Statistics Sweden

Population and Welfare Department

Labour market

SCB

Att. Sofia Löfgren BV/AMÖ

701 89 Örebro

Sweden

The Structure of Earnings Survey (SES) is a 4-yearly survey which provides EU-wide harmonised structural data on gross earnings, hours paid and annual days of paid holiday leave, which are collected every four years under Council Regulation (EC) No 530/1999 concerning structural statistics on earnings and on labour costs, and Commission Regulation (EC) No 1738/2005 amending Regulation (EC) No 1916/2000 as regards the definition and transmission of information on the structure of earnings.

The SES 2018 provides detailed information on relationships between the level of hourly, monthly and annual remuneration, personal characteristics of employees (sex, age, occupation, length of service, highest educational level attained, etc.) and their employer (economic activity, size and economic control of the enterprise).

Regional data is available on NUTS1 level.

The economic activity is coded in NACE Rev. 2 and the occupation is coded in ISCO-08.

Information on the highest successfully completed level of education and training varable is classified using the ISCED2011.

Regional breakdown is based on NUTS.

The statistics cover all economic activities defined in NACE Rev. 2 sections B to S.

The enterprises included employ at least 10 employees and the size classes available are 10 to 49, 50 to 249, 250 to 499, 500 to 999 and more than 1 000. 

Employees are all persons who have a direct employment contract with the enterprise or local unit and receive remuneration, irrespective of the type of work performed, the number of hours worked and the duration of the contract.

Mean annual gross earnings also cover all "non-standard payments", i.e. payments not occurring in each pay period, such as: 13th or 14th month payments, holiday bonuses, quarterly or annual company bonuses and annual payment in kind.

Mean monthly gross earnings in the reference month cover remuneration in cash paid before any tax deductions and social security contributions payable by wage earners and retained by the employer, and are restricted to gross earnings which are paid in each pay period during the reference month.

Mean hourly gross earnings are defined as gross earnings in the reference month divided by the number of hours paid during the same period.

Number of hours paid includes all normal and overtime hours worked and remunerated by the employer during the reference month. Hours not worked but nevertheless paid are counted as paid hours (e.g. for annual leave, public holidays, paid sick leave, paid vocational training, paid special leave, etc.).

The statistical unit is enterprise.

The SES 2018 refers to enterprises with at least 10 employees in the areas of economic activities defined by NACE Rev. 2 sections B to S.

Sweden

The year 2018.

Not applicable.

The public sector

The collection of data in the yearly survey of the public sector is carried out in cooperation with the Swedish Agency for Government Employers (SAGE) and the Swedish Association of Local Authorities and Regions (SALAR). SAGE collects data on individuals from all government authorities with September as reference month. SALAR collects data on individuals from all municipalities and county councils with November as reference month. After processing and editing the data is transmitted to Statistics Sweden.

The private sector

A stratified random sample of enterprises is drawn from the private sector. Both the yearly earnings survey for the private sector and the survey on annual bonuses use the same sample. The frame was set up and the sample was drawn in August 2018.

The yearly earnings survey

The reference month is September and information about the survey was sent out in advance to the sampled enterprises in August 2018. The questionnaire was sent out by the end of September 2018. The respondents had approximately three weeks to send in the requested data. Scanning, coding and validation of the data started as soon as any questionnarie were answered. The data collection was stopped in March 2019.

The survey on annual bonuses

Since annual premiums related to profit sharing depend on the result of the enterprise, it was important to let the respondents finalize the balancing of their accounts. This is often done in the first months of the year. Therefore, the questionnarie was sent out in April 2019, and the respondents were given three weeks to answer. Scanning, coding and validation of the data started as soon as any questionnarie were answered. The data collection was stopped in August 2019.

Sampling

The sampling procedure consists of two stages. First, a stratified random sample of enterprises is drawn from a frame including the private sector. Each stratum is set up by combining size class and economic activity of the enterprise. The public sector (government authorities, municipalities and county councils) is fully covered in the sample. Second, a simple random sample of employees is drawn from the enterprises, authorities, municipalities and county councils drawn in the first stage.

Four-yearly.

The main part of the data concerning the private sector was collected in a yearly earnings survey and a separate survey concerning annual bonuses. These data was collected by using a questionnaire. The yearly earnings survey has been carried out since 1997 and is well established both among users and respondents. The results from the survey are of high importance for the users and the questionnaire has been evaluated and improved during the years. The bonus survey has been carried out since 2002 and some minor changes have been done in purpose to improve the questionnaire.

The data concerning the public sector is collected in corporation with the Swedish Agency for Government Employers (SAGE) and the Swedish Association of Local Authorities and Regions (SALAR). SAGE collects data on individuals from all government authorities with September as reference month. SALAR collects data on individuals from all municipalities and county councils with November as reference month. After processing and editing the data is transmitted to Statistics Sweden.

All data has been validated both on micro and macro level. In the validation process, different kinds of logical tests were carried out. In the yearly earnings survey these tests have been evalutated during the years, and are now considered to reveal most errors in the data. The method for validation of the data from the annual bonuses survey has been changed due to analysis of previous data. In both surveys, respondents were contacted in order to validate errors and collect correct data.

[Not requested]

[Not requested]

The quality of the data is good, the survey for the private sector has been carried out since 1997 and is well established both among users and respondents. For the public sector the surveys have been carried out even longer. Thanks to the identification numbers on individuals and enterprises we can merge different sources and have a good quality.

[Not requested]

Main users are Eurostat and other EU institutions. Other users are researchers, media, National Mediation Office, and Statistics Sweden.

The needs vary between different users. The main part of the Structure of Earnings Survey (SES) 2018 comes from yearly surveys on earnings. The surveys are constructed with consideration for the needs of the users and are well established in Sweden both in terms of using the results and reporting requested data to Statistics Sweden. These yearly surveys on earnings are based on enterprises, not on local units.

There is no publication of SES in Sweden why it is difficult to know who is using the SES and what they think of the data. 

There are information on all mandatory variables.

[Not requested]

[Not requested]

The sampling procedurs consists of two stages. First, a stratified random sample of enterprises is drawn from a frame including the private sector. Each stratum is set up by combining size class and economic activity of the enterprise. The public sector (government authorities, municipalities and county councils) is fully covered in the sample. Second, a simple random sample of employees is drawn from the enterprises, authorities, municipalities and county councils drawn in the first stage.

The coefficients of variation concerning ‘Gross earnings in the reference month’ and ‘Average gross hourly earnings in the reference month’ are presented in the attached document Coefficients of variation.

When calculating coefficients of variation the SAS-program CLAN is used to extract the point- and standard error estimates. The program takes into account both survey design and non-response.

See subheads.

The sample from the private sector was drawn from the Business register at Statistics Sweden in August 2018. The frame is expected to well reflect the target population, since the reference month surveyed is September 2018. Even though the frame is set up close in time to the reference month problems with under- and overcoverage may arise, since updates concerning information in the Business register sometimes takes time. An analysis on the rates of undercoverage has not been done.

The overall rate of overcoverage is 1,4 percent. An enterprise is considered as overcoverage if it did not exist in practice.

[Not requested]

Besides the questionnaire, all respondents receive guidance with explanations to all questions in the questionnaire. To further help the respondents, the guidance includes FAQs and contact information.

All data has been validated both on micro and macro level. In the validation process, different kinds of logical tests were carried out. In the yearly earnings survey these tests have been evaluated during the years, and are now considered to reveal most errors in the data. The method for validation of the data from the annual bonuses survey has been changed due to analysis of previous data. In both surveys, respondents were contacted in order to validate errors and collect correct data.

The municipalities are coded with their main economic activity even if they have more economic activities. They cannot be dissaggregated whith NACE because the data are not based on local units. Some municipalities are coded as P and some as Q even if the sections P and Q are both fairly large in the municipalities.

The variable “Length of service in enterprise” is obtained from a register which extends only back to 1989. Because of this, the maximum value of this variable is 29. For some employees, the figure may be higher.

Other possible sources of errors may be information in administrative registers, at Statistics Sweden or other agencies, which are used in the survey. This may have an impact on for example level of education, gross annual earnings and number of weeks to which the gross annual earnings relate.

Measurement errors have not been evaluated further and it is difficult to estimate the impact they may have on the results of the survey.

Different actions have been taken to reduce the unit non-response rate. First, a reminder by mail. Second, non-respondents were reminded by phone calls. Third, for certain large enterprises, a request of fine was used.

The overall non-weighted response rate was 86 percent, including 1,4 percent overcoverage. The following table shows the unit response rates, by NACE section and size class.

Response rate (percent), by NACE section and size class of enterprise

[Not requested]

[Not requested]

There has not been an overall measurement of the extent of processing errors.

No information.

No adjustment from fiscal year to calendar year is done. The main part of data on earnings related to the reference year is collected from administrative registers at the Swedish Tax Agency and refers to the calendar year.

NACE sections B to S is covered for enterprises, organizations and authoritys with 10 or more employees. Employees surveyed in the enterprises are between 18 and 66 years of age.

Some variables in the SES are collected from administrative registers. In table A, the variables “Size of enterprise to which the local unit belongs”, “Economic activity” and “Economic and financial control” are collected from administrative registers. In table B, the variables “Highest completed level of education and training”, “Length of service in enterprise”, “Citizenship”, “Number of weeks to which the gross annual earnings relate”, “Gross annual earnings in the reference year” and “Annual payments in kind” are collected from administrative registers. To link collected data for an individual or an enterprise with data from registers, identity numbers of the individuals or enterprises are used. Registers may have undercoverage which is considered as item non-response in the SES.

The variable A15 Collective pay agreement is based on a model where all companies as well as the entire public sector that are collected through employers' organizations are assumed to have collective agreements. For the remaining companies, a model is used to determine whether they have a collective agreement or not. From another survey, we have the proportion of companies with collective agreements by size class and can randomize collective agreements for companies.

For the private sector the variable B28 form of employment is modeled using labour force survey. 

[Not requested]

[Not requested]

[Not requested]

[Not requested]

The Swedish Structure of Earnings data for the reference year 2018 was delivered to Eurostat in May 2020.

[Not requested]

[Not requested]

SES was delivered on time to Eurostat in May 2020.

[Not requested]

According to regulations regarding the SES, local units shall be surveyed. Due to problems related to the survey design the Swedish SES collect data on enterprises and not their local units. However, all local units in each enterprise are surveyed. This may have an impact on the results in two general ways. First, data broken down by region might be incorrect. Second, data broken down by NACE might be affected, since local units within an enterprise may belong to a different economic activity than the enterprise. These problems were analyzed when carrying out the Labour Cost Survey 2000 and the analyses indicated that the problem was minimal.

[Not requested]

Statistics Sweden has carried out the SES six times; for the reference years of 1995, 2002, 2006, 2010, 2014 and 2018. The survey design was rather different in 1995, surveying only a portion of employees in the sampled enterprises. In 2002 information from two surveys was used in combination with data from different administrative registers. The survey 2006 was extended to include the public sector.

Comparisons between the surveys should be done with caution, since the survey design has changed since 1995.

Since SES 2010 ISCO-08 is used for classifying occupations. 

SES 2014 and SES 2018 includes employees in the age of 18 to 66 years. SES 2002, 2006 and 2010 include employees in the age of 18 to 64 years. 

[Not requested]

In the table below, the SES variable “gross annual earnings” is compared to "wages and salaries" of the National Accounts (NA) for the private sector. Significant differences between the sources is that sampled units in SES had 10 or more employees whereas the NA data also includes smaller firms and estimates for moonlighting.

Comparisons between SES and NA concerning annual earnings per employee by NACE, private sector 2018

[Not requested]

See 8.3.

[Not requested]

Statistics Sweden does not publish any news release or other publications from SES 2018. 

Statistics Sweden does not publish results from the Structure of Earnings Survey.

Statistics Sweden does not publish results from the Structure of Earnings Survey.

[Not requested]

[Not requested]

No results from the survey have been sent to the reporting units.

There are no documents available in English.

[Not requested]

[Not requested]

[Not requested]

[Not requested]

[Not requested]

[Not requested]

No further comments.