Reference metadata describe statistical concepts and methodologies used for the collection and generation of data. They provide information on data quality and, since they are strongly content-oriented, assist users in interpreting the data. Reference metadata, unlike structural metadata, can be decoupled from the data.
Demographic and Social Statistics Department - Labour Market Statistics Unit
1.3. Contact name
Restricted from publication
1.4. Contact person function
Restricted from publication
1.5. Contact mail address
Av. António José de Almeida
1000-043 Lisboa
Portugal
1.6. Contact email address
Restricted from publication
1.7. Contact phone number
Restricted from publication
1.8. Contact fax number
Restricted from publication
2.1. Data description
“Índice de Custo do Trabalho" is provided to Eurostat according to regulation No 450/2003 of the European Parliament and the Council of 27 February 2003 concerning the Labour Cost Index (LCI). "The objective of this Regulation establishes a common framework for the production, transmission and evaluation of comparable labour cost indices in the Community. Member states shall produce labour cost indices for the economic activities defined in article 4."
Concept, definition: the methodology corresponds to the target methodology of the European Labour Cost Index. The data include wages and salaries [basic salary, regular and irregular bonuses and allowances, overtime payment, payment and benefits in kind (not including canteens, medical centres and Christmas parties), employer's contributions (employer's social security contributions and compulsory insurance for accidents and occupational diseases) and collectively agreed, contractual and voluntary costs (supplementary pension/invalidity benefit, health insurance, life/personal accident insurance, and social benefits paid directly to employees in case of absence on account of sickness).
Statistical concept: The Labour Cost Index is defined as a Laspeyres index of labour costs per hour actually worked, annually chained, based upon a fixed structure of economic activity at NACE Rev.2, section level.
Labour cost categories are defined as indicated in the regulation (EC) Nº 450/2003 of 27 February 2003 (Total labour costs, Wages and salaries, Employer's social contributions, Total labour costs excluding bonuses).
With the publication of the results for the 1st quarter of 2020, Statistics Portugal started to include, in the LCI calculation, regarding the information related to the labour costs obtained from administrative sources, data corresponding to the universe of about 386 thousand entities (instead of the sample of around 3 800 entities), which declared earnings for a total of 4.2 million workers. In 2023 (final data), the information gathered from administrative sources corresponds to the universe of approximately 445 thousand entities which declared earnings to Social Security and to Caixa Geral de Aposentações, for a total of around 4.6 million workers. The information on hours actually worked continues to be obtained by direct inquiry from the entities that belong to the LCI sample (5 040 entities in 2024), and no changes were introduced at this level. For the O, P and Q sections of NACE-Rev.2, estimates of the number of hours actually worked per employee from the Labour Force Survey are used. In 2020, the weight structure for 2018 was updated due to the appropriation of administrative data on labour costs concerning the universe. The data collection for the new series of the LCI covered the four quarters of 2018, to guarantee the comparable basic information for the index computation from the 1st quarter of 2019. Since the publication of the indices for the 1st quarter of 2019, the observation unit is the enterprise instead of the local unit.
With the publication of the results for the 1st quarter of 2023, the Labour Cost Index series (2016=100) were rebased to the year 2020 (2020=100), in line with Eurostat’s procedure for the Labour Cost Index, whenever more up-to-date data of the (quadrennial) Labour Cost Survey become available.
The back series for all indices, from the 1st quarter of 2008, are available at the Official Statistics website.
For this series a new sample was extracted from the "reference population for surveys on enterprises" of Statistics Portugal, being this the sampling frame for the LCI (base 2020). This sample is renewed every year.
The weight structure used in the aggregation of economic activities was updated, resulting from changes in the economy's sectoral composition between 2021 and 2022.
Labour Cost Survey (LCS) data are used to obtain information that are representative of the universe of enterprises between one and nine employees for the hours actually worked as the LCI direct and regular collection is addressed only to enterprises with ten or more employees in order to minimize the statistical burden on enterprises. The LCS is also used to obtain information for the variables that are not available in the administrative sources, namely insurance against accidents and occupational diseases, supplementary pension/invalidity benefits, health insurance, life/personal accident insurance and social benefits paid directly to employees in case of absence on account of sickness.
Data collection for data on hours is carried out through an electronic questionnaire over the Internet (WEBINQ).
Vis-à-vis the statistical operation for the base year 2016 the following remained unchanged:
The legal framework– basic Regulation [Regulation (EC) No 450/2003 of the European Parliament and of the Council of 27 February 2003] and following, setting forth:
The concepts regarding the various labour cost items and the hours actually worked.
The technical specification of the index (Laspeyres index).
The economic activities covered: activities from Sections B to S of NACE-Rev.2.
The quarterly frequency.
As mentioned before, complementary statistical sources are used to ensure compliance with the principle of administrative simplification. These sources are the Labour Force Survey and the Labour Cost Survey 2020 to obtain the hours actually worked for the public administration and, the correction factors to get estimates for hours actually worked for enterprises between one and nine employees, respectively.
2.2. Classification system
NACE Rev.2
Data is compiled and provided to Eurostat for the following aggregates:
Total labour costs, wages and salaries, other costs, and total labour costs excluding bonuses;
Unadjusted data, calendar adjusted data (not seasonally adjusted data), calendar, and seasonally adjusted data.
2.3. Coverage - sector
Sections B to S of NACE Rev.2
Data on labour costs are obtained from administrative sources. All size classes are covered, as well as the private sector and the public administration [Sections O (Public administration and defence; compulsory social security) and the public part of Sections P(Education) and Q (Human health and social work activities).
Labour costs incurred by the employer corresponds to that reported by the entities to the Social Security (Declaração Mensal de Remunerações) and to the Caixa Geral de Aposentações (Relação Contributiva).
The information on hours actually worked is obtained by direct inquiry from a sample, through the quarterly survey "Índice de Custo do Trabalho" and covers enterprises with at least ten employees.
For the O, P and Q sections of NACE-Rev. 2 estimates of the number of hours actually worked per employee from the Labour Force Survey are used.
LCS data are used to obtain information on the hours actually worked that are representative of the universe of local units, namely, to produce estimates for enterprises between one and nine employees, as the LCI direct and regular collection is addressed only to enterprises with ten or more employees.
Number of employees: the persons who during the reference period participated in the business of the enterprise/institution, regardless of the duration of this participation, under the following conditions: a) staff bound to the enterprise/institution by an employment contract, receiving remuneration in return; b) staff which has ties to the enterprise/institution, who, for not being bound by an employment contract, does not receive regular remuneration for the hours worked or the labour supplied (e.g. owner-managers, unpaid family workers, active members of cooperatives); c) staff with ties to other enterprises/institutions who worked at the enterprise/institution and receive remuneration directly from it; d) persons in the above situations, absent for a period of no more than one month due to holidays, labour dispute, vocational training, as well as disease and occupational accident. The following persons are not considered to be staff: i) those in the situations described in a), b), and c) above and who are absent for a period of over one month; ii) workers with ties to the enterprise/institution who moved to other enterprises/institutions, receiving remuneration directly from the latter; iii) workers in the enterprise/institution whose remuneration is borne by other enterprises/institutions (e.g. temporary workers); iv) self-employed workers (e.g. service providers, that use the so-called 'recibos verdes', which is the popular name of the receipt form).
Hours actually worked: the total number of hours that employees on service actually worked. It includes extra hours. It also includes time spent at the place of work on tasks such as preparing work instruments, preparing, and servicing tools, paid dead time at work, time due to sporadic absences from work, the stoppage of machinery or accidents and short coffee breaks. Absence from work is excluded, irrespective of whether it is paid or not.
Compensation of employees: The compensation of employees is defined as the total remuneration, in cash or in kind, payable by an employer to an employee in return for work done by the latter during the reference period.
The components are:
Wage costs: base wage, regular bonuses and allowances, irregular bonuses and allowances (holiday bonus, Christmas bonus, end-of-year bonuses/distribution of profits, other irregular bonuses and allowances), payment of overtime, payment in kind.
Other costs: employers’ legal costs (social security contributions, insurance against accidents and occupational disease), collectively agreed, contractual and voluntary costs (supplementary pension/invalidity benefit, health insurance, life/personal accident insurance, and social benefits paid directly to employees in case of absence on account of sickness), severance payment.
Taxes paid by the employer: All taxes based on the wages and salary bill or on employment.
Subsidies received by the employer: All amounts received in the form of subsidies of a general nature intended to refund part or all the cost of direct remuneration but not intended to cover social security or vocational training costs. It does not include refunds paid to the employer by social security institutions or supplementary insurance funds.
2.5. Statistical unit
Unit of observation: enterprise.
2.6. Statistical population
Active enterprises located in Portugal covered by «Índice de Custo do Trabalho» that, according to NACE-Rev.2, belong to sections B to S.
2.7. Reference area
Mainland and autonomous regions (Madeira and Azores).
2.8. Coverage - Time
Since 2000 to 2nd quarter 2024.
2.9. Base period
Base year 2016 (between 2019 and 2022).
Base year 2020 (from 2023 onwards).
3.1. Source data
The origin of basic data (labour costs) is mainly from an administrative source, namely data provided by the enterprises to the Social Security and by the Public Administration organization to Caixa Geral de Aposentações. The hours actually worked (as well as the number of employees, in order to guarantee consistency) remain collected through the quarterly survey "Índice de Custo do Trabalho" where the statistical unit is the enterprise. For the O, P and Q sections of NACE-Rev.2 estimates of the number of hours actually worked per employee from the Labour Force Survey are used.
Types of data: Basic types of data are basic wages and salaries, regular and irregular bonuses and allowances, overtime payment, payments in kind, employers’ statutory social contributions, collectively agreed, contractual and voluntary costs and others.
Hours worked involve hours actually worked by employees during normal working hours and overtime.
Main internal data sources:
Hours actually worked are collected through the quarterly survey “Índice de Custo do Trabalho” where the statistical unit is the enterprise.
This sample survey is used to collect data on the number of hours worked, the hours not worked and overtime hours to compile the hours actually worked.
Data are collected through an electronic questionnaire over the internet (WEBINQ).
For the O, P and Q sections of NACE-Rev.2 estimates of the number of hours actually worked per employee from the Labour Force Survey are used.
Main external data sources:
Data provided by the enterprises to the Social Security and by the Public Administration organization to Caixa Geral de Aposentações.
Labour Cost Survey.
3.2. Frequency of data collection
The quarterly survey “Índice de Custo do Trabalho” collects data on the number of hours worked, the hours not worked and, overtime hours.
Administrative data are reported monthly.
3.3. Data collection
The LCI is a statistical operation held on a quarterly basis through a sampling survey to collect information on hours. Data collection is carried out through an electronic questionnaire (Computer Assisted Web Interviewing - CAWI).
By economic activity, the LCI covers Sections B to S of NACE-Rev.2. For the O, P and Q sections of NACE-Rev.2, estimates of the number of hours actually worked per employee from the Labour Force Survey are used.
Base period: electronic questionnaire over the internet (WEBINQ).
Quarterly updates: electronic questionnaire over the internet (WEBINQ).
Reference period of data collection on hours: the second month of each quarter (February, May, August, and November).
Administrative sources: data on labour costs reported monthly by the enterprises to the Social Security and by the Public Administration organisations to Caixa Geral de Aposentações.
3.4. Data validation
The quarterly labour cost components data from administrative sources are validated.
The components of hours (duration of work, number of hours worked, the hours not worked and overtime hours) that are collected using the "Índice de Custo do Trabalho" survey are also validated.
Detailed validation criteria (Service levels) are agreed between the unit responsible for the compilation of the LCI and the unit responsible for the data collection. Domain, coherence, and structure validations are carried out at the time of the data collection and entering of the data to identify eventual inconsistencies. After the data collection and the data entry new validations are made in accordance with the rules previously defined to evaluate possible discrepancies (such as reasonability of the magnitude of some ratios, comparison with the levels observed in previous periods, cross checking of the variables). If necessary, the respondents are contacted by the Data Collection and Management Department to clarify or to confirm the data. Similar validations are performed for the administrative data, especially for the average remuneration and the rate of contribution. Final data is also carefully assessed and confronted against the results of other short-term indicators, such as employment, wages and hours worked in industry, construction, and services as well as data from the Labour Force Survey.
3.5. Data compilation
The LCI is a short-term indicator that measures quarterly developments in labour costs per hour actually worked (hourly average cost) incurred by the employer. The index is calculated by dividing the average cost per employee by the number of hours actually worked per employee. For this reason, the evolution of these two variables (labour costs and hours worked) contributes to explaining developments in the LCI.
Labour costs incurred by the employer include the following:
Wage costs:
Base wage;
Regular bonuses and allowances;
Irregular bonuses and allowances (holiday bonus, Christmas bonus, end-of-year bonuses/distribution of profits, other irregular bonuses and allowances);
Payment of overtime;
Payment in kind.
Other costs:
Severance payment;
Employers’ legal costs (social security contributions, insurance against accidents and occupational disease);
Collectively agreed, contractual and voluntary costs (supplementary pension/invalidity benefit, health insurance, life/personal accident insurance, and social benefits paid directly to employees in case of absence on account of sickness).
The components of hours worked are as follows:
Normal hours.
Average number of weekly normal hours per employee.
Number of hours not worked.
Number of hours of paid leave and public holidays;
Number of hours of short paid period’s absence (the number of hours in the normal working period for which the employees in the reference period were paid but did not work, namely resulting from holidays, bank holidays, justified absences, marriage, the death of a family member, the provision of assistance to a family member, wife in childbirth, paternity leave, technical reasons, trade union representation, breast-feeding, medical appointments and sitting examinations. Also includes absences due to illness, occupational accidents and maternity leave when these are fully or partially paid by the company/institution);
Number of hours of absence for temporary suspension of activity or reducing normal working hours (lay-off);
Number of hours not worked and not paid for other reasons (the number of hours in the normal working period in which no work was carried out and for which employees receive no pay, namely due to strike, disciplinary sanctions, and other unjustified absences. Also includes absences due to illness, occupational accidents, and maternity leave in the event that these are not paid by the company/institution).
Overtime.
Hours actually worked
The number of hours actually worked are obtained as follows:
Potentially hours worked in the quarter – Nonworking time held in the quarter + overtime:
Potentially hours worked in the quarter (average number of weekly normal hours per employee x number of employees x 13 weeks).
Nonworking time held in the quarter (number of hours of paid leave and public holidays + number of hours of short paid period’s absence + number of hours for temporary suspension of activity or reducing normal hours + number of hours not worked and not paid for other reasons).
Overtime worked by employees in the quarter.
The variables on duration of work (paid and unpaid) are collected for the reference month of the quarter and then transformed into quarterly data before carrying out the calculations.
Administrative data
Information related to the labour costs components are obtained from Social Security and Caixa Geral de Aposentações.
Base year: 2020 . The choice of this year was based as usual on the reference year for the last available data from the Labour Cost Survey (LCS), conducted every four years. LCS data are used to obtain results that are representative for the universe of enterprises with one or more employees, given that the LCI direct and regular data collection of the hours worked is addressed only to the enterprises with ten or more employees to minimize the statistical burden on enterprises.
Quarterly updates: They are done through the quarterly LCI survey to collect components on hours. The sample in force in the statistical operation for the base year 2020 was extracted from the “reference population for surveys on enterprises” of Statistics Portugal. This sample is renewed every year.
The compilation of the LCI by economic activity (NACE-Rev. 2) is essentially carried out in five steps.
1. Obtain of labour costs and the estimates of hours actually worked (generically referred to as Y variable below) by stratum.
Considering:
i: economic activity of the enterprise
j: region of the enterprise
k: size class of the enterprise
h: stratum defined by the crossing of the variables economic activity, region, and size class
the estimator of the total of the variable Y in the stratum h = (i, j, k) is given by:
Υ = Xh / xh ∑ Υhe, where ∑ betweene=1 to nh
where:
Xh: number of persons of the population in stratum h
xh: number of persons of the answers (sample) in stratum h
e: enterprise belonging to stratum h of the sample
nh: number of enterprises that answered in stratum h
Yhe: variable value Y in enterprise e from stratum h
2. Once the strata are independent, estimates of hours actually worked by economic activity and region are given by the sum of the estimates of the strata that compose them.
3. Estimates on hours actually worked relate to enterprises with one or more persons.
Once the sample of the LCI covers enterprises with 10 or more employees and Regulation (EC) nº 450/2003 of 27 February requires the coverage of all size enterprises, correction factors obtained from LCS 2020 are applied.
Note that the LCS 2020 collects data for local units of all sizes, being compiled separately for establishments with “one to nine persons” and “ten or more persons”. The correction factors for hours actually worked are obtained by dividing the values of the variables corresponding to “one and more persons” (1 to 9 and 10 or more) and “ten or more persons”.
4. Integration of information related to the Labour Force Survey: estimates of the number of hours actually worked by employees for the O, P and Q sections.
5. After applying the correction factors to the data collected through the LCI survey and integration of information on hours actually worked for Public Administration, the final step is the calculation of the indices.
3.6. Adjustment
Working Days Adjusted data
To obtain the working days adjusted series, an adjustment factor is applied to the number of hours actually worked per employee.
Adjustment factor = number of working days in each quarter /number of working days in the corresponding quarter of the base year. This factor is applied for NACE sections data.
Seasonal Adjusted data
Working days adjusted series are used to obtain seasonally adjusted series.
Seasonal adjustments are made using Tramo-Seats (Demetra sofware). Under this method, the seasonal adjustment includes two phases. In the first phase, is made a pre-adjustment of the series in Tramo. In the second phase, the method Seats seasonal adjustment is applied. This method belongs to the family of parametric methods of seasonal adjustment, based on stochastic econometric models (ARIMA models). The choice of the specification of the models is done automatically.
For calendar adjusted and seasonal adjusted data, the indirect method is applied to guarantee consistency in the aggregations. This is performed deriving the unchained LCI for totals from the unchained LCI of the wage and non-wage components. For the LCI 2020 series the indirect method was applied also to the unadjusted data.
4.1. Quality assurance
Parameter variations are defined for each variable. For each variable, levels of service are defined, ranging between 90% and 95%.
4.2. Quality management - assessment
For each variable of the questionnaire ranges were defined. Surpassing the limits, the data should be checked, changed and/or justified by the enterprise.
5.1. Relevance - User Needs
Eurostat/European Central Bank
Both are the main regular users of LCI.
Bank of Portugal (Central Bank of Portugal)
Among other tasks, the Bank of Portugal disseminates the “main economic indicators” for the EURO zone.
Statistics Portugal provides information for wage costs for the total economy and for industry (SDDS).
Enterprises
Enterprises from several sectors of activity use the LCI indicators to follow the effects of labour and monetary policies on labour costs, to forecast and to plan the evolution of wages and personnel costs, as well as an element for indexing the value of contracts for the provision of services.
Internal users at Statistics Portugal
Departamento de Contas Nacionais (National Accounts Department).
Direcção Regional de Estatística da Madeira (DREM, Regional Directorate of Statistics of Madeira).
Serviço Regional de Estatística dos Açores (SREA; Statistics Azores).
Other external users
Satisfaction of the users’ needs
Sometimes other users (journalists) ask for long series.
5.2. Relevance - User Satisfaction
To evaluate the user satisfaction there are the following questionnaires in Statistics Portugal:
Survey on service satisfaction in response to information request;
Survey on the users' satisfaction of Statistics Portugal libraries;
Survey on satisfaction in Statistics Portugal internet website.
In these surveys results, there are no information on the satisfaction of the use of "Índice de Custo do Trabalho" data. Only on the subject where it is integrated, labour market.
There are no specific indications for "Índice de Custo do Trabalho".
5.3. Completeness
All rules of the Regulation are followed by Statistics Portugal.
5.3.1. Data completeness - rate
All rules of the Regulation are followed by Statistics Portugal.
Index series transmitted to Eurostat:
unadjusted (nsa);
working-day adjusted (wda);
seasonally and working-day adjusted (sa).
For each series, indices are transmitted by NACE Rev.2 and aggregates (B_S, B_N, B_E, B_F, G_J, G_N, K_N, O_S and P_S) for the following indicators:
Total labour costs;
wages and salaries;
labour costs other than wages and salaries;
total labour costs excluding bonuses.
Annual weights are also transmitted by indicator and NACE Rev.2.
6.1. Accuracy - overall
Ensures that, for all statistical units, the total corresponds to the sum of wages and salaries and other costs.
6.2. Sampling error
Not applicable.
6.2.1. Sampling error - indicators
Not applicable.
6.3. Non-sampling error
They are not calculated.
6.3.1. Coverage error
They are not calculated.
6.3.1.1. Over-coverage - rate
Not applicable.
6.3.1.2. Common units - proportion
There are common units between the units of the sample and the administrative source. The proportion is higher than 90%.
6.3.2. Measurement error
Before processing, the data are validated. In cases where the reported values exceed the established ranges, companies are contacted to correct or justify reported data.
6.3.3. Non response error
They are not calculated.
Response rate of the quarterly survey "Índice de Custo do Trabalho" that collects data on the number of hours worked, the hours not worked and, overtime hours:
2023 Q1 - 92.0%
2023 Q2 - 93.0%
2023 Q3 - 94.0%
2023 Q4 - 93.0%
2024 Q1 - 92.0%
2024 Q2 - 92.0%
6.3.3.1. Unit non-response - rate
Not applicable.
6.3.3.2. Item non-response - rate
Not applicable.
6.3.4. Processing error
Not applicable.
6.3.4.1. Imputation - rate
Not applicable.
6.3.5. Model assumption error
Not applicable.
6.4. Seasonal adjustment
Working Days Adjusted data
To obtain the working days adjusted series, an adjustment factor is applied to the number of hours actually worked per employee.
Adjustment factor = number of working days in each quarter /number of working days in the corresponding quarter of the base year. This factor is applied for NACE sections data.
Seasonal Adjusted data
Working days adjusted series are used to obtain seasonally adjusted series.
Seasonal adjustments are made using Tramo-Seats (Demetra sofware). Under this method, the seasonal adjustment includes two phases. In the first phase, it is made a pre-adjustment of the series in Tramo. In the second phase, the method Seats seasonal adjustment is applied. This method belongs to the family of parametric methods of seasonal adjustment, based on stochastic econometric models (ARIMA models). The choice of the specification of the models is done automatically.
The template for quality reporting on Seasonal Adjustment of the LCI is duly filled with the results for NACE sections B to S obtained through the JobVacancySurveyQR plugin.
Please note that the lines for activity aggregations have not been filled since they are obtained through the indirect method, outside of Demetra. Only the elementary indices (activity sections) are handled in Demetra.
At national level, there is not a schedule for data revision.
The revisions result from the inclusion of information for the last quarter sent lately by some enterprises of the sample survey on hours worked and from the update of information obtained through administrative sources, namely by data from the Declaração Mensal de Remunerações reported monthly by the enterprises to the Social Security.
The data published in release "Destaque" concerning the last quarters are subject to revisions. These revisions result, on one side, from the inclusion of information for the last quarter sent lately by some enterprises of the LCI sample and, on the other side, from updates of information obtained through administrative sources, namely by data from the Declaração Mensal de Remunerações reported monthly by the enterprises to the Social Security.
With the publication of the results for the 1st quarter of 2020, Statistics Portugal started to include, in the LCI calculation, regarding the information related to the labour costs obtained from administrative sources, data corresponding to the universe of about 386 thousand entities (instead of the actual sample of 3800 entities), which declared earnings for a total of 4.2 million workers. In 2023, the information gathered from administrative sources corresponds to the universe of approximately 445 thousand entities which declared earnings to Social Security and to Caixa Geral de Aposentações, for a total of around 4.6 million workers.
The information on hours actually worked continues to be obtained by direct inquiry from the entities that belong to the above-mentioned sample and no changes were introduced at this level. This change has impact in the indices released since the 1st quarter of 2019. Additionally, the LCI series have been revised in February 2021 because of the detection of a mistake in the “other costs” component since the 1st quarter of 2019.
With the publication of the results for the 1st quarter of 2023, the Labour Cost Index series were rebased to the year 2020 (2020=100), in line with Eurostat’s procedure for the Labour Cost Index, whenever more up-to-date data of the (quadrennial) Labour Cost Survey become available. This procedure does not change, by definition, the year-on-year changes already published. However, this opportunity was taken to fully adjust the method for calculating the indices to the methodology adopted by Eurostat, resulting in small differences from the figures previously published.
The back series for all indices, from the 1st quarter of 2008, are available at the Official Statistics website.
6.6.1. Data revision - average size
These revisions result, on one side, from the inclusion of information for the last quarter sent lately by some enterprises of the LCI sample and, on the other side, from updates of information obtained through administrative sources, namely by data from the Declaração Mensal de Remunerações reported monthly by the enterprises to the Social Security.
See table: Revision of the LCI year-on-year rates of change published (B_S) (working days adjusted data).
8.1.1. Asymmetry for mirror flow statistics - coefficient
Not applicable.
8.2. Comparability - over time
Data from the LCI 2020 series (2020=100) is comparable from the 1st quarter of 2008 onwards.
8.2.1. Length of comparable time series
In 1998: 1995 series (100=1995) - series between 1st quarter of 1996 and 4thquarter 2003 (NACE Rev.1).
In 2004: 2000 series (100=2000) - series between the 1st quarter of 2000 and 4th quarter of 2008 (NACE Rev.1).
In 2009: 2000 series (100=2000) - series were reprocessed on NACE Rev.2, covering data already published for the period 2000-2008; also, data for the 1st 2009 to 3rd quarter 2012.
In 2013: 2008 series (100=2008) - series between 4th quarter of 2012 and the 4th quarter of 2018; back series from the first quarter of 2008 are available.
In 2015: 2012 series (100=2012) - back-data 2000Q1-2008Q4 (NACE Rev.2 sections B to N). From 2008Q1 to 2018Q4 (NACE Rev.2 sections B to S).
In 2019: 2016 series (100=2016) - back data 2008Q1-2018Q4 (NACE Rev.2 sections B to S)
In 2020: 2016 series (100=2016) - back data 2019Q1-2019Q4 (NACE Rev.2 sections B to S).
In 2023: 2020 series (100=2020) - back data 2008Q1-2022Q4 (NACE Rev.2 sections B to S).
Note: The 1995 and 2000 series, on NACE Rev.1, are not comparable, due that the 2000 series adopted a different methodology.
8.3. Coherence - cross domain
Comparability over NACE sections.
Comparison with the evolutions registered in related indicators and infra-annual periodicity such as those obtained from turnover, employment, wages and salaries, and hours worked indices.
8.4. Coherence - sub annual and annual statistics
Not used.
The survey "Índice de Custo do Trabalho" collects data on hours components for monthly reference periods (February, May, August, and November).
8.5. Coherence - National Accounts
The National Accounts use different sources for its data collection, namely, Labour Force Survey, Business Statistics, and administrative sources (Directorate-General of Budget, Simplified Business Information).
There are differences between National Accounts and LCI data.
The latest definitive reference period available data for National Accounts is the year 2021.
8.6. Coherence - internal
Comparison with the evolutions registered in related indicators and infra-annual periodicity such as those obtained from turnover, employment, wages and salaries, and hours worked indices.
9.1. Dissemination format - News release
The national LCI data are disseminated through Press Releases and at the Official Statistics Portal.
Concerning the LCI metadata, some technical notes are available about topics such as geographical and sector coverage, labour costs components, calculation algorithm, and other statistical procedures like references to NSA and WDA series.
The press release analyses changes in the Total LCI and respective items: (1) wage costs versus other costs; and (2) average costs per employee versus number of hours worked per employee.
In the press release the following tables are published:
Table 1: Labour Cost Index (LCI) year-on-year rate of change by component and economic activity (NACE-Rev.2) (working days adjusted data).
Table 2: Labour Cost Index (LCI) year-on-year rate of change by component and economic activity (NACE-Rev.2) (working days non-adjusted data).
Table 3: Labour Cost Index (LCI) year-on-year rate of change by source of variation and economic activity (NACE-Rev.2) (working days adjusted data).
Table 4: Labour Cost Index (LCI) year-on-year rate of change by source of variation and economic activity (NACE-Rev.2) (working days non-adjusted data).
Data analysed in press release are adjusted for working days.
These tables are available on the Official Statistics Portal (Excel and CSV files).
9.2. Dissemination format - Publications
"Índice de Custo do Trabalho” does not have its own publication.
“Índice de Custo do Trabalho” are available in quarterly press release.
"Índice de Custo do Trabalho” data are published in “Anuário Estatístico de Portugal" (Statistical Yearbook of Portugal), a bilingual publication. (Portuguese/English). Disclosure of annual data.
9.3. Dissemination format - online database
Indexes and year-on-year rate of change are available on the Official Statistics Portal. See attached file.
There is a document: Índice de Custo do Trabalho, service levels (size of ratios, crosschecking variables, validation rules). The document lists the questionnaire variables. For each variable, intervals are set. For those cases that exceed the limits, companies are contacted to clarify the data, that are subsequently corrected, or confirmed with justification.
9.7.1. Metadata completeness - rate
Not applicable.
9.7.2. Metadata - consultations
Not applicable.
The average time to fill the questionnaire was not measured.
The questionnaire focuses on about 4270 enterprises (in 2023), which can also be selected for other Statistics Portugal questionnaires.
11.1. Confidentiality - policy
Statistical secrecy is established by Law to safeguard the privacy of citizens in general and to ensure trust in the NSS. All personal data collected by the authorities, for statistical purposes, are considered confidential, being protected by the professional secrecy, either by employees or others who became aware of them by means of their professional duties related with the official statistical activity.
The breach of confidentiality is considered a very serious administrative offence (in case of violation of professional secrecy it implies criminal responsibility), while the non-response to surveys within the period stipulated by the statistical authority is considered a serious administrative offence.
The staff of Statistics Portugal signs legal confidentiality commitments on appointment.
Instructions and guidelines exist for the protection of statistical confidentiality, both at concerns production and dissemination processes.
The National Statistical Law nº 22/2008 of May 13, 2008, establishes the principle of confidentiality under which no individual information about people can be disseminated. Information about corporate enterprises can only be disseminated under very restrictive conditions and always requires prior consent.
11.2. Confidentiality - data treatment
Not available in Labour Cost Index.
Quarterly data are available/disseminate in May (Q1), August (Q2), November (Q3) and February (Q4).
Reference period of "Índice de Custo do Trabalho": Q1 (February), Q2 (May), Q3 (August), Q4 (November) for hours components.
The sample frame of "Índice de Custo do Trabalho" was extrated from the "reference population for surveys on enterprises" of Statistics Portugal.
The stratification was the following:
Economic activity- subsection level of NACE rev2;
Region – NUTS II (2013);
Size classes – 10 - 49 employees, 50 - 249 employees, 250 or more employees.
With the publication of the results for the 1st quarter of 2019, Statistics Portugal starts the dissemination of a new series of the Labour Cost Index (LCI), with the base year in 2016, in line with Eurostat’s procedure for the Labour Cost Index, whenever new data of the (quadrennial) Labour Cost Survey become available. Back series for all indices analysed in this press release, from the 1st quarter of 2008, are available at the Official Statistics website.
In addition to the change of the base year, in the context of the SIMPLEX+ measures, the labour costs information, which was previously collected by direct inquiry, was obtained through administrative sources, namely data reported monthly by the enterprises to the Social Security and by the Public Administration organizations to Caixa Geral de Aposentações. The number of hours worked, the hours not worked and overtime hours, however, remain to be obtained by direct inquiry. The data collection for the new series of the LCI covered the four quarters of 2018, to guarantee the comparable basic information for the index computation from the 1st quarter of 2019. The observation unit changed from local unit to enterprise.
With the publication of the results for the 1st quarter of 2020, Statistics Portugal started to include, in the LCI calculation, regarding the information related to the labour costs obtained from administrative sources, data corresponding to the universe of about 386 thousand entities, which declared earnings for a total of 4.2 million workers. The information on hours actually worked continues to be obtained by direct inquiry for a sample and no changes were introduced at this level. This change has impact only in the indices published as of the 1st quarter of 2019.
With the publication of the results for the 1st quarter of 2023, the Labour Cost Index (LCI) series were rebased to the year 2020 (2020=100), in line with Eurostat’s procedure for the Labour Cost Index, whenever more up-to-date data of the (quadrennial) Labour Cost Survey become available.
“Índice de Custo do Trabalho" is provided to Eurostat according to regulation No 450/2003 of the European Parliament and the Council of 27 February 2003 concerning the Labour Cost Index (LCI). "The objective of this Regulation establishes a common framework for the production, transmission and evaluation of comparable labour cost indices in the Community. Member states shall produce labour cost indices for the economic activities defined in article 4."
Concept, definition: the methodology corresponds to the target methodology of the European Labour Cost Index. The data include wages and salaries [basic salary, regular and irregular bonuses and allowances, overtime payment, payment and benefits in kind (not including canteens, medical centres and Christmas parties), employer's contributions (employer's social security contributions and compulsory insurance for accidents and occupational diseases) and collectively agreed, contractual and voluntary costs (supplementary pension/invalidity benefit, health insurance, life/personal accident insurance, and social benefits paid directly to employees in case of absence on account of sickness).
Statistical concept: The Labour Cost Index is defined as a Laspeyres index of labour costs per hour actually worked, annually chained, based upon a fixed structure of economic activity at NACE Rev.2, section level.
Labour cost categories are defined as indicated in the regulation (EC) Nº 450/2003 of 27 February 2003 (Total labour costs, Wages and salaries, Employer's social contributions, Total labour costs excluding bonuses).
With the publication of the results for the 1st quarter of 2020, Statistics Portugal started to include, in the LCI calculation, regarding the information related to the labour costs obtained from administrative sources, data corresponding to the universe of about 386 thousand entities (instead of the sample of around 3 800 entities), which declared earnings for a total of 4.2 million workers. In 2023 (final data), the information gathered from administrative sources corresponds to the universe of approximately 445 thousand entities which declared earnings to Social Security and to Caixa Geral de Aposentações, for a total of around 4.6 million workers. The information on hours actually worked continues to be obtained by direct inquiry from the entities that belong to the LCI sample (5 040 entities in 2024), and no changes were introduced at this level. For the O, P and Q sections of NACE-Rev.2, estimates of the number of hours actually worked per employee from the Labour Force Survey are used. In 2020, the weight structure for 2018 was updated due to the appropriation of administrative data on labour costs concerning the universe. The data collection for the new series of the LCI covered the four quarters of 2018, to guarantee the comparable basic information for the index computation from the 1st quarter of 2019. Since the publication of the indices for the 1st quarter of 2019, the observation unit is the enterprise instead of the local unit.
With the publication of the results for the 1st quarter of 2023, the Labour Cost Index series (2016=100) were rebased to the year 2020 (2020=100), in line with Eurostat’s procedure for the Labour Cost Index, whenever more up-to-date data of the (quadrennial) Labour Cost Survey become available.
The back series for all indices, from the 1st quarter of 2008, are available at the Official Statistics website.
For this series a new sample was extracted from the "reference population for surveys on enterprises" of Statistics Portugal, being this the sampling frame for the LCI (base 2020). This sample is renewed every year.
The weight structure used in the aggregation of economic activities was updated, resulting from changes in the economy's sectoral composition between 2021 and 2022.
Labour Cost Survey (LCS) data are used to obtain information that are representative of the universe of enterprises between one and nine employees for the hours actually worked as the LCI direct and regular collection is addressed only to enterprises with ten or more employees in order to minimize the statistical burden on enterprises. The LCS is also used to obtain information for the variables that are not available in the administrative sources, namely insurance against accidents and occupational diseases, supplementary pension/invalidity benefits, health insurance, life/personal accident insurance and social benefits paid directly to employees in case of absence on account of sickness.
Data collection for data on hours is carried out through an electronic questionnaire over the Internet (WEBINQ).
Vis-à-vis the statistical operation for the base year 2016 the following remained unchanged:
The legal framework– basic Regulation [Regulation (EC) No 450/2003 of the European Parliament and of the Council of 27 February 2003] and following, setting forth:
The concepts regarding the various labour cost items and the hours actually worked.
The technical specification of the index (Laspeyres index).
The economic activities covered: activities from Sections B to S of NACE-Rev.2.
The quarterly frequency.
As mentioned before, complementary statistical sources are used to ensure compliance with the principle of administrative simplification. These sources are the Labour Force Survey and the Labour Cost Survey 2020 to obtain the hours actually worked for the public administration and, the correction factors to get estimates for hours actually worked for enterprises between one and nine employees, respectively.
Not Applicable
Number of employees: the persons who during the reference period participated in the business of the enterprise/institution, regardless of the duration of this participation, under the following conditions: a) staff bound to the enterprise/institution by an employment contract, receiving remuneration in return; b) staff which has ties to the enterprise/institution, who, for not being bound by an employment contract, does not receive regular remuneration for the hours worked or the labour supplied (e.g. owner-managers, unpaid family workers, active members of cooperatives); c) staff with ties to other enterprises/institutions who worked at the enterprise/institution and receive remuneration directly from it; d) persons in the above situations, absent for a period of no more than one month due to holidays, labour dispute, vocational training, as well as disease and occupational accident. The following persons are not considered to be staff: i) those in the situations described in a), b), and c) above and who are absent for a period of over one month; ii) workers with ties to the enterprise/institution who moved to other enterprises/institutions, receiving remuneration directly from the latter; iii) workers in the enterprise/institution whose remuneration is borne by other enterprises/institutions (e.g. temporary workers); iv) self-employed workers (e.g. service providers, that use the so-called 'recibos verdes', which is the popular name of the receipt form).
Hours actually worked: the total number of hours that employees on service actually worked. It includes extra hours. It also includes time spent at the place of work on tasks such as preparing work instruments, preparing, and servicing tools, paid dead time at work, time due to sporadic absences from work, the stoppage of machinery or accidents and short coffee breaks. Absence from work is excluded, irrespective of whether it is paid or not.
Compensation of employees: The compensation of employees is defined as the total remuneration, in cash or in kind, payable by an employer to an employee in return for work done by the latter during the reference period.
The components are:
Wage costs: base wage, regular bonuses and allowances, irregular bonuses and allowances (holiday bonus, Christmas bonus, end-of-year bonuses/distribution of profits, other irregular bonuses and allowances), payment of overtime, payment in kind.
Other costs: employers’ legal costs (social security contributions, insurance against accidents and occupational disease), collectively agreed, contractual and voluntary costs (supplementary pension/invalidity benefit, health insurance, life/personal accident insurance, and social benefits paid directly to employees in case of absence on account of sickness), severance payment.
Taxes paid by the employer: All taxes based on the wages and salary bill or on employment.
Subsidies received by the employer: All amounts received in the form of subsidies of a general nature intended to refund part or all the cost of direct remuneration but not intended to cover social security or vocational training costs. It does not include refunds paid to the employer by social security institutions or supplementary insurance funds.
Unit of observation: enterprise.
Active enterprises located in Portugal covered by «Índice de Custo do Trabalho» that, according to NACE-Rev.2, belong to sections B to S.
Mainland and autonomous regions (Madeira and Azores).
Not Applicable
Ensures that, for all statistical units, the total corresponds to the sum of wages and salaries and other costs.
Not Applicable
The LCI is a short-term indicator that measures quarterly developments in labour costs per hour actually worked (hourly average cost) incurred by the employer. The index is calculated by dividing the average cost per employee by the number of hours actually worked per employee. For this reason, the evolution of these two variables (labour costs and hours worked) contributes to explaining developments in the LCI.
Labour costs incurred by the employer include the following:
Wage costs:
Base wage;
Regular bonuses and allowances;
Irregular bonuses and allowances (holiday bonus, Christmas bonus, end-of-year bonuses/distribution of profits, other irregular bonuses and allowances);
Payment of overtime;
Payment in kind.
Other costs:
Severance payment;
Employers’ legal costs (social security contributions, insurance against accidents and occupational disease);
Collectively agreed, contractual and voluntary costs (supplementary pension/invalidity benefit, health insurance, life/personal accident insurance, and social benefits paid directly to employees in case of absence on account of sickness).
The components of hours worked are as follows:
Normal hours.
Average number of weekly normal hours per employee.
Number of hours not worked.
Number of hours of paid leave and public holidays;
Number of hours of short paid period’s absence (the number of hours in the normal working period for which the employees in the reference period were paid but did not work, namely resulting from holidays, bank holidays, justified absences, marriage, the death of a family member, the provision of assistance to a family member, wife in childbirth, paternity leave, technical reasons, trade union representation, breast-feeding, medical appointments and sitting examinations. Also includes absences due to illness, occupational accidents and maternity leave when these are fully or partially paid by the company/institution);
Number of hours of absence for temporary suspension of activity or reducing normal working hours (lay-off);
Number of hours not worked and not paid for other reasons (the number of hours in the normal working period in which no work was carried out and for which employees receive no pay, namely due to strike, disciplinary sanctions, and other unjustified absences. Also includes absences due to illness, occupational accidents, and maternity leave in the event that these are not paid by the company/institution).
Overtime.
Hours actually worked
The number of hours actually worked are obtained as follows:
Potentially hours worked in the quarter – Nonworking time held in the quarter + overtime:
Potentially hours worked in the quarter (average number of weekly normal hours per employee x number of employees x 13 weeks).
Nonworking time held in the quarter (number of hours of paid leave and public holidays + number of hours of short paid period’s absence + number of hours for temporary suspension of activity or reducing normal hours + number of hours not worked and not paid for other reasons).
Overtime worked by employees in the quarter.
The variables on duration of work (paid and unpaid) are collected for the reference month of the quarter and then transformed into quarterly data before carrying out the calculations.
Administrative data
Information related to the labour costs components are obtained from Social Security and Caixa Geral de Aposentações.
Base year: 2020 . The choice of this year was based as usual on the reference year for the last available data from the Labour Cost Survey (LCS), conducted every four years. LCS data are used to obtain results that are representative for the universe of enterprises with one or more employees, given that the LCI direct and regular data collection of the hours worked is addressed only to the enterprises with ten or more employees to minimize the statistical burden on enterprises.
Quarterly updates: They are done through the quarterly LCI survey to collect components on hours. The sample in force in the statistical operation for the base year 2020 was extracted from the “reference population for surveys on enterprises” of Statistics Portugal. This sample is renewed every year.
The compilation of the LCI by economic activity (NACE-Rev. 2) is essentially carried out in five steps.
1. Obtain of labour costs and the estimates of hours actually worked (generically referred to as Y variable below) by stratum.
Considering:
i: economic activity of the enterprise
j: region of the enterprise
k: size class of the enterprise
h: stratum defined by the crossing of the variables economic activity, region, and size class
the estimator of the total of the variable Y in the stratum h = (i, j, k) is given by:
Υ = Xh / xh ∑ Υhe, where ∑ betweene=1 to nh
where:
Xh: number of persons of the population in stratum h
xh: number of persons of the answers (sample) in stratum h
e: enterprise belonging to stratum h of the sample
nh: number of enterprises that answered in stratum h
Yhe: variable value Y in enterprise e from stratum h
2. Once the strata are independent, estimates of hours actually worked by economic activity and region are given by the sum of the estimates of the strata that compose them.
3. Estimates on hours actually worked relate to enterprises with one or more persons.
Once the sample of the LCI covers enterprises with 10 or more employees and Regulation (EC) nº 450/2003 of 27 February requires the coverage of all size enterprises, correction factors obtained from LCS 2020 are applied.
Note that the LCS 2020 collects data for local units of all sizes, being compiled separately for establishments with “one to nine persons” and “ten or more persons”. The correction factors for hours actually worked are obtained by dividing the values of the variables corresponding to “one and more persons” (1 to 9 and 10 or more) and “ten or more persons”.
4. Integration of information related to the Labour Force Survey: estimates of the number of hours actually worked by employees for the O, P and Q sections.
5. After applying the correction factors to the data collected through the LCI survey and integration of information on hours actually worked for Public Administration, the final step is the calculation of the indices.
The origin of basic data (labour costs) is mainly from an administrative source, namely data provided by the enterprises to the Social Security and by the Public Administration organization to Caixa Geral de Aposentações. The hours actually worked (as well as the number of employees, in order to guarantee consistency) remain collected through the quarterly survey "Índice de Custo do Trabalho" where the statistical unit is the enterprise. For the O, P and Q sections of NACE-Rev.2 estimates of the number of hours actually worked per employee from the Labour Force Survey are used.
Types of data: Basic types of data are basic wages and salaries, regular and irregular bonuses and allowances, overtime payment, payments in kind, employers’ statutory social contributions, collectively agreed, contractual and voluntary costs and others.
Hours worked involve hours actually worked by employees during normal working hours and overtime.
Main internal data sources:
Hours actually worked are collected through the quarterly survey “Índice de Custo do Trabalho” where the statistical unit is the enterprise.
This sample survey is used to collect data on the number of hours worked, the hours not worked and overtime hours to compile the hours actually worked.
Data are collected through an electronic questionnaire over the internet (WEBINQ).
For the O, P and Q sections of NACE-Rev.2 estimates of the number of hours actually worked per employee from the Labour Force Survey are used.
Main external data sources:
Data provided by the enterprises to the Social Security and by the Public Administration organization to Caixa Geral de Aposentações.
Labour Cost Survey.
Not Applicable
Length of time between data availability and the event or phenomenon they describe: end of reference quarter + 45 days.
