INSEE

DIRECTORATE OF DEMOGRAPHIC AND SOCIAL STATISTICS

Department for Employment and Working Income

Wages and Labour Earnings Division

DG75-F240 - 88 avenue Verdier 92120 Montrouge - France

The Labour Cost Index measures the quarterly variation of hourly labour costs, in compliance with the EC Regulation No 450/2003 (specified by the implementation regulation n°1216/2003 and modified by the Regulation No 224/2007 which changes the scope of NACE sections covered by the indices and by the Regulation n°973/2007 which specifies the NACE version to use).

Indices are available for total labour cost, wages and salaries, and other labour costs, i.e. employers social contributions plus taxes minus subsidies, according to the EC Regulation No 450/2003.

European classification of economic activities: NACE Rev. 2.

France, NACE Rev. 2, sections B to S.

Statistical concepts and definitions follow the CE regulation 450/2003.

Chained Laspeyres prices indices for each section of the NACE Rev. 2 are computed by applying the formula detailed in the European Regulation.

The statistical units are the local units (firm establishment) but only aggregated data by division of the NACE Rev. 2, which serve as inputs to the construction of the indices, are transmitted to INSEE.

All local units with one employee or more in sections B to S of NACE Rev. 2.

France (metropolitan France and overseas departments).

Index series start from the 1^st quarter of 1998 for sections B to N and R to S of the NACE Rev. 2 and from the 1^st quarter of 2008 for sections O to Q.

The labour cost indices are Laspeyres indices in chain-linked prices. As required by Eurostat, the indices are rebased every 4 years. As a result, from the release of the first quarter of 2023, the base year for the labour cost index is now 2020 (vs 2016 previously). More specifically, the average of the four 2020 quarterly indices, weighted by the volume of work (total number of hours paid for each quarter), is equal to 100.

Most of the input data are of administrative origin.

Worked hours (until the fourth quarter of 2019) and some cost elements not available in administrative sources come from two types of statistical surveys: the surveys on labour activity and employment conditions (ACEMO) and the surveys on labour costs (LCS). In 2020, the ACEMO survey has been replaced by an ACEMO-Covid flash survey dedicated to the health crisis, causing an interruption in the availability of hours data in the first half of 2020. To fill this gap, paid hours data from the DSN have been mobilised and have since become a permanent source for this type of information: in fact, they better reflect variations in the volume of work than those previously obtained from the ACEMO survey, which only tracked collective working hours and excluded individual variations such as overtime or partial activity). DSN, or nominative social declarations, are social declarations that all companies are required to send to the social administrations and that INSEE then processes, containing a set of data generally taken from companies' payroll software.

For the non-farming business sector as well as for the private non-business sector (sections B to N + OPQ private + RS), the data on the gross wage bill, the number of employees and the reduction in employers' social security contributions come from the URSSAF Caisse nationale (formerly the Agence Centrale des Organismes de Sécurité Sociale, ACOSS), which centralises this information for the purpose of collecting contributions. Companies covered by the general social security scheme are also required to transmit this information to this body on a monthly or quarterly basis.

For local units belonging to the public sector in sections O, P, or Q of the NACE Rev. 2, wage bills and number of employees, used until the Q2 2022 release, were mainly based on the “summary contributions forms” ("Bordereaux récapitulatifs de cotisations" in French, BRC) sent by the local units to the social security organisations. From the Q3 2022 release onwards, data come from infra annual DSN data, thanks to the generalisation of the DSN to the public sector, starting in 2020.

Social security contributions on behalf of the employer (family, healthcare, unemployment, basic pensions, supplementary pensions, etc.) are first computed by applying the statutory rates. Then, tax and social contribution reduction amounts provided by the URSSAF Caisse nationale are subtracted, together with an estimate of the employer contribution for family allowance reduction. The latter is obtained by applying the statutory reduction rate to an estimate of the total amount of gross wages eligible to that reduction. The apprenticeship tax (including the additional tax), the vocational training tax, taxes on construction and housing (FNAL: National Housing Assistance Fund + part paid to operators collecting employers’ contribution to the construction effort) are also calculated by applying statutory rates. The amounts for transport payments and for accidents at work, variable and therefore not computable by applying the statutory rates, are provided by the URSSAF Caisse nationale.

Since 1^st January 2016, private sector employers have had the obligation to offer complementary health insurance to their employees, and have been required to contribute to, at least, 50 % of the cost. Complementary health insurances (optional ones) are already taken into account in labour cost indices through the LCS information. Since Q1 2016 release onwards, an estimate of the additional cost implied by this extended scope is added.

Since January 1^st, 2019 the tax credit for competitiveness and employment (CICE), set up in 2013, has been abolished and replaced by a 6 percentage point reduction of the health insurance employer contribution (rate cut and base equal to those of the CICE). First, in 2019, this new reduction has been estimated with the CICE firm declarations to the URSSAF Caisse nationale for 2018. Then, from the Q1 2020 release onwards, this reduction has been computed retrospectively for data concerning Q1 2019 and next quarters, based on the healthcare contribution bases provided by the URSSAF Caisse nationale.

Other labour cost components (supplementary social security contribution of behalf of the employer, imputed social contributions (excluding for apprentices), other expenses and taxes on behalf of the employer, such as the payroll tax) are estimated on the basis of the Labour Cost Survey (LCS) information.

The effects of other measures have been also estimated to be included in the calculation of the index. These include the exceptional purchasing power bonus (PEPA) and its successor, the Value Sharing Premium (PPV), as well as support measures introduced in the context of the health crisis (the "1 young person 1 solution" plan to assist in the hiring of young employees, the plan to support the tourism sector via exemptions or aid for the payment of contributions). The impact of the PEPA/PPV bonuses has been estimated directly from the amounts declared in the DSN; the effect of the “1 young person, 1 solution” plan has been estimated from the DARES figures for hiring young people on permanent and fixed-term contracts and on sandwich course contracts; the resulting estimate has been compared with other figures (notably the national accounts), which ultimately provided similar assessments. The other support measures related to the health crisis have been gradually reduced as the health crisis came to an end; however, some employers still have the opportunity of charging to 2022 the remainder of the aid for the payment of contributions to which they were entitled in 2021 and from which they have not yet benefited fully.

Finally, in the context of the health crisis, mainly in 2020 and 2021, the short time working scheme is one of the main business support measures implemented; it requires almost no specific treatment to be taken into account in labour cost indices. Indeed, as short-time working compensations are paid by the public administrations (State, UNEDIC), they are no part of the cost of labour supported by companies. The impact of short-time working compensations on the hourly cost of labour is measured through the concomitant fall in the wage bill paid by employers and in the corresponding hours paid. The data on short-time working compensations, taken from the DSN, were essentially used to compare them with the decreases in the wage bill and paid hours in order to validate the latter's evolution. The impact of the short time working scheme on hourly labour costs has become relatively small again in 2022, but it still affects year-on-year changes in some sectors as a result, especially in the first half of 2022.

Table 1: Variables and sources

The following table contains, by labour cost component, the sources of information, the frequency of updates, and the timelag following the end of the reference period.

Gross wages stand for about 65% of the total cost, employers' social contributions for about 40% of gross wages, i.e. for about 25% of the total cost. These two components are observed and updated on a quarterly basis. The remaining 10% of the total cost consist of components, that are observed and updated annually or every four years.

Table 2: Origin, frequency and timelag of information on labour cost components

* BRC : summary contributions forms (“Bordereaux Récapitulatifs de Cotisations” in French).

** PIPA survey : survey on profit-sharing, participation, company savings plans and employee shareholding.

Q indicates that the reference period is quarterly.

A indicates that the reference period is annual.

The timelag is expressed in days after the end of the reference period.

Note: The data on taxes and allowances on wages are updated every four years according to the LCS surveys. The reference period is the year. The data become available about one year and half (500 days) after the end of the year.

1The following statutory rates are updated: UNEDIC unemployement contributions (unemployment insurance + AGS (insurance against unpaid wages)), AGIRC contributions (AGIRC = General Association of Pension Institutions for Middle Management), ARRCO contributions (ARRCO = Association for Employees' Supplementary Pension Scheme), social security contributions (sickness insurance, maternity, invalidity, death + old age + family allowances + accidents at work + self reliance solidarity contribution), the apprenticeship tax (including the additional tax), vocational training tax, housing construction tax (FNAL + PEEC) and professional and union organisations funding tax (since Q1 2015).

2The URSSAF Caisse nationale provides us with provisional data at best on Q+52, and with final data on about Q+80. Provisional data on public sector (DSN) are available on Q+60.

See table in section 3.1.

Each quarter, a validation procedure analyses LCI data and final indices quality.

By section of the NACE Rev. 2.The denominator is the total worked hours. The numerator adds the total amount of gross wages to the taxes and employer contributions after deduction of subsidies received.

Indices are working day adjusted and seasonally and working day adjusted.

Usually, quality is assessed quarterly through documents detailing the main revisions and changes that have occurred since the previous quarter.

The quality monitoring relies on the traceability of the production with quaterly reports and regular monitoring of legislation changes; both internal and external quality evaluation of statistical methods according to the European Statistics Code of Practice with the support of the INSEE department of statistical methods; comparison with other sources like the national accounts. Further, the LCS, on which the costs components not annually followed are based, is submitted to the Label committee every four years, which judges on the survey quality and can deliver a mandatory reply aspect. Up to now, the LCS has always obtained this label of statistical quality and the mandatory response aspect.

The quarterly quality report analyses and documents labour cost index variations. It contains preliminary checks run on input data (macro data in level, year-on-year and quarter-on-quarter variations) to ensure the quality of input series. Then, labour cost index revisions are calculated and documented.

Labour and social legislation changes are followed continuously over the year. Once a year, all legislation changes are precisely documented. The report is double-checked within INSEE.

The effects of changes of methodology or of important legislation measures are estimated and documented in detailed reports. Synthetized reports are then released on the INSEE web site.

See below.

The labour cost index has three distinct purposes: to inform the public, to facilitate the adjustment of contracts, and to be used in economic analysis.

• GENERAL INFORMATION FOR THE PUBLIC

Since September 2007, an edition of "Informations Rapides" (INSEE publication to disseminate economic indicators) on the labour cost index (LCI) has been published quarterly: "Indice du coût du travail dans l’industrie, la construction et le tertiaire" ("Labour Cost Index in industry, construction and services"). It is released on about Q+75, this date is almost the same as that of Eurostat’s News Release.

• ADJUSTMENT OF CONTRACTS

In France, this function is currently fulfilled by the Revised Labour Cost Index - All Employees (ICHTrev-TS), which has replaced the ICHT-TS since January 2009. The purpose of contract indexation sits uneasily with the adjustability of the Index. The LCI can, indeed, be reviewed. It therefore appeared necessary to publish an index similar to the LCI (using the same sources and based on the same concepts), but not reviewable, by construction (an exception has been made during the 2020 health crisis), this is the ICHTrev-TS.

The European regulations are respected.

100%.

See below.

In addition to some lacks of scope coverage (see 6.3.1), the main LCI shortcoming has been, historically, the lower reliability of the data concerning the O, P and Q sections of the NACE Rev. 2. Indeed, information on those sections, which mixes the private and the public sectors, is provided by several sources. Because of discrepancies between sources, the lower quality of the gross wage data in the BRC, and - until now - the relatively brief history of HERACLES data (infra annual DSN data on the public sector), INSEE has requested for the data of sections O, P and Q not to be released. Expertise work is in progress in order to lift this reservation by using the new administrative source (DSN), available for the whole public sector. For the public sector, employer contributions rates are calculated using national accounts data.

Not relevant (mainly comprehensive data).

Not relevant (mainly comprehensive data).

See below.

The Mayotte department is excluded of most of the sources, due to declarative exemptions in this territory leading to a lack of data.

The ECMO (annual) and the ACEMO (quarterly) surveys cover employees of firms with 10 or more employees. However, Acemo is no longer used from 2020. Firms with 10 employees or more stand for between 55% and 93% of the employees depending on the industry. Therefore, the information coming from these surveys can reasonably be extrapolated to the total scope including firms with fewer than 10 employees.

Table 3: Ratio between the number of employees of firms with 10 employees or more and the total number of employees in the industry, 2019 (source: All employees databases – Base Tous salariés, BTS)

Source: Nominative Social Declarations (DSN) and Annual Declaration of Social Data (DADS) for 2019.

Note: In manufacturing (BE), the number of employees in firms with 10 employees or more according to the BTS stands for 89.3% of the scope.

See 6.3.1.

Not relevant.

•  Worked hours

Prior to September 2007, the labour cost index used a series seasonally adjusted, which monitored the contractual weekly working hours of full-time employees in firms with more than 10 employees or more. This indicator only provided a partial measure of short-term fluctuations in the number of hours worked (as it did not take into account overtime, part-time work, or firms with fewer than 10 employees).

In order to remedy this situation, since September 2007, the labour cost index has used a new series of data on hours worked, which is calculated encompassing all firms (the quarterly results of the ACEMO survey are adjusted from informations on firms with fewer than 10 employees collected in the yearly ACEMO-TPE survey) and taking into account the working time of part-time workers and overtime payment. This indicator does not cover the public sector (sections O, P and Q), for which an estimate is calculated.

In 2017, some parameters which intervene in the computing of the working hours series were updated (share of workers in firms with less than 10 workers, for instance). The series were also rebased on Q1 2018. The scope of the survey on very small firms has been redefined (breaks in terms of the scope of interest of the survey have been corrected).

Finally, a new administrative source for hours is used from 2020 onwards: Nominative Social Declarations (DSN). The paid hours in the DSN used from Q1 2020 onwards also cover cyclical overtime hours and exclude short-time working hours, thereby further improving the coverage of the required hours concept compared to what was done until 2019.

•  Administrative data

The wage bill, provided by the URSSAF Caisse nationale, corresponds to the gross social security base. It excludes employee savings plans (incentive pay, profit-sharing, employer contribution to savings plans), which are not submitted to social security contributions. Apart from this limitation, the wage bill complies with the D11 item of the labour cost classification. Reduction of contributions and exemptions are deducted from the theoretical contributions calculated using the statutory rates and wage bills. They are consistent with European labour cost concepts. Breaks in series have been corrected.

Since 2009, the employee savings plans have been submitted to a new tax, the “Forfait social”. The Forfait social contribution rate was 2% in 2009, and had increased regularly to reach 20%¹ in August 2012. The URSSAF Caisse nationale provides quarterly the Forfait social amounts. Considering that employees saving plans are not covered by LCI indices yet, the Forfait social amounts are not accounted for in the LCI indices neither. The objective is to integrate eventually both employee savings plans and Forfait social amounts in the LCI indices. However, this integration will demand dramatic methodological investments, which may be implemented during a major revision of the index.

1Unless few exceptions with lower rates (8 % or 16 % for some specific payments).

Not relevant.

Not relevant.

Not relevant.

See below.

Employers’ social security contributions are estimated by applying statutory rates. Then, reductions and exemptions provided by the URSSAF Caisse nationale are substracted. Some cost components are estimated when there is no information available in administrative sources or when they are not easily calculated with legal rates. In such cases, the LCS information is used and those cost components are assumed to stand, on average, for the same share of wage as the one observed in the survey.

For the Q1 2023 LCI release, in addition to the change of the base year for the labour cost indices (now 2020, vs 2016 previously), the labour costs index series have been revised for every NACE Rev. 2 sections in order to integrate information from Labour cost survey (LCS), in the same way as for the previous editions. Usually, the LCI results integrated into the labour cost indices include results for both years of collection (for example, the data integrated for 2016 included data relating to LCS 2015 and LCS 2016¹). However, due to the disruption caused by the health crisis in 2020, the LCS data included in the cost of labour in 2023 relate solely to the LCS results for 2019. The imputed employer contribution rates based on LCS surveys are calculated as (D1212 - complementary health insurance mandatory contribution² + D122 - family supplement* + D3 + D4 – Fnal^* – employers’ contribution to construction* – amounts for transport payments* – D5)/(D1111 + D1113 + D1114 + D112). The last edition of LCS available being 2019 (and no longer 2016), the imputed employers’ contribution rates have been revised (in 2023Q1 release) for each year from 2017 onwards: the employers’ contribution rates of years 2017 and 2018 are computed by means of linear interpolation of the rates calculated with LCS - as described above - for 2016 and 2019. For 2019 and the next years, the rates of LCS 2019 are used as imputed rates. This leads to slight upward revisions in average annual changes of the labour cost index over the period 2017 to 2019 (around +0.1 points each year).

When changes in the social legislation on the labour market occur (new measures, substantial modifications or ending of existing measures), their impacts on the labour costs are estimated.In the context of the Covid-19 pandemic, the government has announced various measures to support firms:

• The "1 young person, 1 solution" plan, launched at the end of July 2020, aimed at promoting the professional insertion of young people, provides, under certain conditions, for the payment of a bonus, capped at 4,000 euros, to employers hiring a young person under the age of 26. More specifically, this bonus is paid to employers for the recruitment, between the 1^st August of 2020 and the 31 May of 2021 and for a period of 3 months or more, of a young person under 26 years of age and paid up to 2 times the minimal wage. It amounts to 1000 euros per quarter of presence of the employee in the company for a maximum of 4 quarters. The "1 young person, 1 solution" plan also provides for exceptional aid for the recruitment of a sandwich course student from the 1^st August of 2020. For the recruitment before the 31^st December 2022, aid was capped at 8,000 euros if the student is of age (5,000 euros otherwise); from the 1st January of 2023, it has been capped at 6,000 euros regardless of the age of the student. This bonus is paid for each month of the sandwich course student's presence for a maximum of one year.

These two measures of the "1 young person, 1 solution" plan represent 4.1 billion euros in 2022 (against 5.3 billion in 2021). The reduction in the amounts paid in 2022, compared to 2021, has an upward impact of 0.2 points on the change (as an annual average) in the cost of labour in the non-agricultural market sector between 2021 and 2022.

• The plan to support tourism and cultural and sporting events companies, known as the "Tourism" plan, includes two provisions that had an impact on labour costs from 2020 to 2022:
  • first measure: the exemption from employer social security contributions for the periods from February to May 2020 and from September 2020 to May 2021 (up to Q1 2022 for certain specific activities: cruises and trade fairs). This exemption is planned for companies with less than 250 employees in the sectors most affected by the crisis, identified in the so-called "S1" and "S1 bis" lists, which mainly include trade, accommodation and food service activities, and administrative and support services (the lists of sectors concerned, which were initially drawn up in May 2020, have been expanded in 2021).
  • second measure: aid for the payment of contributions equal to 20% of activity income over the months of February to May 2020 and September 2020 to May 2021 (up to Q2 2022 for certain specific activities), to pay contributions for 2020, 2021 and 2022. The companies concerned by the payment aid are the same as those for the exemption from contributions. However, for companies with less than 11 employees not on the S1 and S1bis lists but whose activity involves receiving the public and has been interrupted, these measures also apply, solely for the months from February to April 2020 as well as for the periods during which their activity was interrupted during the second wave in autumn 2020.

The "Tourism" plan is quantified on the basis of data received from the URSSAF Caisse nationale for an amount of 0.6 billion euros in 2022 (against 2.8 billion euros in 2021). Given the drop in the amounts in 2022, it therefore contributes to an increase of +0.3 points to the change (as an annual average) in the total LCI of the non-agricultural market sector between 2021 and 2022.

It should also be noted that the short-time working scheme, which existed marginally before the health crisis but was reinforced when it emerged, led to the substitution of compensation, which is not considered as wages, for part of the wages. The explosion in the use of short-time working, mainly during the first lockdown in the spring of 2020 and, to a lesser extent, during the successive periods of health-related restrictions, resulted in unprecedented decreases in the wage bill paid by employers and in the corresponding volume of paid hours (and as a result, in increases in these quantities when the use of short-time working diminished). This partly explains the greater volatility of labour cost indices in 2020 and 2021. The rate at which the State covers short-time working allowances (100% at the height of the health crisis) was reduced during the year 2020 or in 2021 in most sectors, a measure that was integrated into the calculation of labour cost indices.

• The exceptional purchasing power bonus (PEPA) voted by Parliament at the end of 2018 was renewed twice (in 2020 and between 1st June 2021 and 31^st March 2022): the scheme provided initially that all companies could pay up to 1,000 euros in bonuses exempt from income tax and social security contributions, per employee whose remuneration was below 3 Smic (minimum wage). At the beginning of 2020, the scheme was renewed only for companies that had signed a profit-sharing agreement, before being extended, in the context of the health crisis, to all companies (the threshold being raised to 2,000 euros for those that had signed a profit-sharing agreement). The scheme in effect in 2020 has been extended on similar terms for the period from 1^st June 2021 to 31^st March 2022. From 1^st July 2022, the Value Sharing Premium (PPV) replaced the special bonus for purchasing power; it is based on the same principles, with the extension of the scheme to all employees (including those earning more than 3 Smic) and a tripling of the payment caps: between July 2022 and December 2023, companies may pay each employee up to 3,000 euros in bonuses per calendar year (and even 6,000 euros for companies with a profit-sharing agreement), exempt from income tax and social security contributions (the bonus is tax-free only for employees earning less than 3 times the minimum wage). These bonuses are exempt from income tax and social security contributions. As such, the payments of theses bonuses are not included in the gross payroll subject to contributions collected by the URSSAF Caisse nationale and used to calculate the index. However, they are provided by another URSSAF Caisse nationale information system: in order to integrate them consistently into the calculation of the LCI, the amount of the PEPA/PPV bonuses is divided by the wage bill provided by this other information system and the bonus "rate" thus obtained is applied to the wage bill in the numerator of the LCI.

In 2022, the amounts of PEPA/PPV bonuses paid were particularly significant, especially in the fourth quarter: for this year, these payments contributed to +0.5 pp on the increase in the LCI – wages and salaries for the entire non-farm business sector (B to S) but, being free of social security contributions, only to +0.2 pp on the increase in the total labour cost.

The URSSAF Caisse Nationale provides the total amounts of actual contributions paid by employers each quarter. The URSSAF contributions cover a more restricted field than the contributions taken into account in the LCI indices. Table 4 shows the annual change in employers' contributions estimated by the LCI production chain and collected by the URSSAF Caisse nationale. The differences are relatively small in 2022, but may have been significant in the past (for example in 2018 and 2019, due in particular to the hiring premium, which is not taken into account in URSSAF Caisse nationale data).

Table 4: Comparison between LCI and URSSAF Caisse nationale employers contributions (changes by source in %, differences in percentage points)

Source: LCI and Urssaf-Caisse nationale

Note: The figures for "Urssaf-Caisse nationale" and "LCI estimate" are changes, expressed in percentage points, from one year to the next.

1 In France, LCS surveys are spread over two years, to smooth the statistical burden.

2^* These cost elements are deducted here because they are already taken into account in URSSAF Caisse nationale data.

Not requested.

From Q1 2023 onwards, the LCI indices are working day and seasonally adjusted using Jdemetra+ software. Previously, these processes were also carried out, but using the SAS X12 procedure. A comparative analysis was carried out, showing only minor discrepancies between the new method and the old one. The new file required by Eurostat documenting the seasonal adjustment is attached to the quality report.

Each component (wage per capita, workforce, and contributions exemptions) is adjusted at NACE 21 level. Then the seasonally adjusted components are aggregated at the level of the main sectoral aggregates.

The table attached to the report presents the results of the seasonal adjustment process of wage per capita at NACE 21 level for the first tab and of the exemptions for the second tab.

Given the sharp and unusual changes in the indices during the health crisis, all points in 2020 and 2021 (so 8 quarters) are manually specified as additive outliers (AO) in the pre-adjustment part.

All published indices could be revised. The URSSAF Caisse nationale wage bill series is recalculated every quarter to take into account late declarations and change of main activity code (APET; if a local unit is re-classified in another sector, the whole series is revised). The scope covered by URSSAF Caisse nationale data could vary from quarter to quarter. The entire LCI series are therefore revised every quarter, although the revisions for earlier quarters are generally very minor.

See below.

Quarter-on-quarter revisions are generally quite small for both B-N and B-S. In 2020 and 2021, a period marked by the health crisis, the magnitude of the revisions is much larger than in previous years. The revision of indices were mainly due to revisions of wage bill and paid hours, which varied significantly over this period. The revisions are much smaller in 2022, particularly in the second half of the year as the health crisis comes to an end.

In addition, the infra-annual profile of the indices for 2022 has, however, been revised more than usual in the publication of the indices for the first quarter of 2023, due to the change in the seasonal adjustment method (use of Jdemetra+, see above). Only the infra-annual profile is affected, which therefore results in revisions to the quarterly changes in the indices (columns Q1-23 in tables 4 to 6); the year-on-year changes in the indices in 2022 are not affected.

Table 5: Revisions of quarterly variations (in %) of the total LCI, sections B to N and B to S, for 2022

Reading note: The Q1-23 column contains the quarter-on quarter variations in the index (for each quarter, marked in rows, from Q1 2022 to Q1 2023) as published when the index for Q1 2023 was released. In this release, labour costs for sections B to N increased by 1.6% between Q4 2022 and Q1 2023 and increased by 1.2% between Q3 2022 and Q4 2022. The latter change has been revised downwards by 0.2 percentage points compared with the previous publication (for Q4 2022), when it was estimated at +1.4%.

Note: the  variations presented here are those of the indices published by INSEE and therefore differ slightly from those of the indices published by Eurostat, which do not incorporate certain effects of the composition of employment by detailed sector of activity, effects that were more important than usual in the context of the health crisis.

Table 6: Revisions of quarterly variations (in %) of the wages and salaries index, sections B to N and B to S, for 2022

Note: As above.

Table 7: Revisions of quarterly variations (in %) of the employer contribution index, sections B to N and B to S, for 2022

Note: As above.

Since 2005, INSEE provides, at Q + 70 days, a LCI, which integrates provisional URSSAF Caisse nationale data available on that date.

2022Q1: Q + 70 days

2022Q2: Q + 70 days

2022Q3: Q + 70 days

2022Q4: Q + 70 days

2023Q1: Q + 70 days

See above.

See 7.1.

The LCI transmission to Eurostat has been made about 70 days after the end of the quarter, or sometimes 1 working day after the scheduled date due to the need for further checks, and the publication has been released at nearly Q+75 days.

See above.

Not requested.

Not requested.

See below.

Since the transition to NACE Rev. 2, the series of employer contributions exemptions begin on Q1 2004. For previous quarters, exemptions have been estimated using the values of exemptions in NACE1. The data on worked hours for 1997 and 1998 were estimated along similar lines. Series on the public sector (sections O, P and Q) begin on 2008.

The series of worked hours based on the DARES’ ACEMO survey begin on September 2007 and end in 2019. They are then extended with the DSN so that the series of hours is seamless (i.e. without breaks).

The European regulation has been adhered to in all respects.

The information is available at NACE Rev. 2 section level.

See 8.5.

Comparison of the LCI - wages and salaries and the total amount of gross wages divided by the number of worked hours quarterly according to the national accounts presents difficulties due to methodological differences. Firstly, the formula for calculating the LCI is different from the one used in the National Accounts. Secondly, the economic classifications used are not identical. Lastly, the process of calibration of the National Accounts using the annual data leads to discrepancies between the annual changes in average wages according to the National Accounts and those according to the index.

a) Differing calculation formulae used in the LCI and in the national accounts

The LCI - wages and salaries are calculated by dividing the wage per employee by the number of hours worked per employee. The wage per employee is equal to the wage bill divided by the workforce. These two variables are provided by URSSAF Caisse nationale and are mutually consistent.

In the national accounts, earnings per hour worked are calculated by dividing the wage bill by the number of hours. The number of hours is calculated as the product of the number of employees and the number of hours worked per employee. The total amount of wages according to the national accounts is calculated using URSSAF Caisse nationale data for the infra-annual profile and the provisional data. The wage base used in the calculation corresponds to the unrestricted wage base. This includes all remuneration, which form the basis for calculating social security contributions, accident at work insurance payments and family allowances, i.e. the basic wage plus statutory or contractual supplements or supplements paid by the employer (commissions, bonuses, overtime, tips and benefits in kind).

In contrast, in the final annual national accounts, the structural business statistics, which themselves are based on firms’ declarations to the tax authorities, are also used. Earnings declared to tax authorities cover a larger scope including profit-sharing and incentive pay, the latter being submitted to the "Generalised Social Contribution (CSG) on Earned Income", but not to social contributions. Profit-related and incentive pay amounts account for about 5% of wages and salaries in 2013¹. The difference in the scope may explain part of the differential observed between LCI wages and salaries concepts and National Accounts hourly wages.

Regarding the PEPA/ PPV bonuses, the way to take it into account is different between the national accounts and LCI : the national accounts include the bonuses in raw data before seasonal adjustment ; conversely, to produce LCI, these bonuses are included after seasonal adjustment, since it is considered as a one-off payment. With the scheme due to be extended from 2023, the way to take it into account in the LCI series will probably change in the future.

At the end, an important part of the discrepancy between changes in average wages according to the national accounts and URSSAF Caisse nationale comes from discrepancies between changes in wage bill according to structural business statistics (used in the national accounts) and URSSAF Caisse nationale.

b) Difference in the economic classification

The discrepancies may be due to a difference in sectoral breakdown. The analysis of the national accounts is done in terms of branches, not industries. The basic classification for the quarterly accounts is the aggregated classification NA, 2008. The hourly labour cost index is calculated for each section of the NACE Rev. 2.

c) Timetable of the integration of CICE tax credit

In LCI, the tax credit for competitiveness and employment (CICE) (which no longer exists since 2019) has been recorded as subsidies received by the employer, since Q1 2013 (date of operative event of the rights). In the national accounts, CICE is also considered as subsidies received by the employer, but it is recorded one year late, when the employer declares the taxes to be refunded. These two rules were validated by Eurostat.

d) Comparison of changes in the LCI wages and salaries and hourly wages according to the National Accounts: see the following tables

The discrepancies between the year-on-year evolution in the national accounts hourly wages and in the LCI - Wages and salaries are significant up to the second quarter of 2022, in line with the chaotic profile and major variations of the underlying variables (in particular the volume of hours not directly measured in the national accounts, but calculated on the basis of the number of employees and an average volume of work per employee) – in the context of the health crisis and the ensuing massive use of the short-time working scheme.

Table 8: Changes in hourly wages according to the national accounts, annual variations

Source: Quarterly National Accounts

Table 9: Changes in LCI - wages and salaries, annual variations

Source: LCI earnings index

Note: The changes presented here are those of the indices published by INSEE. For aggregated sectors (such as B to E or G to N, for example), they differ slightly from those of the indices published by Eurostat, which do not incorporate the effects of the composition of employment by detailed sector of activity, effects that are more important than usual in the context of the health crisis.

1 Source:  Acoss stat n°195, July 2014 (in French).

See 8.5.

Since September 2007, there has been a quarterly publication in the form of an "Informations Rapides", INSEE’s chosen medium for publishing economic indices. The publication concerns only sections B to N. The publication can be consulted on INSEE’s web site.

See 9.2.

Every quarter, the LCI is published on a dedicated web site on which Informations Rapides, historical data, and a summary methodological note are made available (in French and in English). For example, for the first quarter of 2023: https://www.insee.fr/en/statistiques/7635559.

The series of indices for wages and salaries and for total costs are published in the macroeconomic database for sections B to N and the aggregates BE, BN, GN and F: https://www.insee.fr/en/statistiques/series/117242675.

See INSEE's macroeconomic database.

Not relevant.

Not relevant.

The synthetized version of methodological report is available and regularly updated on INSEE web site: https://www.insee.fr/en/statistiques/7635559#documentation or https://www.insee.fr/en/metadonnees/source/indicateur/p1630/description.

Not available.

100%.

Not requested.

All the data are obtained from administrative sources or existing surveys (LCS, ACEMO). There are consequently no specific cost for enterprises.

For INSEE, excluding any potential maintenance operations, the task of producing the LCI usually requires about ten man-days per quarter. During the health crisis, however, the mobilization of new sources to evaluate the volume of work, the taking into account and analysis of the impact of the numerous measures of aid to companies in the context of the health crisis, and the deepening of analysis and validation procedures to certify data that are particularly volatile and subject to revision in this same context have led to a significant increase in the workload to produce these indices in 2020 and 2021 (this workload has been multiplied by 2 to reach about 20 man-days per quarter). In 2022, with the end of the health crisis, production time has tended to return to its pre-2020 level.

Only the data for sections O, P and Q and related totals are subject to confidentiality.

See above.

Not requested.