A Scandinavian success story? Women’s labour market outcomes in Denmark,Finland,Norway and Sweden

Abstract

In current research, the extensive family policies of the Scandinavian countries have been problematized and described as hampering women’s careers. However, mechanisms have been little investigated and the Scandinavian countries are often regarded as a single policy model. Based on an account of institutional variety we study gender gaps in hourly wages and access to authority positions in Denmark, Finland, Norway and Sweden and explore the importance of segregation, skills and work interruptions. The analysis uses pooled cross-sectional data from the European Social Survey (ESS) for 2004 and 2010. The results show that gender gaps vary both in size and regarding the mechanisms producing them. In particular, we find that gender segregation has a radically different impact in the four countries. The analysis suggests that the mechanisms linking family policies to labour market outcomes are more complex than envisaged in the current debate and point to the importance of comparing seemingly similar countries.

Keywords

Authority gender inequality Scandinavia segregation skill wages work interruptions

Introduction

With their longstanding policies aimed at promoting female employment, the Scandinavian countries have been regarded as something of a social laboratory. Generally, the experiments undertaken in Scandinavia have been positively evaluated, but lately several researchers have questioned the notion of Scandinavia as a paradise of gender equality. Rather, they argue, Scandinavian family policies hamper the economic and occupational achievements of women – especially highly skilled women – by institutionalizing female work interruptions and strengthening occupational gender segregation (e.g., Estévez-Abe, 2006; Mandel, 2012; Mandel and Seymonov, 2006). Seen from this perspective, ‘female-friendly’ policies aimed at facilitating work–family reconciliation appear to have a price – a price paid by women themselves.

However, these broad claims – whether positive or negative – may underestimate the differences among the Scandinavian countries. In scholarly typologies of welfare state institutions (Esping-Andersen, 1990), family policies (Korpi, 2000) and labour market regulations (Hall and Soskice, 2001), the Scandinavian countries are referred to as having the same model or regime, yet in all these aspects there are tangible differences among the countries. Against this background, a comparison of women’s situations across Scandinavia is warranted on both empirical and theoretical grounds. First, it opens up the ‘black box’ of regimes and provides a more nuanced picture, which could prove to be relevant for comparative research on welfare states and labour markets. Second, the approach resembles a ‘most similar cases’ design, which allows us to study more closely the importance of relevant institutional differences. Thus, we can shed light on some of the mechanisms highlighted in the current debate on family policies and gendered labour market outcomes, notably female work interruptions, workplace skill development and gender segregation. Ultimately, such insights may also have policy implications.

In this article we describe central aspects of family policies and labour market institutions in Denmark, Finland, Norway and Sweden and discuss their relevance for labour market inequalities.¹ Based on this account, we investigate gender differences on two indicators – access to authority positions and hourly wages – and examine the impact of skill acquisition, work interruptions and occupational gender segregation. The analysis is based on pooled cross-sectional data from the European Social Surveys (ESS) from 2004 and 2010 (n ≈ 1260 – 1660 per country).

Previous research and our contribution

The tale of Scandinavian gender equality may be told in different ways, but commonly begins with the marvel of female employment. By the mid-1980s, the Scandinavian countries already boasted female labour force participation rates that were high by international standards. Thirty years later, Scandinavia is still ahead, as seen in Table 1. For example, the gender gap in participation varies between 4.2 and 6.6 percentage points, compared to 18 for the OECD average.² Additionally, the high employment rates of Scandinavian mothers are prominently featured in the story of Scandinavian gender equality (see below). In short, all statistics indicate that Scandinavian women have gained a firm foothold in the labour market and that they do not have to sacrifice employment during their child-rearing years to the same extent as women in many other countries.

Table 1.

Four Scandinavian countries – an overview

	Denmark	Finland	Norway	Sweden	OECD average^a / EU average^b
Labour force participation (% of population)¹
Women	76.0	72.5	75.6	76.2	61.7^a
Men	82.6	76.7	80.8	81.8	79.7^a
Maternal employment, %²
Youngest child < 3 years	71.4	51.8	81.1	71.9	52.5^a
Youngest child 3–5 years	77.8	76.0	86.4	81.3	65.6^a
Child care enrolment (% of age-group)³
Children < 3 years	65.7	27.7	54.0	46.7	32.6^a
Children 3–5 years	94.1	73.0	96.2	92.9	80.6^a
Female part-time work (% of employed women)⁴
< 35 hours/week	39.5	19.4	40.9	37.1	29.7^b ⁵
Labour market gender segregation⁶
Duncan and Duncan dissimilarity index	0.44	0.60	0.47	0.53	0.49^b
Long-term employment⁷
Employment protection legislation	2.10	2.38	2.23	2.52	2.04^a

¹ Figures for 2010, ages 15–64. Source: OECD (2014a).

² Figures are from 2010 for Norway and from 2011 for the other countries. Sources: Norwegian Labour Force Survey and OECD (2014b).

³ Figures are for 2010. Source: OECD (2014b).

⁴ Derived from European Labour Force Survey (EU-LFS) 2010; women between 18–66 years working at least 1 hour/week (own calculations).

⁵ Derived from European Labour Force Survey (EU-LFS) 2010 including 30 European countries; women between 18–66 years working at least 1 hour/week (own calculations).

⁶ Based on 22 European countries, see Halldén (2014).

⁷ The figures refer to the strictness of the protection of permanent workers against (individual) dismissal. Higher values indicate more strictness. Source: OECD Employment Protection Database, 2013 update. See also OECD (2013).

The success story of female employment is often attributed to the ‘defamilializing’ design of Scandinavian social policy. In the 1970s, female employment was already supported by a welfare regime that granted social and economic rights to the individual rather than to the family (e.g., Saraceno, 1997) and was further stimulated by generous spending on family policies. Just like the welfare state in general, these family policies had a distinct Scandinavian design. Generously compensated parental leave schemes and public child care, provided on a full-time basis also for children under 3, were the common features of the model, labelled by researchers as the dual-earner/dual-career model (Korpi, 2000; Thevenon, 2011).

As the name suggests, the policies were aimed at achieving gender equality both in the labour market and in the family. To date, however, women’s participation in paid work has not been matched by an equal participation of men in the unpaid work of household chores and child care. In addition, despite women’s unusually high level of integration into the labour market, Scandinavia still reports the usual gender inequalities in wages and career advancement (e.g., Bihagen et al., 2014; Boye et al., 2016).

In light of these persistent gender inequalities, the notion of Scandinavia as a ‘women’s paradise’ has been questioned. Lately, prominent scholars – notably, Mandel and colleagues (e.g., Mandel, 2012; Mandel and Shalev, 2009; Mandel and Semyonov, 2005, 2006) – have argued that Scandinavian family policies hamper women’s occupational and earnings achievements. These policies, particularly defamilialized public child care, bring mothers into the labour market on a large scale and create ‘family-friendly’ jobs in the public sector. Meanwhile, institutionalized rights to work interruptions prompt private employers to discriminate against women and, as a result, men and women are segregated into jobs with different qualities and rewards. Gender differences in Scandinavia may be further exacerbated by labour market institutions that encourage long-term employment, as argued by Estévez-Abe and other scholars in the Varieties-of-Capitalism (VoC) school (e.g., Estévez-Abe, 2005, 2006). These institutions, especially employment protection legislation, promote investments in firm-based skill acquisition but due to expected work interruptions, employers are reluctant to invest in women. In sum, Scandinavian family policies contribute to gender inequalities by facilitating work interruptions in labour markets that require long-term, continuous skill development. Notably, even women who are strongly committed to their careers are penalized, because the entitlements to work interruptions and adaptions make employers reluctant to invest in any woman (statistical discrimination).³ In fact, Scandinavian family policies are depicted as particularly harmful to career-minded women because ‘…for jobs with high training costs, [employers] favor more stable and productive employees’ (Mandel, 2012: 243). According to this line of argument, then, the ‘glass ceiling’ barring women from high positions and top wages would be particularly strong in Scandinavia.

These new arguments are relevant not merely for understanding the situation in Scandinavia. Rather, these theoretical propositions have a wide appeal at a time when women throughout the OECD are catching up with men in terms of human capital investments and when issues of work–family reconciliation are rising on the political agenda. Empirically, however, the mechanisms proposed to link family policies to gender inequalities have not been thoroughly explored. Moreover, the main thrust in the field has been to compare the situation of women in Scandinavia with that of women in other regimes. In the critical perspective, as in the more traditional laudatory accounts, Scandinavia is regarded as a model or an ideal-type rather than a set of countries with distinct features. Here, however, we provide a close-up look at the Scandinavian countries. Because these countries are so similar yet differ in several of the dimensions highlighted in current research, we believe that both empirical and theoretical insights may be gained by comparing their institutions and examining empirically the state of gender equality in Denmark, Finland, Norway and Sweden.

The chosen indicators of labour market (in)equality – gender differences in wages and authority – are well established in international research. While there is a substantial body of research on the gender wage gap, the number of studies comparing wage differentials between men and women in the Scandinavian countries is limited, and the reasons for potential differences within this group of countries are very seldom highlighted or discussed. In general, comparative studies of the gender wage gap show that wage compression and centralized wage setting systems are associated with a smaller gender wage gap (Blau and Kahn, 2003). Furthermore, the gender wage differential tends to be larger in liberal market economies than in countries with generous family policies (e.g., Rubery et al., 1997). In a study comprising 20 countries, Mandel and Semyonov (2005) argued that a centralized wage bargaining system compresses the gender wage gap, while generous family policies exacerbate gender inequalities. However, even though the Scandinavian countries are studied separately, the differences among them are not further explored in the Mandel and Semyonov (2005) study. Furthermore, Mandel (2012) showed that family policy interventions increase the wage gap between men and women in higher socio-economic groups and narrow the gap in lower classes. Nonetheless, neither of these studies explore the importance of the proposed mechanisms of skill investments and occupational segregation.

Gender difference in workplace authority is another matter of obvious interest because it captures several key aspects of power, such as potential influence on wage setting, on-the-job training and promotion. The Scandinavian countries have attracted much attention in comparative research on women’s possibilities for reaching high labour market positions (e.g., Abendroth et al., 2013; Halldén, 2011; Mandel and Semyonov, 2006). Studies present partially conflicting results, for example, regarding the connection between female labour force participation and the gender gap in access to authority positions. In addition, results on the relation between the aggregate share of female part-time work and men’s and women’s access to authority have been contradictory. Halldén (2011) showed that a high share of female part-time work is related to a larger gender gap, while Abendroth et al. (2013) arrived at the opposite conclusion. Claims that occupational gender segregation or long, paid maternity leaves are related to gender differences in authority have received little or no support. Also in this field the Scandinavian countries are often treated as belonging to one regime, and none of the studies referred to above has had the specific aim of comparing the Scandinavian countries internally.

In sum, we find that the mechanisms proposed to connect family policies to labour market gender inequalities have not been thoroughly investigated in existing research on wages and authority. Moreover, the Scandinavian countries have either been lumped together in a single regime or discussed using one country, usually Sweden, as representing the whole group.

In this article we strive to fill some of these lacunas by comparing Denmark, Finland, Norway and Sweden. Based on the reasoning above, we argue that the varying combinations of family policies and labour market institutions affect both the gender gaps and the mechanisms behind these gaps differently in the four Scandinavian countries. The contribution of the article is two-fold. First, we provide an account of how family policies and labour market institutions in these countries differ and discuss them in light of the theoretical arguments above. We describe, hypothetically, how institutional patterns can affect gender differences in wages and authority through the mechanisms of work interruptions, skill investments and gender segregation across occupations and sectors. Second, we test our predictions in multivariate regressions on comparable data from the European Social Survey.

Scandinavia: a model with variations

Unquestionably, the commonalities of the Scandinavian countries are well known. They have been identified as encompassing welfare states based on universal social rights (e.g., Esping-Andersen, 1990) and their family policies are the most generous in the OECD (Thevenon, 2011) and share basic traits, as described above.

Nevertheless, the national family policies of these countries differ in design; and although these within-regime differences may be smaller than differences between regimes, they are theoretically relevant. Varying mixes of parental leave entitlements, cash-for-care schemes and child care services – delivered in different labour market contexts – provide parents with diverse possibilities and constraints. We describe below both family policies and labour market institutions in Denmark, Finland, Norway and Sweden, concluding with predictions about gender inequalities.

Using up-to-date information presented by the national social insurance agencies we have examined parental leaves and cash-for-care systems in the four countries.⁴ Parental leave systems differ in several aspects, such as length, compensation levels and the gender-neutrality of entitlements. After considering the different constructions, we conclude that the maximum potential parental leave a mother can take with income-related compensation – provided that the father takes only what is reserved for him – is approximately 48 weeks in Denmark,⁵ 49 in Norway, 47 in Sweden and 44 in Finland. In Sweden and Norway, parental leave schemes are gender-neutral with symmetrical rights assigned to both mothers and fathers.⁶ This means that a certain period is reserved for the father and an equally long period for the mother, while the rest of the leave can be shared between the parents. In Denmark and Finland, part of the leave can be shared equally but the periods reserved for mothers and fathers (maternity and paternity leave) differ. This means that the maximum potential leave is shorter for fathers than for mothers, approximately 34 weeks in Denmark and 35 weeks in Finland.

In Sweden and Denmark, additional leaves, with a low level of compensation, can be taken (by either parent) for 13 and 14 weeks, respectively. In Finland and Norway, longer leaves at flat-rate compensation are provided within the cash-for-care schemes. Cash-for-care can be used (full-time or part-time) after parental leaves and until the child is 3 (Finland) or 2 (Norway) years old. These systems were introduced in 1985 and 1999, respectively, as parts of the national security systems and have been widely used. In 2008, 57% of Finnish children under the age of 3 were cared for at home under this program (Gislason and Björk Eydal, 2011). In Norway, the program had a very high utilization rate when it was introduced, but as public child care facilities were expanded the rate dropped dramatically, and in 2010 only about 25% of all children were in this scheme (Gislason and Björk Eydal, 2011; Pedersen, 2014). In Sweden and Denmark, cash-for-care schemes have played a marginal role. In both countries, such schemes were introduced in the 1990s but were soon eliminated, only to be reintroduced in the 2000s. However, these schemes are voluntary for local authorities and the uptake has been low (Gislason and Björk Eydal, 2011).

Clearly this brief description of the systems does not allow for precise predictions about female work interruptions in relation to childrearing. The parental leave systems also differ in other aspects, such as flexibility. The greatest flexibility (regarding the right to work part-time and the length of the period during which the parental leave must be taken) is provided in Sweden, while Finland has the least flexible system.⁷ Furthermore, the systems are constantly changing and trends in the different components differ. For example, fathers’ parental leave quotas have been eliminated in Denmark, expanded in Sweden and reduced in Norway. Finally, mothers’ work interruptions are also likely to be influenced by the provision of public day-care services – the main alternative to home-based care in the Scandinavian countries.

All four countries have enacted legislation conferring the right to day care for children at the end of parental leave – a right rarely found in other countries (Gislason and Björk Eydal, 2011). However, the rules differ regarding, for example, the age at which a child can be enrolled in day care.⁸ Additionally, day-care facilities have been expanded at different times in the four countries. In the first decade of the 2000s, child care enrolment for children under 3 more than doubled in Norway – from 25% to 54%. In Denmark and Sweden, the major expansion of child-care facilities took place earlier, but the growth has continued at a slower pace; between 2000 and 2010 the share of children under 3 enrolled in formal child care rose from 56% to 66% in Denmark and from 40% to 47% in Sweden. Meanwhile, in Finland, child care enrolment has remained low (Gislason and Björk Eydal, 2011; OECD, 2014b; Pedersen, 2014). As shown in Table 1, Finland stands out with a level of enrolment that is lower than the OECD average. In the other Scandinavian countries, enrolment is considerably higher, in particular among children under 3. However, for this age group, the level of enrolment is higher in Denmark than in Norway and Sweden.

In all four countries, women take the lion’s share of the parental leave (Duvander and Lammi-Taskula, 2011). Thus, the differences in leave entitlements and provision of child care should, by and large, translate into female labour force interruptions, and statistics on maternal employment indeed suggest such a relation. As seen in Table 1, employment levels for mothers with children below the age of three are lower in Finland than in Denmark, Norway and Sweden. Again, Finland is closer to the OECD average than to its Scandinavian neighbours. In sum, the differences in family policies suggest that female work interruptions would be longer in Finland than in the other countries. Presumably, the difference would be particularly clear vis-à-vis Denmark. Although parental leaves are not shorter in Denmark than in the other countries, early enrolment in day care appears to be more common.

However, both the possibilities of work interruptions and their consequences for gender equality are influenced by labour market institutions. In the literature, factors such as working time regimes, occupational and sectoral gender segregation, and long-term employment relations are seen as problematic features of the Scandinavian model. As with parental leaves, however, the picture needs to be nuanced.

In recent research, the Scandinavian countries have been criticized for institutionalizing part-time work by granting parents the right to cut work hours and, more generally, through the expansion of public sector jobs with ‘family-friendly’ part-time contracts. However, using the common definition of part-time as being below 35 hours a week, we found that part-time work is by no means a Scandinavian phenomenon: the average female part-time rate in Europe is almost 30% (Table 1). The incidence of part-time work in Norway, Denmark and Sweden is higher than in Europe at large; but, as discussed above, the share of economically active women and mothers is also much higher. Again, however, Finland stands out. As in the case of maternal employment, the level of part-time work in Finland is well below that of the other Scandinavian countries. Finally, we found that trends in female part-time work diverge. Over the past two decades, the share of women working part-time (< 35 hours) has fallen dramatically in Norway and Sweden, remained almost constant in Denmark and increased slightly in Finland (Wennemo Lanninger and Sundström, 2014).

As noted above, researchers in the VoC school (e.g., Estévez-Abe, 2005, 2006) have argued that female work interruptions are particularly problematic in Scandinavia, where employment protection legislation and other labour market institutions promote long-term employment relations. Long-term employment makes workplace skill development more important, but such training tends to be biased against women, who are more likely than men to interrupt their work life to care for their children and families (Polachek, 1981). Specifically, women tend to be excluded from (or avoid) occupations and jobs that require significant workplace skill development (Estévez-Abe, 2005, 2006; Polachek, 1981). Thus the long-term character of the Scandinavian labour markets – in combination with their extensive family policies – are together seen as exacerbating gender differences in authority and wages, largely through promoting occupational segregation.

According to VoC, institutions promoting long-term employment relations distinguish the ‘coordinated’ economies of Scandinavia and continental Europe from the ‘liberal’ Anglo-Saxon countries. Empirically, this is confirmed by OECD indices showing that, on average, employment protection is stricter in the former group than in the latter (OECD, 2013). However, there are also differences within Scandinavia. In particular, Denmark is often cited as having a more ‘flexible’ labour market than the other countries (e.g., Viebrock and Clasen, 2009), and VoC scholars have argued that, because employment protection is weaker and unemployment benefits more generous, skills would be more industry-based and less firm-specific in Denmark than in other Scandinavian countries, particularly Sweden (Estévez-Abe et al., 2001). As seen in Table 1, this notion is also reflected in the OECD figures on employment protection.⁹

Occupational gender segregation, i.e. the tendency for men and women to work in different occupations, industries and sectors, is commonly depicted as being high in Scandinavia (Charles and Grusky, 2004). In recent research, however, Scandinavian labour markets do not stand out as extremely gender segregated. In addition, the level of segregation varies among the Scandinavian countries. Using data from the EU Labour Force Surveys (EU-LFS) of 2010, Halldén (2014) compared the levels of segregation in 22 European countries. One measure applied was the Duncan and Duncan Dissimilarity Index (Duncan and Duncan, 1955), which can be interpreted as the share of men or women that would have to change their occupational category in order to achieve a gender-balanced labour market. The results of the study place the Scandinavian countries in the middle of the distribution but also point to the differences among them, as seen in Table 1. In Scandinavia, Denmark has the lowest occupational segregation score (0.44), meaning that 44% of men or women would have to change jobs in order to attain a gender-integrated labour market. The highest score, 0.60, is found in Finland, while Sweden and Norway rank in-between, with scores of 0.53 and 0.47, respectively. The average score of the European countries studied is 0.49, with a variation from 0.22 to 0.62.

In sum, we find that the Scandinavian countries differ in aspects that are central to current theories on gender inequalities in the labour market. Below, we will empirically investigate, first, whether there are systematic differences among the countries regarding the size of the gender gap in wages and access to authority positions; and, second, whether these differences can be explained by the mechanisms discussed in the above mentioned theories, namely, work interruptions, workplace skill development and labour market gender segregation.

Based on the account above, we hypothesize that gender differences in wages and authority should be largest in Finland and smallest in Denmark, with other countries falling in-between. Furthermore, we expect all three mechanisms – work interruptions, workplace skill development and gender segregation – to be most important for explaining the gender gap in Finland, where family policies enable mothers to make longer work interruptions and where labour market gender segregation is stronger than in the other countries. In Denmark, in contrast, gender segregation is weaker and mothers’ work interruptions appear to be shorter, as suggested by the enrolment into day care of small children. Moreover, such interruptions should be less problematic in a context where employment is less long-term. Thus, we expect the impact of the mechanisms under study to be weakest in Denmark. Overall, then, the explanatory power of the analysis, i.e. the reduction in gender gaps, should be strongest in Finland and weakest in Denmark. With regard to both the size of the gender gaps and the mechanisms explaining them, Sweden and Norway should fall between these opposite poles.

Data and method

The analysis is based on data from the European Social Survey (ESS). The ESS is an academically driven cross-national survey that has been conducted across Europe since 2001 and is committed to high standards of rigour and comparability in, for example, questionnaire design, sampling and data collection (for further information see europeansocialsurvey.org). For this analysis we used data from the 2004 and 2010 surveys, which focused on family, work and wellbeing. The sample has been confined to individuals aged 20–65 who were employed and working more than 10 hours a week. To increase the number of respondents, data from the two years has been pooled. The final sample comprises 1258 respondents for Denmark, 1418 for Finland, 1655 for Norway and 1543 for Sweden. Design weights are used in all analyses.

The dependent variables in the regressions are authority and wages. Authority is a binary variable. Employees with a supervisory position and five or more subordinates are coded 1, and other employees are coded 0. This definition was chosen on the presumption that the number of subordinates is an indication of organizational power. Hence, we chose the highest cut off point in terms of subordinates that our data permitted. Wages are measured, in line with convention, as the logged gross hourly wage. Using a logarithmic dependent variable in an OLS regression, a change of one unit in the independent variable produces a percentage change in the dependent variable (Allison, 1999). The following estimation is used to calculate percentage change: 100(exp(b)-1) (Ibid.). The wage variable for respondents in 2004 was converted into 2010 values, using the consumer price index of each country, and all values were converted into Euros.

As discussed, the main thrust of the analysis is to examine the importance of three mechanisms proposed to link family policies and labour market institutions to gender inequalities, namely work interruptions, workplace skill development and gender segregation. OLS regressions and, for authority, linear probability models (LPM) are performed separately for the four countries, with a stepwise inclusion of the independent variables.

In the first model, we examine the unadjusted effects of gender. In model 2 we control for parenthood, measured as ever having had a child, which is regarded here as a rough proxy for work interruptions. The motivation for using this proxy is that in the Scandinavian countries, with their extensive parental leave systems, traditional indicators of work experience and employer tenure may not reflect work interruptions as well as in countries where women leave the labour force during the child-rearing years. Obviously, examining the impact of parenthood is also of descriptive interest. Moreover, it should be noted that the theoretical arguments above suggest that in Scandinavia the effects of work interruptions are largely anticipated and generalized to affect all women. If this is the case, parenthood and actual work interruptions should not explain much of the gender gaps.

In models 3 and 4 we include skills, operationalized as years of education and initial on-the-job training, OJT. OJT captures skill acquisition in the workplace and is determined using the question ‘Apart from the competence necessary to get a job such as yours, how long does it take to learn to do the job reasonably well?’ (response categories are recorded as number of months, with a top code of 72 months). This variable captures the initial training requirements inherent in the job, i.e. the basic skill development that comes with the job. Theoretically, on-the-job training is depicted as a factor disadvantaging women in Scandinavian labour markets. The variable used here is well established in previous research, and it has been shown that women have jobs involving less initial on-the-job training – a factor contributing to the gender wage gap (e.g., Grönlund, 2012; Tomaskovic-Devey and Skaggs, 2003).

To supplement our rough parenthood proxy we add, in model 5, a range of variables intended to capture work interruptions and adaptions. Work experience, work experience squared and employer tenure are variables traditionally used to estimate skill development at work. Indirectly, they should also measure work interruptions, but as discussed previously, these indicators may work less well in Scandinavia due to the parental leave systems. Unfortunately, the ESS includes no data on parental leaves. As a proxy we use number of children to capture the number of parental leave periods, although we acknowledge that we do not know the amount of parental leave used for each child. For regressions on authority, the model also includes part-time work, defined as working fewer than 35 working hours a week.

Finally, in model 6 we explore the importance of labour market gender segregation. In line with prior wage research (e.g., England, 1992; Tam, 1997), occupational segregation is measured as per cent F, the percentage of females in the respondent’s occupation. Information on the gender composition of occupations was obtained from the EU Labour Force Surveys (EU-LFS), with pooled data from 2004 and 2010 estimated separately for each of the four countries included in our study and measured through a three-digit occupational classification (ISCO-88). Considering the theoretical arguments, public sector employment may also be relevant when discussing segregation in the Scandinavian countries. Unfortunately, sector was not included in ESS 2004. We include instead a dummy variable indicating whether the respondent is working in the welfare industry, which we use as a proxy for public sector.¹⁰

Results

The aim of the analysis was to examine and compare gender differences in workplace authority and wages in the four countries and explore the impact of mechanisms central to theory and debate. First, we present descriptive statistics of the sample, then we report the results from multivariate regressions carried out separately for the four countries.

The descriptive statistics in Table 2 indicate that in all four countries, women are disadvantaged relative to men in terms of both wages and authority. The hourly wage gaps range between €2.4 and €4.3 and women’s access to authority position varies considerably across the countries. The smallest gender gap is found in Denmark, where approximately 20% of the employed women and 26% of the men hold a managerial position, and the largest gap appears in Finland, where only 7.5% of the women occupy such positions, compared to approximately 20% of the men. There are no evident gender differences in average education and tenure (although Norwegian and Swedish women tend to have slightly higher educational attainments than men). However, men have jobs involving significantly more initial on-the-job training than do women, and Danish men have more labour market experience than women. In all countries but Finland, more women than men are parents, and unsurprisingly, women in all countries more frequently work part-time, in a female-dominated occupation and in the welfare industry (our proxy for public sector).

Table 2.

Descriptive statistics (percentages and averages); employees 20–65 years working > 10 hours/week.

	Denmark		Finland		Norway		Sweden
Sample size	48.9		51.7		44.9		49.5
Women (%)	1258		1418		1655		1543
	Women	Men	Women	Men	Women	Men	Women	Men
Managerial responsibility (%)	20.5	26.0**	7.5	19.6***	19.7	30.0***	11.4	22.7***
Hourly wage in Euros (average)	24.2	27.6***	14.4	18.0***	22.3	26.6***	15.4	17.8***
Years of education (average)	14.7	14.4	14.6	14.3	14.7	14.2*	13.9	13.2***
Initial OJT (average)	9.2	11.3**	8.0	14.8***	7.3	12.1***	6.7	11.2***
Labour market exp. (average)	22.8	24.8**	20.7	20.7	20.7	21.7	21.2	21.4
Tenure (average)	9.3	9.6	11.9	10.8^†	11.3	10.7	10.3	10.5
Part-time among women and men (%)	29.8	4.0***	15.4	5.8***	26.6	6.9***	27.5	6.7***
Parent	82.4	77.8*	71.9	71.8	82.5	75.0**	78.5	70.1***
Number of children	1.0	0.9	0.8	0.9	1.0	1.0	0.9	0.9
% women in the occupation	62.4	33.8***	68.1	33.3***	61.9	35.8***	62.8	35.8***
Welfare industry (average)	61.1	21.3***	52.5	18.8***	55.9	20.9***	55.6	21.7***
N	615	643	733	685	743	912	764	779

^† p < 0.10, * p < 0.05, ** p < 0.01, *** p < 0.001. The stars in the right country column indicate significant gender differences within each country.

Next, we consider our multivariate regression analyses. The findings for authority are reported in Figure 1 (see Tables 3A–3D in the Appendix for full data). First, we note that in all four countries there is a statistically significant gender difference in authority, to women’s disadvantage (model 1). The gender gap in authority is smallest in Denmark. The size of the gender gap is fairly similar for the other countries, though somewhat larger in Finland.

Figure 1.

The gender authority gap (to women’s disadvantage) model 1–6 (left to right) by country. The figure reports beta coefficients for gender (women = 0, men = 1) from linear probability regressions with “managerial responsibilities” as dependent variable. M is an abbreviation of Model with Model 1 including only gender; Model 2: M1 + parenthood; Model 3: M2 + education; Model 4: M3 + initial on-the-job training; Model 5: M4 + indicators of work interruptions; Model 6: M5 + segregation. Sources: Tables 3A, 3B, 3C, 3D.

In all countries, the gender gap increases or remains stable when we control for parenthood (model 2). This indicates that the gaps in authority could not be explained by parenthood per se. The effects of parenthood are positive and significant. To examine whether the effect of parenthood is gendered, we have also carried out analyses including the interaction term women*parent. The results (available upon request) show that in Sweden the positive effect of having children is larger for men than for women. In Finland, parenthood has a negative effect on women’s access to authority positions but has a positive effect for men. In Denmark and Norway, the interaction between gender and parenthood is not significant.¹¹

In model 3 we see that the gender gap in authority cannot be ascribed to gender differences in education, as the gap in authority increases in all countries when accounting for years of education. However, in all countries but Sweden a small part of the gap is explained by women’s lower rate of access to on-the-job training (model 4). In model 5 we find that our indicators on work interruptions and adaptions – i.e. experience, tenure, part-time work and number of children – explain a sizeable part of the gender gap in authority in Norway and Sweden and even more so in Denmark, where the gender coefficient becomes non-significant. In Finland, in contrast, the impact of these variables is modest. The analyses further indicate (see the coefficients in Table 3A–3D) that the difference between countries is due mainly to the varying importance of part-time work. Finally, model 6 shows that when segregation across occupations and sectors is accounted for, the gender gap clearly decreases in Finland and to some extent also decreases in Norway and Sweden.

In sum, we find that parenthood as such does not explain any of the gender gap in authority. However, actual work interruptions and adaptions – mainly part-time work – explains the gender gap in authority in Denmark and provide an important mechanism in Norway and Sweden, but not in Finland. Gender segregation, by contrast, appears to be particularly important in Finland and also to some extent in Sweden and Norway.

Figure 2 displays the regression results for wages (see Table 4A-4D in Appendix for full tables). As the figure shows, there is a significant gender gap in logged hourly wages in all four countries. The unadjusted wage gap (model 1) amounts to approximately 14.0%, 17.6% and 17.2% in Denmark, Finland and Norway, respectively. In Sweden, the unadjusted gap is lower, approximately 12.1%.

Figure 2.

The gender wage gap in percent (to women’s disadvantage) model 1–6 (left to right) by country. Table estimates converted to percent by 100(exp(b)–1). The figure reports beta coefficients converted to percent for gender (women = 0, men = 1) from OLS regressions with “wage” as dependent variable. M is an abbreviation of Model with Model 1 including only gender; Model 2: M1 + parenthood; Model 3: M2 + education; Model 4: M3 + initial on-the-job training; Model 5: M4 + indicators of work interruptions; Model 6: M5 + segregation. Sources: Tables 4A, 4B, 4C, 4D.

The gender wage gaps either increase or remain unaffected when controlling for parenthood in model 2. Further analyses, including woman*parent interaction terms (available upon request), show that the interaction is negative and significant in all countries except Norway. In Finland and Sweden parenthood has a positive wage effect for both men and women, but it is greater for fathers than for mothers. In Denmark there is a slight wage penalty for mothers, while there is a fatherhood premium for Danish men.¹² As with authority, the gender wage gaps tend to increase when accounting for education (model 3). This pattern is found in all four countries, but is somewhat more pronounced in Norway and Sweden. When on-the-job training is entered (model 4), the gender wage gap decreases substantially in Finland and more marginally in the other countries. Accounting for work interruptions and adaptions (model 5) does not explain the wage gap further; in fact the gaps increase slightly in all countries but Sweden. However, occupational segregation and sector of employment explain a large part of the gap in Finland. Here, the gender wage gap decreases from approximately 17% to 11% from model 5 to model 6. In Sweden and Norway, the gap narrows by approximately 4 percentage points and in Denmark the coefficient is basically unaffected.

Overall, the variables included in this analysis explain better the gender wage gap in Finland than in Sweden and Norway and – particularly – Denmark, where the gap is actually higher in model 6 than in model 1. In the final model, rather large gender wage gaps still remain in all countries. This indicates that the major part of the gender wage gap cannot be ascribed to the mechanisms tested here.

To examine whether the size of the gender gaps in authority and wages varies significantly among the countries, we have examined the gender*country interactions in pooled regressions for the first and the last models using clustered standard errors to account for the fact that individuals are nested in countries (Wooldridge, 2009). We also interacted all control variables in model 6 with country dummy variables to allow effects to differ across countries (Ibid.). The results from these pooled regressions (see Tables 5A– 5D in the online Appendix) are in line with the overall results reported here.

Discussion

The aim of this paper was to nuance the common picture of Scandinavian gender equality by comparing gender differences in wages and workplace authority in Denmark, Finland, Norway and Sweden. Such an investigation was warranted by the fact that, despite their similarities, these countries display differences in both family policies and labour market institutions. Presumably, then, a comparison of these countries – frequently regarded as a common model or regime – could further our understanding of the mechanisms put forward in the current debate on policies and gender inequalities in the labour market.

We predicted that the gender gaps would be most apparent in Finland and smallest in Denmark. Furthermore, we anticipated the mechanisms under scrutiny to be of greatest importance in Finland, where family policies enable mothers to take longer work interruptions and where segregation is stronger than in the other countries. In Denmark, we expected the explanatory value of these factors to be lower than in the other countries, both because family policies and labour market institutions make work interruptions less problematic and because segregation is weaker. With regard to both the size of the gender gaps and the mechanisms explaining them, we predicted that Sweden and Norway would fall between these two poles.

Our results show that in all four countries women are clearly disadvantaged in terms of wages and access to authority positions. In line with our hypotheses Finland has a larger raw gender gap in authority than the other countries. Denmark displays the smallest gap and Norway and Sweden place themselves in the middle. However, the pattern changes somewhat when comparing the gender gaps accounting for all control variables, but Denmark still stands out with a low and now insignificant difference between men and women. The raw gender gap in wages is lowest in Sweden, followed by Denmark, while it is somewhat larger in Norway and largest in Finland. The adjusted gender wage gap is also lowest in Sweden but the largest gap is now found in Norway.

With regard to the mechanisms, we note that in Finland occupational and sectoral gender segregation has a strong impact on gender gaps in both authority and wages. For the gender wage gap, gender segregation is also of significance in Norway and Sweden. For Denmark, none of the mechanisms explored had any effect on the gender wage gap. For authority, part-time work explained the gender gap in Denmark and was also important in Norway and Sweden. This was not the case in Finland, where the working-time regime differs from that of the other countries.

Regarding female work interruptions, our predictions were not confirmed, because experience, tenure and parenthood – all proxies for interruptions related to childrearing – did not affect gender gaps to any great extent in any of the countries. However, we note that the limited effect of parenthood is in line with previous Scandinavian wage research. While several studies, based mainly on data from the USA, report a negative relationship between wages and motherhood, labelled the ‘motherhood wage penalty’ (e.g., Gough and Noonan, 2013), studies using data from Finland and Norway (Gash, 2009; Petersen et al., 2014) report no or only a small motherhood wage penalty. Recent Swedish studies have shown that there is, in fact, a wage premium for being a parent, although this premium is significantly higher for men than for women (e.g., Bihagen et al., 2014). Hence, our results corroborate previous findings suggesting that different processes are at play in Scandinavia, with its extensive family policies, compared to the USA where such policies are conspicuously absent.

Finally, it is worth noting that for all four countries, rather large gender wage gaps remain even in the final model. In addition, except for Denmark, the gender gaps in authority remain to a large part unexplained after accounting for occupational gender segregation, parenthood and work interruptions.

With the analysis presented here, we make three main contributions. First, we provide a theoretically driven description and discussion of differences in family policies and labour market institutions across Scandinavian countries, commonly regarded as one regime. Second, we show empirically that gender gaps in wages and authority vary across these countries, both in size and in terms of the mechanisms producing them. In particular, we find that occupational and sectoral gender segregation has a radically different impact on gender inequalities in the four countries. This is also the case for part-time work. In Finland, on-the-job training influences the gender gap, especially in wages but also in authority. On-the-job training is of more marginal importance in the other countries. Clearly, comparative researchers with the intent of making regime-based comparisons should reflect on these empirical differences in institutions and outcomes. However, our analysis also has theoretical implications. The links between extensive family policies and institutions promoting long-term employment on the one hand, and gender inequalities on the other, seem less obvious when we consider the empirical differences among Scandinavian countries.

For the future, we argue that these links should be explored further in comparative research using a larger set of countries, with differences among and within regimes considered simultaneously. Although in principle we agree with Mandel and colleagues that welfare state interventions could be regarded as policy ‘packages’ because they have ‘shared effects that cannot be detached from one another, either theoretically or empirically’ (Mandel and Shalev, 2009: 1879), our analysis points to the importance of unpacking these packages. To understand mechanisms, specific policies should be evaluated in context, not only because work-family strategies are formed at the interface of several institutions but also because, over time, policies will have feedback effects on norms and expectations. In the Scandinavian countries, where dual-earner policies and practices date back to the 1970s, norms of gender equality are widespread (Edlund and Öun, 2016) and, presumably, these norms have also affected employers’ attitudes and the way work is organized.

Our study has several limitations. Due to the restricted availability of comparative statistics on relevant policy entitlements and their usage, our account of institutions is sketchy and explorative. There is a need here for additional data collection based on theoretical reflections on how family policies and labour market institutions interact to affect female work interruptions and the consequences of such interruptions. Clearly, there are several similarities among the countries (e.g., regarding the universal nature of social policies) as well as a range of differences (e.g., in labour market structures, educational systems, wage bargaining institutions and family cultures) that may be relevant for specifying mechanisms further. Regarding our empirical analysis, we note that while the harmonized ESS data ensure comparability in terms of measures, they impose the drawback of small sample sizes. Thus, although the theoretical arguments mainly address the situation for highly skilled women, the sample sizes did not allow us to distinguish between high- and low-skilled individuals. We have not had access to direct information on central factors such as parental leaves and public sector employment; and because the rate of maternal employment varies within Scandinavia our results could be affected by selection into the labour market. However, a deeper exploration of these mechanisms, including issues of selectivity, requires longitudinal designs.

The persistent disadvantages facing women in Scandinavian countries – despite their strong promotion of gender equality – undoubtedly constitute a great challenge for policymakers as well as researchers. Nevertheless, with our detailed examination of these countries, we hope to provide some clues about how to tackle them.

Footnotes

Acknowledgements

We thank participants at the LNU-seminar of the Swedish Institute for Social Research, especially Roujman Shahbazian and Michael Tåhlin, the participants of the Work and Family Researchers Network conference in New York in June 2014 and the anonymous reviewers for helpful comments.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: We gratefully acknowledge financial support from the Swedish Research Council (Vetenskapsrådet, VR), grant number 2010–2083 and 2013–1690 and from the Swedish Research Council for Health, Working Life and Welfare (FAS, currently FORTE) grant numbers 2008–0575 and 2011–0816.

Notes

Appendix

Full data are presented for the Linear Probability Model (LPM) (Tables 3A, 3B, 3C and 3D) and OLS regressions (Tables 4A, 4B, 4C and 4D) for each of the four Scandinavian countries studied. The pooled regressions (Tables 5A, 5B, 5C and 5D) are available in the online Appendix.

References

Abendroth

A-K

Maas

van der Lippe

(2013) Human capital and the gender gap in authority in European countries. European Sociological Review 29: 261–273.

Allison

(1999) Multiple Regression. A Primer. Thousand Oaks, CA: Pine Forge Press.

Bihagen

Nermo

Stern

(2014) The gender gap in the business elite: Stability and change in characteristics of Swedish top wage earners in large private companies, 1993–2007. Acta Sociologica 57: 119–133.

Blau

Kahn

(2003) Understanding the international differences in the gender pay gap. Journal of Labour Economics 21: 106–44.

Boye

Halldén

Magnusson

(2016) Stagnation only on the surface? The implications of skill and family responsibilities for the gender wage gap in Sweden, 1974–2010. British Journal of Sociology, forthcoming.

Charles

Grusky

(eds) (2004) Occupational Ghettos. The Worldwide Segregation of Women and Men. Stanford, CA: Stanford University Press.

Duncan

(1955) A methodological analysis of segregation indexes. American Sociological Review 20: 210–217.

Duvander

Lammi-Taskula

(2011) Parental leave. In Gislason

Björk Eydal

(eds) Parental Leave Childcare and Gender Equality in the Nordic Countries. TemaNord, Copenhagen: Nordic Council of Ministers, p. 562.

Edlund

Öun

(2016) Who should work and who should care? Attitudes towards the desirable division of labour between mothers and fathers in five European countries. Acta Sociologica 59: 151–169.

10.

England

(1992) Comparable Worth. Theories and Evidence. New York: Aldine de Gruyter.

11.

Esping-Andersen

(1990) The Three Worlds of Welfare Capitalism. Cambridge: Polity Press.

12.

Estévez-Abe

(2005) Gender bias in skills and social policies: The varieties of capitalism perspectives on sex segregation. Social Politics 12: 180–215.

13.

Estévez-Abe

(2006) Gendering the varieties of capitalism: A study of occupational segregation by sex in advanced industrial countries. World Politics 59: 142–175.

14.

Estévez-Abe

Iversen

Soskice

(2001) Social protection and the formation of skills: A reinterpretation of the welfare state. In Hall

Soskice

(eds) Varieties of Capitalism: The Institutional Foundations of Comparative Advantage. Oxford: Oxford University Press.

15.

Gash

(2009) Sacrificing their careers for their families? An analysis of the penalty to motherhood in Europe. Social Indicators Research 93: 569–586.

16.

Gislason

Björk Eydal

(2011) Parental Leave, Childcare and Gender Equality in the Nordic Countries. TemaNord Copenhagen: Nordic Council of Ministers, p. 562.

17.

Gough

Noonan

(2013) A review of the motherhood wage penalty in the United States. Sociology Compass 7: 328–342.

18.

Grönlund

(2012) On-the-job training – a mechanism for segregation? Examining the relationship between gender, occupation, and on-the-job training investments. European Sociological Review 28: 408–420.

19.

Hall

Soskice

(2001) Varieties of Capitalism. The Institutional Foundations of Comparative Advantage. New York/Oxford: Oxford University Press.

20.

Halldén

(2011) What’s Sex Got To Do With It? Women and Men in European Labour Markets. Swedish Institute for Social Research Dissertation Series Number 85. Stockholm University, Sweden.

21.

Halldén

(2014) Könssegregering efter yrke på den svenska arbetsmarknaden år 2000–2010. In Kunze

Thorburn

(eds) Yrke, Karriär och Lön - Kvinnors och Mäns Olika Villkor på den Svenska Arbetsmarknaden, Forskningsantologi till Delegationen för jämställdhet i arbetslivet, SOU, Stockholm: Fritzes, p. 81.

22.

Korpi

(2000) Faces of inequality: Gender, class, and patterns of inequalities in different types of welfare states. Social Politics 7: 127–191.

23.

Mandel

(2012) Winners and losers: The consequences of welfare state policies for gender wage inequality. European Sociological Review 28: 241–262.

24.

Mandel

Semyonov

(2005) Family policies, wage structures, and gender gaps: Source of earnings inequality in 20 countries. American Sociological Review 70: 949–967.

25.

Mandel

Semyonov

(2006) A welfare state paradox: State interventions and women’s employment opportunities in 22 countries. American Journal of Sociology 111: 1910–1949.

26.

Mandel

Shalev

(2009) Gender, class, and the varieties of capitalism perspective. Social Politics 16: 161–181.

27.

OECD (2013) Employment Outlook. Paris: OECD.

28.

OECD (2014a) Labour market statistics – Labour force statistics by sex and age: indicators. OECD Employment and Labour Market Statistics (database). http://stats.oecd.org/BrandedView.aspx?oecd_bv_id=lfs-data-en&doi=data-00310-en (accessed 10 December 2014).

29.

OECD (2014b) OECD Family Database. Available at https://www.oecd.org/els/family/database.htm (accessed 29 October 2014).

30.

Pedersen

(2014) Med Staten i Ryggen – Institusjonelle Rammebetingelser og Unge Voksnes Overveielser om å Få Barn. Doktorsavhandling vid en samfunnsvitenskapelige fakultet No 490. Oslo: Universitetet i Oslo.

31.

Petersen

Penner

Hogsnes

(2014) From motherhood penalties to husband premia: The new challenge for gender equality and family policy, lessons from Norway. American Journal of Sociology, 119:1434–1472.

32.

Polachek

(1981) Occupational self selection: A human capital approach to sex differences in occupation structure. Review of Economics and Statistics 58: 60–69.

33.

Rubery

Bettio

Fagan

Maier

Quack

Villa

(1997) Payment structures and gender pay differentials: Some societal effects. The International Journal of Human Resource Management 8: 131–149.

34.

Saraceno

(1997) Family change, family policies and the restructuring of welfare. In OECD (publisher) Family Market and Community: Equity and Efficiency in Social Policy. Social Policy Studies Number 21. Paris: OECD.

35.

Tam

(1997) Sex segregation and occupational gender inequality in the United States: Devaluation or specialized training? American Journal of Sociology 6: 1652–1692.

36.

Thevenon

(2011) Family policies in the OECD countries: A comparative analysis. Population and Development Review 37: 57–87.

37.

Tomaskovic-Devey

Skaggs

(2003) Sex segregation, labour process organization and gender earnings inequality. American Journal of Sociology 108: 102–128.

38.

Viebrock

Clasen

(2009) Flexicurity: A State-of-the-Art Review. Working Papers on the Reconciliation of Work and Welfare in Europe 01/2009. Dissemination and Dialogue Centre, Edinburgh. Available at: http://www.sociology.ed.ac.uk/__data/assets/pdf_file/0013/31009/REC-WP_0109_Viebrock_Clasen.pdf (accessed 12 July 2016).

39.

Wennemo Lanninger

Sundström

(2014) Part-time work in the Nordic region: Part-time work, gender and economic distribution in the Nordic countries. TemaNord 2014:503. Available at: http://norden.diva-portal.org/smash/get/diva2:707663/FULLTEXT01.pdf (accessed 12 July 2016).

40.

Wooldridge

(2009) Introductory Econometrics. A Modern Approach, fourth edition. Mason, OH: South-Western Cengage Learning.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.22 MB