Federal Statistical Office of Germany

Section E16 "Structure of Earnings Survey and Labour Cost Survey, Collective Bargaining Earnings Statistics"

65180 Wiesbaden Germany

The labour cost index is not a survey, but a complex calculation system that links existing statistics. The basic parameter of the calculations are "labour costs per hour worked". The labour cost index reflects the quarterly development of total labour costs. It offers two differentiation options: On the one hand, the development of labour costs can be shown broken down by individual economic sections. In addition to time series for the manufacturing industry and the service sector, time series are calculated for the manufacturing industry or the construction industry, for example. On the other hand, the index makes it possible to divide labour cost development into the two main components of labour costs: the index of gross earnings and the index of non-wage labour costs.

The labour cost index allows current insights into the development of the cost situation of the economy. On the one hand, it contributes to the assessment of Germany's international competitiveness as a production location. On the other hand, it signals risks to monetary stability that can arise from cost increases for producers. The labour cost index was mandated by EU regulations and thus allows a timely comparison of the labour cost developments of all 27 European member states on the basis of a harmonised methodological foundation.

The contribution of a mix of sources to the calculation of the German LCI, implies keeping into account a multitude of factors when evaluating the overall quality of the indicator. This report tries to make an overview of these factors, focusing on the main issues affecting the quality outcomes.

The organization of the report follows the criteria on statistics quality defined in the EU regulations and particularly the annex 1 of LCI Regulation No. 1216/2003.

NACE Rev. 2

NACE Rev. 2 Sections B to S.

Labour costs per hour worked and subdivided in salaries and other labour costs.

Enterprises

Enterprises with ten or more employees

Germany

First quarter of 1996 until fourth quarter of 2022.

2016

The following data sources are used in calculating the labour cost index:

Table 1: Database for calculation the labour cost index

As of the 1st quarter of 2022, the basis for the development of earnings was changed from the Quarterly Earnings Survey to the (monthly) Earnings Survey.

The labor cost survey and the earnings survey cover all economic sections relevant for the labor cost index. In contrast to the (monthly) earnings survey, the labor cost survey has a cut-off threshold: it only surveys companies with ten or more employees. In order to ensure a comparable methodology over time after the changeover from the Quarterly Earnings Survey to the Earnings Survey, only establishments with ten or more employees are included in the model calculation by the Earnings Survey for estimating the earnings parameters.

Quarterly calculation

See chapter 3.1

There are many data sources used to compile the LCI time series (see chapter 3.1). These data sources are processed in an extensive and complex way that ensures all statistical problems are identified and addressed correctly and systematically each quarter. A lot of complex preliminary checks aimed at investigating and possibly correcting errors on administrative and survey data are carried on at micro level. Final key checks on macrodata are carried out.

Source data from the labour cost survey is extrapolated using the other data sources mentioned in chapter 3.1

As regards the working day and seasonal adjustment see concept in chapter 6.4.

Not available.

The good quality of the labour cost index is mainly based on the results of the labour cost survey (see also chapter 3. Statistical Processing). The quarterly earnings and the results from National Accounts also contribute to the quality level achieved. For some labour cost components. However, most of which have a small weight, adequate data sources are not available so that estimates are required.

It is generally possible that the non-sampling errors contained in the basic statistics and included in the calculation of the labour cost index are also contained in the results of the labour cost index. In addition, there may be biases caused by estimated additions and deductions, other estimation procedures or updating of time series. Due to this situation, exact quantification is not possible.

In addition to the European Central Bank and the European Commission, the national users are: the German Bundesbank, several federal ministries, economic research institutes as well as enterprises and organisations representing their interests. Private enterprises and public administration make increasingly use of the labour cost index for sliding-price clauses in contracts.

The need for information on the part of the users of the labour cost index can be met.

The need for a more detailed level of NACE has been expressed.

The labour cost index is available for all NACE sections B to S, starting from 1996q1.

The ratio of the number of data cells provided to the number of data cells required is 100%.

The basis used to calculate the labour cost index are the results of the labour cost survey. These data, which are collected every four years, are updated quarterly using various data sources. When new data become available from the labour cost survey, a benchmarking exercise is performed for the years between two labour cost surveys. The extent to which revision is caused by the benchmarking exercise can be an indicator of the labour cost index quality. As regards benchmarking between 2016 and 2020, the following adjustments were required for manufacturing, which is the largest economic sector. Labour costs had to be revised by an average annual 0.5% and hours worked by 0.1%.

Optional

Optional

It is generally possible that the non-sampling errors contained in the basic statistics and included in the calculation of the labour cost index are also contained in the results of the labour cost index. In addition, there may be biases caused by estimated additions and deductions, other estimation procedures or updating of time series. Due to this situation, exact quantification is not possible.

Labour cost surveys 2000, 2004, 2008, 2012, 2016 and 2020

• Detailed information on structure for D.11 and D.12 for the years 2000, 2004, 2008, 2012, 2016 and 2020
•  Coverage: employees of enterprises with 10 and more employees

Earnings Survey (used from 2022q1 onwards)

• Quarterly data for D.11 + D.1221
• Coverage: As of the 1st quarter of 2022, the basis for the development of earnings was changed from the Quarterly Earnings Survey to the (monthly) Earnings Survey.
•  The labor cost survey and the earnings survey cover all economic sections relevant for the labor cost index. In contrast to the (monthly) earnings survey, the labor cost survey has a cut-off threshold: it only surveys companies with ten or more employees. In order to ensure a comparable methodology over time after the changeover from the Quarterly Earnings Survey to the Earnings Survey, only establishments with ten or more employees are included in the model calculation by the Earnings Survey for estimating the earnings parameters.

Table 2: Proportional share of employees represented by the results of the labour cost survey in 2020 and Earnings Survey in 2022 in the number of employees in accordance with ESA 2010

Optional.

See chapter 6.3.1

Measurement errors of the source data can influence the quality of the labour cost index. The main data sources are subject to quality criteria of official statistics.

No data collection is carried out for the labour cost index.

Optional

Optional

Optional

Optional

Optional

The quarterly data are adjusted for seasonal and calendar variations using the internationally common X-13 JDemetra+ analysis method. As a special user service, the Federal Statistical Office also publishes calendar and seasonally adjusted analysis results obtained with the Berlin method (BV 4.1). The two procedures are based on different mathematical-statistical methods and can therefore produce different results.

The working day and seasonally adjusted German LCI time series are produced according to an indirect approach: the total labour cost aggregate by section, as well as the totals of each labour cost component are derived by summing up the related adjusted component series. A hierarchical calculation sequence must be followed: firstly the total labour cost by sections are calculated from earnings and non-wage costs and secondly their aggregation brings to the LCI totals.

See also chapter 6.6 "Data revision"

A revision covers the processes of reviewing and updating results, for instance, by incorporating new data, statistics and/or methods in the accounting system. In this context, a distinction is made between continuous revisions which concern smaller corrections in the data of individual quarters or years and major or “large“ revisions. The latter are basic revisions of the time series as a whole.

Reasons for major revisions are, for example, new statistical bases for calculation which have not been used so far.

The data sources used to calculate the labour cost index (national accounts, employment accounts and the working time and volume of work measurements of the Institute for Employment Research) are again accounting systems using various data sources. At regular intervals, possibly even quarterly, the results are revised. Such revisions are taken over, at least partly, for the labour cost index.

Table 3: Growth rate of the unadjusted time series on the same quarter a year earlier Date of publication: 1st quarter 2020 to 4th quarter 2022

Table 4: Impact of seasonal adjustment on the level of revisions of the Index of labour costs Revision Original-, Sections B to S of NACE Rev. 2, Growth rate on the same quarter a year earlier in % Date of publication 1st quarter 2020 to 4th quarter 2022

Table 5: Explanations for the larger revisions of the years 2007 to 2022

Table 6: Index of labour costs for sections B to S (NACE Rev. 2) by time of release;
Growth rates of unadjusted time-series on the same quarter a year earlier in %

According to the Commission Regulation (EC) No. 1216/2003 of 7 July 2003, Member States shall transmit the quarterly results of their LCI statistics to the Commission (Eurostat) no later than 70 days after the end of the reference period.

Table 7: Timeliness and Punctuality in the delivery of LCI

1) Since the delivery date was a Saturday, Eurostat also accepted 13.03.2023 as a punctual delivery date.

Optional

Optional

See Table 7

See chapter 7.2

With the revision taking effect from the 4th quarter 2004, the computation system of the labour cost index was geared to the results of National Accounts (NA). That caused a high coherency of the two statistics, which is reflected by only small deviations between highly aggregated unadjusted time series of the compensation of employees (NA) and labour costs (labour cost index). Starting with the first publication of the 1st quarter 2010 the calculation of the LCI was changed retrospectively from the 1st quarter 2007: the Quarterly Earnings Survey became the new main data source. Therefore differences have occurred between the LCI and compensation of employees from 2007 onwards (with an exception: from 2012 to 2013 again National Accounts were the fundamental data source for the calculation of the LCI.)

The definitions of the cost components which the labour cost index is based on correspond to those used in the context of the labour cost survey. The definitions of the cost components used in the labour cost survey are harmonized with those of the European System of Accounts (ESA). Hence, there is consistency between national accounts and the labour cost index. The EU regulations concerning the labour cost index form the basis for obtaining results that are comparable at the European level. Regional results for Germany are not available.

Optional

There are conceptual and methodological differences concerning the main data source used to compile LCI between the time series of different quarters. Main data source:

• 1996 - 2006: National Accounts
• 2007 - 2011: Quarterly Earnings Survey
• 2012 - 2013: National Accounts
• 2014 - 2021: Quarterly Earnings Survey
• From 2022 Earnings Survey onwards

Comparable time series are available starting from 1996q1.

See chapters 8.4, 8.5 and 8.6

There are no differences between the growth rates of the 4-yearly labour cost surveys (LCS), the growth rates of the quarterly LCI and the growth rates of the yearly estimations of labour costs. The LCI time series are adjusted by the results of the 4-yearly-LCS (see also chapter 6.1)

Table 8: Cohärenz labour cost index and National Accounts

The development in the 1st and 2nd quarters of 2022 deviate slightly from each other. This can presumably be attributed to the methodological basis changeover (see table 5).

The results of the 4-yearly labour cost survey, the yearly estimations of labour costs and the quarterly LCI are consistent.

Germany published the labour cost Index for the first time on 9 June 2005 when the data were supplied for the 1st quarter 2005. Since then all time series as well as meta data have been accessible free of charge in the data base GENESIS-Online at http://www.destatis.de. Further background information and an example table are available at http://www.destatis.de (Follow: Themes -> labour -> labour costs, non-wage costs).

The publication of the results (70 days after the end of the reporting quarter) is accompanied by a press release.

Not available.

Detailed results of the labour cost index can be obtained free of charge from the Genesis-Online database (http://www.destatis.de/genesis, Code: 62421).

Optional

Optional

Optional

Methodological documentation is available in this report as well as through the national quality report and short description on destatis' website.

Not available.

Optional

Optional

As the labour cost index is based on an accounting system which processes results of existing primary and secondary statistics, respondents are not additionally burdened.

See sub-points

Pursuant to Article 16, para. 1 of the Federal Statistics Law (BStatG), the bodies of German official statistics must keep individual data secret. Exceptions are individual data which cannot serve to identify a respondent or aggregated together with the data of other respondents (tables). The labour cost index is a set of secondary statistics whose data basis largely consists of data published earlier in other statistics.

The labour cost index is largely calculated on the basis of other statistics whose results were published earlier. In this respect, all relevant confidentiality and data protection principles were observed before. Individual data of these statistics are not used.

Not available.