiveness of measures to ensure protection of a worker claiming enforcement of his/her rights to equal pay, companies and businesses were far more likely to state that certain measures were effective, for example, the reversed burden of proof on employers (see Figure 4). In contrast, no substantial differences across stakeholder groups were observed as regards effective penalties on employers, which were considered less effective. To some extent the responses to this question were influenced by the high number of ‘no such measure in my country’ responses, which primarily stem from Germany.

Figure 4. Proportions of ‘very effective’ or ‘somewhat effective’ to responses concerning measures to protect workers by stakeholder group

In accordance with the foregoing conclusions, companies and businesses were around twice as likely — compared with other stakeholders — to believe that in the past 10 years progress had been made in the areas of enforcement of equal pay rights and the protection of workers claiming these rights.

Furthermore, these findings on EU-level measures are mirrored by the answers given by companies and businesses on national measures facilitating the respect of equal pay rights for women and men and the protection of a worker claiming his/her rights. Companies and businesses are far more likely to have favourable views on the level of the employee awareness of pay gaps and whether victims of gender-based discrimination in relation to pay have increased access to justice (approx. 80%) compared to EU citizens and trade unions and NGOs (approx. 40% each).

The survey also gathered input on the broader impact of the EU legislation, which shows that EU legislation stimulated the debate on equal pay. More than 70% of all respondents agreed that the EU legislation increased public debate about the need to ensure pay transparency and the need to introduce effective measures. Opinions concerning the extent to which public debate was increased varied somewhat by stakeholder group, but not substantially.

Pay transparency – benefits, risks and costs

Respondents had an opportunity to provide opinions on the benefits and risks of pay transparency measures and to share their views on the benefits in relation to the costs of implementation.

Figure5. Proportion of ‘rather important’ and ‘very important’ responses by select stakeholder groups

The divergence between stakeholder groups observed in the results of the general opinion on equal pay between men and women (see above) is mirrored by the level of importance given to each of the seven different benefits of pay transparency the respondents were asked to comment on. While the proportion of trade unions and NGOs who considered the benefits to be important ranged from 53% to 83% depending on the type of benefit, the proportion among companies and businesses oscillates between 13% and 24% (see Figure 5).

The highest level of the response ‘no such measures in my country’ was observed for ‘employees’ rights to request pay information enables their right to equal pay for work of equal value’ and ‘the employers’ duty to provide and publish pay information helps to create an atmosphere of trust’.

On the perceptions concerning the risks of pay transparency measures, it is not surprising that companies and businesses were far more greatly concerned about the risks of pay transparency measures, especially the risk of administrative burdens and costs for employers (at about 85%) than trade unions and NGOs (about 10%).

Respondents struggled to provide a viewpoint on the level of the implementation costs and of the benefits of EU legislation on ‘equal pay’. In full contrast to the other stakeholders, companies and business believe that their costs exceeded the benefits of EU legislation with only 13% agreeing that the benefits outweigh the implementation costs. These findings are consistent with those concerning risks of pay transparency measures where the majority of employers cited administrative burdens and costs as examples of such risks.

Enforcement measures and protection of victims of gender pay discrimination

Finally, the public consultation allowed respondents to provide their opinions on how effectively measures are enforced and on measures that could encourage victims of pay discrimination to come forward. It also solicited input on the need for additional EU action to reduce the occurrence of gender pay discrimination.

At least 20% of respondents, mostly those from Germany, Italy and Austria, believe that such measures are not in place in their countries.

Regarding factors that may influence a victim’s decision whether or not to seek redress in the case of alleged pay discrimination, there are substantial differences between the different stakeholder groups on several specific factors (see Table 2 ): The most obvious gap between the views of companies and businesses and those of EU citizens and trade unions/NGOs respectively concerns the fear of professional consequences and that no action will be taken to remedy a wrongdoing. This substantially diverging subjective perception suggests that employers have little awareness of the factors that keep people from seeking redress in cases of alleged pay discrimination and points to need for more transparent communication on these points.

Table 2. Proportion of respondents from select stakeholder groups who choose ‘rather important’ or ‘very important’ in response to factors that may influence a victim’s decision whether or not to seek redress in case of alleged discrimination.

This conclusion is supported by the insights that emerge from the opinion of companies and businesses on measures that could encourage victims of gender-related pay discrimination to seek redress. They are overall less likely to hold a favourable view on any measure, especially those concerning: i) employers, the State and equality bodies having to provide clear information and raise awareness; and ii) the burden of proof being put on employers.

Finally, when it comes to the need for further EU-level action to address gender-related pay discrimination, the vast majority of companies and businesses believe that existing measures at EU or national level in place are sufficient (85%). Only about 9% believe that new legislative measures are needed.

EU citizens/trade unions and NGOs

When analysing stakeholders’ positions the overall impression is that EU citizens and trade unions/NGOs often do not differ a lot from each other. This can be explained with the fact that both sub-groups generally represent the employees’ perspective rather than that of employers. Thus these two sub-groups are evaluated and analysed together in the following part.

General part

EU citizens and trade unions/NGOs are substantially more likely to believe that men and women are not paid equally for the same work or work of equal value (55% and 57% respectively) than companies and businesses. This finding is mirrored when it comes to the evaluation of the effectiveness of the implementation of the equal pay principle, where the majority of the EU citizens and trade unions/NGOs reported that it is effectively implemented in the public sector but less so in the private sector.

This distinction between private and public sector continues within the responses of EU citizens and trade unions/NGOs on the effectiveness of the individual measures. Regarding the access to information on pay levels, private sector measures were considered less effective than public sector measures. Enforceable rules on equal pay measures were perceived similarly.

The responses on effective measures to ensure protection of a worker claiming enforcement of his/her rights to equal pay confirm the foregoing observations. In the opinion of EU citizens and trade unions/NGOs none of the measures in their respective countries are ’very effective’ or even ‘somewhat effective’. It is noteworthy that respondents did not consider either the reversal of the burden of proof or penalties on employers as effective.

As has been analysed for companies and businesses, the responses on the impact of national measures to facilitate the respect of equal pay rights mirror the above described findings on EU-level measures. Only around 40% of the stakeholders in this section agree that national measures increase i) employee awareness of gender pay gaps and ii) the access to information on pay levels for victims of pay discrimination. Furthermore, fewer EU citizens and trade unions/NGOs than companies and businesses were of the opinion that progress had been made in past 10 years in facilitating the enforcement of equal pay rights and especially the protection of workers claiming these rights.

Pay transparency — benefits, risks and costs

Trade unions and NGOs considered each of the seven different benefits of pay transparency as significant, while their level of significance ranged from 53% to 83% depending on the type of benefit. As regards EU citizens, at least 60% considered the benefits in question to be rather or very significant.

In contrast to companies and businesses, trade unions/NGOs considered risks of pay transparency measures such as the risk of administrative burdens and costs for employers as not a big concern (about 10%). Correspondingly, 82% of the EU citizens and 92% of trade unions/NGOs surveyed in this consultation felt that the benefits stemming from the implementation of pay transparency measures exceed the implementation costs.

Enforcement measures and protection of victims of gender pay discrimination

As pointed out in the above analysis, for EU citizens and trade unions/NGOs the most significant factors that may influence a victim’s decision whether or not to seek redress in the case of alleged pay discrimination, are the fear of professional consequences, the perception that no action will be taken to remedy a wrongdoing and the lack of support from professional bodies. They attributed a far greater significance to these factors than companies/businesses. Only on the answer option ‘lack of knowledge on how/where to complain’ did all stakeholders somewhat agree on the level of significance.

Consequently, EU citizens consider clear information from the State and equality bodies as well as the burden of proof on employers to be the most effective measures to encourage victims of gender-related pay discrimination to seek redress. Trade unions and NGOs were more likely to value clear information and awareness raising from employers.

Finally, only a few EU citizens (15%), trade unions and NGOs (11%) find the current measures sufficient and they were more likely than companies and businesses to believe that new legislative measures at EU or national levels were needed (52% and 57%).

Analysis of stakeholders’ position papers

As part of the consultation, respondents were invited to send in any additional evidence or material illustrating the stakeholder position and recommendations to ensure equal pay for women and men. Out of 16 files received (after removing the duplications), only six presented the positions of different EU-level stakeholder organisations 229 (two employer/business organisations, two NGOs, one trade union and one organisation representing local and regional governments). One document presenting the initiative of the Business and Professional Women Europe, and four documents on national action plans did not provide a position on equal pay. In the same vein, four documents provided national level statistics or a comparative analysis on the pay gap rather than conveying a specific position. One file contained an unidentified document titled ‘Commission Consultation to Review Equal Pay’.

The analysis of the position papers sent by stakeholders draws a similar picture as found in the responses to the public consultation.

Business associations find the current legislation sufficient and appropriate and opt for a stronger focus on collaborating with social partners as well as implementing the Commission’s suggestions on a voluntary basis.

Trade unions and NGOs state that the biggest obstacles to achieving equal pay still are the lack of effective measures on pay transparency and insufficient access to judicial redress. In addition to the position analysed in the answers, NGOs demand that other drivers of the gender pay gap be addressed outside the scope of the legislation, e.g. affordable childcare services, reconciliation of work and care and tackling the motherhood penalty.

Conclusion

The replies to this public consultation fed into the European Commission’s evaluation of the EU legal framework on equal pay.

Summing up the analysed responses from the different stakeholders, the majority of the companies and businesses are already satisfied with the status quo as regards the implementation of both EU-level and national measures and the perceived progress that has been made. They believe the legislative measures are being implemented effectively, especially in the private sector and feel that employees are much more aware of their equal pay rights and access to justice.

Furthermore, on the one hand companies and businesses accord low importance to the benefits of pay transparency and on the other perceive high risks of administrative burden and costs for employers as a result of implementing pay transparency measures. Finally, they feel that the implementation costs mostly exceed the benefits. Overall, they do not seem to consider it necessary to make changes to the current legislation.

In contrast, the majority of respondents among EU citizens and trade unions/NGOs, representing the employees’ perspective, stated that the fear of professional consequences prevents them from seeking redress and that clearer information from both the employer and the public equality bodies could encourage them to take legal measures.

This divergence points to the opportunity to raise employers’ awareness of the factors that may influence a victim’s decision to seek redress in a case of alleged pay discrimination, which could usefully be done by intensifying social dialogue.

Furthermore, when it comes to the question whether new EU legislation is needed, the stakeholders’ opinions also vary widely.

Public consultation on gender equality in the EU

In addition, the open public consultation on gender equality in the EU 230 took place from 8 March until 31 May 2019 to collect views on the current state of gender equality in the EU as well as on future priorities. The consultation received 1,335 replies from a broad range of stakeholders, including academic/research institutions, business associations, company/business organisations, consumer organisations, EU citizens, non-EU citizens, NGOs, public authorities, trade unions and others. Among other things, the survey’s participants were asked which specific goals they would prioritise for EU action to increase women’s participation in the labour market and to tackle the gaps in gender employment, pay and pensions. Introducing measures to further support the principle of ‘equal pay for equal work or work of equal value’, such as gender-neutral job classifications and pay transparency (54% of all respondents) was the option that was most often marked by the respondents for this policy area 231 .

Annex 3 — Methodological issues on the gender pay gap indicator and its decomposition using the Structure of Earnings Survey microdata

The (unadjusted) gender pay gap is computed as the difference between average gross hourly earnings of male paid employees and of female paid employees as a percentage of average gross 232 hourly earnings of male paid employees 233 . This is a simple indicator of wage inequalities and it is a handy tool for communication and policy-making for this very reason. Important specificities of this indicator must however be taken into account in its interpretation.

The notion of pay in Article 157 TFEU and the Directive 234 (i.e. ‘the ordinary basic or minimum wage or salary and any other consideration, whether in cash or in kind, which the worker receives directly or indirectly, in respect of his/her employment from his/her employer’) does not correspond to earnings as expressed in the gender pay gap indicator. The reference data set, the Structure of Earnings Survey (SES) 235 , excludes bonuses, irregular overtime and any 13th month payment, which are relevant indicative tests for gender equality in the labour market as they often favour men. Moreover, SES data exclude companies with fewer than 10 employees 236 and cannot take into account differences between part-time and full-time workers. Finally, the indicator is computed for the whole economy defined as industry, construction and services except public administration, defence and compulsory social security 237 .

It is important to stress that the gender pay gap (GPG) gives an overall picture of the differences between men and women in terms of pay and, therefore, measures a broader concept than the concept of equal pay for equal work. As with any statistical indicator, it only provides factual information and it does not have the capacity to distinguish between differences due to discriminatory treatment or legitimate deviations according to the Directive; it cannot therefore be used as a reference point for the implementation of the anti-discrimination framework.

A more precise picture is obtained by decomposing the indicator 238 into explained and unexplained parts. The explained part is the gap that can be linked to the differences in the average characteristics (sector of activity, age, occupation, etc.) of male and female employees. The unexplained part is the difference between earnings of men and women with the same characteristics that can be linked to unobservable variables (such as negotiating skills) or unavailable variables (such as career breaks as the SES dataset only contains information on employees’ tenure in the current job). However, the model cannot account for possible feedback effects between the two components. For instance, access to some occupations might be gender-biased. Finally, comparisons of the indicator across countries must also take into account differences in the structural conditions of the labour market and the relative social value of occupations.

A statistical working paper (SWP) from Eurostat analysed in more detail the GPG data through this decomposition. 239 The adjusted (explained) GPG is lower in most countries, better reflecting average characteristics of men in terms of sector, occupation and age, though still pointing to differences in earnings. However, in some countries it is the other way around, i.e. women have better average characteristics. In particular, this is the case for countries with a low employment rate of women. In this situation, the adjusted GPG highlights (by being higher that the unadjusted GPG) a selection bias (i.e. only women more likely to earn a relatively higher salary will work, or, in other words, mostly higher-educated women work in higher-paying sectors) and provides a better illustration of the GPG. The adjusted GPG enables segregation effects on the basis of overall differences in average characteristics of men and women to be grasped, e.g. in economic activities or occupations. However, it is not yet a precise indicator because the inclusion in the model of omitted variables could still change the results of the decomposition. These limitations must be specifically taken into account in the interpretation of the unexplained (residual) part of the GPG decomposition which has the same limitations and, therefore, is in itself not an indicator or even an upper bound of discrimination on the grounds of sex in the labour market.

Figure 1. Unadjusted gender pay gap by Member State; 2006, 2014 and 2017

Source: Eurostat, sdg_05_20.

Table 1. Gender pay gap, by Member State and EU overall

Source: Eurostat (sdg_05_20).

Decomposition of the gender pay gap indicator using SES microdata

The approach followed that taken by a methodological statistical working paper 240 carried out by Eurostat in 2018 241 . In that working paper, Eurostat employed a ‘Blinder — Oaxaca decomposition’ to break down the unadjusted gender pay gap into an explained and an unexplained component using microdata from the 2014 SES. The analysis builds on this approach by applying a consistent methodology to break down the unadjusted gender pay gap across time looking at microdata from SES 2006, SES 2010 and SES 2014 (data for 2018 were not available at the time of writing and data for 2002 are not fully comparable). Eurostat results are based on the full dataset only available at Eurostat’s ‘Safe centre’ in Luxembourg. The data released to the research community undergoes an anonymisation procedure which prevents the same granularity. Therefore, the reference samples, and the gender pay gap indicator computations, are slightly different from those computed by Eurostat. The technical details are available in the Annex of the support study 242 .

The table below presents the findings from the decomposition analysis for the baseline (2006).

Table 2. Decomposition of the results for the baseline (2006)

Source: Support study

a The indicator was recomputed for the evaluation and might differ from official Eurostat data.

For countries with more than one wave of data available, the table below present the development of the unadjusted gender pay gap, the explained and the unexplained gender pay gap.

Table 3. Trends in the gender pay gap decomposition (2006, 2010 and 2014)

Source: Support study

a The indicator was recomputed for the evaluation and might differ from official Eurostat data.

Annex 4 — Intervention logic

Annex 5 — Evaluation questions

The evaluation assessed the relevant equal pay provisions of the Directive, strengthened by the Recommendation, against the evaluation dimensions of effectiveness, efficiency, coherence, relevance and EU added value based on these questions:

Effectiveness

1.How has the introduction of the equal pay principle at the EU level contributed to Member States applying the concept in their national legislation and practice? What are the monitoring and enforcement mechanisms in individual Member States and what is their role in making the equal pay provisions effective? In cases where national legislation was adopted, how effective was it in reducing the gender pay gap in its adjusted and unadjusted form 243 ?

2.To what extent have the relevant provisions of the Directive as regards the implementation of the equal pay principle, reinforced by the Recommendation, triggered actions at the national level to increase wage transparency? In cases where national legislation was adopted, how effective was it in reducing the gender pay gap in its adjusted and unadjusted form?

3.To what extent have the specific provisions of the Directive as regards the implementation of the equal pay principle, reinforced by the Recommendation, been effective in bringing to light pay discrimination?

4.To what extent have the relevant provisions of the Directive as regards the implementation of the equal pay principle, reinforced by the Recommendation, contributed to allowing a victim of pay discrimination to effectively bring a case to court? What is the impact on effectiveness of the provisions relating to compensation to victims and sanctions? Do the latter have a dissuasive character?

5.To what extent have the relevant provisions of the Directive as regards the implementation of the equal pay principle, reinforced by the Recommendation, been effective in inciting companies to take the initiative to analyse their pay structures in order to determine pay discrimination?

6.To what extent have the relevant provisions of the Directive as regards the implementation of the equal pay principle, reinforced by the Recommendation, been effective in inciting the use and diffusion of gender-neutral job evaluation and classification systems resulting in the implementation of the equal pay principle?

7.To what extent have the relevant provisions of the Directive as regards the implementation of the equal pay principle, reinforced by the Recommendation, been effective in ensuring enforcement at national level by competent bodies such as national equality bodies, labour inspectorates, and other enforcement bodies?

Efficiency

1.What are the economic costs and benefits incurred from applying the present EU legislation regarding the equal pay principle?

2.To what extent have the specific objectives linked to equal pay of the Directive, reinforced by the Recommendation, been achieved at a reasonable cost in terms of financial and human resources deployed at administrative and regulatory level by Member States’ administrations and by employers?

3.If there are significant differences in costs or benefits between Member States, what is causing them?

4.How are these costs and benefits distributed among the different stakeholders affected?

5.Did countries which implemented measures proposed by the Recommendation experience some significant upwards wage pressure linked to this?

6.What is the interplay between the short-term and longer-term impacts e.g. are the potential short-term costs incurred justified and proportionate to the potential longer-term benefits from gender equality and from enlarged access and participation to the labour market?

Relevance

1.To what extent do the relevant provisions of the Directive, reinforced by the Recommendation, correspond to the need to prevent pay discrimination and ensure ‘equal pay’ between women and men in the EU? Have they been relevant for the different stakeholders affected?

2.To what extent do the relevant provisions of the Directive, reinforced by the Recommendation, contribute to mainstreaming pay transparency in national policy?

3.To what extent do the relevant provisions of the Directive, reinforced by the Recommendation, provide for relevant actions for improving the prevention of pay discrimination and realisation of pay transparency?

4.To what extent do the relevant provisions of the Directive, reinforced by the Recommendation, contribute to enforcing antidiscrimination measures relating to the pay of women and men at national level through national equality bodies and other enforcement bodies such as labour inspectorates?

5.To what extent do the relevant provisions of the Directive, reinforced by the Recommendation, contribute to ensuring access to justice for victims of sex discrimination in relation to pay?

6.To what extent have the provisions of the Directive on the equal pay principle, as reinforced by the Recommendation, been applied by national courts and non-judicial authorities? Have the provisions affected the amount of cases prosecuted or court judgments issued on pay discrimination? Have they affected court practice on granting compensation to victims of pay discrimination? Is information available on the length of proceedings or the availability and use of legal aid in pay discrimination cases? Is information available on the enforcement of pay discrimination judgments?

Coherence

1.To what extent is the Directive, reinforced by the Recommendation, coherent or not with other interventions which have similar objectives of reducing gender discrimination and contributing to gender equality at EU and national level?

2.To what extent is the Directive, reinforced by the Recommendation, coherent or not with wider EU or national policies? (In particular, please consider the relationship between the data protection and privacy legislation and provisions of the Recommendation referring to the right to request pay information).

3.To what extent is the Directive, reinforced by the Recommendation, coherent with international obligations and standards (such as those defined in the International Covenant on Civil and Political Rights, the International Covenant on Economic, Social and Cultural Rights and the Convention on the Elimination of Discrimination Against Women (CEDAW))?

EU added value

1.What is the additional value resulting from the application of the Directive, reinforced by the Recommendation, compared to what could have been achieved in terms of equal pay by Member States solely on the basis of Article 157 TFEU?

2.To what extent do the issues of equal pay addressed by the Directive, reinforced by the Recommendation, continue to require action at EU level?

Annex 6 — Provisions within the scope of the evaluation

The scope of the evaluation covers Article 157 of the TFEU, selected provisions of the Directive and the Recommendation. Article 157 of the TFEU requires that each Member State ensure that the principle of equal pay for female and male workers for equal work or work of equal value is applied. The selected provisions of the Directive that fall within the scope of the evaluation are highlighted in the table below.

Table 1 — Provisions of the Directive

Annex 7 — Relevant CJEU-case law

Case C-262/88, Douglas Harvey Barber v Guardian Royal Exchange Assurance Group, [1990] ECLI:EU:C:1990:209.

Case C-109/88, Handels- og Kontorfunktionærernes Forbund i Danmark v Dansk Arbejdsgiverforening, agissant pour Danfoss, ECLI:EU:C:1989:383

Case C-171/88, Rinner-Kulhn v FWW Spezial-Gebaudereinigung GmbH [1989] ECLI:EU:C:1989:328.

Case C-147/02, Alabaster v Woolwich plc and Secretary of State for Social Security [2004] ECLI:EU:C:2004:192.

Case C-278/93, Freers and Speckmann v Deutsche Bundepost [1996] ECLI:EU:C:1996:83.

Case C-12/81, Eileen Garland v British Rail Engineering Limited, [1982] ECLI:EU:C:1982:44.

Case C-360/90, Arbeiterwohlfahrt der Stadt Berlin e.V. v Monika Bötel, [1992] ECLI:EU:C:1992:246.

Case C-33/89, Maria Kowalska v Freie und Hansestadt Hamburg, ECLI:EU:C:1990:265.

Case C-400/93, Specialarbejderforbundet i Danmark v Royal Copenhagen, ECLI:EU:C:1995:155.

Case C-309/97 Angestelltenbetriebsrat der Wiener Gebietskrankenkasse v Wiener Gebietskrankenkasse, ECLI:EU:C:1999:241.

Case C-381/99, Brunnhofer v Bank der österreichischen Postsparkasse AG, ECLI:EU:C:2001:358.

Case C‑427/11 Margaret Kenny and Others v Minister for Justice, Equality and Law Reform and Others, ECLI:EU:C:2013:122, para. 28.

Case C-61/81 Commission of the European Communities v United Kingdom of Great Britain and Northern Ireland, ECLI:EU:C:1982:258.

Case C-43/75, Gabrielle Defrenne v Sabena. C-320/00 Lawrence, ECLI:EU:C:2002:498.

Case C-96/80, Jenkins v Kingsgate (Clothing Productions) Ltd, ECLI:EU:C:1981:80.

Case C-333/97, Susanne Lewen v Lothar Denda, ECLI:EU:C:1999:512.

Case C- 236/09, Association Belge des Consommateurs Test-Achats ASBL and Others v Conseil des ministres, ECLI:EU:C:2011:100.

Case C-342/12, Worten — Equipamentos para o Lar SA v Autoridade para as Condições de Trabalho (ACT), ECLI:EU:C:2013:355.

Case C-407/98, Katarina Abrahamsson and Leif Anderson v Elisabet Fogelqvist, ECLI:EU:C:2000:367.

Case 129/79, Macarthys v Wendy Smith, ECLI:EU:C:1980:103.

Case C-200/91, Coloroll Pension Trustees Ltd. Russell, ECLI:EU:C:1994:348.

Case C-196/02 Nikoloudi v Organismos Tilepikoinonion Ellados AE, ECLI:EU:C:2005:141.

Annex 8 — Concept and measurement of gender pay discrimination

Discrepancies in pay do not necessarily point to pay discrimination. Many factors might influence observed disparities and identifying them all is already a challenging task. For instance, discrimination might result in pay differences only because people are deprived of the opportunity to perform challenging tasks that would facilitate promotions, or to participate in training, etc. Finding a causal link between all these factors and the disparity in pay is very challenging. Moreover, a significant feature of pay discrimination is its cumulative nature. As Justice Ruth Bader Ginsburg, writing for the dissent, argued in a controversial landmark decision of the US Supreme Court 244 , in contrast to specific discriminatory acts such as termination, ‘[p]ay disparities often occur in small increments’ and ‘cause to suspect that discrimination is at work develops only over time. Moreover, comparative pay information is often hidden from the employee’s view.’

It is difficult to frame the concept of pay discrimination in this context in order to measure it, even leaving out the broader implications of gender bias before actually entering the labour market that might shape educational and career choices.

Since disparities do not necessarily indicate discrimination, the debate on the gender pay gap is largely influenced by opinions on the extent to which it is evidence of discrimination or simply a result of individual attributes and choices that might on average differ for women and men 245 . 

Definition

Broadly speaking discrimination 246  refers to the unequal treatment of people or groups because of status characteristics that are unrelated or irrelevant to the outcome. This can happen either through a direct differential treatment or an equal treatment based on rules and procedures that result in indirectly favouring one group over another. This definition differs from (negative) prejudices, attitudes, stereotypes and sexism that might or might not cause actual discriminatory behaviour. The selection of a specific definition for discrimination has consequences on the methodology applied to identify it and measure it.

Methods for measuring pay discrimination

Since discriminatory behaviour is rarely observed directly and identifying association between gender and pay disparities can be relatively straightforward, it is difficult to infer a causal relationship between the two in order to argue that a discriminatory treatment caused the difference in pay.

A proper measurement of pay discrimination should answer the counterfactual question: how much would a woman have been paid had she been a man? If the answer shows that she would have been paid more it would mean that there is a causal relationship between gender and the amount of pay. Unfortunately, answering this question is physically impossible because it is not possible to recreate a situation where the only difference between two people is gender.

There are four main methods to measure discrimination. They are used by a large number of disciplines and applied to various types of discrimination, though none is without flaws. These are: 1. laboratory experiments, 2. field experiments, 3. analysis of observational data and natural experiments, and 4. analysis of survey and administrative record reports 247 .

Laboratory experiments try to recreate, in a controlled environment, a counterfactual situation in which one discrimination variable can be manipulated. They are helpful in identifying the situations and the characteristics of people that trigger discrimination which can then be used as working hypotheses to be tested on observational data. Similarly, larger scale experiments in the field could provide results that can be more easily generalised. These studies are known as audit studies and have been used, for instance, for housing or job-seeking with a comparable pair of people sending applications. They try to approximate the counterfactual situation by assessing differences based on gender if the pair are called back by landlords when they enquire about renting a house or if they are called to an interview by prospective employers when they receive their CVs. The famous analysis carried out by Goldin and Rouse 248 of blind auditions, despite providing mixed results, is another example. These experiments can prove causality in their own setting (internal validity), but it is difficult to generalise the results to a larger population (external validity) in the absence of additional information. Generalisation would in particular require replicability which is difficult to achieve.

The most used method to measure discrimination is the statistical analysis of the difference in a given outcome (pay) between different groups using a regression model. This implies, like for any econometric model, that the mechanism that causes discrimination is widely understood and that there is a large availability of reliable data. A lack of these two elements translates into omitted variable bias and sample selection bias, as in the case of the gender pay gap. Statistical decomposition methods 249 (which can assume that some variables (e.g. education) have the same effect for women and men or allow for interaction between gender and other variables) like the Blinder-Oaxaca used in the support study for this SWD and which is the most common, are most beneficial for descriptive use but they are not capable of analysing causality. Economic modelling focused on the labour market, namely analysing individual actions as result of prejudice 250 or of rational process deprived of discriminatory animus (known as ‘statistical discrimination’) 251 .

Finally, surveys and administrative records also provide useful data. Surveys are normally done through interviews using techniques that allow hidden bias to be discerned. Since attitudes do not necessarily translate into behaviours, prevalence of active discrimination can normally not be measured using data on attitudes, though they can provide useful indications. Perception surveys are very important for analysing patterns of discrimination because of the high correlation with health issues such as anxiety or depression, and with outcomes other than pay; however, since events may be misperceived or overlooked, these surveys might over- or underestimate the actual incidence of discrimination. Moreover, perceptions might change over time and across countries. Nevertheless, these surveys offer useful insight and opportunities for generalisation or to investigate specific aspects related to discrimination.

The same applies to administrative records, such as those of complaints filed for alleged pay discrimination. The scarcity of court cases does not reflect the extent of pay discrimination because it could actually reflect lack of trust in the judicial system, flaws in the victims’ protection system, and inadequate length of the limitation period, as this SWD has showed. Furthermore, the lack of legal action might simply reflect lack of awareness and information on the part of victims on how much other workers are paid for the same work or work of equal value.

Empirical findings

In a 2019 study conducted for Australia by KPMG, gender discrimination was revealed as the most significant component contributing to the gender pay gap, accounting for 39% of the gender pay gap in 2017 252 . In the report, gender discrimination is understood as the element of the gender pay gap that would remain if men and women had the same levels of the other factors taken into account 253 . This is the same percentage linked to the ‘years not working due to interruptions’, largely above the effect of ‘occupational and industrial segregation’ which represent 17% of the gap and is on a decreasing trend. These interruptions might be due to different reasons but for women they tend to be related to the ‘care- penalty’, a factor that might also contribute to statistical discrimination. In the EU, the Directive on parental leave might have a positive impact on these aspects; unfortunately, as mentioned in Annex 3, comparable data at EU level on career breaks are not currently available.

A similar decomposition 254 for the UK finds that 38% of the pay gap is due to direct discrimination and differences in the labour market motivations and preferences of women as compared with men while gender differences in life-time working patterns account for 36%. The remaining part is due to labour market rigidities (namely segregation) and to a lesser extent women attaining fewer qualifications in the past.

(1) The principle of equal pay between women and men has been enshrined in the Treaties since 1957 (Article 157(1-2)). Article 4 of the Directive 2006/54/EC implements the principle of equal pay by providing that, for the same work or for work of equal value, direct and indirect discrimination on grounds of sex is prohibited in all aspects and conditions of remuneration.

(2) Foubert P., The enforcement of the principle of equal pay for equal work or work of equal value. A legal analysis of the situation in the EU Member States, Iceland, Liechtenstein and Norway, ELLN for the European Commission, 2017. Available at: https://op.europa.eu/en/publication-detail/-/publication/834d34ab-2d87-11e8-b5fe-01aa75ed71a1  

(3) The gender pay gap is the difference between average gross hourly earnings of male and female paid employees, expressed as a percentage of the former. Source: Eurostat [ sdg_05_20 ], See also http://ec.europa.eu/eurostat/statistics-explained/index.php/Gender _pay_gap_statistics .

(4)   https://ec.europa.eu/info/policies/justice-and-fundamental-rights/gender-equality/gender-equality-strategy_en .

(5) 18% of women over 65 had an income below the relative poverty threshold in 2018 compared to 13.4% of men [Eurostat, ilc_li02 ].

(6) See https://composite-indicators.jrc.ec.europa.eu/social-scoreboard/

(7) Proclaimed by the European Parliament, the Council and the Commission on 17 November 2017 Available at: https://ec.europa.eu/commission/priorities/deeper-and-fairer-economic-and-monetary-union/european-pillar-social-rights_en

(8) The European semester is the framework for the coordination of economic policies across the EU. See https://ec.europa.eu/info/business-economy-euro/economic-and-fiscal-policy-coordination/eu-economic-governance-monitoring-prevention-correction/european-semester_en

(9) The Regulatory Fitness and Performance Programme, or REFIT, is the Commission’s programme for ensuring that EU legislation remains fit for purpose and delivers the results intended by EU lawmakers.

(10) Directive 2006/54/EC of the European Parliament and of the Council of 5 July 2006 on the implementation of the principle of equal opportunities and equal treatment of men and women in matters of employment and occupation (recast), OJ L 204, 26.7.2006. Available at: https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32006L0054

(11) Commission Recommendation of 7 March 2014 on strengthening the principle of equal pay between men and women through transparency. OJ L 69/112, 8.3.2014. Available at: https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32014H0124

(12) Commission’s Action Plan 2017-2019: tackling the gender pay gap, adopted on 20 November 2017, available at: https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX% 3A52017DC0678

(13) Report from the Commission to the European Parliament and the Council on the application of Directive 2006/54/EC of the European Parliament and of the Council of 5 July 2006 on the implementation of the principle of equal opportunities and equal treatment of men and women in matters of employment and occupation (recast) — COM(2013) 861 final. Available at: https://eur-lex.europa.eu/legal-content/en/ALL/?uri=CELEX:52013DC0861 .

(14) Report from the Commission to the European Parliament, the Council and the European Economic and Social Committee on the implementation of Commission Recommendation on strengthening the principle of equal pay between men and women through transparency — COM(2017) 671 final. Available at: https://ec.europa.eu/transparency/regdoc/rep/1/2017/EN/COM-2017-671-F1-EN-MAIN-PART-1.PDF

(15) European Commission, Overview of landmark case law of the Court of Justice of the European Union, 2019, Available at: https://op.europa.eu/en/publication-detail/-/publication/8ecbe429-fb90-11e9-8c1f-01aa75ed71a1

(16) S.Burry, National cases and good practices on equal pay, ELLN for the European Commission, 2017, Available at: https://ec.europa.eu/info/sites/info/files/ equalpaygoodpractices.pdf , p. 34

(17) European Commission, Better Regulation: Guidelines and Toolbox. https://ec.europa.eu/info/better-regulation-guidelines-and-toolbox_en .

(18)   https://ec.europa.eu/info/law/better-regulation/initiatives/ares-2018-3415794_en

(19) Case 43-75, Gabrielle Defrenne v Société anonyme belge de navigation aérienne Sabena (Defrenne II), ECR 1976/455, 8 April 1976, para. 8-10.

(20) In 2006, 25 Member States were subject to the Directive. Bulgaria and Romania joined the EU in 2007 and Croatia in mid-2013. All 28 Member States were subject to the Recommendation.

(21) Directive75/117/EEC on equal pay; Directive 86/378/EEC, as amended by Directive 96/97/EC, on equal treatment in occupational social security schemes; Directive 76/207/EEC, as amended by Directive 2002/73/EC, on equal treatment of men and women; Directive 97/80/EC, as amended by Directive 98/52/EC, on the burden of proof in cases of discrimination based on sex.

(22) An overview of landmark case-law on equal pay is provided in Annex 7.

(23) Proposal for a Directive of the European Parliament and of the Council on the implementation of the principle of equal opportunities and equal treatment of men and women in matters of employment and occupation (recast version), COM(2004) 279 of 21.4.2004.

(24)  The Treaty uses the term ‘equal work’, while the Directive - ‘same work’.

(25) Recitals 8 and 10 of Recommendation 2014/124/EU of 7 March 2014 on strengthening the principle of equal pay between men and women through transparency, OJ L 69, 8.3.2014 and Report from the Commission to the European Parliament and the Council Report on the application of Directive 2006/54/EC on the implementation of the principle of equal opportunities and equal treatment of men and women in matters of employment and occupation, COM/2013/861 final.

(26) The Recommendation also invites Member States to provide the Commission (Eurostat) each year with statistics on the gender pay gap broken down by gender, economic sector, working time (full-time/part-time), economic control (public/private ownership) and age, and data on the number and types of pay discrimination cases. The Recommendation invites Member States to clarify the concept of ‘work of equal value’, following the guidelines established by the European Court of Justice; to promote the development and use of gender-neutral job evaluation and classification systems, and to ensure that the implementation of the equal pay principle is consistently monitored and that all available remedies to combat pay discrimination are enforced. Finally, the Recommendation includes suggestions on the role of equality bodies. These already had the competence to act on matters of equal pay, including pay discrimination, on the basis of the Directive. The Recommendation aimed at strengthening their action by inviting Member States to ensure that equality bodies can handle questions and claims relating to pay transparency. More specifically, equality bodies should have access to pay reports prepared by employers and pay audits and they should be able to represent victims in pay discrimination cases. Finally, importantly, equality bodies’ action should be strengthened by establishing a close cooperation and coordination with national bodies that have an inspection function in the labour market (labour inspectorates).

(27) Council Directive 75/117/EEC of 10 February 1975 on the approximation of the laws of the Member States relating to the application of the principle of equal pay for men and women, OJ L 045, 19/02/1975 P. 0019 – 0020.

(28) Council Directive 76/207/EEC of 9 February 1976 on the implementation of the principle of equal treatment for men and women as regards access to employment, vocational training and promotion, and working conditions, OJ L 039, 14/02/1976 P. 0040 – 0042.

(29) Council Directive 98/52/EC of 13 July 1998 on the extension of Directive 97/80/EC on the burden of proof in cases of discrimination based on sex to the United Kingdom of Great Britain and Northern Ireland, OJ L 205, 22.7.1998, p. 66-66.

(30) The International Covenant on Civil and Political Rights (ICCPR), the International Covenant on Economic, Social and Cultural Rights (ICESCR) and the Convention on the Elimination of all forms of Discrimination Against Women (CEDAW) and its Optional Protocol (See Section 5.4. for details).

(31) See https://ec.europa.eu/archives/growthandjobs_2009/

(32) See for instance Beveridge, F., & Velluti, S., Gender and the open method of coordination, Dartmouth: Ashgate, 2008.

(33) Employment and social policy, 2392nd Council meeting, Brussels, 3 December 2001, doc. 14762/01 (Presse451-G) 2001.

(34) The indicator therefore does not capture differences in net earnings that may result from provisions such as joint taxation.

(35) A decrease of average gross earnings for males therefore also reduces the gender pay gap. The impact of this on poverty rates is obviously different than an increase in female average gross wages would be.

(36) Excluding Ireland and Greece. See national data in Annex 3 .

(37) Eurostat calculates the at-risk-of-poverty rate as the proportion of people with an equivalised disposable income (after social transfer) below 60% of the national median equivalised disposable income after social transfers. Being a relative measure, the indicator does not necessarily imply a low standard of living.

(38) For example, see Clark, A., ‘Job satisfaction and gender: Why are women so happy at work?’ in Labour Economics, 4(4), 1997, pp. 341-372; Vlaedisavljevic, M. and Perugini, C., Gender inequality and the gender job satisfaction paradox in Europe, 2018, http://hermes-ir.lib.hit-u.ac.jp/rs/bitstream/10086/29083/1/wp2017-9.pdf

(39)  European Commission, Discrimination in the European Union, Special Eurobarometer, 2007, Fieldwork: June — July 2006, http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_263_en.pdf .

(40) Study to support the evaluation of the relevant provisions in Directive 2006/54/EC implementing the Treaty principle on ‘equal pay for equal work or work of equal value’. Available at: https://op.europa.eu/en/publication-detail/-/publication/fd8a4b90-5848-11ea-8b81-01aa75ed71a1/language-en/format-PDF/source-117998205

(41) Eurostat, Gender pay gap statistics, 2020, available at: https://ec.europa.eu/eurostat/statistics-explained/index.php/Gender_pay_gap _statistics .

(42) A recent example of possible discriminatory practices comes from the first phase of the implementation of the French Index A recent example of possible discriminatory practices comes from the first phase of the implementation of the French Index. It showed that: (i) 1 out of three companies had not respected the legal obligation to grant to mothers returning from maternity leave the same increase in the salary granted to colleagues in the meantime (in undertakings of more than 1000 employees with a score below 75/100) and (ii) 1 out of five companies had not respected the legal obligation to grant to mothers returning from maternity leave the same increase in the salary granted to colleagues in the meantime (in undertakings between 250 and 1 000 employees with a score below 75/100), Available at: http://hauts-de-france.direccte.gouv.fr/sites/hauts-de-france.direccte.gouv.fr/IMG/pdf/dp_egapro-ok-02-bd.pdf  

(43) Eurobarometer, ‘Gender Equality 2017 – Summary Report’, Special Barometer No 465, 2017.

(44) A difference that about 90% of citizens consider ‘unacceptable’, varying from over 95 % in the Netherlands, Luxembourg, France, Sweden and Spain, to around 80 % in Poland, Austria and Romania (see also section 5.3).

(45) Although the Blinder-Oaxaca (BO) methodology is the prevailing approach used in the empirical work on wage gaps it has some limitations: First, the relationship between characteristics and wages is not necessarily linear, and recent data have been found to violate key implications of the Mincerian model (i.e. a single-equation models that explains income as a function of schooling and experience), which is the key input of the BO decompositions. Second, BO is informative only about the average wage gap decomposition, providing no clues about the distribution of the differences in pay. Third, BO fails to restrict its comparison to comparable individuals, which is likely to substantially upwardly bias the estimators for unexplained differences in pay. Other methodologies to decompose the gender pay gap (with their own limitations) are illustrated in Annex 8.

(46) Parts of the literature use a different approach for the decomposition of the gender pay gap (see Olsen, W.K.; Walby, S., ‘Modelling Gender Pay Gaps’, Manchester: Equal Opportunities Commission, EOC Working Paper Series, 2004, p. 80. According to this estimate, in the UK, 38% of the gender pay gap is due to direct discrimination and differences in the labour market motivations and preferences of women as compared with men. Other estimates provide similar results. (See Annex 8 for details).

(47) A gender gap in unpaid working time is visible in all EU Member States, ranging from 6 to 8 hours in the Nordic countries to over 15 hours in Italy, Croatia, Slovenia, Austria, Malta, Greece and Cyprus. See Eurofound (2018), Striking a balance: Reconciling work and life in the EU, available at: https://www.eurofound.europa.eu/publications/report/2018/striking-a-balance-reconciling-work-and-life-in-the-eu

(48) Directive (EU) 2019/1158 of the European Parliament and the Council of 20 June 2019, OJ L 188/79, 12.7.2019.

(49) The Gender gap in part-time employment from 23.6 to 21.1 as percentage of total employment from 2006 to 2018. Pay gaps can be analysed from the perspective of part-time however information at this level of detail is not available for all EU Member States (See http://ec.europa.eu/eurostat/product?code=earn_gr_gpgr2wt&language=en&mode=view ).

(50) Report from the Commission to the European Parliament and the Council on the application of Directive 2006/54/EC of the European Parliament and of the Council of 5 July 2006 on the implementation of the principle of equal opportunities and equal treatment of men and women in matters of employment and occupation (recast) — COM(2013) 861 final. Available at: https://eur-lex.europa.eu/legal-content/en/ALL/?uri=CELEX:52013DC0861.

(51) CZ, DK, EE, EL, FR, IT, CY, LT, PL, PT, SI, SK, SE, UK. Source: Report on the application of Directive 2006/54/EC of the European Parliament and of the Council of 5 July 2006 on the implementation of the principle of equal opportunities and equal treatment of men and women in matters of employment and occupation (recast), COM(2013) 861 final.

(52) AT, BE, DE, ES, FI, IE, LV, LU, HU, MT, NL. Source: Report on the application of Directive 2006/54/EC of the European Parliament and of the Council of 5 July 2006 on the implementation of the principle of equal opportunities and equal treatment of men and women in matters of employment and occupation (recast), COM(2013) 861 final.

(53) In line with Article 31 of the Directive.

(54) BG, CY, CZ, EE, EL, HR, HU, LV, MT, PL, RO, SI, SK.

(55) Report from the Commission to the European Parliament, the Council and the European Economic and Social Committee on the implementation of Commission Recommendation on strengthening the principle of equal pay between men and women through transparency — COM(2017) 671 final.

(56) Foubert P., The enforcement of the principle of equal pay for equal work or work of equal value A legal analysis of the situation in the EU Member States, Iceland, Liechtenstein and Norway, ELLN for the European Commission, 2017.

(57) See, for example, Case C-58/81, Commission of the European Communities v Grand Duchy of Luxembourg, ECLI:EU:C:1982:215; Case C-171/88, Rinner-Kulhn v FWW Spezial-Gebaudereinigung GmbH, ECLI:EU:C:1989:328; Case C-147/02 Alabaster v Woolwich plc and Secretary of State for Social Security, ECLI:EU:C:2004:192; Case C-342/93, Gillespie and Others ECLI:EU:C:1996:46; Case C-278/93, Freers and Speckmann v Deutsche Bundepost, ECLI:EU:C:1996:83; Case C-12/81, Eileen Garland v British Rail Engineering Limited, ECLI:EU:C:1982:44; Case C-360/90, Arbeiterwohlfahrt der Stadt Berlin e.V. v Monika Bötel, ECLI:EU:C:1992:246; Case C-33/89, Maria Kowalska v Freie und Hansestadt Hamburg, ECLI:EU:C:1990:265.

(58) Recital 13 of the Directive based on CJEU case C-262-88, Douglas Harvey Barber v Guardian Royal Exchange Assurance Group, ECLI:EU:C:1990:209.

(59) European Commission, Better Regulation: Guidelines and Toolbox. https://ec.europa.eu/info/better-regulation-guidelines-and-toolbox_en.

(60) AT, FI, IT, LV, SE, UK

(61) National country fiches.

(62)  For instance, despite the restrictive definition of pay in the Latvian Labour Code, the Supreme Court interpreted the concept of pay according to the case-law of the CJEU within the meaning of equal pay in a Decision of the Supreme Court (15 December 2010) in case No. SKC-694/2010. Available at http://www.at.gov.lv/files/archive/department1/2010/694-10.pdf

(63) European Network of Legal Experts in Gender Equality and Non-discrimination (2016), A comparative analysis of gender equality law in Europe.

(64) Case C-320/00, A.G. Lawrence and Others v Regent Office Care Ltd, Commercial Catering Group and Mitie Secure Services Ltd, ECLI:EU:C:2002:498.

(65) Although there are no definitions laid down in national law, jurisprudence has defined the concept and detailed the parameters allowing comparison of jobs of equal values. This jurisprudence is further detailed in the box below ‘Good practice examples’.

(66) The Polish Supreme Court, in its ruling of 9 February 2007 (I PK 222/06), has specified the parameters to take into account in the comparison. According to the Court, ‘it is possible to compare the remuneration for the work of the claimant with another person employed on the position of an accountant, while taking into account many other circumstances (professional qualifications, previous work experience, scope of tasks associated with the working position, responsibility, physical workload or amount of work) serving as a criteria of evaluation of their work.’ This jurisprudence is sufficiently clear and allows specific parameters to be laid down, thus preventing a lack of understanding of the concept in practice.

(67) CJEU Case C-400/93 Royal Copenhagen, ECLI:EU:C:1995:155; Case C-309/97 Angestelltenbetriebsrat der Wiener Gebietskrankenkasse, ECLI:EU:C:1999:241; Case C-381/99 Brunnhofer, ECLI:EU:C:2001:358; Case C‑427/11 Margaret Kenny and Others v Minister for Justice, Equality and Law Reform and Others [2013] ECLI:EU:C:2013:122, paragraph 28.

(68) CJEU Case C-61/81 Commission of the European Communities v United Kingdom of Great Britain and Northern Ireland, ECR 1982 -02601.

(69) European network of legal experts in gender equality and non-discrimination, Report on pay transparency in the EU: A legal analysis of the situation in the EU Member States, Iceland, Liechtenstein and Norway, 2016. Available at: https://publications.europa.eu/en/publication-detail/-/publication/329c3e47-2bd8-11e7-9412-01aa75ed71a1/language-en .

(70) Case 129/79 Macarthys (1980) ECR 1275 and Case C-200/91 Coloroll Pension Trustees Ltd v Russell (1994) ECR I-4389.

(71) EE, EL, FI, HR, IE, LT, LU, NL, MT, PL, SK identified by national experts involved in the evaluation.

(72) AT, DK, ES, FR, HU, IT, MT, PT, RO, SE, UK identified by national experts involved in the evaluation.

(73) BE, BG, DE, CY, LV, SI.

(74) NL, SK, EE, HR, MT, EL and PL.

(75) Barnard, C., EU Employment Law – Fourth Edition, 2012.

(76) Asda Stores Ltd v Brierley CA [2019] EWCA Civ 44 the Court of Appeal Currently, the CJEU is looking at a similar case - C-624/19 Tesco.

(77) Member States where the use of confidentiality clauses is allowed and routinely used are DE, EE, IE, HR, HU, LV, MT, LV.

(78) This was the case in the following judgments: Cass. Soc. 28 September 2004 No 03-41825; CA 13 July 1995, n°16893; CA 7 December 2015, n°39457; CA 14 July 2016, n°41026; ICF (2019), Country fiche.

(79) Barnard, C., EU Employment Law – Fourth Edition, 2012.

(80) Opinion of the Advisory Committee on Equal Opportunities for Women and Men on the effectiveness of the current legal framework on equal pay for equal work or work of equal value in tackling gender pay gap, June 2009.

(81) FR, LV, MT, SK.

(82) AT, BG, CZ, DK, EE, EL, HU, IT, LT, SI, FI.

(83) The Spanish expert stated that it was not possible to evaluate the implementation and understanding of the legal clarify of the concept of equal pay for equal value, given recent legal developments in 2019.

(84) CY, DE, ES, HR, PT, SE, UK. Note that in Croatia, despite a clear definition in the legislation and an extensive legal framework detailing the obligation stemming from the concept of equal pay for work of equal value, it is yet not understood in practice.

(85) BE, IE, LU, NL, PL. Note that although the legal definitions of the concepts are clear under national law, they are very restrictive in theNetherlands. Equal work or work of equal value can only be identified with a comparator within the same company of the alleged victim, asunderlined in Article 7(1) of the Equal Treatment Act, available at: https://wetten.overheid.nl/BWBR0003299/2015-07-01 (accessed 19 June2019).

(86) Motion for a European Parliament Resolution (2015), on the application of Directive 2006/54/EC of the European Parliament and of the Council of 5 July 2006 on the implementation of the principle of equal opportunities and equal treatment of men and women in matters of employment and occupation (2014/2160(INI)) accessed at: http://www.europarl.europa.eu/doceo/document/A-8-2015-0213_EN.html  

(87) Burri. S., Annex I — Legal aspects and direct and indirect discrimination, in the Implementation Assessment of the Directive 2006/54/EC (‘Recast Directive’) for the European Parliament Committee on Women’s Rights and Gender Equality FEMM, 2015, p. 39.

(88) S.Burry, National cases and good practices on equal pay, ELLN for the European Commission, 2017, Available at: https://ec.europa.eu/info/sites/info/files/ equalpaygoodpractices.pdf , p.9.

(89) Case C-61/81 Commission of the European Communities v United Kingdom of Great Britain and Northern Ireland, ECR 1982-02601.

(90) Article 17 (defence of rights), Article 18 (effective, proportionate and dissuasive compensation or reparation), Article 19 (shared burden of proof) and Article 24 (protection against dismissal or other adverse treatment).

(91) Council Directive 97/80/EC of 15 December 1997 on the burden of proof in cases of discrimination based on sex.

(92) Case C-109/88, Judgment of the Court of 17 October 1989, Handels- og Kontorfunktionærernes Forbund I Danmark v Dansk Arbejdsgiverforening, acting on behalf of Danfoss, ECR 1989-03199. See also Case C-262/88, Douglas Harvey Barber v Guardian Royal Exchange Assurance Group, [1990] ECR I-1889 (Barber), paragraphs 33-34.

(93) ICF Country Fiches (2019) and Questionnaires of the Network of Legal Experts on Gender Equality and Non-discrimination supporting the following Study: Foubert P., The enforcement of the principle of equal pay for equal work or work of equal value A legal analysis of the situation in the EU Member States, Iceland, Liechtenstein and Norway, ELLN for the European Commission, 2017.

(94)  Annex 2 provides a more refined analysis of the replies to the public consultation. The latter by their own nature are however not representative of the general population or subcategories of respondents.

(95) In Bulgaria, general courts and the equality body procedures are exempt from costs, both state fees and expenses. However, administrative case-law does not respect this, and parties are ordered to pay each other’s costs if the case was lost, as well as court fees. In addition, the equality body (CPfD) demands that it be awarded costs when complainants appeal its decisions before the court.

(96) AT, BE, BG, CY, CZ, DE, EE, EL, ES, HR, IT.

(97) For example, CY, PL, ES.

(98) Commission’s European Experts Legal Network, The enforcement of the principle of equal pay for equal work or work of equal value, A legal analysis of the situation in the EU Member States, Iceland, Liechtenstein and Norway, for the European Commission. See also an overview of national case in the Network’s 2019 report: https://www.equalitylaw.eu/downloads/5002-national-cases-and-good-practices-on-equal-pay .

(99) European Trade Union Institute, ETUC Annual Gender Equality Survey, 2019.

(100) An expert reported that the possibility to impose administrative fine exists in Estonia, but is not applied in practice by the labour inspectorate due to limited resource to enforce equal pay provisions.

(101) Imprisonment can be up to 6 months (CY, FI, MT), 1 year (BE) or 3 years (LT).

(102)  In Ireland, companies can be required to carry out a full equality audit along with the implementation of equality procedures within a certain period and the publication of the decision; in the UK, the tribunal is required to order the employer to carry out an equal pay audit.

(103) Some experts stressed other effective types of enforcement mechanisms. For instance, in Belgium, one trained staff member responsible for discrimination is allocated per regional district (there are 24 in total). In practice, however, they have very few cases, so they cannot be considered professionals in discrimination cases.

(104) BE, BG, CY, DK, EE, FR, FI, HR, HU, IE, LT, LV, MT, NL, PT, SE, SI, SK, UK.

(105) This is the case in 12 Member States, namely BG, EE, FR, HR, HU, IT, LV, MT, NL, PT, RO, SE.

(106) European Network of Legal Experts, The enforcement of the principle of equal pay for equal work, 2018.

(107)   https://ec.europa.eu/info/policies/justice-and-fundamental-rights/combatting-discrimination/tackling-discrimination/network-experts-field-anti-discrimination_en .

(108) SWD(2013) 512 final.

(109) Job classification is one of the possible job evaluation systems. Jobs are classified by assigning the various functions in a company or sector to existing categories and structures. Since there is in principle no detailed analysis of the job content, this approach risks the possible influence of gender prejudices and stereotypes and might result in predominantly female jobs being undervalued. Job evaluation systems can go further in ranking functions based on a range of factors such as skill, effort, level of responsibility and working conditions.

(110) Barnard, C., EU Employment Law – Fourth Edition, Oxford University Press, 2012.

(111) European Network of Legal Experts in Gender Equality and Non-discrimination, Report on pay transparency in the EU: A legal analysis of the situation in the EU Member States, Iceland, Liechtenstein and Norway, 2016. Accessible at: https://publications.europa.eu/en/publication-detail/-/publication/329c3e47-2bd8-11e7-9412-01aa75ed71a1/language-en .

(112) Luxembourg ICF Country Fiche (2019).

(113) Convention Collective de travail numéro 25 du 15 octobre 1975 sur l’égalité des rémunérations entre les travailleurs masculins et les travailleurs féminins, modifiée par les conventions collectives de travail numéro 25 BIS du 19 décembre 2001 et numéro 25 TER du 9 juillet 2008. Accessible at: http://www.cnt-nar.be/CCT-COORD/cct-025.pdf

(114)  Guide pour une évaluation non discriminante des emplois à prédominance féminine (Défenseur des droits, 2013a).

(115) B. Grésy, M. Becker ‘Guide pour la prise en compte de l’égalité entre les femmes et les hommes dans les systèmes de classification’, 2017. Accessible at: https://www.egalite-femmes-hommes.gouv.fr/guide-du-conseil-superieur-de-legalite-professionnelle-csep-pour-la-prise-en-compte-de-legalite-entre-les-femmes-et-les-hommes-dans-les-systemes-de-classification/ (Accessed 20 May 2019).

(116) Binet, 2017; Union des employeurs de l’économie sociale et solidaire (UDES), 2016.

(117)  HR, MT and FR.

(118)

 Women and Equalities Committee [WEC] (2016), Gender pay gap, Commons select committee inquiry .

(119)

United Kingdom, Government Equalities Office (2016), Closing the gender pay gap: Government response to the consultation.

(120) AT, BE, DK, DE, ES, IT, FI, FR, LT, LU, PT, SE, UK.

(121) In Germany, employers with at least 500 employees and bound by collective agreements are obliged to carry out pay reporting. The legislation encourages, but does not mandate, pay audits.

(122)

The Equality Ombudsman (DO), Sakligt motiverad eller koppling till kön? En analys av arbetsgivares arbete med att motverka osakliga löneskillnader mellan kvinnor och män’, No 2016:1, prepared by J. Kumlin, Stockholm, 2016. The estimated fines were converted from SEK (SEK 100,000-2,000,000). The government launched a governmental inquiry in 2018 to determine if increased sanctions were needed to achieve greater compliance with the legislation. The report is due October 2020.

(123)

 United Kingdom, Government Equalities Office (2018b), Gender pay gap information regulations: Summary of reported data for 2017/18.

(124) Dromey J. and Rankin L., The fair pay report: How pay transparency can help tackle inequalities, IPPR, 2018, Available at: http://www.ippr.org/research/publications/the-fair-pay-report   http://www.ippr.org/research/publications/the-fair-pay-report .

(125) Nyberg, A. (2019), ‘Including Equal pay in Collective Bargaining: Experience in Sweden’, Presentation at the SAAGE Thematic Seminar, Brussels May 2019.

(126) Eurofound (2018), Pay transparency in Europe: First experiences with gender pay reports and audits in four Member States, Publications Office of the European Union, Luxembourg.

(127)

Unionen, Lönekartläggning lönar sig, Stockholm, 2016.

(128) Eurofound, Pay transparency in Europe: First experiences with gender pay reports and audits in four Member States, Publications Office of the European Union, Luxembourg, 2018.

(129) Scheele, A. (2019), ‘The Pay Transparency Act in Germany: Idea and Outcomes’, Presentation at the SAAGE Thematic Seminar, Brussels May 2019. Considering these figures, the efficiency section made the assumption that 20% of companies comply with the legislation in Germany.

(130) Eydoux, A., ‘The French Gender Equality Index’, Presentation at the SAAGE Thematic Seminar, Brussels May 2019.

(131) For the illustrative calculations in the efficiency section, 50% compliance was assumed in Belgium. However, as explained above, the true compliance rate is unknown.

(132) The Ministry for Women in Austria commissioned an evaluation in 2015 – see country fiche.

(133)   Bennedsen, M., Simintzi, E., Tsoutsoura, M. and Wolfenzon, D., ‘ Do Firms Respond to Gender Pay Gap Transparency? ’ , NBER Working Paper No 25435, 2019.

(134) Pay reports disclosed in AT, BE, DK, ES, IT, FR, LT, PT and the UK, and pay audits disclosed in DE, FI, FR, PT and SE.

(135) Austria, Denmark and Finland examples from Eurofound, Pay transparency in Europe: First experiences with gender pay reports and audits in four Member States, Luxembourg, Publications Office of the European Union, 2018.

(136) Interview with the Equality Body in Belgium.

(137) The French index initiative also includes a number of corrective measures. In particular, the Labour Inspectorate carries out 7,000 specific inspections per year, with companies still below the required threshold after 3 years facing a fine of up to 1% of the total payroll. A preliminary analysis of the results at the beginning of the third phase of implementation is available here: https://travail-emploi.gouv.fr/IMG/pdf/dp_egapro-ok-02-bd.pdf .

(138) Dromey, J. and Rankin, L., Fair Pay Report: How Pay Transparency can Help Tackle Inequalities, Institute for Public Policy Research, 2018, https://www.ippr.org/files/2018-11/the-fair-pay-report-november18.pdf .

(139) Information received through an interview with a stakeholder in France. It is not clear if the remaining 6% were fined.

(140) For example, Nyborg, A., ‘Including Equal Pay in Collective Bargaining: Experience in Sweden’, Presentation at the SAAGE Thematic Seminar, May 2019. Statistics taken from a study by the equality body, JämO: Arbetsvärderingssystem i Sverige, Stockholm, 2008a; Bennedsen, M., Simintzi, E., Tsoutsoura, M. and Wolfenzon, D. (2019), ‘Do Firms Respond to Gender Pay Gap Transparency?’, NBER Working Paper No 25435.

(141) EIGE (2017c), Gender Equality Index 2017: Measuring gender equality in the European Union 2005-2015. Main Findings.

(142) Cipollone, A., Patacchini, E. and Vallanti, G., Women’s Labour Market Performance in Europe: Trends and Shaping Factors. CEPS Special Report No 66, 2012.

(143) For example, the proposal for a directive of the European Parliament and of the Council on improving the gender balance among non-executive directors of companies listed on stock exchanges and related measures, COM/2012/0614 final — 2012/0299.

(144) In a statistical sense, ‘explained’ does not necessarily mean ‘justified ‘or ‘non-discriminatory’ as the two components might actually be interrelated e.g. the choice of education, sector and occupation might have been influenced already by gender discrimination. See among others Grimshaw, Damian and Rubery, Jill., The Adjusted Gender Pay Gap: a Critical Appraisal of Standard Decomposition Techniques , 2002.

(145) The assessment of the impact of a given measure would ideally require carrying out a ‘counterfactual analysis’ with an experimental design. The intuitive concept involves comparing the outcomes of an intervention with the outcomes that would have been achieved if it had not been implemented. This is done by looking at the impact on two groups (treatment and control group) that differ only by whether or not they were subjected to the intervention. The typical reference context is the testing of medical treatments. However, since the Recast Directive did not add major changes to the existing legislation of most Member States, it was unlikely that any impacts would be found following its transposition. Similarly, the limited likelihood of impacts being generated limits the potential to compare Member States that introduced legislation against those that did not. Finally, the three EU countries that were not subject to the Recast Directive (BG, RO, HR) could have been used as a control group, but beyond the mere lack of data for Croatia, there are no Member States with a similar trend in the gender pay gap or in other labour market indicators between 2002 and 2006 that also have a similar social and labour market context. In any case, the results were unlikely to be robust with such a small sample. Referring to the Recommendation was also not considered likely to produce robust results, as implementation began too recently (since 2014) and since Member States adopted a wide range of measures, resulting in considerable heterogeneity in the treatment group.

(146)

Bennedsen, M., Simintzi, E., Tsoutsoura, M. and Wolfenzon, D., ‘Do Firms Respond to Gender Pay Gap Transparency?’, NBER Working Paper No 25435, 2019; Huet-Vaughn, E., ‘Striving for status: A field experiment on relative earnings and labour supply’, Job Market Paper, Berkeley: University of California, 2013.

(147)

Baker, M., Halberstam, Y., Kroft, K., Mas, A. and Messacar, D., Pay Transparency and the Gender Gap No. w25834, National Bureau of Economic Research, 2019; Cullen, Z. and Pakzad-Hurson, B., Equilibrium Effects of Pay Transparency in a Simple Labour Market, Mimeo, 2018.

(148)

See Scott, D. (2018). Pay Communications and Fairness: An Employee Perspective. Compensation & Benefits Review, 50(1), 5–20. https://doi.org/10.1177/0886368718809774

(149)

See for instance, A. Chamberlain, “Is salary transparency more than a trend?”, Research report, 2015, available on the Glassdoor website ; Arnold, A., Fulmer, I.S., Sender, A., Allen, D.G., Staffelbach, B., & Perkins, S.J. (2018). International study on compensation and pay transparency practices. Lucerne, Switzerland: Center for Human Resource Management, University of Lucerne. Available at https://www.unilu.ch/fileadmin/fakultaeten/wf/institute/hrm/dok/Forschung/PayTransparency_IntReport.pdf ; Belogolovsky, E., Bamberger, P., Alterman, V. et al. Looking for Assistance in the Dark: Pay Secrecy, Expertise Perceptions, and Efficacious Help Seeking Among Members of Newly Formed Virtual Work Groups. J Bus Psychol 31, 459–477 (2016). https://doi.org/10.1007/s10869-015-9427-4 ; E. Belogolovsky and P. A. Bamberger, Signaling in Secret: Pay for Performance and the Incentive and Sorting Effects of Pay Secrecy, Academy of Management Journal, Vol. 57, No. 6, available at https://doi.org/10.5465/amj.2012.0937 ; S. Marasi and R.J. Bennett Pay communication: Where do we go from here?, Human Resource Management Review, Volume 26, Issue 1, March 2016, Pages 50-58, available at https://doi.org/10.1016/j.hrmr.2015.07.002 ; Shelly Marasi, Alison Wall & Rebecca J. Bennett (2018) Pay openness movement: Is it merited? Does it influence more desirable employee outcomes than pay secrecy?, Organization Management Journal, 15:2, 58-77, available at: https://doi.org/10.1080/15416518.2018.1471978

(150)

S. Cooney, “Should You Share Your Salary With Co-Workers? Here’s What Experts Say”, Times 14 August 2018, available at https://time.com/5353848/salary-pay-transparency-work/ ; S. McLaren,” Why These 3 Companies Are Sharing How Much Their Employees Make”, 14 February 2019, LinkedIn, available at https://business.linkedin.com/talent-solutions/blog/trends-and-research/2019/why-these-3-companies-are-sharing-how-much-their-employees-make ; V. Bolden-Barrett, Top HR academic programs now teach students to advocate for pay equity, March 23, 2017, HRDIVE, available at https://www.hrdive.com/news/top-hr-academic-programs-now-teach-students-to-advocate-for-pay-equity/438718/ https://www.hrdive.com/news/top-hr-academic-programs-now-teach-students-to-advocate-for-pay-equity/438718/

(151) ETUI, 2019.

(152) ETUI, 2019.

(153) Deutsche Bundestag, Entwurf eines Gesetzes zur Förderung der Transparenz von Entgeltstrukturen, Gesetzentwurf des Bundesregierung, Drucksache 18/11133.

(154) Nyberg, A., ‘Including equal pay in collective bargaining. The experience of Sweden,’ Presentation at the SAAGE Thematic Seminar, May 2019.

(155) Dromey J. and Rankin L., The fair pay report: How pay transparency can help tackle inequalities, IPPR, 2018, http://www.ippr.org/research/publications/the-fair-pay-report

(156) Smith, M., ‘The Costs and Benefits of Pay Transparency’, Presentation at the SAAGE Thematic Seminar, Brussels, May 2019.

(157) Source: Stakeholders’ interviews conducted as part of national research.

(158)   https://ec.europa.eu/info/publications/eu-mutual-learning-programme-gender-equality-equal-pay-iceland-27-28-may-2019_en

(159) Payscale (2019), Does Pay Transparency Close The Gender Wage Gap? Pay Equity For Men And Women, available at: https://www.payscale.com/data/pay-transparency#quit-form  

(160) As mentioned before, a recent study in Denmark found that the introduction of pay transparency measures led to a 7% reduction in the gender pay gap over a three-year period.

(161) Dromey J. and Rankin L. (2018), The fair pay report: How pay transparency can help tackle inequalities, IPPR, available at: http://www.ippr.org/research/publications/the-fair-pay-report

(162) Nyberg, A., ‘Including equal pay in collective bargaining. The experience of Sweden,’ Presentation at the SAAGE Thematic Seminar, May 2019. For Belgium, it was not possible to collect evidence on the actual level of compliance. A representative of the Belgian equality body pointed out that as there is no effective monitoring mechanism in place, it is not possible to retrieve data on the number of firms that comply with the legislation (interview conducted as part of the national research in Belgium).

(163)  Leythienne, D. and Ronkowski, P. A decomposition of the unadjusted gender pay gap using Structure of Earnings Survey data, Eurostat, Statistical working papers, 2018.

(164)  The statistical working papers collection is used to publish innovative methodological work to inform and stimulate discussions among statisticians. Therefore, the data and statistics in SWPs should be considered as work in progress.

(165) Gender pay gap statistics are not available for Croatia, Ireland and Greece.

(166) The following examples come from Eurofound (2018), Pay transparency in Europe: First experiences with gender pay reports and audits in four Member States. Available at: https://www.eurofound.europa.eu/publications/report/2018/pay-transparency-in-europe-first-experiences-with-gender-pay-reports-and-audits-in-four-member  

(167) Eurobarometer (2017), Gender Equality 2017. Special Eurobarometer 465 — Wave EB87.4 — TNS opinion & social.

(168) Ibid.

(169) Ibid.

(170) Ibid.

(171) Ibid.

(172)   https://ec.europa.eu/info/policies/justice-and-fundamental-rights/gender-equality/gender-equality-strategy_en .

(173)   https://ec.europa.eu/commission/priorities/deeper-and-fairer-economic-and-monetary-union/european-pillar-social-rights_en .

(174) European Commission (2013).

(175)   https://ec.europa.eu/newsroom/just/item-detail.cfm?item_id=607452

(176) Impact assessment accompanying the proposal for a Council recommendation on access to social protection for workers and the self-employed COM(2018) 132 final — SWD(2018) 71 final.

(177)

Cited in EIGE, Tackling the gender pay gap: not without a better work-life balance, 2018. The pay calculator is available at: https://www.monitor-entgelttransparenz.de/

(178)  Eurobarometer ( 2017 ).

(179)  Eurobarometer ( 2017 ).

(180)  European Commission (2013), Implementation report.

(181)  European Commission, See Justice Scoreboard (2018), Figure 31,  http://publications.europa.eu/webpub/com/refit-scoreboard/en/priority/7/7-20.html

(182)  Foubert P., The enforcement of the principle of equal pay for equal work or work of equal value A legal analysis of the situation in the EU Member States, Iceland, Liechtenstein and Norway, ELLN for the European Commission, 2017.

(183) Ibid.

(184) European Commission, 2018, Standards for Equality Bodies C(2018) 3850 Final, Available at: https://ec.europa.eu/info/sites/info/files/2_en_act_part1_v4.pdf .

(185) See Section 3.4.4.2 for supporting sources and analysis.

(186) OJ L 6, 10.1.1979, p. 24–25.

(187) For more details see the Commission staff working document on the evaluation of Directive 79/7/EEC, SWD(2019)450 final.

(188) Case 236/09, Test-Achats, see below, in relation to coherence with Directive 2004/113/EC.

(189) OJ L 348, 28.11.1992, p. 1–7.

(190) OJ L 373, 21.12.2004, p. 37–43.

(191) CJEU Case 236/09 Association Belge des Consommateurs Test-Achats ASBL and Others v Conseil des ministres ECLI:EU:C:2011:100.

(192) OJ L 68, 18.3.2010, p. 13–20.

(193)   http://data.europa.eu/euodp/en/data/dataset/S2185_470_ENG  

(194) OJ L 188, 12.7.2019, p. 79–93.

(195)  OJ L 180, 15.7.2010, p. 1–6.

(196)  OJ L 14, 20.1.1998, p. 9–14.

(197) The second para of Clause 3(2) of Directive 97/81/EC establishes that where there is no comparator in the same establishment, the comparison should be made by reference to an applicable collective agreement or, where there is none, in accordance with national law, collective agreements, or practice.

(198) CJEU Case 43/75, Gabrielle Defrenne v Sabena. C-320/00 Lawrence ECLI:EU:C:2002:498: the comparator has to be identified in a single source (not two different employers), otherwise inequality cannot be restored.

(199) CJEU C-320/00 Lawrence ECLI:EU:C:2002:498.

(200) Case 129/79 Macarthys (1980) ECR 1275 and Case C-200/91 Coloroll Pension Trustees Ltd v Russell (1994) ECR I-4389.

(201) Case C-381/99 Brunnhofer v Bank der Österreichieschen Postsparkasse AG (2001) ECR I-4961 para. 48.

(202)  CJEU Case 96/80 Jenkins ECLI:EU:C:1981:80; is fundamental here, as is a report provided by the European Legal Network of Gender Equality Experts, Countouris N. and Freedland, M., The Personal Scope of EU Sex Equality Directives, European Commission, 2012, available at: http://ec.europa.eu/justice/gender-equality/document/index_en.htm#h2-7.

(203) See Judgment in Joined Cases C-395/08 and C-396/08 - Istituto nazionale della previdenza sociale (INPS) v Tiziana Bruno and Massimo Pettini (C-395/08) and Daniela Lotti and Clara Matteucci (C-396/08), ECLI:EU:C:2010:329.

(204) Directive (EU) 2019/1152 of the European Parliament and of the Council of 20 June 2019 on transparent and predictable working conditions in the European Union, OJ L 186, 11.7.2019, p. 105–121.

(205) OJ L 186, 11.7.2019, p. 105-121.

(206)  CJEU Case 236/09 Association Belge des Consommateurs Test-Achats ASBL and Others v Conseil des ministres ECLI:EU:C:2011:100.

(207) CJEU Case 318/13, X Case ECLI:EU:C:2014:2133.

(208) This is the case, for instance, in CZ, EL, IT, MT, and the UK.

(209)  Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation), OJ L 119, 4.5.2016, p. 1-88.

(210) CJEU, C-342/12, Worten — Equipamentos para o Lar SA v Autoridade para as Condições de Trabalho (ACT), 30 May 2013, para. 45.

(211) See Section 3.3.3.2 of the Study supporting the evaluation.

(212) See for example, C-407/98 Katarina Abrahamsson and Leif Anderson v Elisabet Fogelqvist ECR [2000] I-05539.

(213) Today: Article 157(1-2) TFEU.

(214) Van der Vleuten A., The Price of gender Equality. Member States and Governance in the European Union, 2007, p. 54.

(215) Eurobarometer (2017).

(216) Directive (EU) 2019/1158 of the European Parliament and of the Council of 20 June 2019 on work-life balance for parents and carers and repealing Council Directive 2010/18/EU, OJ L 188, 12.7.2019.

(217) Equinet, How to build a case on equal pay. An Equinet Handbook, 2017.

(218)   https://www.eesc.europa.eu/en/our-work/opinions-information-reports/opinions/eu-action-plan-2017-2019-tackling-gender-pay-gap .

(219) European Commission (2017a), Communication EU Action Plan 2017-2019. Tackling the gender pay gap, COM(2017) 678 final.

(220)   http://publications.europa.eu/webpub/com/refit-scoreboard/en/priority/7/7-20.html

(221) The SWPs’ collection is used to publish innovative methodological work to inform and stimulate discussions among statisticians. Therefore, the data and statistics in SWPs should be considered as work in progress.

(222)   https://www.getresponse.co.uk/

(223)

http://ec.europa.eu/newsroom/just/document.cfm?action=display&doc_id=48360 .

(224) https://ec.europa.eu/info/law/better-regulation/initiatives/ares-2018-3415794_en .

(225) For further detail, see chapter 5.

(226) In each of these countries, less than five responses in total were received.

(227) ‘company/business organisation’ and ‘business association’ were grouped together, referred to as ‘companies and businesses’.

(228) ‘trade union’ and ‘NGOs’ were grouped together, referred to as ‘trade unions and NGOs’.

(229) CEEP — European Centre of Employers and Enterprises providing Public Services and Services of general interest (employer and business organisation); HOTREC — Association of Hotels, Restaurants, Bars and Cafes in Europe (employer and business organisation); EUROCADRES — Council of European professional and managerial staff (trade union); AGE Platform Europe — European network of non-profit organisations of and for people aged 50+ (network organisation); MMM Europe — Make Mothers Matter EU Delegation (NGO); CEMR — The Council of European Municipalities and Regions (association of local and regional governments).

(230)   https://ec.europa.eu/info/law/better-regulation/initiatives/genderequalitybrp/public-consultation_en .

(231) This was marked by 73% of NGOs/trade unions, 65% of academic/research institutions, 63% of public authorities, 51% of EU citizens and 35% of employers’ organisations.

(232) Taxation systems might, therefore, further penalise the worker with lower salary in the household (typically the woman) who will be subject to the tax rate of the higher-earning worker.

(233) A decrease of average gross earnings of males, therefore, also reduces the GPG and has different impact on poverty rates than an increase in female average gross wages.

(234) In Article 157(2) TFEU and Article 2(e) of Directive 2006/54/EC.

(235) SES microdata are collected every 4 years as of 2002. National estimates are computed for the years in-between.

(236) Participating countries are required to collect data on companies with at least 10 employees operating in the economic activities of the European Community (NACE — Rev 2) ranging from B to S excluding O (public administrations). Nevertheless, countries can voluntarily collect information on section O and on companies with 10 employees. Currently data for all companies are only available for Czechia.

(237) Data by ownership of economic authority are available in another time series for most countries.

(238) Using the ‘Blinder-Oaxaca decomposition’ — see later.

(239) See Denis Leythienne, Piotr Ronkowski. A decomposition of the unadjusted gender pay gap using Structure of Earnings Survey data. Eurostat, 2018. The ‘Statistical Working Papers’ (SWPs) collection is used to publish innovative methodological work to inform and stimulate discussions among statisticians. Therefore, the data and statistics in SWPs should be considered as work in progress.

(240)  The ‘Statistical Working Papers’ collection is used to publish innovative methodological work to inform and stimulate discussions among statisticians. Therefore, the data and statistics in SWPs should be considered as work in progress.

(241)  D. Leythienne, P. Ronkowski. A decomposition of the unadjusted gender pay gap using Structure of Earnings Survey data. Eurostat, 2018.

(242)   https://data.europa.eu/doi/10.2838/823955  

(243) http://ec.europa.eu/eurostat/cache/metadata/en/sdg_05_20_esmsip2.htm.

(244) The lawsuit had been filed by Lilly Ledbetter who had been working for Goodyear for nearly 20 years when she discovered that she was being paid between 15 and 40% less than her male counterparts. The Supreme Court eventually held that the time within which an employee may file a discrimination charge ‘is triggered when a discrete unlawful employment practice takes place.’

(245) See J.Altonji and R. Blank, Race and gender in the labor market , 2017, in J.Altonji and R. Blank (eds.), Handbook of Labor Economics, vol 3, Part C, Elsevier for an overview of the debate.

(246) See, for instance, the seminal work from Allport, The nature of prejudice, 1954 Reading, Mass.:Addison-Wesley.

(247) See the review presented in National Research Council. 2004. Measuring Racial Discrimination. Panel on Methods for Measuring Discrimination , R.M. Blank, M. Dabady, and C.F. Citro, eds. Committee on National Statistics. Washington, DC: The National Academies Press.

(248) C. Goldin and C. Rouse, ‘ Orchestrating Impartiality: The Impact of ‘Blind’ Auditions on Female Musicians ’, The American Economic Review, Vol. 90, No 4 (Sep., 2000), pp. 715-741.

(249) There are three main decomposition techniques: the Oaxaca-Blinder (and its extensions), Juhn-Murphy-Pierce and Walby and Olsen simulation approach, each with its own limitations. For a summary and comparison see Cassells et al. (2009), ‘The impact of sustained gender wage gap on the Australian economy’ Report to the office for Women, department of Families, Community services, Housing and indigenous affairs, Appendix A.

(250) Becker formalised this model referring to ‘taste’, i.e. at least some members of the dominant group prefer not to interact with members of the minority group (G. Becker, (1957 and 1971. The Economics of Discrimination. Series: (ERS) Economic Research Studies). The existence of this prejudice is then treated as an individual preference. Even if members of the disadvantaged group have the same productivity as those in the other group, the firm will not employ them to avoid the disutility of being in contact with them (if it has to pay them the same wage) or will employ them only if it can pay them less. In perfect competition the discriminating employers would be driven out of the market as their profits are reduced.

(251) The seminal references of this part of the literature are E. Phelps, The Statistical Theory of Racism and Sexism, 1972 in American Economic Review, Vol 62.and K. J. Arrow, (1973), ‘The Theory of Discrimination,’ in Discrimination in Labor Markets, ed. by O. Ashenfelter and A. Rees, Princeton, NJ: Princeton University Press. The literature originating from their contribution consider inequality to be caused by exogenous differences between groups and employers’ imperfect information about workers’ productivity. Since employers have limited information about the skills and turnover propensity of applicants they will use an observable characteristic, i.e. gender (or race) to ‘statistically discriminate’ among workers. For instance, based on past experience employers might consider that there is a high likelihood that young female workers will have lower labour market attachment than men and therefore offer them fewer career incentives (training, positions …).

(252) KPMG, ‘She is Price(d)less: The Economics of the Gender Pay Gap’, summary report, 22 August 2019, p. 7. The study uses the Olsen-Walby decomposition mentioned above.

(253) In the report, other factors constitute ‘occupational segregation’(males per 100 workers), ‘part-time employment’, ‘age’, ‘employed by the Government or NGOs’, ‘years of not working due to interruptions’, ‘industry segregation’ (males per 100 workers), ‘unpaid and carer work’ (proxied by housework), and ‘tenure with current employers (years)’.

(254) See W.K. Olsen and S.Walby,  Modelling Gender Pay Gaps , EOC Working Paper Series, Manchester: Equal Opportunities Commission, 2004.