Last updated:
France Population Projections 2026: Immigrants, Descendants and the 2050 Outlook
INSEE's central projection gives France 68.1 million people on 1 January 2070, up from 67.4 million in 2021 (INSEE). Pew Research Center projected a Muslim share of 12.7% in 2050 with zero migration and about 17.4% with medium migration (Pew). In an illustrative model on this page, immigrants plus second-generation descendants reach 30.8% of the population in 2050 and 36.1% in 2070 in the central case; this is a calculation, not an official forecast.
No official body publishes a projection of the immigrant-origin share of France's population. This page sets out the official projections, the published religion-based projection, and a transparent illustrative model. It also summarises the "Great Replacement" debate, with the numbers each side uses. For current data, see immigration statistics.
Key figures
| Figure | Value | Year | Source |
|---|---|---|---|
| INSEE central projection, total population | 68.1 million | 2070 | INSEE |
| INSEE assumptions | fertility 1.80; net migration +70,000 a year | 2021–2070 | INSEE |
| Observed fertility and net migration | 1.56 children per woman; +176,000 | 2025 | INSEE |
| Pew, Muslim share, zero migration | 12.7% (8.6 million) | 2050 | Pew |
| Le Bras, immigrants including descendants | 19%–32% | 2050 | Connexion |
| Illustrative model, immigrants + descendants (low / central / high) | 28.3% / 30.8% / 33.7% | 2050 | Section below |
| Illustrative model, immigrants + descendants (low / central / high) | 31.7% / 36.1% / 41.2% | 2070 | Section below |
Definitions: immigrant, foreigner, descendant
- Immigrant (immigré): a person born abroad as a foreigner and living in France. Naturalised immigrants are still counted as immigrants.
- Foreigner (étranger): a person without French nationality.
- Descendant of immigrants (second generation): a person born in France with at least one immigrant parent.
Many descendants have one French-born parent and are French citizens. In 2024, 55% of second-generation descendants had only one immigrant parent, and 34% of immigrants held French nationality (INSEE). The starting point for any projection is therefore a choice of definition, which is why results differ between sources.
What do official projections say about France's population?
INSEE's central scenario, published in November 2021, projects 68.1 million people on 1 January 2070 and a peak near 69.3 million in 2044. Deaths exceed births from 2035 (INSEE; INED summary). The earlier 2016 exercise had 76.5 million for 2070.
INSEE's assumptions were a total fertility rate of 1.80 (revised from 1.95) and net migration held at +70,000 a year because of high uncertainty (INSEE English note). Observed values have drifted from both. Fertility was 1.56 in 2025, below 1.80. Net migration was estimated at +176,000 in 2025, above +70,000 (INSEE). INSEE's published projection does not give an immigrant-origin share for 2070.
Eurostat's EUROPOP2023 offers a baseline and five sensitivity tests, including zero net migration for 2023–2100. Eurostat describes these as what-if scenarios, not forecasts. Across the EU, population falls by 157 million over 2025–2100, with migration adding about 104 million. For France, the changes are described as modest (Eurostat).
What share of France will be Muslim in 2050?
Pew's 2017 report projected a Muslim share of 12.7% (8.6 million people) in France in 2050 under zero future migration, from 8.8% (5.7 million) in 2016 (Pew).
| Pew scenario for France, 2050 | Muslim share | Assumption |
|---|---|---|
| Zero migration | 12.7% | No new migration; growth from existing age structure and fertility |
| Medium migration | about 17.4% | No more refugees, continued regular migration |
| High migration | about 18% | 2014–2016 record refugee flows continue indefinitely |
The medium and high values for France come from secondary reporting, not from Pew's own table, so treat them as approximate (Maldita.es; Le360). Pew assumed that Muslim women in Europe have about 2.6 children, against 1.6 for non-Muslims. For Europe as a whole, Pew's medium case is 11.2% and its high case is above 14% (Pew). Fact-checkers have noted that even the highest Pew figure for France is far below 50% (Maldita.es).
France has no census data on religion, and a projection by religion depends on survey-based starting points. Current survey estimates are on the statistics page.
What would the share of immigrants and descendants be in 2050, 2070 and 2100?
In the central case of the illustrative model below, immigrants plus second-generation descendants account for 30.8% of the population in 2050, 36.1% in 2070 and 43.1% in 2100.
ILLUSTRATIVE MODEL, not a forecast and not official. This is a one-year-step, three-group stock-flow model built for this site's research, using the assumptions below. It uses crude rates and ignores age structure inside groups and emigration of the native-born. It counts a person of immigrant origin as an immigrant or a second-generation descendant only; the third generation counts in the residual group, which understates origin broadly defined. Treat any result beyond about 2050 as a sensitivity exercise. All inputs are in the tables.
Starting point (2025)
| Group | Number | Share | Source / method |
|---|---|---|---|
| Total population | 68.85 million | 100% | INSEE series |
| Immigrants (I) | 7.97 million | 11.6% | INSEE series |
| Descendants, second generation (D) | 8.0 million | 11.6% | INSEE 2024, held for 2025 |
| Everyone else (N) | 52.88 million | 76.8% | residual |
| Immigrants + descendants (I+D) | 15.97 million | 23.2% | sum |
Assumptions
| Parameter | Value | Basis |
|---|---|---|
| Net migration, all added to group I | Low 70,000 / Central 150,000 / High 250,000 a year | Low is INSEE's central assumption (INSEE); Central lies between that and +176,000 in 2025 (INSEE); High is a stress case |
| Immigrant birth rate converges toward | Low 12.0 / Central 13.0 / High 14.5 per 1,000 by 2040 | Reflects the fertility gap of 2.3 vs 1.7 narrowing as arrivals age (INSEE) |
| Descendant birth rate | converges to the native rate over 40 years | Second-generation completed fertility 1.90 vs 1.86 (INSEE) |
| Ageing drag on native and descendant birth rate | down 0.4% a year, capped at 20% | Calibration so the Low case gives 65.4 million in 2070 |
| Mixed births | A child counts as D if the mother is I or the father is an immigrant; 14% of births to non-I mothers in 2025 | 2018: 31% of births had at least one foreign-born parent (INSEE) |
| Naturalisation | no effect | INSEE definition: immigrants stay immigrants |
| Overall fertility level | held at the 2025 level (1.56) | INSEE |
Birth and death rates were calibrated so that the model reproduces about 643,000 births (actual 645,000) and about 646,000 deaths (actual 651,000) in 2025 (INSEE). The share of births to immigrant mothers (about 21%) is an assumption of the model, not a published figure, and is not verified.
Each year: I(t+1) = I + net migration − deaths of I. D(t+1) = D + births to I mothers + mixed births − deaths of D. N(t+1) = N + other births − deaths of N.
Results: share of total population
| Year | Scenario | Total population | Immigrants (I) | Immigrants + descendants (I+D) |
|---|---|---|---|---|
| 2030 | Low (70,000 a year) | 69.1 M | 11.5% | 24.4% |
| 2030 | Central (150,000) | 69.5 M | 12.0% | 24.8% |
| 2030 | High (250,000) | 70.0 M | 12.7% | 25.4% |
| 2050 | Low | 68.4 M | 11.6% | 28.3% |
| 2050 | Central | 70.7 M | 13.8% | 30.8% |
| 2050 | High | 73.5 M | 16.3% | 33.7% |
| 2070 | Low | 65.4 M | 11.9% | 31.7% |
| 2070 | Central | 69.6 M | 15.4% | 36.1% |
| 2070 | High | 75.1 M | 19.2% | 41.2% |
| 2100 | Low | 59.9 M | 12.7% | 35.8% |
| 2100 | Central | 67.0 M | 17.7% | 43.1% |
| 2100 | High | 76.4 M | 22.5% | 50.6% |
| Reference: zero net migration | 2050: 66.6 M; 2100: 54.3 M | 9.6% (2050); 7.1% (2100) | 26.2% (2050); 28.2% (2100) |
The zero-migration row uses the Low fertility inputs. The share of immigrants plus descendants rises from 23.2% to about 26–28% even with no arrivals, because the existing second generation is young and the rest of the population is ageing.
Will immigrants become a majority in France?
In this illustrative model, immigrants alone do not approach 50% in any scenario (the maximum is 22.5% in 2100), and immigrants plus descendants stay below 50% in the Low and Central scenarios through 2100. Only the High scenario crosses it, around 2098 (50.6% in 2100). This is a result of the stated assumptions, not a forecast.
- Immigrants alone (I): to reach 50% by 2100, the model needs net migration of about 2.2 million a year, roughly twelve times the 2025 level.
- Immigrants plus descendants (I+D): the break-even level of net migration for 50% in 2100 is about 240,000 a year in this model. The High scenario (250,000 a year for 75 years plus a higher birth rate) sits almost exactly on that line.
- Stress tests: 500,000 a year gives 39.7% in 2050, 50.4% in 2070 and 62.2% in 2100. One million a year gives 49.0% in 2050. These are far outside any historical or official assumption.
- Fertility sensitivity: a 15% shift in overall fertility changes the 2100 I+D share by under one point (High case: 50.1% instead of 50.6%), but changes total population by 5–6 million. In the model, migration matters more than fertility for the origin share.
The "majority" question also depends on definition. Counting only French-nationality status would give a very different result from counting origin, because many descendants are French citizens.
What would change these results?
- Net migration: the single largest driver. INSEE's +70,000 assumption is already below the +176,000 estimated for 2025.
- Fertility: INSEE assumed 1.80, and 1.56 was observed in 2025.
- Definition: counting the third generation as of immigrant origin, as Michèle Tribalat does, would add several points to every row. Her three-generation figure for under-60s is near 30%, against 24% for two generations in the same data (Espace Populations Sociétés). This figure was read from a summary and is not verified against the paper.
- Mixed unions: 66% of second-generation descendants live with a partner with no migratory ancestry (INSEE Première 1910), and how their children are classified moves every result.
How does the model compare with other sources?
| Source | Measure | Result | Comparable? |
|---|---|---|---|
| This model, Central | Immigrants + descendants, 2050 | 30.8% | Different definition from Pew |
| Le Bras (Connexion) | Immigrants including descendants, 2050 / 2100 | 19%–32% / 22.5%–42% | Closest; model range 28.3%–33.7% for 2050 is in the upper half of his range; 2100 range of 35.8%–50.6% overlaps then exceeds his upper bound |
| Pew 2017 (Pew) | Muslims, 2050 | 12.7% zero / about 17.4% medium / about 18% high | Religion, not origin |
The model sits at the high end of Le Bras's range for 2050 because it counts every child with an immigrant father or mother as a descendant, and because arrivals exceed INSEE's +70,000.
What is the "Great Replacement" debate, and what do demographers say?
The term "Grand Remplacement" comes from the writer Renaud Camus, around 2010. The idea of "replacement migration" appeared earlier, in a 2000 UN report on population ageing (Hérodote). The claim has a quantitative part (numbers and fertility) and a qualitative part (cultural change). Demographers treat the two separately. The positions below are summarised as their authors state them.
Sceptical view (François Héran, Hervé Le Bras)
- François Héran, holder of the Collège de France chair on migrations and societies from 2017 to 2025, argues that the public debate overstates flows and that policy should rest on measured data (Collège de France).
- Hervé Le Bras, in Il n'y a pas de grand remplacement (2022), argues the thesis fails because immigrants are about 10% of the population, many are naturalised, and children of mixed unions hold French nationality. In his scenarios French nationals remain a majority (Slate). He projects immigrants including descendants at 19%–32% of the population in 2050 and 22.5%–42% in 2100, depending on definition, and considers 19% for 2050 reasonable. His assumptions include about one extra child for immigrant women and about 170,000 arrivals against 120,000 departures of French-born people a year (Connexion).
- Supporting figures cited: immigrants 11.6% and descendants 12%; 34% of immigrants are French citizens (INSEE); second-generation completed fertility of 1.90 (INSEE Première 1939); 66% of second-generation descendants live with a partner with no migratory ancestry (INSEE Première 1910).
Critical view (Michèle Tribalat, Observatoire de l'immigration et de la démographie, Jean-Paul Gourévitch)
- Michèle Tribalat argues that counts based on nationality or on immigrants alone understate the population of foreign origin. Her estimates, from 2011 data, are 19.2% of metropolitan France over two generations and near 30% of under-60s over three generations. She has also argued that demographic projections are very unreliable (Espace Populations Sociétés; Sud Radio). The figures were read from a summary and are not verified against the paper, and no 2050 figure from her was found.
- The Observatoire de l'immigration et de la démographie (OID, a think tank) presents the replacement idea as having two parts. It argues that flows and fertility differences support the quantitative part (OID). No OID projection with numbers was found.
- Jean-Paul Gourévitch has published a book on the question and estimates of irregular migrants and costs (Critiques libres). The Jean-Jaurès Foundation faults his cost study for relying on estimates rather than official expenditure data (Fondation Jean-Jaurès). His own demographic projection figures were not retrieved.
Where the camps agree and differ
The current stock is not disputed: immigrants are about 11–12% of the population and descendants about 12% (INSEE). The disagreement concerns definitions (whether a French citizen of immigrant origin counts), the future (INSEE's +70,000 migration assumption against the +176,000 estimate for 2025), and whether origin is the right category at all.
The debate has also reached the institutions that produce the data. A June 2026 Le Figaro article criticised INED and François Héran, and Héran replied in July 2026 (La Grande Conversation). This is a live dispute and is not settled.
Related pages
- Immigration statistics for France: current data on immigrants, permits and asylum
- Nahel riots 2023
- Recent riots and unrest
- Macron government record on immigration
- Immigration in Europe
- Frequently asked questions