Central Statistics Office, Cork, Ireland

Earnings Analysis section

Labour Market and Earnings Division 

Earnings Analysis Section

Labour Market and Earnings Division

Central Statistics Office,
Cork Ireland

The Structure of Earnings Survey (SES) is a 4-yearly survey which provides EU-wide harmonised structural data on gross earnings and hours paid, 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 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).

The "economic activity" is coded in NACE Rev. 2 (General industrial classification of economic activities within the European Communities) whereas the "occupation" is coded according to the Commission Recommendation of 29 October 2009 on the use of the International Standard Classification of Occupations (ISCO-08).

Information on the 'Highest successfully completed level of education and training' variable is classified using the International Standard Classification of Education, 2011 version (ISCED11).

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

The enterprises size classes (corresponding to the number of employees) available are 1 to 9, 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 (full or part-time) and the duration of the contract (fixed or indefinite).

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 payments in kind.

Mean monthly gross earnings cover remuneration in cash paid before any tax deductions and social security contributions payable by wage earners and retained by the employer. Monthly earnings is calculated as weekly earnings x 4.333.  Weekly earnings is calculated as annual earnings divided by the number of weeks worked.

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

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

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

Ireland

Reference year is 2018.

Reference month is October.

N/A

The SES 2018 is based on Administrative data sources (including employee tax data) and the Structure of Earnings survey 2018 which was carried out to collect additional information not available from Administative sources, for a sample of employees.

Administrative data related to earnings is available on an annual basis. Earnings analysis from these sources are reported on in the Earnings Analysis using Administrative Data Sources (EAADS).

The Structure of earnings survey 2018 was carried out in 2019 for reference year 2018, to supplement the administrative data available.

The SES 2018 is based on Administrative data sources (including employee tax data) and the Structure of Earnings survey 2018 which was carried out to collect additional information not available from Administative sources, for a sample of employees.

The Structure of Earnings survey 2018 surveyed a sample of employees engaged by local units in NACE sectors B-S.

The sampling frame for the survey is the Earnings Analysis using Administrative Data sources (EAADS), the main source for which is the Revenue Commissioner’s P35L file (details of income Tax returns). The P35L gives details for each employee of their Gross Annual Income, No. of weeks worked in the year, earnings per week, date of employment and Employer’s Unique enterprise no. (CBR) and NACE code (industrial sector). 

A number of observations are excluded, such as pension payments, outliers (very low/high earnings) and the data is restricted to employments that were active in Ocober of the reference year.

From the EAADS a sample of employees are selected. The sampling method is a one stage stratified random sample. Stratification is based on:

• Main Economic Sector (NACE section)
• Gender
• Earnings bands (15 bands based on average weekly earnings, determined using the cumulative root frequency method).

 Employees responded primarily (73%) through an e-form questionnaire, with the remainder responding by paper questionnaire. 

Data entry validation rules are included in eForm questionnaire where possible (e.g. specific range of values allowed). 

Data collected were subjected to plausibility testing at both micro and macro levels. The first step was for the survey data to be examined and corrected at micro level. After weighting the data, aggregates were subject to coherence checks at a macro level to ensure the statistics were reasonable and consistent. 

Data from administrative data sources is combined with variables collected through survey process. Where required data from these sources are used to derive other variables required. 

Grossing factors derived from the sampling process are adjusted for non-response and calibrated to the Earnings Analysis using Administrative Data Sources (EAADS).

The calibration for the SES 2018 involved calibrating to the total number of individuals and the total weekly pay of the following groupings:

Sex (2) by age group (6) – 12 groups

Sex (2) by NACE (13) – 26 groups

Nationality – 4 groups

Firm Size – 6 groups

Sector – 2 groups (public/private)

Area – Dublin, Rest of Leinster, Connaught, Munster – 4 groups

The independence of the CSO is enshrined in the Statistics Act, 1993 and reflects best international practice for the organisation of Official statistics. The CSO subscribes fully to the principles set out in the European Statistics Code of Practice and to the UN Fundamental Principles of Official Statistics.

The Quality Policy for the Office is set out in “Standards and Guidelines, Volume 1 (Quality in Statistics)" which is available on the CSO website. It provides information and recommendations on best practice and contains clear guidelines to ensure that the quality of our processes and outputs are of the highest standard. The Office also promotes a culture of continuous improvement through the use of regular Business Process Improvement reviews so that the quality of our core processes and outputs are further enhanced.

Before transmission to Eurostat, the results of the SES are checked internally at both a micro and macro level to ensure their coherence.  Reponse rates by sector are considered.  Coefficients of variations are calculated to measure variability in key indicators.  The results also undergo a series of validation checks with Eurostat.

SES microdata are checked for completeness at a micro level and aggregates are checked at a macro level against other statistics to ensure coherence.

The main users of the SES survey are the following:

• Government Departments
• European Union/Eurostat
• International Organisations e.g. OECD, UN ILO
• Research institutes – e.g. ESRI, Universities, academics
• Professional Bodies (e.g. Trade Unions)
• The general public
• Trade Unions
• Other CSO sections

User Needs

The main user needs are a breakdown of average earnings (hourly, weekly, annual) in the main Classifications e.g. NACE, Occupation etc.; Median earnings and Gender Pay Gap requirements.

Researcher’s main needs are analysis of the Research Microdata files (RMFs) for macroeconomic research. Extensive research has been published from RMFs for the previous versions of this data (National Employment Survey (NES)) on Gender; Nationality; Public/Private Wage Gap; Earnings levels in the economy; etc.

Researchers are assisted with infrastructural support to work on the RMFs. Staff in the Labour Market and Earnings Division and the Research Coordination Unit (RCU) liaise with the researchers and provide technical support. Researchers are very satisfied with the level of support and a large body of research has been carried out by researchers involved in policy issues and macroeconomic research. Most researchers renew their access to the RMFs on an annual basis. 

Researchers are assisted with infrastructural support to work on the RMFs. Staff in the Labour Market and Earnings Division and the Research Coordination Unit (RCU) liaise with the researchers and provide technical support. Researchers are very satisfied with the level of support and a large body of research has been carried out by researchers involved in policy issues and macroeconomic research. Most researchers renew their access to the RMFs on an annual basis.

All the obligatory variables required by Eurostat regulations are provided.

All mandatory variables required by relevant Eurostat regulation are provided.

Accuracy is described in 6.1 to 6.6.

See 6.2.1

In general, the coefficients of variation are low, although there are some exceptions. 

Please see the attached document for detailed figures on the Coefficients of variation.

See below.

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

The initial dataset used is the Revenue Commissioner’s P35L file (details of income Tax returns). This dataset gives details for each employee of their Gross Annual Income, No. of weeks worked in the year, earnings per week, date of employment and Employer’s Unique enterprise no. (CBR) and NACE code (industrial sector). 

A number of observations are excluded, such as pension payments, outliers (very low/high earnings) and the data is restricted to employments that were active in Ocober of the reference year.

There is no measurement of over-coverage of the survey.

All employees included in the survey are on the Administrative data sources. 

The Revenue Commissioner’s P35L file (details of income Tax returns) includes pay details for employees for whom taxable income has been correctly notified.

The majority of responses were through e-form which contained controls for data entry including specific characters or ranges of values.

Non-response adjustment

Grossing factors (design weight) from the sampling process are adjusted to create non-response adjusted weights.

The weights are then rescaled to ensure they sum to the population totals. The calibration for the SES 2018 involved calibrating to the total number of individuals and the total weekly pay of the following groupings:

Sex (2) by age group (6) – 12 groups

Sex (2) by NACE (13) – 26 groups

Nationality – 4 groups

Firm Size – 6 groups

Sector – 2 groups (public/private)

Area – Dublin, Rest of Leinster, Connaught, Munster – 4 groups

See 6.3.4.1 (Imputation - rate)

Although processing errors may occur there has not been an overall measurement of the extent of processing errors.

There is no imputation done on observations, only on some of the variables included in the survey. 

Variables for which values imputed were:

• Occupation: 1.75%
• Education: 0.73% 
• Annual Leave: 4.89%
• Full-time/Part-time: 1.6%

Non-response adjustment

Grossing factors (design weight) from the sampling process are adjusted to create non-response adjusted weights.

The weights are then rescaled to ensure they sum to the population totals. The calibration for the SES 2018 involved calibrating to the total number of individuals and the total weekly pay of the following groupings:

Sex (2) by age group (6) – 12 groups

Sex (2) by NACE (13) – 26 groups

Nationality – 4 groups

Firm Size – 6 groups

Sector – 2 groups (public/private)

Area – Dublin, Rest of Leinster, Connaught, Munster – 4 groups

No seasonal adjustment takes place.

Provisional data provided to Eurostat in June 2020 (T+18 months) was subject to validation before results based on final data were published by Eurostat in November 2020.

No revisions have been made to the 2018 data since validated.

No revisions have been made to the 2018 data since validated.

Data for 2018 reference year provided to Eurostat in June 2020 (T+18 months).

Data validated and results published by Eurostat in November 2020 (T+23 months).

Data for 2018 reference year provided to Eurostat in June 2020 (T+18 months).

Data validated and results published by Eurostat in November 2020 (T+23 months).

Data for 2018 reference year provided to Eurostat in June 2020 (T+18 months).

Data validated and results published by Eurostat in November 2020 (T+23 months).

Data for 2018 reference year provided to Eurostat in June 2020 (T+18 months), in line with deadline set out in Regulation.

Data for 2018 reference year provided to Eurostat in June 2020 (T+18 months), in line with deadline set out in Regulation.

National concepts have been defined as close as possible to European concepts.

Classifications

There are no differences between the national classifications and the Eurostat classifications.

Not applicable

For the 2014 SES a Structure of Earnings Statistics Administrative Data Project (SESADP) was developed which matched data from the Employee Level Tax data with a range of other administrative sources. Hours for the SESADP were imputed using data from the Earnings, Hours and Employment Costs Survey (EHECS), which is conducted to meet EU requirements in relation to Labour Costs

For the SES 2018 an employee survey was carried out to collect data on a range of variables, including hours worked. 

Given these changes in methodology results are not directly comparable over time.

For the 2014 SES a Structure of Earnings Statistics Administrative Data Project (SESADP) was developed  which matched data from the Employee Level Tax data with a range of other administrative sources. Hours for the SESADP were imputed using data from the Earnings, Hours and Employment Costs Survey (EHECS), which is conducted to meet EU requirements in relation to Labour Costs

For the SES 2018 an employee survey was carried out to collect data on a range of variables, including hours worked. 

Given these changes in methodology results are not directly comparable over time.

Number of employments/employees - Coherence with Labour Force Survey

The SES is weighted and calibrated for population totals to the EAADS which is based primarily on administrative data (employee tax data). It is restricted to employments that were active in October 2018.

The LFS is a household survey and covers persons who worked in the week before the survey for one hour or more for payment or profit, including work on the family farm or business and all persons who had a job but were not at work because of illness, holidays etc. in the week.

One significant difference between the SES and the LFS is that the unit of observation for the SES is 'employment'. An individual may be included in the SES more than once if they have multiple employments with different employers. By contrast the LFS is based on 'employee' and an individual will only be counted once.

Also the LFS categories for NACE codes is based on the respondents definition of the NACE, whereas the SES is based on the activity of the organisation as defined in the LCI, for example a respondent in LFS may state they work in construction which is in NACE F, but if they work for a Public sector organisation they are classified as NACE O in the SES.

Gross Annual Earnings - Coherence with EHECS

The values for gross annual earnings presented below for the SES 2018 only include employees working 50 or more weeks per year. 

This differs from the LCI (Labour Cost Index) where the average annual earnings is calculated by dividing the wage costs for the enterprise by the average number of employees and grossing this figure up to the NACE Sector to get the quarterly earnings; the average of the four quarterly earnings are then used to compile the average annual earnings.

* EHECS (Earnings Hours and Employment Costs Survey; the CSO's Quarterly earnings Survey)

Not applicable

Results have been checked for coherence with both LFS and ELC, both of which feed into the National Accounts processes.

Results by NACE section for Mean weekly earnings, Mean paid hours and mean hourly pay compared to internal sources where applicable (Earnings & Labour Costs(ELC), Labour Force Survey(LFS))

Administrative data related to earnings is available on an annual basis. Earnings analysis from these sources are reported on in the Earnings Analysis using Administrative Data Sources (EAADS).

Results for the SES 2018 for Ireland were published by Eurostat in November 2020.

Results for the SES 2018 for Ireland were published by Eurostat in November 2020.

Administrative data related to earnings is available on an annual basis. Earnings analysis from these sources are reported on in the Earnings Analysis using Administrative Data Sources (EAADS).

All details of national publications on the CSO website http://www.cso.ie/en/statistics/earnings/

SES 2018 results are available on Eurostat database.

Administrative data related to earnings is available on an annual basis. Earnings analysis from these sources are reported on in the Earnings Analysis using Administrative Data Sources (EAADS).

All details of national publications on the CSO website http://www.cso.ie/en/statistics/earnings/

Data published nationally are also available on PXTAT (CSO statistics database).

Not available

Requests for micro-darta access are considered on a case by case basis and subject to CSO policies on data for researchers as set out here https://www.cso.ie/en/aboutus/lgdp/csodatapolicies/dataforresearchers/

All details of national publications on the CSO website http://www.cso.ie/en/statistics/earnings/ , data from which are typically also available on PXTAT (CSO statistics database).

Details for the SES Survey are provided on www.CSO.ie

http://www.cso.ie/en/methods/earnings/

The CSO's Quality statement and the Quality policy for the office are available on the CSO website, as is further information on the CSO's Quality Management Framework.

Codebook will be provided to researchers when data is made available through the CSO's Research Micro Data process.

Not available

Where possible data avaialble from Administrative data sources have ben used in order to reduce the cost and the burden on respondents.

The time required to complete the employee questionnaire is estimated at 15 minutes.

Safeguards in place to protect data collected by the CSO include:

Legal: All information obtained by the CSO under the Statistics Act, 1993 is strictly confidential and may only be used for statistical purposes.  This is specified in Sections 32 and 33 of the Act; any breach of those sections is an offence subject to significant penalties. 

These national statistical confidentiality provisions are reinforced by the following EU legislation:

• Council Regulation (EC) No 223/2009 on European statistics for data collected for EU statistical purposes.

Personnel: All staff of the CSO are Officers of Statistics under the Statistics Act 1993 and have signed a Declaration of Secrecy under the Act.  The data provided to the CSO may only be processed by Officers of Statistics and only for statistical purposes.  All staff must attend regular training on statistical confidentiality and data security, including mandatory refresher courses.

Governance: The CSO’s governance structure for data protection comprises: the Confidentiality and Data Security Committee (CDSC) which reports to the Management Board; the Data Protection Officer at Assistant-Director level who oversees compliance with statistical confidentiality and data protection requirements; and the Data Office which provides support in relation to policies, awareness, training and compliance.

Data Office: The CSO’s Data Office has the pivotal role of managing policies in relation to data protection and statistical confidentiality, promoting awareness, providing training, and assuring compliance.  The Data Office provides advice to CSO statistical areas on all issues related to data protection and statistical confidentiality.

Policies: The CSO has a comprehensive suite of policies in relation to roles, responsibilities and corporate rules in relation to statistical confidentiality, data security and data protection.  These policies are consolidated into a single Data Management Policy, which is a central reference point for all statistical processing.

Data Classification Scheme:  This is a corporate confidentiality classification system that allocates a confidentiality level (A to D) to all data held by the CSO.  Detailed management rules and procedures are assigned depending on the level given. Statistical micro-data has the highest level of security (A) and the strictest rules and procedures with regard to processing of the data.

Access to statistical data: All access to statistical data is restricted and limited to relevant staff that have a legitimate business reason for that access. This access is monitored on an ongoing basis.

Access limitations: Staff access to CSO buildings is controlled electronically.  Visitors to the CSO must be signed in and accompanied at all times.

IT Security: Staff access to IT systems is password-controlled.  Staff only have access to the data or systems relevant to their work and these access rights are regularly reviewed.  The IT systems and hardware are thoroughly protected by firewall and anti-virus security.  No external access is allowed.

Tables and publications: The CSO implements thorough Statistical Disclosure Control procedures to ensure that the tables and reports it publishes do not identify any individual or business. 

Administrative Data Centre: All administrative data received by the CSO is transmitted via a secure IT link to the CSO’s Administrative Data Centre (ADC).  In the ADC, identifying details such as name, address, date of birth, PPSN are removed before the data is used for statistical analysis.  The statisticians processing the data work with anonymous data and do not see the person’s identity. 

Data Lifecycle: When survey forms or other data records are no longer required for statistical purposes, they are securely destroyed.  The only identifiable records retained by the CSO are Census of Population forms – under Section 35 of the Statistics Act, these become public records after 100 years.  All other forms are securely destroyed.

Data Linkage: All statistical projects that involve linkage between administrative data sources are subject to the CSO’s Data Linkage Policy and must meet a Privacy Impact Assessment.  The CSO publishes a register of all projects in which it links administrative data.

Formal Assessment Processes: The CSO undertakes a Data Necessity and Proportionality Assessment for each source of personal data used in the production of statistics.  In addition, the CSO undertakes a Privacy Risk Assessment for each stage of statistical collection and processing of personal data, using the UN GSBPM as a standard template. The aim of these two assessment processes is to identify risks at each stage and make sure that confidentiality and data protection are respected and well-managed at every stage of collecting and producing statistics. 

Before performming any statistical analysis, the CSO removes all identifying personal information including the Personal Public Service Number (PPSN). The PPSN is a unique number that enables individuals to access social welfare benefits, personal taxation and other public services in Ireland. The CSO converts the PPSN to a Protected Identifier Key (PIK). The PIK is an encrypted and randomised number used by the CSO to enable linking of records across data sources and over time which is internal to the CSO. Using the PIK enables the CSO to link and analyse data for statistical purposes, while protecting the security and confidentiality of the individual data. 

Age is calculated from year of birth (specific date of birth is not on the dataset).

Other characteristics such as Occupation, highest level of education etc. are reported at a grouped level. 

N/A