Principle 9: Non-excessive Burden on Respondents

The range and detail of European Statistics demands is limited to what is absolutely necessary.

In order to ensure that the scope and detail of any new survey is limited to what is strictly necessary, a two-tier system has been put in place. The National Council for Statistical Information (Ouvrir dans un nouvel ongletCNIS) examines all new survey projects relating to Official Statistics. Based on the opinion of appropriateness issued by the National Council for Statistical Information (CNIS), the Ouvrir dans un nouvel ongletOfficial Statistics Quality Label Committee issues an opinion on compliance, ensuring that the scope, detail and costs of statistics are commensurate with needs and are therefore limited to what is absolutely necessary.

In addition, INSEE complies with the obligation to minimise the collection of personal data imposed by the General Data Protection Regulation (GDPR). In particular, a data protection compliance analysis file (DCPOD) is established for any new processing of personal data. The variables are listed and analysed to verify the method through which they are obtained, their purposes and therefore the justification for their collection, as well as the recipients to whom they are addressed.

The response burden is spread as widely as possible over survey populations and monitored by the statistical authority.

The Ouvrir dans un nouvel ongletOfficial Statistics Quality Label Committee ensures that the survey projects submitted to it meet the statistical quality criteria in terms of collection and sampling: it ensures that the burden placed upon the respondents is not excessive. In particular, it examines the relevance of the questions in light of the objectives and information available elsewhere (other surveys, administrative data) as well as the length and duration of the questionnaire, which are estimated during the testing phase or calculated during previous editions. It also verifies the manner in which the sample has been compiled and ensures that its size is proportionate to the precision objectives.

In some cases, particularly for business surveys, INSEE uses negative sample coordination methods which make it possible to spread the response burden. On this basis, INSEE draws up an assessment of the burden on companies based on the statistical use of companies’ response times to surveys of the French Official Statistical Service. The assessment makes it possible to understand the response burden and how it is spread across the economy. The largest companies, which are few in number, are frequently surveyed, because their weight in the economy justifies the need to understand their situation. In contrast, the burden on the smallest companies is limited as they are surveyed more infrequently.

In the case of household surveys, INSEE undertakes not to survey the same households, dwellings or individuals twice within five years, except in the case of panel follow-up.

The data sought from businesses is, as far as possible, readily available from their accounts and electronic means are used where possible to facilitate its return.

In order to facilitate responding to business surveys, INSEE is committed to the digitisation of exchanges, using the Portail entreprises company portal, which acts as a single point of access for the companies surveyed. There are two digital ways to respond: the electronic form and a file deposit service. The first of these uses a technical mixed-mode collection platform (Coltrane) to which the surveyed companies are able to connect. In 2021, 90 % of business surveys (INSEE or Ministerial Statistical Offices (MSOs)) used this system. The second is a digitisation protocol set up by INSEE through a file deposit and withdrawal service used by companies to respond to surveys directly from their information system (management software).

Administrative and other data sources are used whenever possible to avoid duplicating requests for data.

INSEE is seeking to reduce the burden of responding to surveys for respondents. Thus, the Institute performs evaluations on administrative and private data sources in order to determine whether they can meet needs identified by Official Statistics and, if so, to suggest using them. Some sources have been in use since the 1950s.

The ÉSANE programme (Elaboration of annual statistics of companies) is a striking example of this. The main source of this system is composed of the annual declarations on profits sent each year, by companies, to the Directorate-General for Public Finance (DGFiP): other administrative sources are also used, such as social security returns filed by employers. However, as this linking of administrative sources does not fully cover the information needs identified for this system, two additional surveys are necessary: the Annual sectoral survey (ESA) and the Annual production survey (EAP). Compared to the system that was in place to produce annual business statistics prior to ÉSANE, the number of variables requested from companies in these surveys has been reduced by about half: this reduction is a result not only of the reduction in size of the questionnaire made possible by the increased use of administrative sources, but also of work to optimise the sample.

Other examples illustrate this desire to use alternative data sources to surveys or additional data sources to surveys to reduce the burden placed upon respondents to the extent possible. For a number of years, INSEE has drawn on:

In addition, the National Council for Statistical Information (Ouvrir dans un nouvel ongletCNIS), to issue its opinions on appropriateness, ensures that the information collected through a survey that is submitted to it is not already available (in an administrative source, for example). Finally, the Ouvrir dans un nouvel ongletOfficial Statistics Quality Label Committee, to issue an “opinion on compliance”, verifies that the questions comply with the terms of the opinion on appropriateness.

Data sharing and data integration, while adhering to confidentiality and data protection requirements, are promoted to minimise response burden.

As it is explicitly encouraged by the National Council for Statistical Information (Ouvrir dans un nouvel ongletCNIS), matching between data sources (surveys and/or administrative sources) is increasingly used within the French Official Statistical Service (SSP), in strict compliance with the protection of personal data.

INSEE and the Ministerial Statistical Offices (MSOs) work together on an ongoing basis to develop the sharing and integration of data, while respecting the rules on confidentiality. Relations between INSEE and the MSOs are formalised by agreements within the framework provided for in Article 7bis of French Ouvrir dans un nouvel ongletAct No 51-711 of 7 June 1951 on Legal Obligation, Coordination and Confidentiality in Statistical Matters. Many matching operations are thus in progress or planned within the French Official Statistical Service, based on files mainly located within INSEE (Permanent demographic sample (EDP) panels, All employees file (“base tous salariés”), ÉSANE, etc.).

Working in partnership is also on the rise between entities of the French Official Statistical Service and public stakeholders that hold data. Several surveys conducted by Ouvrir dans un nouvel ongletDREES, the French Research, Studies, Evaluation and Statistics Directorate, are thus matched by INSEE with tax data (from the Directorate-General for Public Finance) and social security data (from pension funds and social security benefit organisations).

To facilitate links between sources in compliance with confidentiality rules, INSEE works on the use of identifiers, such as the “code statistique non signifiant” (CSNS), a non-identifying statistical code, which is key to performing matching in a secure manner that does not allow for identification. As another focal point for the development of data sharing and integration, a programme for the creation of a statistical directory of households, dwellings and individuals from administrative sources was launched by the INSEE Executive Committee in 2020.

Beyond matching, many systems rely on the combination of aggregated data and/or data on individuals from multiple sources. For example, the quarterly and annual National Accounts produced by INSEE are based on the use of a very large number of sources produced by the French Official Statistical Service and other administrative bodies. Similarly, employment estimates are calculated by sharing information between INSEE, the Direction de l’animation de la recherche, des études et des statistiques (Ouvrir dans un nouvel ongletDARES), the French Research, Studies and Statistics Directorate, and the Caisse nationale de recouvrement des cotisations de sécurité sociale et d’allocations familiales (URSSAF-Caisse nationale), the French Social Security and Family Allowance Contribution Collection Office, in compliance with confidentiality and the protection of personal data.

Statistical authorities promote measures that enable the linking of data sources in order to minimise response burden.

INSEE and the Ministerial Statistical Offices (MSOs) are working together on an ongoing basis to develop systems to link databases in order to reduce the survey burden. This working in partnership is formalised by agreements within the framework provided for in Article 7bis of French Ouvrir dans un nouvel ongletAct No 51-711 of 7 June 1951 on Legal Obligation, Coordination and Confidentiality in Statistical Matters.

In relation to demographic and social statistics, the framework will be developed. For years now, the use of the National Registration Number (NIR), a registration number in the register of natural persons, has been subject to very strict rules on the confidentiality of personal data and the possibilities for matching between files are reduced. INSEE is part of this framework and carries out the processing made possible by law and regulations. Thus, in application of the French Ouvrir dans un nouvel ongletOrder of 6 August 2014, as amended, the Permanent demographic sample (EDP) is created by linking multiple sources, particularly by using the NIR: civil status, census data and socio-fiscal data.

In order to develop matching between databases containing data on individuals while ensuring full compliance with confidentiality and respect for personal data, French Ouvrir dans un nouvel ongletAct No 2016-1321 of 7 October 2016 for a Digital Republic states that, prior to its statistical use by the French Official Statistical Service, the NIR must undergo a cryptographic operation to replace it with a statistical code that is not significant (CSNS). Once it has been implemented into all files, this code will be the secure and encrypted privileged matching key that can be used within the French Official Statistical Service. French Ouvrir dans un nouvel ongletDecree No 2016-1930 of 28 December 2016 simplifying the advance formalities relating to processing for statistical or research purposes designates INSEE as the body responsible for the introduction of the CSNS and specifies the service to be provided.