Welcome to issue “N4” of the Courrier des statistiques journal in its new, digital format. We are continuing with our 2018 ambition, with the aim of exploring the various aspects of the professions and methods associated with statistics, whether public or private, as well as continuing to work towards several goals. Firstly, an educational requirement: although the subjects are very complex and technical, the articles are aimed not only at public statisticians, but also potentially at students, academics and, more generally, well informed citizens, who need to be able to understand the principles of the methods used. Secondly, we must show openness, by inviting foreign statistics institutes, producers of data of interest for statistical use and academics looking into issues of data processing and data quality. In the same vein, we are also ensuring that the articles push people outside their comfort zone, for example by insisting on the need for benchmarks, especially at international level.

This fourth edition of Courrier des statistiques begins with a subject that has not yet been addressed in the journal, with three articles dedicated to microsimulation. Questioning whether the methods used in microsimulation fall within the scope of official statistics is just one of the subjects highlighted in the first article. In it, Didier Blanchet defines the concepts of microsimulation, whether static or dynamic, the methodology applied in practice, and the main components of a microsimulation model. He explains the reasons why this method was the obvious choice for official statistics in France. His piece is a general article, which, despite citing the DESTINIE model several times, also offers, to some extent, a framework, which will be indispensable for future work. The other two papers each discuss a specific microsimulation model.

Firstly, Pierre Cheloudko and Henri Martin tell the story of the TRAJECTOiRE model, which simulates the pension system and therefore naturally includes a strong time dimension. It is a dynamic model in which the notion of simulation covers both the (deterministic) simulation of legislation and the (stochastic) simulation of future careers. The article show the stages that have led to the current scheme, starting with an aggregate model touching on case studies, then a second block modelling retirements, which brings a dynamic dimension, and finally TRAJECTOiRE, which models careers in their entirety, with data at individual level.

Simon Fredon and Michaël Sicsic present INES, a statistical model used to simulate social and fiscal policies. They highlight the various areas of work involved when developing and maintaining a microsimulation model: closely monitoring and taking into consideration developments in legislation, developing solid databases using several sources and characterising the schemes to be simulated. The piece also mentions the power of these models, which can be used in various ways: upward or downward effect on social and fiscal transfers, effects of reforms on inequalities, etc.

The other articles in this issue have one point in common: in one way or another, they discuss “benchmarks” of official statistics.

Firstly, the nomenclature of professions and socio-professional categories: this major instrument for social statistics has just been overhauled: a rare event. Thomas Amossé discusses the background to this, after recounting the history of socio-professional classifications and presenting the concept of nomenclature. This is not just a new version, but also a new logic that takes into account the fact that there are multiple uses; it must therefore be sufficiently malleable so as to adapt to these different uses.

The SIRENE register is another frame of reference, a pivotal element for the production of business statistics and for administrative simplification, with its unique identifier, which gives structure to the entire administrative ecosystem around businesses. Contrary to the previous article, which presented an achieved outcome, the article from Christophe Alviset positions itself further upstream, introducing the SIRENE4 programme. The aim of the project is not to undertake a comprehensive review of the SIRENE system, but to improve it in several key areas in order to industrialise processing by automatically identifying establishments, automatically codifying activities, standardising data exchanges and through checks. The challenge will also be to support companies through the digital administrative revolution.

The need for benchmarks not only applies to the data, but also to the processes used to produce statistics. For some 10 years, the GSBPM has been the standard at European level, and is now broadly shared. In this issue, the choice was made to give the floor to a foreign statistics institute, Statistiska centralbyrån (SCB) in Sweden, as it has very broadly and systematically applied an approach that is very similar to that used by the international statistics community with the GSBPM. Johan Erikson explains the comprehensive lessons learned from this practice: the success factors, the obstacles and the developments envisaged.

When carried out face-to-face, household surveys require very sophisticated sampling techniques associated with the practical need to facilitate the movement of the interviewers, while respecting the restrictions associated with inclusion probabilities and optimising accuracy. In order to avoid facing complex methodological questions during each sampling process, we have a benchmark, which we keep for 10 years, from which we draw the samples for each new face-to-face survey: the master sample. As Patrick Sillard et al. explain, this has now been in use since the start of 2020. The novelty stems from using fiscal sources and coordinating with the Labour Force survey sample.

The last article, contrary to the preceding ones, which discuss transverse frames of reference, focusses on one specific area: housing. Here too, it reflects on the benchmark, looking into what exactly the reference data and statistics comprise. Jérôme Harnois and Pierre Lamarche address the question in all its aspects: construction, housing facilities, price, cost of occupation, state of housing and poor housing, etc. The article does not simply point out this or that administrative survey or source; it takes an original perspective by targeting a systemic approach, structured by the lifecycle of housing but constrained by the limits of the data sources available.