The french use of housing tax in the new census method
- 6 Years Ago
- 4. Rolling census
By the National Institute of Statistics (INSEE), France, 2010.
1. France has implemented, since 2004, a rolling census: a survey is conducted annually on a
portion of the territory. At the beginning of each year, about 14% of the population completes a
census form. At the end of each year, an official population (used in used in many financial and
regulatory areas ) is calculated, for each commune, based on the last 5 surveys. This “legal
population” is calculated for each commune (of which there are 36,682 in France) regardless of the
survey’s date in its territory.
2. The method is mainly based on population surveys. But an administrative source, the
housing tax, is used to put back to the same reference date data collected within different years. This
paper aims to present the details of this "statistical” use of administrative data.
3. When the survey in a commune under 10 000 inhabitants has taken place one or two years
before the reference date of the census, the number of homes subject to the tax allows to extrapolate
the population. The administrative data is used as a trend factor -not a level factor-, to refresh the
data from the last survey. (Part 1)
4. The first step is collecting data from the tax administration. The data is constructed
according to rules and practices that meet the fiscal targets. We had to understand their way of
elaborating and schedule update. A second phase of expertise then led to the conclusion that the
separation between occupied dwellings and other dwellings (seasonal or secondary used, vacant),
was not sufficiently reliable in the tax databases. So, the final method takes in account the evolution
of the total number of dwellings from one year to another. A correction coefficient, from the figures
collected during the previous census, takes into account the trend in the number of persons per
dwelling in the town.
5. Finally it was necessary to establish “correction” rules in the very few situations where the
data changes from year to year is clearly erroneous, with significant decreases one years followed by
increases of similar magnitude following year. (Part 2)
6. Finally, this process has been operated for 2 years (legal populations have been computed
and published in December 2008 and December 2009), and has raised no objections, especially from
the 21 000 municipalities whose population has been calculated by this method. It however requires
significant permanent human resources. (Part 3).
(Paper presented at the Joint UNECE/Eurostat Expert Group Meeting on Register-Based Censuses, The Hague, The Netherlands, 10-11 May 2010)
[Available in French and English]
Also In This Category
Optionally provide your comments to help us improve this article...
Thank you for your feedback!
Created: 6 Years Ago
Last Modified: 6 Years Ago
Article has been viewed 3.3K times.
Article not rated yet.