Czech Statistical Office (CZSO)

Labour Market and Equal Opportunities Statistics Department

Na padesatem 81
100 82 Praha 10
Czech Republic

This quality report is about the Czech Labour Cost Index (LCI), reference year 2022.

The legal basis for the LCI is Commission Regulation (EC) N° 1216/2003 of 7 July 2003 implementing Regulation (EC) N° 450/2003 of the European Parliament and of the Council concerning the Labour Cost Index.

NACE Rev. 2

NACE Rev. 2 sections B to S

The Labour Cost Index (LCI) is defined as an index of labour costs per hour worked. The labour costs are the total quarterly costs incurred by the employer in the employment of labour. Labour cost indices are provided separately for the four labour cost categories:
(a) Total labour costs;
(b) Wages and salaries;
(c) Employers' social contributions plus taxes paid by the employer less subsidies received by the employer;
(d) Total labour costs excluding bonuses.

The LCI is calculated every calendar quarter. No specific statistical survey on labour cost index has been organized in the Czech Republic. The index is calculated strictly by matching data from relevant sources. At present, a combination of two data sources is used:
• Enterprise Report on Employment and Wages (ER);
• Labour Cost Survey (LCS).

The first survey is organized regularly with the quarterly periodicity, while the LCS is an annual one in the Czech Republic.
The LCI is broken down by NACE Rev 2 section.

Local unit

All local units with a main activity in sections B to S of NACE Rev. 2.

The Czech Republic

Data are available since the 1^st quarter of 1996 for NACE Rev 1.1 (until the 4th quarter of 2008)  and since 1^st quarter of 2000 for NACE Rev 2.

2016 = 100

This section contains information about the statistical processing made by CZSO.

The Labour Cost Index is calculated every calendar quarter. No specific statistical survey on labour cost index has been organized in the Czech Republic. The index is calculated strictly by matching data from relevant sources. At present, a combination of two data sources is used:

• Enterprise Report on Employment and Wages (ER);
• Labour Cost Survey (LCS).

The ER is organized regularly with the quarterly periodicity, while the LCS is an annual one in the Czech Republic. The Table shows the frequency of collecting data from source/supporting surveys.

The dates in the table show when the final results became available. The most problematic term is the fourth quarter, when the gap between availability of the earnings data and the LCI deadline is very narrow.

Depends on statistical source.

Depends on statistical source.

The ER is used for numbers of employees for volumes of wages/salaries and for numbers of hours worked. The LCS is used for structures of total labour costs (non-wage costs are derived).

No methods used to estimate/model the missing information on groups of employees or cost items have been created. Survey samples are supposed to represent the whole national economy.

Data are internally validated before dissemination. Herefore we use the national validation procedures.

The Czech Republic fulfills the requirements of Regulation (EC) N° 450/2003. The quality of the Labour Cost Index depends on the quality of the source data, especially the ER. It is quick but the results are regularly revised (see more in 6.5.).

The main and regular user of the Labour Cost Index has been Eurostat.
The Czech National Bank (Central Bank) has requested some results of LCI occasionally for monitoring of inflation pressures. The satisfaction of its needs and the relevance has not been expressed.
A shortened time series of the labour cost indices by NACE Rev 2 section have been included to the Indicators of Social and Economic Development of the Czech Republic (see the Annex).

No response from the users on the relevance has been delivered.

The Labour Cost Index in the Czech Republic has been prepared strictly according to the Regulation (EC) N° 450/2003. The compliance with the regulation is perfect.
In the past, the weaknesses of our calculation scheme were working time series. The average number of hours worked per week was taken from the LFS and multiplied by the number of employees from the quarterly enterprise survey and the number of weeks in the quarter. Originally, it was the best idea because we did not have proper statistics on the working hours of businesses. However, the use of two completely different sources caused problems because the results of the household survey did not always respond correctly.

Now we've managed to get relevant information from business statistics. As a result, the LCI calculation system was reconstructed to a new source. For this reason, new hours worked have been revised to 2003 in order to maintain consistency in the time series of the final indicator. We believe that the new LCI will generally be of better quality. This change should also lead to a smoother development of all LCI components, resulting in more complicated seasonal / WDA adjustments.

100 %

The Labour Cost Index is calculated every calendar quarter. No specific statistical survey on labour cost index has been organized in the Czech Republic. The index is calculated strictly by matching data from several relevant sources. At present, a combination of three data sources is used:

• Enterprise Report on Employment and Wages (ER);
• Labour Cost Survey (LCS).

No sampling takes place.

Not applicable, see 6.2.

This section describes the non-sampling errors.

The Table shows the numbers of employees and the percentages of the employees represented by the Enterprise Report (ER) as well as Labour Cost Survey (LCS) on the overall number of employees in the national economy according to ESA 2010; NACE B-S. Detailed information broken down by NACE Rev 2 section groups is in the Annex.

The ESA 2010 figures are estimated using the Social Security Register. The distribution of employees into industries is not made using surveys on employment only, other auxiliary variables such as gross value added and average earnings in the industry have been taken into account as well. This is why the comparison by NACE Rev 2 sections is distorted (see the Annex) nevertheless the aggregate figures show very good coverage, as only a small part of economy (shadow economy) is not covered by the most important source (ER). The coverage rates for all the years are analogous, as the statistical system has not encountered any changes.

The LCS covers the sample of enterprises/legal persons with 1 or more employees (the survey is exhaustive for units with 500 or more employees and a sample is drawn from all units with less than 500 employees). Sample survey results are grossed up to the universe of all active units incorporated in the Business Register maintained by the CZSO including household sector.

The coverage of the ER differs by economic branch and sector and period. Figures for the whole national economy are estimated using a number of both statistical and administrative sources.

The Labour Cost Index should therefore represent the whole economy.

See 6.3.1.

No common units are detected.

As it was mentioned earlier the LCI derives data from several statistical surveys, so the information on measurment and non-response errors should be included in quality reports from these individual surveys.

Not applicable, depends on the source.

See 6.3.3.

See 6.3.3.

See below 6.3.4.1.

Estimation

No methods used to estimate/model the missing information on groups of employees or cost items have been created. Survey samples are supposed to represent the whole national economy. The annual Labour Cost Survey covers all cost items.
Some estimation is used for the internal calculation of the source data on non-wage costs. They are calculated strictly on “ceteris paribus”. In our case, it virtually means to make two presumptions: 1/ the structure of labour costs remains stable; until new LCS results, the structure of the preceding reference year remains unchanged; 2/ the wages development changes neither at present nor in near future; some estimation of the future average quarterly wage needed for calculating the whole year average should be derived by means of the latest y-o-y wage index known:
est WQn = (WQn-1/WQn-5) * WQn-4
where
WQn = Quarterly average wage in (future) quarter n
WQn-1 = Quarterly average wage in quarter n-1 (i.e. the latest known figure)
WQn-5 = Quarterly average wage in quarter n-5
WQn-4 = Quarterly average wage in quarter n-4
The influence on the final results is minimal.

Administrative data

No administrative data on labour cost have been used.

No model used.

TRAMO-SEATS is used as application for the seasonal adjustment. National specific holidays are taken into account for the working day adjustment.

The LCI results are never considered final, but practically they have changed (at most) two years backwards.

The Annex shows the revision history in the released year-on-year growth rates for the total labour costs (TOT) using the unadjusted series for the last 12 quarters; broken down by NACE Rev 2 section.
Generally, revision policy is as follows: updating based on a new structure of labour costs is made when new LCS results become available and always after the updating of the Enterprise Report on Employment and Wages (ER) - practically every quarter.
The LCS update had greater influence on the LCI results than the ER update, which uses to be petite. It is especially visible in the NACE Rev 2 sections where significant change in the labour cost structure happened (for example NACE E; see the Annex), then the influence of the y-o-y growth rate would be in percentage points.

The LCI results are never considered final, but practically they have changed (at most) two years backwards.

See 6.6.

2022Q1: t+ 70 days

2022Q2: t+ 70 days

2022Q3: t+ 70 days

2022Q4: t+ 70 days

See 7.1.

See 7.1.

The deadline has been kept in all quarters of the years 2020, 2021 and 2022.

A shortened time series of the LCI have been included to the Indicators of Social and Economic Development of the Czech Republic, t + 90 days.

Time comparability is unproblematic - the same concepts are valid for every period since the 1st quarter 2003.
There is full comparability between the European standard NACE Rev 2 and the Czech national classification CZ-NACE.

Internal coherence

Full comparability

No mirror flow statistic.

Full comparability

Since the 1st quarter 2003.

There is full comparability between the European standard NACE Rev 2 and the Czech national classification CZ-NACE.

Annual LCI data are calculated by Eurostat.

The graph in the Annex shows annual unadjusted growth rates of the total labour cost index (TOT) and of the National accounts (ESA2010) compensation of employees per hours worked by NACE Rev 2 sections.

The Table shows y-o-y unadjusted growth rates of the total labour cost index (TOT) and of the national accounts (ESA2010) compensation of employees per hours worked by NACE Rev 2 sections.

There are some inconsistencies that predominantly stem from different ways of distribution of employees into industries in the national accounts and the ER (see more in 6.3.1.).
The National Accounts (NA) have produced figures on volumes of employees compensations, ER results on the average hours worked per employee have been used for the denominator, correspondingly to the method of LCI calculation. The increase in differences between NA and LCI is due to a different approach to counting subsidies provided in connection with COVID 19, especially in section I.

Internal coherence (e.g. between nace sections) is guaranteed by using the same data sources and the same methods.

The Labour Cost Index is published in the CZSO Social and Economic Development Indicators of the Czech Republic. Data are available on the web-pages in the form of time series (latest 32 quarters) by NACE Rev 2 sections, seasonally adjusted figures as well as raw series. The short methodological description of the Index is also available there.

There is no national News release for the LCI.

See the Annex.

There is no LCI on the online database.

See 9.3.

There is no microdata access for the LCI.

No other format is used.

Further information on definitions and methods are available at the dedicated section of Czech Statistical Office website.

Eurostat publishes the LCI metadata.

100 %

Optional

LCI is an hybrid index using existing data only.

No confidental data are involved.

The Czech confidentiality rules are presented on the website of CZSO.

The LCI contains aggregated figures, so identification of individual bussiness is not possible.

See the other sections for more information.