Quality Strategy 2022–2027 (July 2022) The quality1 of the French Official Statistical Service (SSP) is ensured on three levels: • An SSP quality policy, namely "Integrating quality into processes with a view to security and efficiency" • A quality strategy, defined as the route to achieving the target set in the quality policy. It also forms part of a logic of continuous improvement and is based on guidelines (axes and themes); • Roadmaps, which break the axes and themes of the Quality Strategy down into operational actions. The Quality Strategy 2022–2027 is based on four non-hierarchical axes. Each axis is broken down into themes, introduced by a short text summarising the context and the challenges. Each theme is linked to one or more of the principles of the European Statistics Code of Practice. For each theme, actions for improvement are included in the INSEE and Ministerial Statistical Departments (SSM) roadmaps. These actions are taken from the European action plan established following peer recommendations during the third review and the areas for improvement identified during discussions with the services. The latter can be either pooled (undertaken by the SSP as a whole) across the entirety of the SSP or generic (applicable for a given stakeholder in a particular context). A common feature of many of the actions could be to seek out and benefit from experience and practices at both national and international level. In this regard, success stories will benefit from greater value with a view to increasing their knock-on effect. The INSEE and MSD roadmaps will follow the structure of the Quality Strategy, integrating the relevant actions from the European action plan and actions specific to the entity concerned. The latter could be included in additional themes or axes not considered in the current framework. Finally, a clause requiring a mid-term review will allow the actions to be updated depending on the context. Warning: the proposed axes and themes are levers for improvement and in no way imply that no action has been taken in this regard. Conversely, there is also other work that contributes to the quality of official statistics that is not included here. 1 Quality is understood here within the meaning of the European Statistics Code of Practice (CoP) and the challenge is to improve our compliance with its principles through our practices.
Joint investment in administrative and private sources Professionalising the use of administrative data In the context of the consistent and growing use of administrative data, increasingly demanding requirements from users with regard to the freshness of information and limited resources, while also being bound by quality standards for official statistics production, the Official Statistical Service (SSP) must ensure that it:
- secures the supply of administrative data, i.e. obtaining from its suppliers regular deliveries of data covering its information needs that are as easy as possible to exploit for its statistical purposes (by seeking in particular to be more involved in the design, modification and abandonment of their administrative files);
- guarantees efficient data processing, i.e. generating high-quality statistical products with controlled consumption of resources by means of the standardisation and/or pooling of processing processes incorporating active metadata. Promoting access to private data and their use in the SSP The use of private data, such as scanner data and the web scraping of commercial websites for the consumer price index, or credit card transactions to shed light on the economic situation, or even mobile phone data to measure the territorial distribution of the population, have brought the Official Statistical Service (SSP) into the era of big data. This offers it new opportunities: new information, exhaustiveness in certain areas and high frequency guaranteeing that the data are up to date. In order to make the most of this data pool, the SSP must now commit to providing greater and more reliable access to private data, but must also ensure that it shares its experiments and work in progress to allow collective progress to be made in their use.
Documented, optimised and secure processes Documenting in order to capitalise on knowledge The existence of (statistical production or other) process documentation that is as complete and detailed as necessary without being excessive, but also structured, accessible, legible and up-to-date, is essential to the preservation of knowledge and the control and continuity of production, which sometimes becomes fragile as a result of frequent moves. In this regard, the Official Statistical Service (SSP) will therefore continue its efforts in the area of internal documentation, working towards standardisation by drawing inspiration from best practices. Particular emphasis will be placed on the role of stakeholders, the rules to be complied with for computer programs and the checking and validation of processes and results, as well as on documents intended to facilitate transitions during moves. Adapting and developing quality approaches and process reviews The INSEE Internal Audit Unit has identified processes for which risks must be controlled in order to ensure the continuity of the Institute's activities. Quality approaches offering a process analysis methodology using official and proven tools help to secure processes: formalisation helps to secure the execution of processes, and the shared and documented vision secures the transmission of competences. Beyond the essential processes, this methodology must now be adapted according to the context and must therefore be modular to allow the entirety of the Official Statistical Service (SSP) and the various different types of operation to benefit from it. In addition, in order to promote its deployment, communication with regard to the approaches will further enhance the contributions and progress that they will help to make. Continuing investment in data validation and control systems Data validation and control systems are key to the efficient management of statistical production activities. A self-evaluation operation for some of these systems was carried out by their project managers at the Official Statistical Service (SSP) in 2021. The findings drawn up upon completion highlighted the quality of the processing performed with a view to ensuring the consistency of the sources, from an internal or temporal point of view and by means of comparison with other existing sources. However, areas for improvement have emerged:
- explain the concept of expected quality at the launch of an operation;
- extend the self-evaluation system to the processing of individual data with a view to achieving a more complete overview of the systems. Beyond production, the attention paid to control and validation systems will also be extended to cover studies with the aim of securing the work performed by research officers.
Improving the quality of statistical metadata The Statistical Metadata Repository (RMéS) allows for the centralisation of statistical metadata (concepts, statistical operations, nomenclatures, codes, quality reports, etc.) in a single location with a view to providing all users with consistent reference documentation. The standardisation of their descriptions makes it easier for users to use them and therefore encourages the sharing of knowledge. It offers tools that facilitate their reuse by both man and machine, whether internal or external. It is now a question of improving the completeness of the repository and the associated service offer. Boosting statistical confidentiality, secrecy and data security In May 2018, the General Data Protection Regulation brought about profound change in the rules that are to be complied with when processing is carried out for statistical purposes. With a view to supporting the consideration, innovation and implementation of measures concerning the protection and security of personal data, efforts to develop the skills of agents will be intensified. In order to secure our practices and to protect ourselves from risks likely to undermine public confidence, the Official Statistical Service (SSP) will continue its methodological investment in order to guarantee secrecy and statistical confidentiality and to ensure that agents are able to benefit from ad-hoc training that is tailored to new techniques.
A strong focus on the public Assuring the public of the proportionate, relevant and confidential use of data collected The SSP informs its audiences of the measures taken in order to comply with the legislation in force (GDPR and the Law on Data Processing, Data Files and Individual Liberties) in connection with the protection of the personal data used to draw up official statistics, in particular by the INSEE website. However, there is a need for a more pedagogical presentation of this complex subject, drawing inspiration from foreign examples, for instance. Against the backdrop of an expansion of the data being used by the SSP, it is useful to develop means of boosting public confidence in the use of the data collected, particularly with regard to maintaining confidentiality and ensuring their social acceptability. Better understanding the needs of users of official statistics Knowledge of user requirements is a prerequisite for defining effective and relevant production and dissemination within the meaning of the Code of Practice. It assumes that the key characteristics of the main categories of users have been identified, with the data being gathered by means of the channels that they use to express their needs. The mid-term exercise of the National Council for Statistical Information (CNIS) 2024–2028 will provide guidelines that will inform the thinking of the Official Statistical Service in this regard. The collective approach that will be initiated will allow users to express their needs and producers to present their work programmes.
Strengthening links with researchers In recent years, the amount of micro-data accessible to researchers and the associated access protocols have been significantly improved by official statistics, in particular though the use of a Secure Data Access Centre (CASD) and the Quetelet-Progedo Diffusion network. The role of the Statistical Confidentiality Committee has also been strengthened with regard to the approval of requests to access data from anywhere within the Official Statistical Service (SSP) and, where appropriate, administrative data held by other administrations. However, efforts should be continued to allow researchers to benefit from even greater access to the databases used within the SSP. Collaborations between the SSP and research will be further developed (chairs, NRA projects, etc.), drawing inspiration from good practices identified internationally, particularly when developing the dialogue. Diversifying the systems used to measure user satisfaction Measuring satisfaction is an essential step in identifying needs and assessing the quality of the service provided. The system for listening to and consulting users at INSEE has multiple aspects and is based on actions carried out at both national and regional level. It uses a number of different channels (website, INSEE Contact service, meetings with public stakeholders, etc.) and concerns publications, statistical data, navigation and access to data on the website, etc. However, there are some areas for improvement. The Official Statistical Service (SSP) will in particular undertake actions to develop and vary the systems used to measure satisfaction, making use of new quantitative and/or qualitative methods targeting some or all of the different categories of users (institutional partners, local authorities, researchers, the general public, etc.).
Broad, modern and transparent dissemination and communication Drawing up and applying shared and transparent dissemination rules In order to provide users, who are placed on an equal footing, with accurate and reliable information, the Official Statistical Service (SSP) must complete the process that it commenced following the second peer review, which aims to develop clear, strict and transparent shared dissemination principles by means of the homogenisation of practices. More precisely, this involves:
- consolidating the implementation of the rules for disseminating embargoed indicators;
- extending the application of the underlying principles to the rest of its statistical publications;
- promoting its dissemination policies more widely with a view to ensuring their strict application. Continuing to improve online access to data With a view to allowing for the widest and easiest possible reuse of the information it produces, INSEE and the Ministerial Statistical Departments (MSDs) have a dissemination strategy based on their websites, accompanied by constant efforts to improve access conditions. In particular, the insee.fr website has a responsive design and acts on accessibility recommendations issued by the World Wide Web Consortium (W3C). Since 2018, developers and advanced users have been able to access data via APIs that allow them to automate the collection of streaming data. In a changing context and environment, marked by the development of open data, the explosion of available data and strong competition between disseminators, the Official Statistical Service (SSP) will continue its investments aimed at improving and diversifying access to data, which will include in particular improvements to the search engine and navigation. Developing multi-channel communication strategies Social networks have taken on a decisive role in communication and make it possible to reach multiple audiences. They require the usual practices to be adapted and editorial lines tailored to each medium must be set. When shared, the experience gained by INSEE in the use of these networks will allow any Ministerial Statistical Departments (MSDs) that are interested to develop their own communication strategy using these channels.