For our analyses, we use the (2010) wave of the European Structure of Earnings Survey (hereafter SES). The SES is a large-scale European matched employer-employee dataset. It consists of inter-country harmonized microdata, as the survey is conducted in all the countries according to a common methodology. Thus, cross-country comparisons are developed on a strictly comparable basis with a rich set of harmonized information. This data set is particularly suited for our purposes for two reasons. First, it provides highly reliable and extensive salary information which is a clear advantage over household survey data, which provide comparatively coarse wage information which further bears a high risk of measurement error. Second, the SES data offers a wealth of workplace and employer-related variables whereas this sort of information is regularly rather limited in household survey data. The SES is conducted every four years through questionnaires sent to a representative sample of enterprises. The sample is drawn through a two-stage random sampling. First, a random sample of companies is drawn, and thereafter, a random sample of employees is drawn within each selected company. The national statistical institutes are responsible for selecting the sample, preparing the questionnaires, conducting the survey and forwarding the results to Eurostat. The SES provides detailed information on the relationships between the level of remuneration and individual characteristics of employees (sex, age, occupation, length of service, highest educational level attained, etc.) and those of their employer (economic activity, size and location of the enterprise). As mentioned earlier, the large set of explanatory variables is a plus of this dataset. In comparison to other data, it allows us to include details on temporary contracts and firm characteristics.

The sample regularly covers workplaces with at least ten employees in sections C (Manufacturing) to S (Other services) of the Statistical Classification of Economic Activities in the European Community (NACE rev. 2). Public administration, defence and compulsory social security (section O) is not available for some of the countries in our analysis. This induces us to drop employees from this sector in all countries for the sake of consistency.⁷ Note that also in the official statistics from Eurostat addressing the gender pay gap in the public versus private sector and drawing on the Structure of Earnings Survey, section O is excluded (Eurostat 2017). However, section P (education) which together with section O is normally referred to as the ‘public sector’ (e. g. BMFSFJ 2009), is included. The partial exclusion of the public sector in our analysis will probably impact our results, as we know from the literature, that employment in the public sector plays a considerable role in explaining gender pay gap differences between countries (Mandel/Semyonov 2014;BMFSFJ 2009; Estévez-Abe 2005, 2006). One reason is that wage differentials tend to be lower in the public sector than in the private sector (Arulampalam et al. 2007). However, the likely impact of the exclusion of section O on our results is not straightforward to tell, as the relative magnitude of the gender pay gap in the public vs. the private sector varies between countries.⁸ As further restrictions, self-employed are excluded from our analyses and information on sectors and occupational groups are only available at a limited level of disaggregation.

Given that data availability concerning individual and job-related characteristics differs to some extent between countries, we had to weigh the aim of accounting for as many insightful characteristics as possible against the need to preserve a sufficient number of countries for our analysis. In the end, we were left with 22 countries (21 EU countries plus Norway).⁹ The total number of observations in our aggregated sample is 8,829,191. Note that the estimation results for the country-pooled sample are particularly shaped by large countries (Czech Republic, Norway, and Germany). For example, the Czech Republic (845,009 observations) is 48 times as large as Lithuania (17,491 observations). To account for the heterogeneity of gender wage gaps within Europe, country-specific estimations are indispensable.

The dependent variable is the log average gross hourly wage (in euro). In addition to regular payments, the wage includes overtime payments, shift premiums, bonuses, commissions, allowances for teamwork, night work, weekend work, family allowances, payments to employees’ savings scheme and other gratuities in cash fixed by collective agreements or voluntarily agreed. Not included are advances or pay for holiday, periodic bonuses and gratuities not paid regularly at each pay date, statutory family allowances, allowances for work clothes or tools, reimbursements or payments for travel, subsistence etc., and expenses incurred in carrying out the employer’s business.

As individual worker characteristics, age and education were included. Age proxies experience, i. e. work experience (which as such is not available in our data at hand) and is therefore a broader measure of human capital, including also general, non-firm-specific components (contrary to firm tenure). Age is given in terms of six categories, where the youngest group comprises the 20–29 years old workers and the oldest group the more than 60 years old. The measure of education is derived from an aggregation of ISCED levels into three categories (ISCED 0-2, ISCED 3-4, ISCED 5-6). As job-related characteristics, contract type, firm tenure, hours of work, occupational group as well as industry, ownership and size of the enterprise were taken into account. Contract type is captured by a dummy variable that is equal to one for temporary and zero for permanent contracts. Firm tenure is split into four time spans (0–1 years, 2–4 years, 15–24 years,>24 years) and is based on the self-reported length of service in the enterprise. As noted earlier, firm tenure is not identical with work experience since firm tenure focuses on years of employment with the same employer whereas work experience is the aggregate sum of years of employment. Wage premiums of firm tenure measure returns to firm-specific human capital whereas work experience premiums measure market returns to general human capital. Work experience is not available in the dataset (cf. Foster-McGregor et al. 2013).¹⁰ This data shortcoming will not distort the aggregate level of (unadjusted) pay gaps since unobserved gender differences in characteristics are captured by the unexplained part of the gap, but the latter will be higher (and the explained part will be lower) as it would be the case if actual work experience was observable in the data.

Information on hours of work is available with respect to the share of a full-timer’s normal hours. We differentiate between two dummy variables indicating that workers work 60–99 % (large part-time) and those who work less than 60 % of a full-time worker’s normal workload (small part-time). This measurement restriction is not imposed by us, but by SES itself. Hence, we are unable to detect any potential non-monotonicity properties in the marginal effect of working time within these groups. Nevertheless, as our argumentation (and also the discussion within the literature) is focused on the general dichotomy between part-time and full-time work rather than on modelling a comprehensive functional relationship between wages and work hours, we consider this delimitation useful for our purposes. Moreover, in existing studies, information on precise working hours are commonly generated based on workers subjective self-assessment and are therefore prone to measurement error. Occupational groups are identified based on the ISCO-08 classification at the two-digit-level, discriminating between 42 different groups.

The industry of the enterprise is assigned based on an own aggregation of the NACE-Rev.2- classification, motivated by the need for cross-country harmonization. It allows us to distinguish between 16 different sectors. Concerning the impact of ownership, we include a dummy variable that is set equal to one if the firm is under public control. This is defined to be the case if a share of more than 50 % is in public ownership.

Finally, the size of the enterprise is measured by its number of employees, broadly categorized into enterprises with less than 50 and others with at least 50 employees. As aforementioned, firms with less than 10 employees (very small firms) are not included in the SES 2010. Since women are overrepresented in those firms and at the same time, these firms regularly pay less, this sample restriction is likely to underestimate the role of firm size for the gender pay gap. Analyses that build on the EU-SILC data including very small firms show that this is indeed the case (Boll et al. 2016).

For a detailed description of all variables, see Table 2 in the Annex.

Table 3 based on the aggregate sample, in the Annex depicts summary statistics for men and women, observation numbers by country and gender are reported in Table 4. Average nominal wages of women differ between 1.86 Euros in Bulgaria and 26.19 Euros in Norway. For men, the range is between 2.15 Euros in Bulgaria and 31.50 Euros in Norway. In 16 out of 22 countries, the share of persons with tertiary education is higher among female workers than among males. Among the exceptions, Germany stands out with the highest lead of men over women. Furthermore, the share of employees that exhibit firm tenure from 5 up to 14 years is higher among women than among men in 18 out of 22 countries (with Bulgaria and Hungary marking the most important exceptions), and the same holds true for the share of employees with 15 years and more in 12 countries (with Norway, Slovakia and Finland as the most prominent exceptions). In general, the share of workers aged 50 or older is higher among males than among females.¹¹

More than 20 % of employees work in temporary jobs in Portugal, Poland and Spain and this also holds true for Czech female workers. Generally, temporary jobs are more frequent among women than among men (among the 7 exceptions, Latvia, Poland and Estonia stand out with the highest lead of men over women). Percentage shares of part-timers are higher among female workers than among male workers. This holds true for both types of part-time and throughout Europe. However, country differences in magnitude become apparent.¹²

Concerning horizontal segregation, everywhere in Europe the female share exceeds the male share on employees in health and social work activities and education (with the sole exception of Hungary in the latter case), whereas men dominate throughout Europe in the chemical and metal industry as well as in the field of construction. However, the magnitude of segregation varies by sector and country. Furthermore, women work more often in enterprises under major public control than men do. Publicly controlled firms hold a share of more than 30 % on male employees only in Hungary, Croatia, Norway and Poland, whereas in the named countries plus Norway and Sweden, publicly controlled firms hold shares of more than 50 % on female workers.

To sum up, it becomes evident that women’s formal and also firm-specific human capital outperforms men’s in most countries. On the other hand, traditional patterns of segregation prevail all over Europe. However, from the perspective of mean wages, women are still unable to reap the full returns in terms of earnings. Several reasons might account for this finding.

In what follows, we present selected results on wage regressions. As mentioned above, the nation- and gender-specific wage regressions inform us about the wage returns that men and women achieve for single characteristics. Table 5 in the Annex depicts the full results.

For females, part-time work, disregarding its scope, is associated with significant¹³ hourly wage penalties compared to full-time work in all countries except France, Hungary, Poland and Croatia, where both types of part-time yield wage premiums. Remarkably, in the Netherlands, where part-time is most widespread in the European context, penalties are restricted to low-scale part-time. In Spain, Portugal and Finland, low-scale part-time even yields a wage premium. Similar to females, part-time premiums are observed for males in France and Hungary irrespective of its scope, but with partially insignificant effects in Poland and Croatia. Male patterns resemble female ones also for the remaining countries. Note however, that for both types of part-time, wage penalties tend to be more pronounced for men than for women (with the UK marking the sole exception for low-scale and Slovakia for large-scale part-time), whereas wage premiums are mostly in favour of men (with Hungary marking the sole exception).¹⁴

Working with an enterprise under major public control is related to a wage bonus in roughly one half of countries and to a malus in the other half for females. By contrast, male workers receive a premium in 16 out of 22 countries. Holding a temporary contract (compared to a permanent contract) and working in a small firm with less than 50 employees is both related to wage penalties for women and men in all countries (except for Latvian men in case of temporary work). More years of firm tenure are associated with a wage reward, this holds true for both women and men. At the same time, women benefit from higher premiums for staying 15 years or longer with the same firm than men in Germany, Spain, France, Italy, Belgium and the United Kingdom. Similar to tenure, a higher educational level is associated with higher earnings. This holds for both genders, although educational premiums are differently distributed across genders and countries.¹⁵

Before addressing the level of single endowments, we split the overall gender pay gap (subsequently termed the unadjusted gap) in an explained and an unexplained part. For our cross-country sample, the unadjusted gap is 15.0 % (2010). About one third of the gap can be traced back to the role of the explanatory factors included in our analysis. A wage difference of 11.0 % remains as the unexplained (adjusted) gap. Hence, the source of the largest part of the gap is not a difference in measured worker attributes but in unobservables. At country level, the picture however varies drastically, as shown in Table 1.

Concerning the unadjusted gap, figures range from 3.6 % for Poland to 25.1 % for Estonia. From a geographical perspective, it is noticeable that most Central and Eastern European states are exhibiting gaps clearly below average, with the Czech Republic, Slovakia and Estonia marking the exceptions. Among the West European countries, only Italy is exhibiting a very small gap (4.5 %). Further country variation is revealed by the decomposition results. First, the country ranking with respect to the adjusted gap changes substantially compared to the unadjusted gap. Second, the role of gender differences in average worker features is in some countries not only more pronounced than in others, it also works in opposite directions. For instance, it is striking that the two countries with the smallest raw gap (Poland and Italy) both exhibit negative explained gaps.¹⁶ Hence, the average female worker in these countries is endowed with better characteristics than her male counterpart, at least concerning those characteristics included in our dataset. The reason why also in Poland and Italy female workers nevertheless have lower average earnings is exclusively to be found in the unexplained residual.

Moreover, this unexplained part is nowhere identified to be negative. It doesn’t even get lower than five percent. In most countries, it is thus this term that comprises the bulk of factors that prevent women from catching-up. The only two countries where the explained gap exceeds the unexplained part are Germany and Norway (see for the same result Foster-McGregor et al. 2013; Simón 2012¹⁷). As explained above, it consists of two different kinds of effects. First, it accounts for the fact that the same endowment could be evaluated differently by the market, depending on whether the person is male or female. Second, it includes the impact of gender differences in those market-relevant characteristics which are not controlled for in our model. This second aspect is of special relevance, as our dataset does not allow us to assess potentially important gender differences related to e. g. actual work experience (as mentioned in Section 4.2). It is interesting to see that some of the countries with negative explained gaps like Poland and Portugal perform worse than the country average when it comes to the unexplained gap. Apparently, from the fact that women outperform men in attributes like education one cannot conclude on a lower pay gap. This provides justification for a more disaggregated analysis of the sources of the gender pay gap.

Figure 1 documents which share of the explained part of the gender pay gap can be attributed to which measured characteristic. Precise numbers can be found in Table 6 in the Annex.¹⁸ While some features show similar effects across countries, the role of others is highly heterogeneous.

Figure 1:

Decomposition of the gender pay gap (in %), 2010.Sources: SES (2010), own calculations (see Boll et al. 2016).

Starting with education, the role of schooling tends to contribute to wage convergence, confirming the first part of our first research hypothesis. Female workers in most countries exhibit a higher average level of education than their male counterparts, at least when measured on our three-level scale.

Combined with the fact that higher educational levels are associated with higher earnings (see Section 5.2), the consequence is a diminution of the cross-country gender gap by 1.0 %, in its magnitude clearly exceeding previous results by Simón (2012)¹⁹ and a diminution of national gender pay gaps where men have higher educational endowments. In two countries, Poland and Portugal, the diminution even exceeds 3 %, foremost due to large gender differences in the shares of college graduates. On the other hand, we witness with Germany a case where differences in schooling further nourish the wage gap by 1.6 %. This is due to German men’s lead over women with respect to the employee share with tertiary education which is most pronounced in the European comparison.

As discussed earlier, firm tenure is another predictor for human capital, focusing on firm-specific skills and knowledge, acquired by training-on-the job via staying with the same firm for years. In the aggregate sample, differences in tenure raise the gender pay gap by merely 0.1 pp., which is significantly lower than the 0.5 pp. estimated by Simón (2012) for his dataset of nine European countries. This positive relationship between wages and tenure is confirmed at the country level.²⁰ In line with basic intuition, literature findings (Brown 1989) and descriptive statistics (Section 5.1), longer job tenure is associated with higher earnings in the cross-country regression. This can both be explained by a mechanism of self-selection (higher wages imply higher job satisfaction, thus workers stay longer) and the productivity-enhancing accumulation of job-specific human capital over time (Topel 1991). In our cross-country sample, women exhibit a slightly larger average tenure than men, contradicting the finding of Macpherson and Hirsch (1995) who identify average tenure to be lower in female-dominated occupations. However, we find that gendered endowments vary at country level. The data confirm the second half of our first hypothesis, stating that a lead of women with respect to firm tenure should manifest in a decreasing effect of this characteristic on the gender pay gap for the respective country. An example is Bulgaria, where the impact of tenure on the wage gap is measured to be – 1.2 pp., reflecting a high local share of female workers (43.3 %) with very long tenure of more than 15 years which is particularly pay-attractive. Only 9.4 % of Bulgarian men display such a long firm tenure. An even higher wage decreasing effect of tenure is observed for Lithuania with -3.1 pp. and Latvia with -1.5 pp., supporting the findings of Simón (2012). On the other hand, firm tenure contributes to the overall wage gap with 0.4 pp. in Finland, where 28.7 % of males but only 8.6 % of females exhibit a firm tenure of more than 15 years. A general finding that accords with Simón (2012) is that in CEEC countries, tenure is more likely to be to women’s advantage than in the rest of Europe.

Our second hypothesis addresses the role of atypical work for gendered pay. From Section 5.1 we know that women are more prevalent among part-time workers and also among workers with temporary jobs. From Section 5.2 we learned that part-time work is associated with wage penalties (compared to full-time) in most countries, and the same holds true for temporary compared to permanent jobs. Therefore, it does not come as a surprise that the factor ‘hours of work’ predominantly raises the wage gap, supporting the first part of our second research hypothesis. According to our findings, women’s higher frequency of part-time work contributes to a widening of the cross-country gender pay gap by 1.6 %. This fits recent evidence by Goldin (2014) for the US, who assigns working time arrangements a key role for explaining the incomplete gender convergence on the US labour market. An outlier in our study concerning the magnitude of this effect is Germany, where the part-time effect reaches a level of 5.1 %, the second largest of all measured characteristics effects in this country. This result originates from a dominance of low-scale part-time (as shown in Section 5.1, 37.3 % of German female part-timers assume a job of low-scale part-time), combined with a severe wage penalty of 8.5 % (compared to full-time). Moreover, German women’s part-time affiliation is severely persistent over the life course. 58.3 % of working women with a youngest child at teen age (12+) holds a part-time job, whereas the EU-28 average is 32.1 % (Eurostat 2015). This results in a 34.0 % contribution of part-time work to the German gender gap in accumulated earnings even at age 50 (Boll et al. 2017). If we look at the Netherlands for comparison, the share of female workers in low-scale part-time is even higher (43.1 %), but the wage penalty is much lower (2.6 %) and is not even significant for large-scale part-time, whereas German female workers suffer a 4.9 % penalty in this case. As a result, part-time adds to the Dutch gender pay gap with 3.1 pp.²¹ By contrast, In Hungary and France, where we observed part-time premiums for both genders, part-time work is to the advantage of women, decreasing the pay gap, since more women than men benefit from the premium. Throughout Europe and also in these two countries, women are more likely to work part-time than men. But, as our data show, a more widespread use of part-time does not necessarily correspond to a lower penalty, as the examples for Germany and Latvia show (see for the same result Foster-McGregor et al. 2013: 50). However, in a geographical perspective, CEEC countries stand out with a negative or close to zero contribution of part-time work to the pay gap, what is in line with the findings of Simón (2012) with the same data source eight years earlier. In this context, the sectoral structure of part-time employment also matters. Publicly-controlled sectors sometimes display high shares of part-time employment and rather low part-time penalties, e. g. Education in Poland or Health Services in Sweden (Foster-McGregor et al. 2013: 51).

Another aspect of atypical work is temporary contracts. Like part-time work, temporary work tends to widen the gender gap, confirming the second part of our second research hypothesis. Working in a temporary position reduces the expected earnings (compared to permanent work) in all countries and for both genders except for Latvian men (see Section 5.1). This is consistent with general findings of the literature (Booth et al. 2002). Temporary workers have less incentives to accumulate job-specific human capital, as they face the risk of depreciation when the contract is not prolonged. For the same reason, employers are also less inclined to give them access to internal training. At the same time, temporary jobs are more frequent among women than among men (see Section 5.1). This seems intuitive in the presence of self-selection: facing a higher risk of career interruptions through child birth, women on average are less inclined to commit to a certain career path and employers are for the same reason less willing to offer them permanent positions. Nevertheless, the overall effect remains of low magnitude. In our cross-country sample, temporary work widens the wage gap by only 0.1 pp. (which is identical with the finding by Simón 2012). At country level, the maximum contribution is 0.5 pp. (Finland). Cases where the effect goes in the other direction comprise those countries where the gender distribution of temporary work is reversed. In Poland and Portugal, this implicates a modest reduction in the gender pay gap by 0.2 pp.

Something that can be noticed for all countries is that the selection of male and female workers into different industries contributes to the existence of wage differences, supporting our third research hypothesis. Hence, a significant part of the gender gap is due to the fact that women are over-represented in industries with low pay levels (and accordingly under-represented in well-paid industries). Sector affiliation contributes to the gap in the cross-country sample with 5.2 %. Sector represents the factor with the greatest effect size for the cross-country sample, a finding that is consistent with recent results by Simón (2012) for a smaller sample of the 2002 wave of the same dataset. As shown in the descriptive statistics in Section 5.1, women are particularly over-represented in Education as well as in Health and Social Work Activities. At the same time, they are highly under-represented in Construction and in manufacturing sectors such as Chemical Products, Metal Products or Electric and Transport Equipment.

In a country comparison, the largest effects of sectoral distribution are measured for Romania and Latvia, where its contribution to the overall gender gap amounts to 11.3 pp. and 9.7 pp., respectively. In both countries, the comparatively small presence of women in well-paid jobs in the area of Manufacturing and Construction is again responsible for this result. At the other extreme, there are two countries where the industry effect remains fairly marginal: the Netherlands (0.3 pp.) and Croatia (1.1 pp.). In the Netherlands, manufacturing sectors as well as wholesale trade are an important part of the explanation. Dutch women show a lower participation in these sectors than in the cross-country average. At the same time, these sectors offered, all else being equal, a comparatively low remuneration compared to other sectors in the Dutch economy, a fact that primarily concerned men.²² Related to industry is firm size. The fact that the gender distribution of workers differs with firm size mitigates the wage gap which is consistent with the results by Simón (2012). As descriptive statistics reveal, the share of workers working in firms with more than 50 employees is higher among female than among male workers (except in Bulgaria, Poland and the Czech Republic). From wage regressions, we further know that the payment level in large firms is ceteris paribus higher throughout Europe, a result that is well documented in the labour economics literature (Oi and Idson 1999). Explanations could be the occurrence of productivity gains through a higher division of labour or the need to pay compensating differentials due to the unpleasantness of working in an impersonal atmosphere (Masters 1969). As a consequence, the gender pay gap is reduced by 0.6 pp. in the cross-country estimation. The only conflicting evidence at country level is obtained for Bulgaria, Poland and Greece. However, we have to bear in mind that the negative effect of firm size most likely accrues from the fact that firms with less than ten employees are excluded. Since women are over-represented in very small firms, the effect sign of firm size could reverse if very small firms are taken into account. Analyses based on the EU-SILC data (2013) confirm this suggestion (Boll et al. 2016).

Public control over the firm (which is, by definition, the case when a share of more than 50 % is in public ownership) is another factor that predominantly reduces the gender wage differential, confirming previous results from Arulampalam et al. (2007) and Simón (2012). In all observed countries in our dataset at hand, female workers were over-represented in publicly controlled firms. This result accords with findings of Gornick and Jacobs (1998) and may be explained with attractive employment conditions the public sector offers for mothers, due to the high degrees of protection, time flexibility and tolerance towards periods of absence (Kolberg 1991). At the same time, we find in the majority of countries a higher conditional remuneration in publicly than in privately controlled firms, implying a reduction of the wage gap by 1.1 pp. in the aggregate and up to 2 pp. (Romania) at country level. Gregory and Borland (1999) argue that these differences in wage structure are not surprising given that wage setting in the public sector occurs in a political environment, whereas private-sector decision making occurs in a market environment. Moreover, anti-discrimination legislation may be more aggressively enforced in the public sector. However, the public control premium applies more for men than for women: Whereas male workers receive a premium in 16 out of 22 countries, on the side of women this holds true for around half of all countries. Moreover, countries with a public control penalty for both genders exist. In Finland and the Netherlands, working in a publicly controlled firm implies a wage penalty for both genders, yielding an increase in the gender gap by 2.7 pp. and 2.3 pp., respectively.²³

Workers’ age distribution does hardly impact on the aggregate wage gap, the net effect of age differences is practically zero (−0.1 %). Effects of the single age groups are of a similar magnitude. A look at the wage regressions shows that this is not due to an irrelevance of the factor age in wage setting. Compared to the reference group of 40–49 years old workers, workers in most other age groups are estimated to earn significantly less in the cross-country regression for male workers, reproducing the typical inversely U-shaped wage evolution from the literature (Skirbekk 2004) for the aggregate sample. Rather, differences in the age distribution of male and female workers are simply too small to let this affect the wage gap. However, the practically zero effect for the aggregate sample masks somewhat higher effects in single countries (in line with Simón 2012). The highest contribution of age to the pay gap is observed for Greece and Estonia.²⁴

Finally, the characteristic causing the most heterogeneous effects is occupation. Its contribution to the gender gay gap in the aggregate sample is – 0.9 %. Hence, at the time of observation, women tended to cluster in the better paid occupational groups (from a male perspective). At a first sight, this seems to reject the theories linking occupational segregation to gender pay differences laid out previous Section 2. However, we need to remain cautious with our interpretation, due to several data limitations. First, we merely distinguish between 43 occupational groups, thereby not capturing the full extent of gender heterogeneity in occupational sorting. Second, we can expect a high degree of correlation between occupational choice and sector, up to the point that some occupations are only observed within some sectors. Thirdly, with the occupational classification at hand, it is not possible to adequately control for vertical hierarchy. This is an important point since the different allocation of women and men to hierarchical positions within occupations (and sectors) is a robust finding in the literature (e. g. Bettio/Verashchagina 2009). Last but not least, employment selection matters: in some countries, tasks associated with a female image are still largely executed outside the formal labour market (Bettio 2002).

Referring to these particularities, the moderate effect measured for occupational endowment achieved from the aggregate sample appears a bit less striking, especially since it does not stand out in the literature (cf. Bettio/Verashchagina 2009; Ministère du travail, de l’emploi, de la formation professionelle et du dialogue social 2015 for France). Moreover, the overall effect hides tremendous heterogeneity across countries. In Spain, Estonia and the UK, occupational differences are measured to contribute more than 3.5 pp. to the overall wage gap, implying this to be the prime factor responsible for the existence of a positive explained gap in these countries. Also for Finland, France, Greece, Hungary, the Netherlands and Norway a positive impact is observed. By contrast, we witness a massive negative impact e. g. in Italy and Poland, reaching levels of – 9.6 pp. and – 8.3 pp., respectively, nourishing the result that endowment differences in total work in favour of women. Lithuania displays a negative impact of occupation as well, advantaging women (consistent with the finding of Simón 2012). A detailed analysis of the contributions of the single occupational groups reveals that this is mainly the outcome of a strong concentration of female workers in the group of teaching professionals. In both Italy and Poland, the wage bonus of teaching professionals compared to the reference category is much higher than in the cross-country sample. At the same time, the degree to which women are over-represented in this occupational group is higher than in other countries.

Occupational affiliation has to be interpreted in the context of employment selection. Figure 2 depicts gender gaps in employment rates (employment rate males – employment rate females) together with unadjusted pay gaps in the observed European countries.

Figure 2:

Relationship between gender pay gap and gender employment gap in SES.Sources: Eurostat (2015), SES (2010), own calculations (see Boll et al. 2016).

The pattern documents a clear negative relationship between the two measures: countries with high employment gaps tend to exhibit low pay gaps and vice versa. Poland and Italy obviously belong to the group of European countries with low wage gaps and at the same time comparatively large employment gaps due to low female labour market participation. Apparently, this is a reflection of the fact that some typically low-paid service tasks like nursing and cleaning, which have traditionally been viewed as women’s work, are in these countries to a large part still not delegated through formal work contracts, but mostly executed within households. More generally, the picture supports the notion of a non-random selection of women into the labour market, in the form that women with low returns to paid jobs choose to stay out, thus narrowing the gender gap in average wages (Olivetti/Petrongolo 2008).

Results in Table 1 have shown that the unexplained gap is everywhere positive. Furthermore, it makes up the largest part of the overall gender wage gap in almost all countries under observation (with Germany and Norway marking the exceptions). Given the unavoidable data limitations, this does not come as a surprise. Foremost, this should result from the lack of a measure for actual work experience.²⁵ Endowment effects resulting from these differences are implicitly included in the constant as part of the residual (adjusted) gap. A study with German data shows that gender differences in actual experience account for 5.6 percentage points of the overall gap (Boll/Leppin 2015). As we hypothesize in the beginning of the paper (thesis 5), the constant should be of notable extent for this reason in all countries under investigation. Similar unobserved effects captured by the constant could stem from factors like personal abilities and negotiating skills.

Moreover, the adjusted pay gap might encompass effects of hierarchical and occupational sorting, which cannot be captured by the precision and aggregation level of an occupation measure like ours. As our fourth research hypothesis postulates, we expect that these gender differences show up at the sectoral level, i. e. in different sector premiums of women and men. In what follows, we split up the unexplained part of the gender pay gap to verify our hypotheses. Figure 3 plots the contributions to the unexplained gap at country level. Precise numbers can be found in Table 7 in the Annex.

Figure 3:

Decomposition of the unexplained part of the gender pay gap (in %), 2010.Sources: SES (2010), own calculations (see Boll et al. 2016).

As it is the case for the characteristics effect, also sources of the coefficients effect differ substantially between countries. Nevertheless, some major patterns can be identified. First, industry is estimated to exert a sizeable positive coefficients effect in almost all countries (except Sweden and UK), confirming our fourth research hypothesis. For the aggregate sample, this effect equals 4.8 %. In Belgium, the Czech Republic, Latvia and Romania, the magnitude even exceeds 10 %. That is, in all countries except the two named above, male wage premiums are particularly pronounced in those sectors where female workers are over-represented. This indeed hints at considerable intra-sectoral gender heterogeneity with respect to the sorting into occupations and hierarchical positions. Apparently, much of the sorting takes place within rather than between industries.

The second consistent pattern is the negative sign of the age composition coefficient. It reduces the gender pay gap by 1.0 pp. in the aggregate sample. It is also negative throughout the single country estimations, but not always significant. The negative effect of part-time premiums (decreasing the pay gap) hints to the fact that men suffer higher wage penalties from part-time work than women (see Section 5.2). An exception (for low-scale part-time) is the United Kingdom, resulting in an increasing effect of this factor on the UK gender pay gap. France and Hungary are among the few countries displaying wage premiums for part-time work compared to full-time. Whereas the premiums are higher for men in France (resulting in a positive impact on the French gender pay gap), they are higher for women in Hungary (resulting in a negative impact on the Hungarian gender pay gap).

Finally, the constant term represents a major contributing factor in the majority of countries, which partly confirms our fifth research hypothesis (Bulgaria and Romania are exceptional in this context). With a contribution of 14.8 pp. in the cross-country sample, the constant is almost exclusively responsible for the existence of an unexplained wage variation. It captures the influence of unobserved variables. As discussed earlier, we assume that gender differences in actual work experience over the lifecycle make up the bulk of this amount. The wage-reducing effect of a temporary labour market absence of women due to birth and childcare is nowhere explicitly accounted for in our approach. Moreover, Becker (1985) and Fuchs (1989) speculate that most of the wage gap not attributable to experience is due to unmeasured differences between men and women in their commitment to parenting which points to the importance of gender roles.

However, the striking performance of Bulgaria and Romania as still quite recently acceded EU member states asks for a detailed institutional analysis. This points to the natural limits that large-scale cross-country studies like ours face in disentangling the socioeconomic interrelations behind a certain observation.