EurWORK European Observatory of Working Life
Report examines gender pay gap
The annual report of the Swedish National Mediation Office, issued in February 2004, includes a special examination of the statistical evidence on pay differentials between women and men. The official wage statistics indicate that on average women's pay is around 82% of men's. However, weighting the statistics for factors such as occupation, age, working time, education and sector brings women's average pay to 92% of men's.
The third annual report of the National Mediation Office (Medlingsinstitutet), set up in 2000 (SE0105195F), was presented in February 2004. The report concentrates as usual (SE0203105F) on bargaining and wage formation in the previous year, in this case 2003. However, bargaining was generally scarce in 2003, as the three-year collective agreements signed in most sectors in 2001 (SE0105102F) do not expire until spring 2004, when the next 'big' bargaining round starts. However, the National Mediation Office has taken the opportunity to look in more detail in its report at some of the other issues supervised by the Office. One of these is equal opportunities, and specifically the question of the gender wage gap and the interpretation of the official wage statistics on this point. The report examines what information from these statistics may be used when examining 'non-objective' wage differences between women and men in contravention of the Equal Opportunities Act (Jämställdhetslagen, SFS 191:433).
In the Swedish public debate, it is often stated that women’s wages are about 18% lower than men’s, and have been so for at least the last 10 years. This assertion is in itself correct, it it is borne in mind that the relevant information from the official wage statistics concerns the average gender wage differential in the whole labour market. In this context, all wages are recalculated on the basis of full-time employment. However, looking at the wage differentials within broad sectors of the labour market, the picture is different. The largest wage differential in Sweden in 2003 was in the city council sector, where women’s wages amounted on average to around 69% of men’s. The smallest differences was found in the municipal sector, where women’s average wages were 90% of men’s. For the private sector and central government sector the difference was the same - female employees' wages were on average 84% of men's.
Factors in gender wage gap
The gender wage gap depends on many different factors, the National Mediation Office points out. These include occupation, age, working time, education and the part of the labour market where the job is carried out. In order to draw up a more nuanced picture of the gender wage situation, these variables must be taken into account. This can be done through a so-called 'standard weighting' (standardvägning) of the different factors. For example, this method is used in the Statistics Sweden (Statistiska Centralbyrån, SCB) Wage Statistics Yearbook 2002 (Lönestatistisk Årsbok 2002). Instead of arriving at the normal wage differential of 18% across the whole labour market, this method of calculation put the overall, difference at only 8%. By sector, the differential is reduced to: 1% the municipal sector; 8% in the city council sector and central government sector; and 10% in the private sector it is 10%. For all sectors, this method puts women's average pay at of 92% of men's (in both 2000 and 2001). Without weighing the factors, the figures are 82% (2000) and 83% (2001).
The most important factor explaining the largest part of the wage differential is found to be 'occupation'. The current Swedish statistical classification of occupations has proved to be too crude, and 355 sub-groups or occupations will be added to the current 114 groups. This work, being carried out by Statistics Sweden, should be completed in 2004. A more refined classification of occupations will give a better picture of the wage differentials between women and men, the National Mediation Office states. The differences will, however, still not be fully explained as, for example, information about organisational responsibilities and professional experience still will be lacking.
The example of the graphical sector
There are currently two collective agreements between the Graphical Workers’ Union (Grafiska Fackförbundet) and the Swedish Graphic Companies' Federation (Grafiska Företagens Förbund) employers' association. The 'packaging agreement' (Förpackningsavtalet) applies to the packaging, wrapping and wallpaper industry. The 'book house agreement' (Civilavtalet) applies to the graphical industry. The two agreements between then cover 17,000 employees. Women make up 20% of workers covered by the packaging agreement and about 25% of those covered by the book house agreement. The social partners asked the statistical department of the Confederation of Swedish Enterprise (Svenskt Näringsliv) to analyse wage differences between the female and male employees covered by the two pay agreements. The analysis was carried out using September 2002 wage information for the workers concerned, and 319 companies with 12,359 employees were surveyed. In December 2003, the findings were presented.
'Standard weighting' of different factors influencing pay was used in the wage analysis. In the area covered by the packaging agreement there are 72 occupations, while in the graphical industry there are 50. Before the weighting exercise, the wage differential between women and men was 6% in packaging and 9% in the graphical industry to the advantage of the male workers. Counting all the factors, the final result of the analysis showed that the women’s wages were 98% of men’s in the graphical industry and 97% in the packaging industry.
Effects of Equal Opportunities Act amendments
In October 2002, the newly formed minority government (SE0210102F) of the Social Democratic Party (Socialdemokratiska Arbetarepartiet, SAP) government, along with its cooperation partners the Left Party (Vänsterpartiet) and the Green Party (Miljöpartiet de Gröna), decided (among '121 points for a safer, fairer and greener Sweden') that the Equal Opportunities Act should be overhauled in 2004, three years after its last amendments in 2001 (SE0102179N). These amendments included a new legal definition of 'work of equal value' and an obligation on employers to conduct an annual wage survey.
With this in mind, the National Mediation Office decided during 2003 to conduct a limited study as to whether the 2001 amendments of the Act have had an effect that can be detected in the wage statistics. Statistics Sweden carried out the study, which encompassed parts of the government, municipalities and city council sectors. The study investigated if there had been any changes in wage differentials between women and men that have occurred on the legal grounds of 'work of equal value'. The 2000-2 statistics for 100 municipalities, 20 city councils and 13 governmental areas were examined. The finding of this 'pilot' survey is, in brief, that it is not possible to show that wage differentials between men and women have diminished as a consequence of the changes in the Equal Opportunities Act. Women still have on average lower wages than men. There are reasons for a large part of the difference. The main explanation is that men in the public sector work in occupations that are more highly valued. Men dominate among doctors, for example, who have the highest salaries. The occupation of speech therapist is dominated by women, and has the lowest wage level.
Generally, there are no evident signs of wage discrimination within occupations in the public sector, the National Mediation Office states. Women and men mainly have the same wage in the same occupation. Gender wage differences often receive public attention in terms of the differences between different occupations. For example, in Sweden low-paid female-dominated occupations have often been compared with better-paid male-dominated occupations. In order to show gender wage discrimination in the legal sense it is required that the work tasks compared may be seen as being of equal value and that they be carried out at the same employer. Statistics are of no help when it comes to the question of whether two jobs are of equal value or not (SE0103187F).
The Swedish labour market is gender segregated. Men most often work in the private sector, while women workers are evenly divided between the private and the public sector. They dominate especially in the municipal and city council sectors, where they make up 80% of the workforce. There were about 3.8 million employees in Sweden in 2003. Of these, about 2.5 million worked in the private sector (65%). About 1.1 million (29%) worked in the municipal and city council sectors, and a little more than 220,000 worked in the central government sector (6%).
The National Mediation Office is the central government body responsible for mediating in labour disputes. It is also itself part of the wage formation system, being required to promote an efficient wage formation process. A part of this latter task is looking into gender discrimination matters.
Sweden’s official wage statistics show the wage levels for women and men in various sectors of employment. Over the past 10 years, the gap between the sexes has remained largely unchanged, the National Mediation Office states in latest annual report. The differences primarily reflect gender segregation in the labour market and the fact that jobs traditionally dominated by women are lower paid. The statistics do not show whether the pay gaps are unwarranted or discriminatory as defined by the Equal Opportunities Act. This is because the official statistics and the Act operate at different levels. The statistics show the differences in pay between the various occupational categories (roughly) and across whole sectors but not pay relations at individual workplaces. The aim of the Equal Opportunities Act is to eliminate unwarranted differences in pay between employees performing equal work or work of equal value for the same employer. (Annika Berg, Arbetslivsinstitutet)