INSEE - Business statistics Directorate

Short-Term Activity Indices division

INSEE - Direction générale
88 avenue Verdier
CS 70058
92541 Montrouge Cedex

Industrial production indices provide a short-term indication of changes in the activity in industry.

Sources : They are calculated thanks to branch surveys conducted by INSEE (Institut National de la Statistique et des Études Économiques), the SSP (Service de la statistique et de la prospective), the SDES (Service de la Donnée et des Etudes Statistiques) and professional bodies. They also rely on prices indices for some activities.

NACE Rev. 2.

Activities covered: Sections B to E of NACE Rev. 2.

The production indices measure the evolution of output (value added) in volume. The measure of production at finest levels rely on physical units, deflated turnover or worked hours (taking yearly changes in productivity into account).

Reporting unit: Legal unit.
Observation unit(s): Kind of activity unit.

Units employing 20 employees or more or having a turnover superior than 5 million euros in one industrial branch.

France

Time series cover the period back to 1990.

 Base (reference) year: 2015.

Index.

Month.

At national level:

Law No. 51-711 of 7 June 1951 on the obligation, coordination and secrecy of statistics.

At European level:

Regulation (EU) 2019/2152 of the European Parliament and of the Council on European business statistics (EBS-Regulation) and its General Implementing Act (Regulation (EU) 2020/1197)

Agreements for collecting process with professional bodies and with the statistical services of the Ministry of Ecological Transition (data on energy) and the Ministry of Agriculture (data on food and beverages).

At the national level, Article 6 of Act No. 51-711 of 7 June 1951, as amended, on the obligation, coordination and secrecy of statistics determines what statistical secrecy is, its limits and the conditions for its application. These rules apply to surveys conducted by the official statistical service (SSP) whether or not they are mandatory. The derogations provided for in this article are governed by the Act. As it stands, the only exemptions that remain applicable are those relating to the status of public archives for surveys and censuses, which authorise the disclosure of individual information contained in the questionnaires and relating to personal and family life and, in general, to the information contained in the questionnaires. family life and, in general, to private facts and behaviour, after a period of 75 years for individuals and 25 years for legal entities. According to the Act, this communication may not be used for tax control or economic repression.

The obligations relating to statistical confidentiality also apply to administrative data that INSEE or ministerial statistical services may have access to under the terms of Article 7 bis of the aforementioned Act, as well as to private data communicated under the terms of Article 3 bis. under the terms of Article 3 bis of the Act. Generally speaking, with regard to access to public data, confidentiality obligations relating to the protection of privacy or business secrecy and the protection of personal data are guaranteed by Act (Article 1 of the Act for a Digital Republic).

At the European level, the confidentiality of statistical information is affirmed by Article 338 of the EU Treaty. < >. Statistical confidentiality is also the subject of Chapter V of Regulation 223/2009 as amended and Implementing Regulation No 557/2013 as regards access to confidential data for statistical purposes.

A Statistical Confidentiality Committee ensures that these statutory guarantees are maintained. Its powers are set out in Article 6 bis of Act No. 51-711 of 7 June 1951, as amended, on the obligation, coordination and secrecy of statistics and Chapter II of Decree No. 2009-318 of 20 March 2009 on the National Council for Statistical Information and the Statistical Confidentiality Committee. Statistical Confidentiality Committee. It is called upon to give its opinion on any question relating to statistical secrecy, and gives its opinion on requests for communication of individual data collected by means of a statistical survey or transmitted to the official statistical service, for the purpose of establishing a statistical report. for statistical purposes. Researchers can also ask the committee to give an opinion on access to various administrative data other than public statistics.

Chaired by a State Councillor, it includes representatives of the National Assembly and the Senate. The composition and operating procedures of the committee are set by decree in the Council of State.

The beneficiaries of data communications resulting from ministerial decisions taken after the opinion of the Statistical Confidentiality Committee undertake not to communicate these data to anyone.

Any breach of the provisions of this paragraph shall be punishable by the penalties provided for in Article 226-13 of the Criminal Code.

The main rules for turnover indices are : no dissemination of data if they are based on the compilation of less than 3 enterprises or if a single entreprise represents more than 85 % of the turnover of the field covered by the serie (dominance rule and p% rule).

Confidential treatment through TAU ARGUS (software designed to protect statistical tables) and confidential indices are not disseminated.

Link:

Software Tau-Argus (fr)

A monthly calendar is produced for the four following months. It can be found on the INSEE website: https://www.insee.fr/fr/information/1405540?conjoncture=3&debut=0

On the INSEE Website:

https://www.insee.fr/fr/information/1405540?conjoncture=3&debut=0

Simultaneous release to all interested parties: Yes.
Identification of internal government access to data before release: No.
Transmission to Eurostat and further use of the statistics: The transmission to Eurostat is made using the SDMX format.

Monthly.

Not available.

National paper publications: "Informations Rapides". Monthly publications for economic indicators.

The results are available in the indices and time series category of the < > section on insee.fr. The data can be downloaded in xlsx or csv format. They can also be retrieved via a web service, available on the api.insee.fr portal and compliant with the SDMX standard.

Not available.

Transmissions to Eurostat using eDamis.

"Documentation on methodology" on the INSEE website: https://www.insee.fr/en/information/5349872

Specific treatments have been implemented for the first month of the Covid crisis in 2020 and are explained in a methodological note ( https://www.insee.fr/en/statistiques/4488313#documentation)

Not available.

Since 2005, the European Statistics Code of Practice has been the reference for assessing the quality of the output of national statistical institutes. Periodic reviews by European peers are organised to ensure that the principles of this reference framework are implemented and to ensure that each institute is committed to continuous improvement. Within this framework, INSEE has adopted a process-based approach. A range of tools, pooled within the Official Statistical Service (SSP), has been created to describe statistical production processes, analyse their strengths and weaknesses, assess the risks involved, examine their documentation (metadata) or assess a particular stage (analysis of users' needs, data validation, etc.). The diagnoses resulting from these << quality approaches >>; lead to the establishment of action plans that are regularly monitored in the context of << process reviews >>. In addition, INSEE regularly conducts satisfaction surveys on the indicators and data it produces. The results of these surveys are available on the insee.fr website.

Eurostat requirements are fulfilled and the variables used to calculate the indices are accurate and of good quality.

The Industrial production indices are meant to be used by many clients (i.e. users) including Eurostat, the official statistical service (INSEE and the SSM) and the general public (inseenauts, private or public institutions, media).

Different types of products are produced by the team on Industrial production indices:

• monthly << InfosRap >> publications;

• aggregated data files and series available on the on-line database;

• specific files responding to user needs (National Accounts, Business Cycle Analysis, Structural Surveys, SSM);

• metadata: annual quality reports sent to Eurostat and made available on insee.fr;

• weights.

This process is not subject to a specific satisfaction survey for external users but internal users are satisfied with the quality and punctuality of indices.

Eurostat’s requirements in terms of time series are fulfilled.

Statistical error and bias are small.

Not available.

Response rate: Around 90%.
Actions to speed up or increase the rate of response: Phone or mail recall, (rarely) penalties and increased use of internet for data collection.

Non-sampling errors are due to the firms’ delayed reponses or the non exhaustive coverage (small firms are not included in the statistic population).

35 days after the end of the reference period (except for the publication on July and November, 40 days after the end of the reference period)

Deadlines are respected and data have so far always been published on time.

Methods are comparable to those of other European. They are compliant with the european regulation (EBS).

Consistency over time is ensured through the use of stable methods. In the case of methodological changes, or change of base period, past indices are backcast to ensure comparability over time.

Indices and microdata are confronted with the results of the annual production survey "Enquêtes annuelles de production " (structural business statistics). 

A benchmarking with other STS data, structural data or national accounts can be carried out. In addition, national accountants use Industrial production indices for quarterly estimates or provisional annual estimates and provide feedback on consistency with past account data.

Internal consistency is ensured through the aggregation method.

Burden: 27 216 hours / Cost: 40 000 hours.

These figures correspond to the costs for Insee and the burden for enterprises to calculate IPI for Eurostat and for national purposes.

The same revision policy is applied to STS data released nationally and transmitted to Eurostat.

All revisions are taken into account.

Routine revisions are mostly due to available additional information, especially overdue respondents at the survey. Then SA-WDA data are also revised each month as a consequence of the new calculation of SA-WDA coefficients. These revisions are taken into account each month.

Major revisions are mainly due to rebasing, new weights or update of productivity factors.

No revision calendar exists.

Only the last vintage is available on the website. Users have to download monthly the data to obtain a real-time database.

Routine revisions are explained in the national publication. There is a dedicated paragraph entitled “revision of variations”.

Warnings are published in the national publication "Informations Rapides" and on the web site www.insee.fr when the revisions are major (revisions like after rebasing or changes in methodology).

Generally, initial and final values are close, there are few revisions.

In the whole industry, the mean revision between january 2020 and december 2022 is -0.02 point for the year-on-year growth rate of the calendar adjusted serie (see Figure below).

In the whole industry, the mean absolute revision between january 2020 and december 2022 is 0.21 point for the year-on-year growth rate of the calendar adjusted serie (see Figure below).

These revisions do not include the impact of rebasing or major methodological changes.

Figure : revisions in the total industry between january 2020 and december 2022 (36 months)

Annual rebasing process

The 2015 rebasing implemented an important innovation with the introduction of annually chained indices instead of fixed-weights indices. In addition, since the implementation of the new basis 2015 in March 2018, the industrial production indices (IPI) are now reviewed at the rate of one fifth of the series each year. The aim of this annual update is to improve the relevance of the IPI with regards to the evolution of technologies and production processes, by adapting the way of measuring a product, by including emerging products or, on the contrary, removing old products. Each year, the annual review is implemented with the publication of March concerning the indices of January. The adjusted series are backward-reviewed. At an aggregated level, the revisions on indices are moderate. Nethertheless, some indices can be significantly revised resulting from an improving coverage or a modification of the measurement variable (for instance production sold in euros instead of quantities).

Type of source: Statistical survey. Data for energy are received from the SDeS, the statistical service of the Ministry of Ecological Transition. Some data are also obtained from professional bodies.

Sample for industry except manufacture of food products and beverages: about 4800 firms; the sample is selected in the almost 30000 firms of the Annual Production Survey (with thresholds of 20 employees and 5 million euros of turnover); in average, the sample covers 70% of the value added of the stratum.
Sample for the manufacture of food products and beverages: more than 7000 firms. The survey is conducted by the SSP (Service de la Statistique et de la Prospective), the statistical service of the Ministry of Agriculture.

Monthly.

Questionnaires used in the survey: "Enquêtes de branche" (branch surveys).
Planned changes in national questionnaires: Yearly.
Data collection media: Mainly (95 %) electronic (Web questionnaire).
Planned changes in data collection methodology: Overseas departments are included since 2014.

The data are validated before being sent to Eurostat. This validation is done in two stages: the checking (and if necessary correction) of individual data and then the checking of the aggregated indices. This second check may lead to further corrections of the individual data. We use selective editing.

Estimates for non-response: Data of the previous period multiplied by the average rates of change of responding units.

Type of index: Annual chain-linked Laspeyres indices with 2015 as reference year

Method of weighting and chaining: The weightings of the branches are based on raw value added at basic prices and are updated annually (chain-linked index). Within each branch the index is based on a representative sample of activities weighted by the sampling weights. The indices have a mean of 100 in 2015.

Planned changes in production methods : regular work is done to improve the construction of the sample and then the robustness and accuracy of estimates. In particular, work is carried out on the optimal choice, depending on activities, between cut-off methods and stratified sampling.

Actions to speed up or increase the rate of response: Phone or mail recall, (rarely) penalties and increased use of internet for data collection.

The raw indices are seasonnaly and working-day adjusted (SA-WDA) using the X13 ARIMA program available in JDemetra +. The WD adjustment (trading days, leap year) and the seasonal adjustment decomposition are calculated at the 4-digit level of the NACE Rev. 2. The upper levels are obtained by aggregating the series (indirect method), in the same way as the agregation of raw data.

The Reg ARIMA calendar adjustment is used by constructing working day regressors based on the French national calendar (which takes into account working days specific to France).
Outliers (additive outliers, temporary changes, level shifts, seasonal outliers) are fixed in the past and are detected automatically on the past 12 months onwards. The critical value for outlier detection, the filter length and the model/filter selection depend on the series and may have to be changed manually to improve the quality of the seasonal correction. This was the case to neutralize some particular points associated with the 2020-2021 health crisis (lockdowns for example), which would have induced an unjustified distortion of the seasonal coefficients over the past.
Either additive or multiplicative decomposition can be used. The seasonal adjustment models are reexamined every year (favouring stability) and the parameters are re-estimated every month.

Each month the SA-WDA data are revised from 2012. For the seasonal adjustment of indices in the recent past, the models are now estimated over a reduced sub-period (from 2005 onwards), in accordance with Eurostat guidelines, and in order to reinforce the robustness of the seasonal adjustment. The data before 2012 are fixed in evolution, in accordance with Eurostat's guidelines (avoid revisions over a too long period).

Not available.