Segregation and the gender wage gap. Helsinki 2007. 127pp. (Reports of the Ministry of Social Affairs and Health, Finland, ISSN 1236-2115; 2008:26) ISBN 978-952-00-2622-6 (pb), ISBN 978-952-00-2623-3 (PDF)

This is the final report of the project “Segregation and the gender wage gap" funded by the European Social Fund and the Ministry of Social Affairs and Health. The project has been carried out jointly by Statistics Finland, the Research Institute of the Finnish Economy and the Labour Institute for Economic Research.

In the first part of the project various methods of measuring occupational segregation by gender were examined and a suitable indicator was selected for the Finnish context. A product of the research is the recommendation to use a conventional Index of Dissimilarity (ID) to measure occupational segregation by gender. The index has the particular strength of unambiguity, and the calculation of the index in connection with regular statistics production was considered feasible. According to the results, occupational segregation has decreased slightly during the period 1995-2004. A close examination of the private sector shows that the amount of segregation varies according to the characteristics of individuals and enterprises. In light of the results for the under 30-year-olds it would seem that segregation does not automatically abate after the older generations retire.

The second part of the research project examined the relationship between segregation and the gender wage gap. There are no major differences between sectors in the years 1995 and 2004; the average gender wage gap is roughly 17 and 20 per cent in all sectors. In the public sector occupational segregation explains roughly 40 to 50 per cent of the total wage gap, while in the private sector it explains just under 30 per cent of it. Most of the overall gender wage gap originates from pay discrepancies within individual collective labour agreements. The joint effect of segregation due to different collective agreements and occupational segregation explains roughly 40 to 50 per cent of the wage gap in the private and government sectors. In the local government sector the explanatory power of this so-called joint segregation rises up to 80–95 per cent of the total wage gap. The impact of segregation on the gender wage gap does not appear to have decreased in 1995–2 004.

As regards the follow-up monitoring of the wage gap impacts of segregation, the collective agreement classification method developed during the project showed most promise. This method pays close attention to the heterogeneity of wage determination of individuals, which is a result of the large number of collective agreements. The method produces estimates on the share of the total wage gap which is explained by the different collective agreements and different occupations of men and women as well as what share of the average wage gap is explained by other factors. The incorporation of this method into statistics production, however, requires further development.

Analyses focusing on the technology industries show that vertical gender segregation is a key mechanism producing gender wage gaps. Gender differences related to the hierarchical position of individuals explain 50 to 60 per cent, that is, at least one-half, of the average gender wage gap.

The selection process related to hierarchical positions appears to involve inefficiencies in the sense that upper-level employees do not always get placed at the hierarchical level which would maximise their productivity. For example, characteristics which are typically taken to promote career development, such as education and work experience, have a surprisingly small impact on the selection of employees to the highest level, that is, the management positions.

When gender wage gaps are examined from the perspective of the attained hierarchical level, the gaps are at their widest – but still rather moderate – in the less demanding tasks. Tailored remuneration practices contribute somewhat to these wage gaps. Gender wage gaps are at their narrowest in establishments with high productivity.

Segregation as such does not create gender wage gap. Instead it is caused by the placement of men and women in positions with hich or lower wages. In theory it is possible that differences in the wages of women and men are small even under conditions of large-scale segregation.

The project also developed practical models for the local government sector which serve wage surveys and assessed possibilities of examining wages across more than one collective agreement.

Statistics Finland’s total data on monthly-paid local government personnel were used in the development work. These data do not contain information on the remuneration systems applied by municipalities or the competence assessments of different tasks, which limits the development of models.

The results indicate that when conducting comparisons between various tasks, attention should be paid not only to classes by basic pay but also to the average wage formation in different tasks and fields with different agreements.

On the basis of data for the whole country, tasks have been divided into classes which have taken both the above-mentioned issues into account. The variance of total earnings has then been examined by gender, taking into account the average wage formation and basic pay of the tasks. In light of the inquiry conducted in the local government sector we can deduce that at least a part of municipalities and joint municipal boards have a need for the development of a competence assessment based on uniform criteria which would apply to the entire personnel. In addition, practical models serving the wage survey should be developed for local authority employers. If there is an intention to use broad statistical data files for the development of models or the production of reports in future, they should include data on the assessment methods applied by the employer and the competence classifications of the tasks.

Keywords : equal pay, gender equality, pay differentials, pay system, segregation