For many years, business statistics produced by the French Official Statistical Service have been largely governed by European regulations. European Union (EU) institutions have required economic indicators, particularly on businesses, to promote the construction of Europe and further integration. The Community survey on industrial production was the subject of a regulation as early as 1991, while the first European regulation on structural statistics was introduced in 1996 and the European framework for short-term indicators was adopted in 1998. Historically, however, European business statistics have been governed by “sectoral” regulations, by domain, based on an “organ pipe” structure. In the early 2010s, Eurostat launched several draft “framework” regulations relating to social statistics, agricultural statistics and business statistics with the aim of streamlining and rationalising the new legislative architecture in the spirit of the Lisbon Treaty (2009), of making it easier to adapt it to future developments and of improving coherence and efficiency (note).

Adopted in 2019, the framework regulation on business statistics, formally known as FRIBS (Framework Regulation Integrating Business Statistics), is the result of a long process of preparation and negotiation at the European level.

As early as 2009, Eurostat laid the foundations for a strategy for the decade ahead (Ouvrir dans un nouvel ongletEurostat, 2009) based on a new generation of cross-domain legislation in which technical topics are governed by implementing and delegated acts. The principle was that such an architecture would help to reduce cross-domain inconsistencies and redundancies, both of which create inefficiencies for the National Statistical Institutes (NSIs) and Eurostat and increase the response burden on respondents.

The stated aims of the FRIBS project launched in 2011 are clear:

“The objective of the framework regulation is to integrate business statistics within the EU, to streamline and rationalise them, to make them more flexible and to reduce the burden of responding to statistical questionnaires. FRIBS will provide common infrastructure tools for the production and compilation of business statistics while improving their relevance and overall quality. The operational objective of the project is to put in place a common legal framework for the systematic collection, compilation, transmission and dissemination of European statistics related to the structure, economic activity, competitiveness, global transactions and performance of the business sector. However, this is not a simple consolidation of existing legal texts in the various domains related to business statistics, but an ambitious project based on a critical review of existing data requirements and methodological and implementing provisions and on an assessment of user needs” (Eurostat, 2012).

The origins of FRIBS thus lie in the idea of a framework regulation designed to replace several existing sectoral regulations. Because of this, the regulation must be based on common methodological principles, definitions and quality criteria. It aims to allow for increased flexibility to adapt to economic changes and new emerging statistical needs. It also seeks to promote cross-domain synergies and simplifications of the data requirements to alleviate the response burden on businesses.

The FRIBS project has also been enriched by the findings of the MEETS programme (standing for Programme for the Modernisation of European Enterprise and Trade Statistics) covering the period 2009-2013. Drawing on methodological and conceptual work carried out within the ESSnets, the programme focused in particular on the following issues:

Several projects forming part of Eurostat’s “ESS Vision 2020” then took over. The ESBR (European System of Interoperable Statistical Business Registers) project has provided a basis for developing topics related to the common European framework for statistical business registers and the EGR, including greater harmonisation of national registers, improvements in the quality of the EGR, and thus of the statistics on enterprise groups and globalisation, and pooling of services (e.g. identification) between NSIs. The SIMSTAT and REDESIGN projects tested the feasibility of exchanging micro-data on intra-EU trade in goods and involved a cost-benefit analysis of the different options for redesigning the production of statistics on intra-EU trade in goods. The work undertaken as part of these projects have provided significant input into the process of preparing FRIBS.

FRIBS replaces ten sectoral regulations or decisions of the European Parliament and of the Council currently in force (Box 1). These texts cover a wide range of statistical operations, as shown in the following list (Figure 1):

Figure. 1 Domains Covered in FRIBS

BD : Business demography ou Démographie d’entreprises

FATS : Foreign Affiliates Statistics ou Statistiques sur les filières étrangères

GVC : Global Value Chains Statistics ou Statistiques sur les chaînes d’activité mondiales

ICT : ICT Statistics ou Statistiques sur l’usage des TIC et le e-commerce

INN : Innovation Statistics ou Statistiques sur l’innovation

ITGS : International Trade in Goods Statistics ou Statistiques sur le commerce international de biens

ITSS : International Trade in Services Statistics ou Statistiques sur le commerce international de services

PCM : Prodcom ou Production communautaire

RD : Research & Development Statistics ou Statistiques sur la R&D

SBS : Structural Business Statistics ou Statistiques structurelles sur les entreprises

STS : Short Term Statistics ou Statistiques de court terme

TEC : Trade by enterprise characteristics ou Statistiques sur le commerce selon les caractéristiques de l’entreprise

TIC : Trade by invoicing currency ou Statistiques sur le commerce par devise de facturation

Without repealing existing Regulations, FRIBS also covers:

In its early versions, FRIBS also covered statistics on foreign direct investment, a domain taken out of scope following the CMFB opinion.

The question also arose as to whether tourism statistics should be included. The latter ultimately remained out of scope and remains governed by a 2011 regulation covering both statistics on tourism supply and tourist attendance and statistics on the departure behaviour of households.

In addition, FRIBS does not cover “business-based social statistics” such as on continuing vocational training in enterprises, labour costs and the structure of earnings. The possibility of a FRIBS2 Regulation covering these domains was considered at one point but is no longer on the table.

Given the many domains covered and the cross-cutting nature of FRIBS, the regulation applies to several of France’s statistical authorities and extends well beyond the confines of INSEE's remit. As well as INSEE, several Ministerial Statistical Office are also concerned by the regulation, including agriculture, foreign trade, sustainable development, higher education, research and innovation, labour and the Banque de France.

The current “organ pipe” approach (Figure 1) raised issues of cross-domain coherence and redundancy. For example, the statistical units used in different domains tended to vary, with the Kind-of-Activity Unit and the Enterprise being predominantly used (Box 2). In addition, definitions of concepts (e.g. turnover) often varied from one domain to another, as did both the field covered and the breakdowns required (e.g. by economic activity or size class). The same data could be required under two different regulations, leading to different responses from Member States.

A typical example is the number of enterprises, required both within the scope of structural statistics and in the context of business demography. These inconsistencies also limited the possibilities to link micro data.

More than harmonisation between countries or harmonisation through inputs, the primary objective of FRIBS is therefore to increase cross-domain coherence and to enhance the role of common reference frameworks such as classifications and business registers, as well as to develop statistics on specific topics and sectors, including the globalised economy following the Riga Memorandum adopted by the European Statistical System Committee (ESSC, 2014), enterprise groups and the tertiary sector (trade and services).

FRIBS and its implementing measures define specific outputs that must be transmitted (aggregate variables, indices, etc.). To calculate them, national statistical institutes are free to use several types of data sources specified in the Framework Regulation provided the results meet certain quality criteria. Permitted sources include surveys, administrative records, exchanges of micro-data and any other relevant innovative sources, methods or approaches. For example, FRIBS includes no pre-defined survey questionnaire, although in the case of topics such as ICT, innovation and global value chains, the variables required in the implementing legislation are so detailed that they effectively resemble survey questions. But these are exceptions rather than the rule. There is also no requirement for any particular information to be collected directly from businesses, nor is there a requirement that surveys, if conducted, be mandatory. With some exceptions, the data to be transmitted are aggregated data. The only exceptions are mandatory exchanges of micro-data, which relate on the one hand to business registers and the EuroGroups register and on the other to intra-EU trade in goods.

A provision was also introduced to allow exchanges of confidential data, collected or calculated within the scope of FRIBS, between the national statistical authorities of Member States, their central banks, the European Central Bank and Eurostat “to safeguard the quality and comparability of European business statistics or national accounts”. The specific conditions for applying this provision, originally introduced to authorise exchanges of confidential data on multinational enterprise groups and described at this stage as an “enabling clause”, remain to be specified.

In short, the framework regulation creates a common legal framework for business statistics. As well as the common framework it establishes, the regulation also introduces a number of changes and innovations, the most important of which are listed in Box 3 and Box 4.

France has been a strong supporter of the FRIBS project throughout its preparation. INSEE and the Foreign Trade Ministerial Statistical Office have also been actively involved in the projects and ESSnets that designed some of the regulation’s key building blocks, in particular on the profiling of large multinational groups, the European framework for registers and intra-EU trade in goods.

In addition, many of the changes required under FRIBS were anticipated by France’s Official Statistical Service and have already been implemented in recent years. While in some cases France’s statistical monitoring system was already capable of responding to new demands, developments at the European level also provided a decisive impetus to enhance and complete the range of indicators at the national level prior to the implementation of the Framework Regulation. This is particularly the case for the new short-term volume indicators (retail trade volume index and index of services production) and for the extended scope and changes in periodicity of short-term indicators (Scherrer, 2019).

However, a number of adaptations are still required at the national level. Among the more structural changes needed is the transition to effective dissemination based on the concept of Enterprise for statistics on ICT, innovation, R&D, global value chains and the implementation of the micro-data exchange approach on intra-EU exports of goods.

Dissemination at the Enterprise level still requires an assessment of the most suitable protocol: should data be collected directly at this level and, if so, which unit(s) should be surveyed to ensure relevant and high-quality information is compiled? Should we survey legal units within Enterprises and then aggregate their responses, in a context where the data collected are mainly qualitative variables? Drawing in particular on experiments conducted in recent years, work is currently underway on these issues at both INSEE and the SIES.

On the second subject, many points remain to be defined or fine-tuned within the scope of delegated or implementing acts designed to complement the framework regulation. A secure exchange system needs to be set up for very detailed confidential data on intra-EU exports, and each country wishing to receive individual data from other Member States must be certified. Furthermore, the quality of statistics based on the exchange of micro-data has not yet been thoroughly assessed. The actual use to which exchanges of micro-data are put will be determined by the level of statistical quality, with the potential for such exchanges to supplement or even replace the collection of statistics on intra-EU imports at the national level. With respect to trade in services, the detail of data requirements within the scope of the FRIBS application package has yet to be specified. Depending on what is ultimately decided, further substantial work may be required on the part of the Banque de France. Existing methods of collaboration between data producers may also need to change. For example, new aggregates on the economic performance of businesses, combining data produced by INSEE on the non-financial market sectors and by the Banque de France on the financial sector, will need to be transmitted, requiring increased cooperation across departments.

The framework regulation defines the general framework and its key elements. However, the specific details of many domains will be set out in delegated and implementing acts. For statistics at the country and regional levels and on international activities, the Framework Regulation defines a list of general topics (e.g. for short-term statistics: prices, output and performance, labour, etc.) and detailed topics (producer prices; import prices, etc.). However, the legal architecture provides that variables and data requirements may be defined (and modified) at an even more detailed level by means of implementing acts. Similarly, the periodicity of data requirements is set by the Framework Regulation, although the scope and breakdowns required will be defined in other texts.

In the case of FRIBS, the package of texts drafted to accompany the Framework Regulation were prepared in parallel with the regulation itself. They make up the “FRIBS package”, which has already been extensively discussed by various thematic working groups and by the group of directors of business statistics, although some domains, relating primarily to international trade, remain to be clarified.

The Framework Regulation provides for implementation of the provisions of FRIBS from 2021 (the reference year), except for intra-EU trade in goods (2022). In the case of short-term statistics, while implementation is scheduled for 2021, it will only be completed in 2024, several years of data being necessary to calculate seasonally adjusted data for the new indicators. As in the case of social statistics, several calls for proposals issued by Eurostat are intended to provide financial support to countries in implementing the innovations introduced by FRIBS. France has so far not responded positively to the calls issued to date since they related to innovations that it had already implemented.

While FRIBS serves to complement the existing supply of statistics, one of its main objectives was also to improve cross-domain coherence. When three different definitions are replaced by just one, when countries are required to provide a single “number of enterprises” or when size classes are harmonised across domains, consistency is necessarily enhanced. There will also be more scope for linking data from different domains.

However, there is still room for improvement:

FRIBS will improve coherence and comparability, but does not address all these issues.

FRIBS already incorporates requests made by users, mainly the political departments of the European Commission or the European Central Bank (ECB), in particular on globalisation and enterprise groups and on short-term indicators.

For future developments, the Framework Regulation includes safeguard clauses (or safeguards) designed to contain and manage the number of possible changes, notably on detailed topics and variables, over rolling five-year periods.

These safeguard clauses, as well as the legal instruments allowing for changes to data requirements, were key domains of negotiation between Eurostat and the NSIs throughout the FRIBS preparation process. The European Parliament and the Commission pushed for a widening of the scope of the delegated acts, which could have resulted in increased costs for the NSIs and businesses. A compromise was therefore reached in an attempt to reconcile the requirement for flexibility and adaptability with changes in demand and guaranteed cost control.

Ultimately, changes in variables and detailed topics will remain possible, but within defined boundaries. They will be easier to introduce in some domains. In their search for a compromise, Eurostat and the directors of business statistics introduced a distinction in 2015 between “dynamic statistics” (ICT, Innovation, Global Value Chains) and “stable statistics”. Compared to the latter, the former have less stringent safeguard clauses attached to them and therefore offer greater flexibility. Registers are the domain where the scope for change and development is most limited since the vast majority of countries have supported the role of registers as the backbone of business statistics – a role that requires great stability.

Future developments will also involve pilot studies. Some topics have already been identified in the framework regulation. These reflect major needs expressed by users, including on real estate, with the new subject of commercial real estate, international trade in services (as requested by the Commission and the ECB) and financial indicators on businesses, climate and the environment (requested by the European Parliament). It can therefore be expected that pilot studies on these topics will be launched, leading to new data requirements in the coming years. The process has already begun on commercial real estate and trade in services by business characteristics and method of supply.

FRIBS encourages the use of data sources other than surveys and eliminates the practice of multiple requests for identical data. However, new data requirements have also been included. Ultimately, it is hard to see how the regulation significantly reduces the burden for both NSIs and businesses. One of the domains where a significant alleviation of the burden was announced is intra-EU trade in goods. A slight reduction in the coverage rate has been decided, although the reduction in burden will only be significant if the introduction of the micro-data exchange approach serves to alleviate the burden at a national level. On this subject, the effectiveness of the proposed changes remains to be proven.