Population estimates

Dernière mise à jour le : 24/01/2017

Présentation de l'opération

Organisme producteur :

- Direction des statistiques démographiques et sociales (DSDS)
- Direction de la diffusion et de l'action régionale (DDAR)

Type d'opération :

Summary of census results, civil status statistics for births and deaths, and administrative statistics on migratory flows.

Objectifs :

- To estimate the national population by gender, age and marital status on 1st January of each year;

- To estimate the regional and departmental population by gender and age group on 1st January of each year.

These population estimates are mainly used to calculate different demographic indicators: birth rate, mortality rate, total period fertility, death probability, life expectancy.

Thèmes abordés :


Champ de l'opération

Champ géographique :

 - France (mainland and overseas departments): population by gender, age and marital status.

- Metropolitan France: population by gender, age and marital status;

- Regions and departments: population by gender and age group.

Champ(s) statistique(s) couvert(s) :

The entire population living on the territory

Caractéristiques techniques

Périodicité de l'opération :


Historique :

The population census is used as a base for annual population estimates. It sets the reference levels. Between two censuses, provisional estimates are carried out.

These become definitive once the results of a new census are known. After the 1999 total population census, total population estimates, by gender and age and marital status were therefore revised for the period from 1st January 1991 to 1st January 1999.

The population estimates published subsequently were based until 2008 on the results of the 1999 census, updated every year from civil status statistics (concerning births and deaths) and an estimate of net migration.

At the end of 2008, the definitive results of the new census became available, relating to 1st January 2006, and as a result, population estimates for 1st January 2000 and the following years were revised.

From now on, new definitive census results will be made available every year, changing the calendar of estimates.

New calendar of estimates:

In January of year n estimates "until the end of n-1" are published:

- national, regional and departmental population estimates relating to 1st January of the year n-3 are revised and become definitive and are than combined with the latest available results from the census;

- national population estimates for 1st January n-2 and 1st January n-1 are then revised, but remain provisional;

- regional population estimates for 1st January of year n-2 are revised and remain provisional;

- The first national population estimates for 1st January n are published, along with the first departmental population estimates for 1st January of year n-2 and the first regional estimates for 1st January of year n-1. At this point these estimates are provisional.

Méthodologie :

The breakdown of the population by age is according to age in years passed on the 1st January of the year in question.

The population census is used as a base for annual population estimates.

It is the only source which provides exact details of the population by gender, age and marital status at the different geographical levels.

Between two censuses, or pending availability of the data from the census of the year in question, population level and distribution are estimated. When the results from a new census become available, estimates that have already been made are revised to take these into account.

They then become definitive.

Calculating the population pyramid with the new census method:

With the new census method, which summarises information from five successive years of data collection, two population pyramids can be constructed:

- the first pyramid, produced and published by the Census, is obtained directly from the age of the people being counted.

This is a smoothed pyramid because it uses information collected over five years and therefore corresponds to mean population numbers across the period.

For example, those who are aged 10 in number of years passed are obtained from the number of people counted every year who are aged 10 in number of years passed, thus people counted in 2004 born in 1993, people counted in 2005 born in 1994, etc.

- the second pyramid is also based on census information, this time taking into account the year of birth.

For the pyramid concerning 1st January 2006 in number of years passed, the number of people aged 10 is obtained from numbers of people born in 1995 and counted for all five years when data was collected.

Minimal corrections are then applied so that those surveyed in 2004 and 2005 but who died before 1st January 2006 are not counted in 2006 or, on the contrary, to ensure that people who died in 2006 and 2007 are counted, because although they could not be surveyed during the annual census surveys in 2007 and 2008, they were alive on 1st January 2006.

It is this second pyramid, which is consistent with the population pyramid obtained from earlier general population censuses (last edition in 1999), and which is produced in population estimates by gender and age at national, regional and departmental levels.

It is produced for the census years, then used as a reference for subsequent population estimates from the latest available census.

Method for estimating national population and distribution by gender, age and marital status:

At national level, the population and its distribution by gender and age are estimated by updating the last available population pyramid with the civil status statistics relating to births and deaths by gender and age, and with an estimate of net migration by gender and age.

Birth and death statistics are obtained from information given to INSEE by the town halls.

Net migration, the difference between those entering and those leaving the territory, is estimated by INSEE from fragmentary information. Administrative statistics provide information on nationals from countries outside the European Union entering the territory.

Foreigners leaving the territory, movements by French nationals and nationals from the European Union are not systematically counted by the administration.

INSEE therefore has to rely on trends observed from the censuses.

Statistics on divorces and marriages by gender and age are also used to update distribution of marital status by gender and age.

Method for estimating regional and departmental populations and their distribution by gender and age group:

At regional and departmental levels, net migration is often a larger component of the population trend than at national level, especially as at these levels, it also includes internal migrations between the different departments of France.

However, there is not enough statistical information available to give a sufficiently robust estimate for net migration at departmental or regional level.

Thus the methodology used at national level cannot be broken down and applied to these other geographical levels.

Firstly, the population at departmental and regional levels is estimated globally. Civil status statistics provide the natural increase at these levels.

For net migration, using previous population censuses an apparent net migration can be calculated by subtracting the natural increase from the population change.

The observed trends are then extended.

Finally, an adjustment is made so that the population for all the departments or regions coincides with the national population estimate.

Secondly, the population at departmental and regional levels is distributed by age group, using civil status statistics by gender and age and extending the observed trends across the migratory flows by gender and age with the population censuses.

Voir aussi :