STATISTICS AUSTRIA

Directorate Social Statistics

Living Conditions, Social Protection

Guglgasse 13

1110 Vienna
AUSTRIA

The Labour Cost Index (LCI) has been prepared according to the Regulation (EC) No 450/2003 for the demanded NACE Rev. 2 sections B to S. Different data sources for sections B to F (Short Term Statistics, Labour Cost Survey) and for sections G to S (administrative data, Labour Force Survey, Labour Cost Survey) are used.
Evolution over time of the LCI:

• Standard calculation based on NACE Rev. 1 classification from 1996 until 2008Q1.
• Quality improvements were performed from 2008Q2 to 2008Q4 (transmitted with 2008Q4 by discontinuing two quarters).
• The transition to NACE Rev. 2 was completed in 2009Q1 (2000Q1 to 2009Q1).
• Integration of gross wages data for an additional group of employees due to a law change in December 2009.
• In 2011Q3 the method in sections O and P has been changed and consequently the entire series changed.
• In 2013Q4 imputations in section O have been installed and consequently the entire series changed.
• In 2020Q3 data concerning short-time work allowance were included to correct the labour costs between 2020Q2 and 2022Q4.

For details on calculation procedures, see the Austrian 'Standard Dokumentation' (available in German only) published on the Statistics Austria website.

According to the Regulation (EC) No 450/2003 the following classifications are used:

• NACE Rev. 2 sections B to S
• types of labour costs (total labour costs, total labour cost excluding bonuses, wages and salaries, labour costs others than wages and salaries)
• unadjusted data, calendar adjusted data (not seasonally adjusted data), calendar and seasonally adjusted data

According to the Regulation (EC) No 450/2003 the LCI cover NACE Rev. 2 sections B to S.

The raw statistical data / source data differs depending on the NACE Rev. 2 section. 

Table: Components of the source data

*) Information regarding division between voluntary and compulsory payments to employees leaving the enterprise.
**) Transmitted by the Austrian Business Register.
***) Transmitted by the Main Association of Austrian Social Insurance Institutions.
****) Used till the second quarter of 2011. 

*****) Used between the third quarter of 2020 and the fourth quarter of 2022. 

NACE Rev. 2 sections B to F (industry and construction): Kind-of-Activity unit (KAU).
NACE Rev. 2 sections G to S (trade and services): Enterprise.

All enterprises incurring labour costs for employees.

Austria.

NACE Rev. 2 sections B to N start with 2000, sections O to S start with 2009.

The LCI use the year 2016 as reference year till the fourth quarter 2022 (afterwards the year 2020).

The LCI for NACE Rev. 2 sections B to F is calculated mainly on the basis of the existing survey 'Short Term Statistics'. For sections G to S, mainly administrative data are used. Beside the following description on calculation procedures, you can find information in the Austrian 'Standard Dokumentation' (available in German only).

Characteristics of the source data

Labour Force Survey
The Labour Force Survey provides quaterly information about private households including the hours actually worked in a week.

Monthly wage tax
Monthly wage tax data provide two variables

• Taxes on wages and salaries
• Employers’ contribution to the family burdens equalisation fund (DB )

Whereas the taxes on wages and salaries are only used for occasional crosschecking, our main variable for estimating wages and salaries is DB. DB is calculated on the basis of the sum of wages and salaries of all employees of one employee. We do not need to distinguish between workers, clerks or civil servants because DB is uniformly 3,9% (2017 it was 4.1% and before 2017 4,5%) of the taxable base.

Yearly pay slip statistic
Statutory social security contributions (D.1211) and severance pays (D.1223) based on the system till end of 2002 are estimated by using yearly pay slip data. Pay slip statistics are only available with a substantial time lag.

Social Insurance Data
Social insurance data are used for estimating the number of persons employed in each NACE Rev. 2 section (from the Austrian Business Register) and payments to the new severance pay scheme (from the Main Association of Austrian Social Insurance Institutions).
Information regarding the new severance pay scheme:
In 2003 a new severance pay scheme was introduced affecting all new employees. Although employees who were hired before 2003 can switch from the old to the new scheme not many employees use this possibility. The financing of the new severance pay scheme is defined by a compulsory contribution system. The employers contribute 1.53% of the monthly salary and any additional payments. This amount will be collected by the health insurance carrier and forwarded to the responsible Employee Benefits Fund. The Main Association of Austrian Social Insurance Institutions provides data about total payments made in the new scheme on a monthly basis.

Data on short-time work allowance

This data comes monthly from the Austrian unemployment office and includes the monetary support received and the number of employees for whom the support was paid, per company. The data are used to correct the cost for the employers.

Table: Frequency of data availability

*) Used till the second quarter of 2011.

**) Used between the third quarter of 2020 and the fourth quarter of 2022.

All data necessary for the LCI are secondary data (see '3.1. Source data') and available within Statistics Austria in the demanded time. There is no data collection especially for the LCI. The data will be transmitted electronically.

Data are continuously monitored during the calculation process regarding their plausibility.

Two examples:

• Companies in NACE Rev. 2 sections G to S are excluded if the gross wages per employee are under the first or above the 99 percentile within the same section.
• In NACE Rev. 2 section O missing values of either gross wages or number of employees and additional implausible gross wages per employee get imputated (for detailed information see 6.3.4.1. Imputation - rate).

The data compilation depends on the NACE Rev. 2 sections. The LCI for NACE Rev. 2 sections B to F is calculated mainly on the basis of the existing survey 'Short Term Statistics', for sections G to S mainly administrative data are used.

NACE Rev. 2 sections B to S: 

During the corona pandemic data concerning short-time working allowance from the Austrian unemployment service were used to correct cost components. These data were available starting from the 3rd quarter of 2020 till the 4th quarter 2022 and are used to correct cost components starting with the 2nd quarter of 2020 till the 4ht quarter 2022. 

NACE Rev. 2 sections B to F (industry and construction):

All cost elements and hours actually worked can directly be used from the survey 'Short Term Statistics'. Only the information regarding division between voluntary and compulsory severance payments are taken from the Labour Cost Survey.

NACE Rev. 2 sections G to S (trade and services):

In NACE Rev. 2 sections G to S the variables/data derive from different sources.

Hours actually worked (Labour Force Survey = LFS)
The hours actually worked are used for calculating the volume of work as follows:
'Average number of hours worked per employee in the reference week (LFS)' multiplied by 'Number of weeks in the quarter' multiplied by 'Average number of employees in the quarter (the average number of employees is taken from the Social Insurance Data)'

Wages and Salaries (D.11) (Monthly wage tax)
Employers contribution to the family burdens equalisation fund (DB) is our main variable for estimating wages and salaries. DB is calculated on the basis of the sum of wages and salaries of all employees engaged by one employer. DB is uniformly 3,9% (2017 it was 4.1% and before 2017 4,5%) of the taxable base.
Wages and salaries can therefore be estimated by the simple formula:
Wages and Salaries = DB / 0.039 (the calculated wages and salaries include here also the guarented remuneration in the event of sickness (D.1221))
However, there are some restrictions that have to be taken into account:

• Some cost components are not included in the taxable base. Among these the most important are severance pays (D.1223).
• If the contribution basis in one calendar month does not exceed the amount of 1460 Euros, the basis decreases by 1095 Euros.
• By law enterprises do not have to pay DB in their first year of existence.
• For workers older than 60 no DB has to be paid since 2004. This restriction gets corrected with information from the 'Yearly pay slip statistic'.
• For civil servants no DB had to be paid until May 2008.
• For 'holder of a non-standard contract' no DB was to be paid until December 2009.

Statutory social security contributions (D.1211) and severance pays (D.1223) (Yearly pay slip statistic)
To estimate actual social security contributions we take the following factors into account:

• changes in the rates of contribution for social insurance separately for civil servants, civil servants under contract, wage earners, salary earners and apprentices.
• changes in the maximum assessment basis.

Additionally we assume that the development is similar to the development of wages and salaries.
Severance pays based on the system till 2002 are only available as an annual sum, so the amount is distributed quarterly over the year.

Number of employees and payments to the new severance pay scheme (Social Insurance Data)
Social insurance data are used for estimating the number of persons employed in each NACE Rev. 2 section (from the Austrian Business Register) and payments to the new severance pay scheme (from the Main Association of Austrian Social Insurance Institutions). The Main Association of Austrian Social Insurance Institutions provides the total payments made in the new scheme started with 2003 (explanation at point '3.1. Source data') on a monthly basis, unfortunately not separated  according to NACE Rev. 2 sections. To estimate the distribution to the NACE sections we use information on the distribution of payments in 2007.  

In addition to these previous descriptions on calculation procedures you can find information in the Austrian 'Standard Dokumentation' (available in German only) on the Statistic Austria website.

See point '3.5. Data compilation'.

Information relating to quality can be found on the Quality Guidelines.

Data quality is high, because data collecting and methodological implementation are based on the Quality Guidelines. 

The LCI is for instance used by the special interest groups of entrepreneurs and of employees to update employment contracts and for wage negotiations and by the Austrian Central Bank (Oesterreichische Nationalbank) for various analyses.

No direct measurement is implemented but the feedback suggests that the satisfaction of the main users is high at the most part.

At the delivery of the LCI 2010Q4 the index excluding bonuses in the NACE Rev. 2 sections B to F was sent for the first time. For the remaining NACE Rev. 2 sections G to S it is still not possible to calculate the index 'total labour cost excluding bonuses' (TXB) due to a lack of necessary data sources (according to the result of the feasibility study in regulation (EC) No. 450/2003, article 10).
Beside these limitations Austria is in line with regulations (EC) No. 450/2003, No. 224/2007 and No. 973/2007.

By taking all NACE Rev. 2 sections and different types of labour costs (TOT, WAG, OTH, TXB) as required cells (72) only NACE Rev. 2 sections G to S TXB (13) are not provided. Consequently the ratio is 82%.

The Austrian LCI is based on administrative data and surveys (see point '3.1. Source data') and these data sources have different levels of accuracy.

• The administrative data represent the total population of interest with very high level of accuracy, some process-related reductions of cases can reduce the accuracy.

• The accuracy of the different surveys are considered in general as very high and described in the related descriptions ("Standard-Dokumentation", only available in German) chapter "3.2 Genauigkeit" ("3.2 Accuracy"), added as links below.

Optional

Optional

See following points.

This section examines the divergence between the employment contracts according to National Accounts (NA), ESA 2010 (target group) and the LCI (frame population).

Because of the different data sources (NACE Rev. 2 section B to F: Short Term Statisics, G to S: Austrian Business Register) for calculating the LCI, the coverage (percentage of employment contracts) varies depending on the NACE Rev. 2 section observed.

Table: Share of employment contracts used for LCI compared to National Accounts in 2022

*) “frame population” either from the Short Term Statistics (NACE Rev. 2 sections B to F) or the Austrian Business Register (NACE Rev. 2 sections G to S). Data from the Short Term Statistics are preliminary.

**) Employment contracts in sections O, P and Q are difficult to allocate, therefore these sections are reported together. 

Explanations NACE Rev. 2 sections B to F (industry and construction): 

As demonstrated in the previous table, the percentage of recorded employees in the NACE Rev. 2 sections B to F is by trend slightly lower (particularly in NACE Rev. 2 sections E and F) than in NACE Rev. 2 sections G to S (even if NACE Rev. 2 sections L and S have similar rates). This effect is caused by the fact that the Short Term Statistics is a sample (sample description - only in German available) and the Austrian Business Register represents in principle the basic population.
The lower percentage in sections B to F (based on Short Term Statistics) is adjusted to the slightly higher percentage in sections G to S (based on administrative data) during the calculation process (for detailed information see '8.3. Coherence - cross domain'). 

Explanations NACE Rev. 2 sections G to S (trade and services):

Administrative data (monthly wage tax data, yearly pay slip statistic and data from the Austrian Business Register) theoretically cover the whole economy. But some enterprises are sorted out during the calculation process because of matching problems, late data transmission and plausibility checks, this effect is particularly visible in the NACE Rev. 2 sections L and O to S.

The inclusion of the 'holders of a non-standard contract'* for compiling the LCI and the exclusion of such employees from the national account (ESA 2010) leads to a slight overestimation in the comparison in point 6.3.1. This effect is obviously in section K. By definition the 'holder of a non-standard contract' has more elements of an employee as a free lancer and belongs therefore rather to the frame population of the LCI.

*) Holders of a non-standard contract ('Freie Dienstnehmer', loosely translated 'Free employees') are a special group of persons employed, whose working contracts comprise elements of employees and of free lancers.

Not available.

Not available.

Not available.

Not available.

Not available.

Because the methods of calculation of the LCI differ according to the NACE Rev. 2 section, the explanations are divided into two categories.

NACE Rev. 2 sections B to F (industry and construction):

The main data source for the sections B to F is the monthly Short Term Statistics, which covers all cost items except the information on the segmentation of voluntary and compulsory severance payments (this information arise from the Labour Cost Survey).
The Short Term Statistics includes mainly enterprises with more than 19 employees (some enterprises with less employees are included regarding special rules, see sample description (only in German available)). Up to now there are no estimation procedures implemented for enterprises that are excluded from the sample used, this leads to an error due to the assumption that the excluded companies have the same personnel costs than the included companies.

NACE Rev. 2 sections G to S (trade and services):

The monthly wage tax data (contribution to the “Family Relief Funds”) are used for estimating wages and salaries (D.11). By law employees aged older than 59 are excluded from this kind of wage tax. To counteract this missing part we estimate the percentage of this group from the yearly pay slip statistic of the previous year and add this rate to the recent amount of wages and salaries. This rate differ between the years approximately up to +/-1.9 percentage points depending on the NACE Rev. 2 section - this is the ratio of the possible error.
Data of the Labour Cost Survey is used to estimate voluntary social security contributions (D.1212) and subsidies (D.5). So far, the most reliable results are achieved by assuming that the development of both components is similar to the development of wages and salaries – this means both components are estimated as a constant proportion of D.11. The relative size is taken from the latest Labour Cost Survey and is then held constant for four years.

Only in NACE Rev. 2 section O the following imputations are executed (since the fourth quarter of 2013):

• Missing values of either gross wages or number of employees were imputated by the median of the gross wages per employee of the whole NACE Rev. 2 section O.
• If enterprises have very low or very high gross wages per employee, gross wages and numbers of employees are replaced by the median of NACE Rev. 2 section O. The limits of gross wages per employee are under the 1st percentile and over the 99th percentile multiple 1.5 (hence more lower than higher wages got replaced).

These imputations reduce the number of implausible records and increase the number of companies in the data pool. The ratio of imputated companies in the whole data pool of NACE Rev. 2 section O in 2022 is about 4,8% (varies a little depending on the quarter).

Optional

Seasonal and calendar influences were adjusted with X13-Arima-Seats provided by the US Bureau of the Census. 

Adjustments were performed directly at the level of labour cost components ('wages and salaries' = 'WAG' and 'labour costs other than wages and salaries' = 'OTH'), deriving total costs (='TOT') indirectly by aggregation of the adjusted components. Time series were only adjusted if the regressors were significant and the effect was plausible. Since 2016 trading-day adjustment (6 regressors) was performed for NACE Rev. 2 sections B to G. For short time series (NACE Rev. 2 sections O to S) a one regressor working-day adjustment was performed. In fact, currently working-day adjustment for short time series was only performed for NACE Rev. 2 section O. 

For the adjustment, models and regressors in the program were kept constant for one year. However, parameters were re-estimated each quarter, meaning that all adjusted series changed back to 2000Q1 respectively 2009Q1 (section O) every time trading-day adjustment was performed. Every third quarter models and regressors were adapted, because then revised original data became available for the previous year. 

Further details see in the attached file.

NACE Rev. 2 sections B to S (industry, construction, trade and services):

At each publication the previous quarter is revised. With the third quarter usually all four quarters of the previous year and the two quarters of the recent year get revised.

In 2020, revisions up to two quarters back were possible due to the partially delayed data delivery caused by the corona pandemic.

NACE Rev. 2 sections G to S (trade and services) except O:

Every four years information from the Labour Cost Survey are included. This causes a revision every four years in the third quarter back to the year of the survey. The LCS does not cover businesses from section O, so this section cannot be updated.

The revision practice differs between NACE Rev. 2 sections B to F (industry and construction) and G to S (trade and services) as their data bases are different.

NACE Rev. 2 sections B to F (industry and construction):
The main data source for sections B to F is the monthly Short Term Statistics. For data processing of the Short Term Statistics a three-step procedure is applied:

Step 1: 60 days after the end of the quarter (T+60) only about 60% of all enterprises in the survey have responded and not all plausibility checks have been made yet. Thus, the data are considered preliminary (this concerns the last month in the quarter).

Step 2: 90 days after the end of the quarter (T+90) almost 100% of the enterprises have delivered data – the data are still provisional but more reliable. The exchange of the T+60 data with the T+90 data at the calculation of the index in the following quarter causes the revision of the previous quarter.

Step 3: The final data are available 9 months after the end of the year. The implementation of these data causes the revision of the whole previous year at the calculation of the 3rd quarter.

NACE Rev. 2 sections G to S (trade and services):
In principle, revisions are caused by updates of the following data sources:

• Monthly wage tax statistics: By law enterprises have to pay the contribution to the “Family Relief Funds” 15 days after the end of a month. Many payments are done later which causes revisions of the index to the previous quarter.
• Annual pay slip statistics: Data are available about 9 months after the end of the year. This causes a revision of the whole previous year in the third quarter of the current year.
• Every four years information from the Labour Cost Survey (share of D1212 and D5 compared to D11) are included. This causes a revision every four years in the third quarter back to the year of the survey. The LCS does not cover businesses from section O, so this section is not updated.
• Data from the Austrian Social Insurance Institutions provide information on the number of employees per enterprise. Statistic Austria receives the data about 50 days after the end of the quarter. No updating is done.
• The Labour Force Survey provides information about the number of hours actually worked. Due to the fact that the final version of the corresponding datasets is available 90 days after the end of each quarter, we have to use preliminary data. However, from experience we know that the data we use are hardly different from the final ones and therefore no updating will be executed.

The tables in the added excel file show the revisions in the published labour cost index using the unadjusted series. Revisions are highlighted in grey.

Revisions in the third quarter are done according to the revision procedure described in previous point 6.6. 

Extraordinary 'revision' in 2019Q1:

In the first quarter of 2019, the reference year was changed from 2012 to 2016, which resulted in a change in the entire series. However, this was not a revision, it was just a shift in the index series.

Revision in 2020Q3, 2021Q3 and 2022Q3:

Due to the usual revision procedure in the third quarter the index values were revised back to 2020Q1 respectively 2021Q1 respectively 2022Q1.

Revision in 2020Q4:

Due to the updating of the data on short-time working allowance, some sections have been revised two quarters back till 2020Q2.

See point 7.1.1.

The index are published at the website of the Statistic Austria approximately 78 days after the end of the quarter, depending on the weekday.

NACE Rev. 2 sections B to F (industry and construction) and O:

Final results are available 6 quarters after the first publication at the latest. For instance the values of the publication 1Q2012 get revised the last time with the publication of the 3Q2013.

NACE Rev. 2 sections G to S (trade and services) except O:

Final results are available 10 quarters after the first publication at the latest. For instance the values of the publication 1Q2012 get revised the last time with the publication of the 3Q2014.

Further details about the reasons for revisions see section '6.6. Data revision - practice'.

Target date to deliver data to Eurostat are 70 days after the end of quarter.

2012Q1: t+ 69 days

2012Q2: t+ 68 days

2012Q3: t+ 71 days (deadline on a weekend)

2012Q4: t+ 70 days

2013Q1: t+ 67 days

2013Q2: t+ 71 days (deadline on a weekend)

2013Q3: t+ 65 days

2013Q4: t+ 65 days

2014Q1: t+ 66 days

2014Q2: t+ 65 days

2014Q3: t+ 65 days

2014Q4: t+ 68 days

2015Q1: t+ 70 days

2015Q2: t+ 69 days

2015Q3: t+ 63 days

2015Q4: t+ 63 days

2016Q1: t+ 69 days

2016Q2: t+ 67 days

2016Q3: t+ 68 days

2016Q4: t+ 67 days

2017Q1: t+ 68 days

2017Q2: t+ 67 days

2017Q3: t+ 66 days

2017Q4: t+ 67 days

2018Q1: t+ 67 days

2018Q2: t+ 72 days (deadline on a weekend)

2018Q3: t+ 66 days

2018Q4: t+ 64 days

2019Q1: t+ 67 days

2019Q2: t+ 65 days

2019Q3: t+ 65 days

2019Q4: t+ 65 days

2020Q1: t+ 65 days

2020Q2: t+ 65 days

2020Q3: t+ 68 days

2020Q4: t+ 68 days

2021Q1: t+ 69 days

2021Q2: t+ 68 days

2021Q3: t+ 68 days

2021Q4: t+ 67 days

2022Q1: t+ 69 days

2022Q2: t+ 67 days

2022Q3: t+ 66 days

2022Q4: t+ 67 days

2023Q1: t+ 66 days

See above

Statistics Austria publishes only on national level. Eurostat compiles a comparison of the LCI between all participating countries.

Not available.

Comparability between two consecutive quarters
In order to compare two consecutive quarters it is essential to view the history of the LCI:

• Standard calculation based on NACE Rev. 1 classification from 1996 until 2008Q1.
• Quality improvements were performed from 2008Q2 to 2008Q4 (transmitted with 2008Q4 by discontinuing two quarters). This led to a revision of the entire time series (1996Q1 to 2008Q4) of the LCI based on NACE Rev. 1.
• The transition to NACE Rev. 2 was completed in 2009Q1 (2000Q1 to 2009Q1).
• Integration of gross wages data for an additional group of employees due to a law change in December 2009.
• In 2011Q3 the method in sections O and P has been changed and consequently the entire series was changed.
• In 2013Q4 imputations in section O has been installed and consequently the entire series was changed.
• In 2015Q1 the 'hours actual worked' changed in NACE Rev. 2 sections G to S and consequently the entire series was changed.
• In 2017Q1 the 'hours actual worked' changed in NACE Rev. 2 sections G to S and consequently the entire series was changed.
• In 2018Q1 the matching table get updated in NACE Rev. 2 sections O and consequently the year 2016 and 2017 changed.
• In 2021Q1 the question 'hours actual worked' from the labour force survey was asked more precisely.

This history of adaptation in the calculation resulted in limited comparability between several consecutive quarters. In the following part the background of the major adaptations will be outlined.

Transition to NACE Rev. 2

Due to the implementation of the NACE Rev. 2 two methods for back casting had to be combined - the micro-approach (2007Q4 to 2004Q1) and the macro-approach (2003Q4 to 2000Q1). Therefore the comparability between these two time periods is limited.

Change in method in NACE Rev. 2 sections O and P

In 2011Q3 the method in sections O and P has been changed based on an analysis of the past time series. The revised method does not include corrections between sections O and P anymore. Thus, it is coherent with the method used for G to N and Q to R. As a consequence we accept that a certain number of statistical units related to section P (e.g. primary schools) are now assigned to section O. Therefore, the comparability of the LCI in sections O and P before and after the publication 2011Q3 is limited. 

Change in method in NACE Rev. 2 section O (see also point '6.3.4.1. Imputation - rate')

• Missing values of either gross wages or number of employees were calculated by using the median of the gross wages per employee of the whole NACE Rev. 2 section O.
• If enterprises have very low or very high gross wages per employee, gross wages and numbers of employees are replaced by the median of NACE Rev. 2 section O. The limits of gross wages per employee are under the 1st percentile and over the 99th percentile multiple 1.5 (hence more lower than higher wages got replaced).

Therefore, the comparability of the LCI in section O before and after the publication 2013Q4 is limited.

Change of data source in NACE Rev. 2 sections G to S

Due to changes in the weighting procedure of the micro-census (introduced 2014Q4, implemented 2015Q1) the used variable for the LCI 'hours actual worked' changed as well. This led to a revision of the entire LCI series back to 2000Q1 in NACE Rev. 2 sections G to N respectively 2009Q1 in NACE Rev. 2 sections O to S.

Therefore, the comparability of the LCI in sections G to S before and after the publication 2015Q1 is limited.

Change of matching method in NACE Rev. 2 sections G to S

Due to technical changes in the business register the method to merge different data sets had to be changed in 2017Q1; this change affects all series and all quarters 2016 in NACE Rev. 2 sections G to S.

Therefore, the comparability of the LCI in 2016 sections G to S before and after the publication 2017Q1 is limited.

Updating of matching table in 2018Q1, section O:

Due to the implementation of the NACE Rev. 2 two methods for back casting had to be combined - the micro-approach (2007Q4 to 2004Q1) and the macro-approach (2003Q4 to 2000Q1). Therefore the comparability between these two time periods is limited.
The NACE Rev. 2 sections O to S start with 2009, all other sections start with 2000.

The average annual change in labour costs per hour worked (see attached table) was 2.8% in the LCS and 4.0% in the LCI (difference: 1.2 percentage points) between 2016 and 2020 as a whole for the economic sectors presented (sections B to S except O). The annual rate of change differed most in section 'Real estate activities' (L) with 4.3 percentage points.

Reasons for different rates of change: 

Differences in the definition of labour costs (D.2 vocational training costs and D.3 other expenditure paid by the employer do not feature in the LCI) and statistical unit (LCS publish/calculate at 'local unit' and LCI at 'KAU' or 'enterprises', depending on the NACE Rev. 2 section (see table above)) have some effect on labour costs per hour worked. Methodological peculiarities of the LCI, such as the inclusion of micro-enterprises and 'holders of a non-standard contract'*⁾, may cause differences by comparison with the LCS. The impact of restricting the LCI in NACE Rev. 2 sections G to S to enterprises which were also represented in the previous year's data pool and of KAUs/enterprises that change NACE Rev. 2 section is discernible but cannot be measured. The reasons for the difference mainly stem from the fact that the LCI uses other data sources. In the industries and construction sector, data from the Short Term Statistics are used, where the sample of enterprises (20 and more employees and other sample criteria) are extrapolated using administrative data, and in the services sector it is administrative data that are used to estimate labour costs and data from the Labour Force Survey for hours worked. The LCS is used only to a small extent for calculating the LCI; the LCS levels are not used in the LCI, as the LCS does not cover enterprises with fewer than 10 employees and is available only every four years.

*) Holders of a non-standard contract ('Freie Dienstnehmer', loosely translated 'Free employees') are a special group of employed persons, whose working contracts comprise elements of employees and of free lancers. By definition the 'holder of a non-standard contract' has more elements of an employee as a free lancer and is therefore referred to the frame population by calculating the LCI.

The average annual change in labour costs per hour worked (see annex) was 2.8% in the Labour Cost Statistics (annual) and 3.3% in the LCI (difference: 0.5 percentage points) between 2008 and 2022 as a whole for the economic sectors presented (B-N, P-S). The average annual rate of change differed most in sections 'Education' (P) with 2.9 followed by 'Real estate activities' (L) with 2.0 percentage points. 

The reasons for different rates of change are the same like mentioned at point 8.3. because the Labour Cost Statistic (annual) is mainly based on the Labour Cost Survey.

The coherence of the cost element D1 between LCI and NA according to ESA 2010 in the years 2009 till 2022.

Table: Coherence yearly growth rates between LCI Total unadjusted and National Accounts (based on ESA 2010) compensation of employees (D.1) per hours worked

The general trend in the chart of the yearly growth rates at the NACE Rev. 2 aggregate is similar for both sources, but with some differences between the growth rates (more visible in NACE Rev. 2 sections with few companies). There are several explanations for discrepancies between the two indicators: The displayed table is additional available in the added Excel file. This Excel file includes also charts who shows the developments of these two data sources of the aggregate B to S and some important NACE Rev. 2 sections.

• Definitions to create LCI and NA are slightly different.
• Data sources to create LCI and NA are different.
• Estimation methods are different.

Optional

No news release is published.

Data are available on the website of Statistics Austria 'www.statistik.at' > 'Statistics' > 'Labour market' > 'Labour costs and index of agreed minimum wages' > 'Labour Costs' > 'Labour cost index'.

Some of the working day adjusted NACE Rev. 2 sections and aggregates (B-N, B-E, F, G, I, H) are available at 'STATcube – statistical database' > 'Guest Login 'Login'' > 'Accept terms and conditions' > 'Pbulications and services' > 'Austrian Economic Atlas' > select categorie of interest, for instance '07 Key data industry' > 'Standardtabelle / Default table (defaulttable_dewatlas7)' > 'Open Table' > in this example column 'Labour Cost Index (2016 = 100; ÖNACE B-E)' shows the LCI.

https://www.statistik.at/en/databases/statcube-statistical-database/free-access

Optional

Optional

Optional

Descriptive text and methodological information see the Austrian 'Standard Dokumentation' (describtion of the methodology available in German only, overview of Standard-documentation also in english available).

Statistics Austria is committed to ensuring the highest quality with respect to the compilation of statistical information. Further information at the website of Statistics Austria 'www.statistik.at' > 'About us' > 'Responsibilities and principles' > 'Standards' > 'Statistics Austria's quality Guidelines' > 'Quality Guidelines'.

All suitable points were filled out.

Optional

Cost
Only secondary data is used. Therefore, the costs are mainly generated by the staff of the project.

Burden
No additional burden on the respondents are generated because of using administrative data and existing surveys (overview of data sources see table at '3.1. Source data').

The staff of the Statistics Austria is bound by strict confidentiality guidelines, described in the SUMMARY of Quality Guidelines.

Beside these guidelines the number of observed units in each NACE Rev. 2 section is at least 164 (smallest section is B), therefore no economic entity can be identified. No micro data are available to users.

The staff of the Statistics Austria is bound by strict confidentiality guidelines, described in the SUMMARY of Quality Guidelines.

Beside these guidelines the number of observed units in each NACE Rev. 2 section is at least 164 (smallest section is B), therefore no economic entity can be identified. No micro data are available to users.

No comment.