Sweden: New research on the gender pay gap
Men’s wages are still higher than women’s in Sweden. But the gender wage gap is shrinking, and over the past decade the gap has decreased by 3.1 percentage points, reaching a record low of 13.2%. Around one-third of the decrease is due to rises in pay, and the rest to the altered composition of the workforce. Collective agreement type was also recently found to have no significant effect on the gap.
Although there is general consensus in Sweden agrees that the gender pay gap is a serious problem, employers, trade unions and the government tend to disagree on how to close it. The government, which generally has a very limited role in wage-related issues, appointed a delegation in 2011 to investigate gender equality in working life with a special focus on pay. The delegation’s final report on equality in working life (in Swedish, 4 MB PDF) was published in June 2015. The Swedish National Mediation Agency (MI) also published two reports on the same topic, both in the summer of 2015. The first report aims to explain the decrease of the wage gap (in Swedish, 900 KB PDF) and the second looks at whether different types of collective agreement differently affect the gap (in Swedish, 1.9 MB PDF).
Women’s wages catching up with men’s
The annual study on the wage gap was conducted by the MI using regression analysis, based on the structure of earnings statistics. The data are collected annually from all of the public sector, every company in the private sector with more than 500 employees, and a random sample of small and medium-sized companies. In the data, salaries are listed as full-time and include solid supplements and a large number of variable increments in addition to the fixed salary. The MI found that women’s wages amount to approximately 86.8% of men’s pay; in other words, the unadjusted pay gap is 13.2%.
However, to avoid the estimated wage difference being determined by the fact that women and men have different backgrounds and are found in different occupations, a number of control variables were included in the regression analysis, such as age, level of educational attainment, sector, industry (NACE), size of the business, extent of employment and occupation. The MI found the adjusted gap to be 4.2%.
The analysis showed that occupation was the single most important variable to explain the wage difference. Furthermore, the results showed that the size of the adjusted wage gap varied between sectors. The largest was among white-collar workers in the private sector (8.1%), and the smallest among municipal workers (0.4%).
The report also showed that the wage gap is shrinking. The unadjusted pay gap is 0.2 percentage points smaller than the previous year and since 2005 it has decreased by a total of 3.1 percentage points. The development of the adjusted gap could unfortunately not be calculated with reliable results due to changes in how the data on occupations are collected.
The MI found two reasons to explain the shrinking gap:
- women’s wages have increased more than men’s wages;
- the composition of the workforce has changed.
During the most recent bargaining rounds, collective agreements covering workforces with a large share of women have set higher wage increases compared to the ‘cost-mark’ set by the industry agreement. In 2007, the increase was between 1.1 and 1.6 percentage points higher than the cost mark, and in the following three years the increase was between 0.4 and 0.7 percentage points higher. The results also indicate that the labour market has become less gender-segregated. Over the past few years, the share of women has fallen in traditionally female occupations and has correspondingly increased in male-dominated occupations. The MI found that around 30% of the decrease is due to the wage increase effect, while the workforce composition effect explains 60%–70% of the gap closure.
Gap not affected by type of collective agreement
Many employers who support ‘numberless’ agreements, that is, central collective agreements that leave wage-setting to be done at local level, have often stated that this can even out the wage gap (in Swedish). However, this claim has been criticised by, among others, the Swedish Metalworkers’ Union, which argues that more centralised wage-setting decreases the risk of gaps emerging (in Swedish).
In light of this debate, the government asked the MI to investigate whether the type of collective wage agreement has any effect on the gender wage gap. The MI used official wage data for the analysis. However, because the data are organised according to occupation rather than collective agreement, occupations were categorised into different agreements based on the MI’s knowledge of the structure of the labour market. By comparing the seven different agreement construction types (in Swedish, 155 KB PDF), ranging from those where wage increases and their distribution is decided at the sectoral level, to those where all wages are set at the local level, the MI has analysed the wage development for so-called ‘identical individuals’ between 2009 and 2013.
The results of the study indicated that factors other than the type of collective wage agreement offer better explanations for shifts in the wage gap. Within all forms of agreements there are occupations with both higher and lower rates of wage increase. There are bigger differences within the types of agreements than between them and, in conclusion, the MI found no significant support for the hypothesis that one form of agreement is preferable to another when trying to decrease the gender wage gap.
Men’s work is valued more highly
In its final report, the Delegation for Gender Equality in Working Life calculated the structural wage differentials. The delegation grouped occupations using a tool developed by the Swedish Equality Ombudsman to compare equivalent work. The tool takes education, problem-solving, responsibility, social skills and physical and mental working conditions into account. One example of a female-dominated occupation and a male-dominated occupation that were grouped together were elementary school teacher and mechanical engineer.
The delegation found that, with few exceptions, male-dominated occupations were more highly paid than the equivalent female-dominated occupations. Furthermore, this pattern was present in all age groups. This imbalance is caused partly by the supply and demand of labour, and partly by differences between salaries in the private and the public sectors. The delegation suggested that this could be considered ‘value discrimination’ (in Swedish). But Anna Hedborg, Chair of the delegation, said this was a controversial issue as there is no consensus on whether this unadjusted type of difference is worth investigating. The fact that pay varies between different sectors is generally seen as a less serious problem than the gap that is left when occupation has been controlled for.
The delegation also found that in many female-dominated professions the wage gap between the lowest-paid and highest-paid person was relatively small. This means that it can be difficult for employees within these professions to increase their salaries, and that there is little, if any, financial reward for professional development.
There is a wide range of views on what should be done to close the gender pay gap. In Sweden’s Gender Equality Barometer report for 2015 (in Swedish), the Swedish Trade Union Confederation (LO) criticised the way the wage gap is usually measured. When comparing men and women’s wages, the analyses are often based on full-time wages. Because women more often work part time, the results can be misleading. Thus, the ‘real’ gap is wider than reported and the problem of unequal pay is more severe than thought. Among LO's female members, more than half work part time (compared with 15% of men). LO argues that this has to be rectified in order to shrink the gap. It says that ‘actual monthly wages’ should be used when comparing wages. LO has also proposed that a three-part parental insurance system should be introduced. However, LO’s recipe for equal wages is equal working time, and value discrimination is not seen as a major issue.
Even within LO, opinions are divided as to what should be done to shrink the gap. The Swedish Municipal Workers' Union (Kommunal) argues that men should accept smaller wage increases in order to leave room for larger increases in occupations where wages have been inhibited by value discrimination. The Swedish Metalworkers’ Union (IF Metall) argues that the solution is to break the gender segregation of the labour market and to make sure women receive more training to increase their skills levels.
For the employers, the Confederation of Swedish Enterprise (Svenskt Näringsliv) stated recently that equal pay for equal work is a must, but competence must be rewarded financially and should be what ultimately decides one’s wage. In conclusion, in order for the gender wage gap to shrink further, the social partners will have to agree on a joint course of action. Such a plan would require consensus on whether ‘value discrimination’ is indeed a problem.