1 Every year, Population publishes an article on recent demographic trends in France. This edition focuses on the year 2021. The COVID-19 pandemic affected demographic behaviours less markedly than in 2020, but the various indicators—births, abortions, unions (marriages and PACS civil unions), and deaths—have still not returned to pre-crisis levels, signalling a recovery that remains partial and an ongoing demographic impact of the health crisis. Migration data for 2021 are not yet available, but this article provides new evidence of the effects of the pandemic on 2020 migration flows of third-country nationals who require a residence permit to reside in France. This year, the various indicators are analysed at subnational level, that of the former regions (NUTS Level 2 of the European nomenclature), to reveal geographical contrasts, but also to judge the effects of the health crisis on spatial variations in demographic behaviours.
2 On 1 January 2022, the population of France was 67.8 million, 187,000 more than on 1 January 2021. Natural increase totalled 81,000 in 2021, a figure below the INSEE estimates of net migration, before or after adjustment.
3 At the European level, France was one of the nine EU-27 countries with positive natural increase in 2021 and is still the main driver of growth in the European Union, whose overall population fell by 170,000 in 2021. While Germany still has Europe’s largest population, ahead of France and Italy, its natural increase is negative, and total population growth was just 82,000. In Italy, the population declined by 253,000.
4 French population dynamics vary markedly across regions, in both the overseas territories and metropolitan France. The number of regions where populations have declined has risen from five between 2011 and 2016 (representing 9.3% of the population) to 12 between 2016 and 2021 (representing 29.2% of the population).
5 In 2020, the most recent year for which statistics are available, 211,255 third-country nationals (who must hold a residence permit to reside in France) entered France, a decrease of 17.8% with respect to 2019 (257,137). This drop is due mainly to the travel restrictions imposed during the COVID-19 pandemic. The decrease in admissions for employment reasons was especially strong, but admissions of students fell only slightly. Student numbers from Africa even increased. The share of asylum seekers remained stable (19%). The Paris region (Île-de-France) received the largest share of inflows to metropolitan France (38.4%), while Mayotte accounted for 50.2% of inflows to the overseas territories.
6 In 2021, 742,500 children were born in France, 7,300 more than in 2020 but 10,900 fewer than in 2019, the year preceding the health crisis. The recovery in births was only partial and did not begin until the summer of 2021. This increase was due almost entirely to an increase in fertility: the total fertility rate increased slightly from 1.80 in 2020 to reach 1.83 children per woman in 2021. It was women aged 30–34 and those aged 35–39 who had most children on average. France is still the country with the highest fertility in Europe but is followed closely by Romania. Age-specific fertility patterns vary across the French regions. Four different profiles can be distinguished: regions with relatively early fertility (peak at ages 25–29), mainly located in the north and centre of metropolitan France (Picardie, Nord-Pas-de-Calais, Centre-Val de Loire, Burgundy, Poitou-Charentes); those with relatively late fertility (peak at ages 30–34), mainly along the Atlantic coast in the south of metropolitan France (Provence-Alpes-Côte d’Azur, Brittany, Aquitaine, Midi-Pyrénées, Languedoc-Roussillon); the overseas departments (with higher rates at ages 20–24 than elsewhere); and the atypical departments of metropolitan France (Corsica with very low rates and a plateau at ages 25–34; Île-de-France with a sharp peak at ages 30–34). The prevalence of births outside marriage and child naming practices also vary across the regions, with generally higher levels of non-marital births and children with double names in the west than in the east.
7 Numbers of abortions remained relatively stable (224,200 in 2021 vs. 222,300 in 2020, down from 233,300 in 2019). The proportion of surgical abortions is declining each year, while that of medical abortions, often managed outside a hospital setting, is increasing (77% in 2021). While a majority of abortions in 2021 (65%) took place in a health facility, the share of non-hospital medical abortions continued to increase (35% vs. under 20% in 2015). There are wide geographical disparities in the prevalence of abortion. Rates are above average in most of the departments of Île-de-France, in Provence-Alpes-Côte d’Azur, Languedoc-Roussillon, and in the overseas territories. Women living in Île-de-France, Rhône-Alpes, Provence-Alpes-Côte d’Azur, in certain departments of Midi-Pyrenées, and in Aquitaine more frequently have non-hospital medical abortions. The proportion of surgical abortions remains high in the departments of Indre-et-Loire and Loire-Atlantique (53% and 45%).
8 After reaching a historical low in 2020, the number of marriages rebounded in 2021 but without returning to its 2019 level. The number of new PACS unions in 2021 is still not available, but PACS unions outnumbered marriages for the first time in 2020. Same-sex unions accounted for 2.7% of marriages in 2021, a decrease of 0.3 percentage points with respect to 2020 and, for the second consecutive year, more marriages between women were registered than between men. Beyond the effects of the choice of wedding location, marriage intensity varies considerably across regions. While marriages are more frequent in the east (from Alsace down to the Mediterranean), they are outnumbered by PACS unions along the Atlantic coast and in the south-west. Same-sex unions also follow a geographical pattern. Unions between men are more concentrated in Île-de-France, while unions between women are more evenly distributed across the country.
9 The COVID-19 pandemic made it more difficult to celebrate new unions, but also to initiate divorce. While available divorce statistics remain incomplete, the number of divorces fell again in 2020 (57,400), a drop of 13% with respect to 2019. Divorce and PACS dissolutions are more frequent in the south and east of France, and much less so in the north-west.
10 In 2021, 657,500 deaths were registered, 1.7% fewer than in 2020, but still 7.2% more than in 2019. Between 2020 and 2021, life expectancy increased by 2.4 months for men (79.3 years) and 3.6 months for women (85.4 years), after losing 7 months and 5 months, respectively, between 2019 and 2020. This means that France has still not returned to pre-COVID mortality levels; excess mortality in 2021 is estimated at 6.3% (vs. 7.5% in 2020). COVID-related excess mortality has not modified the position of France in Europe. It still has low mortality at ages above 60 and high infant mortality relative to other European countries.
11 The disparities across French regions—between Mayotte, where mortality is highest, and Corsica, where it is lowest—are very large. The year 2020 was by no means an ‘ordinary’ year. The COVID-19 pandemic disrupted the traditional mortality rankings of the French regions, and those where mortality was highest in 2019 were not those with the highest excess mortality in 2020, in metropolitan France at least. While overall mortality tends to be low in Île-de-France and the south-east, excess mortality was high in these regions. Lorraine, Nord-Pas-de-Calais, Basse-Normandie, and Picardie, where initial mortality levels were relatively high, were also severely affected.
I – General population trends and age structure
1 – Slow population growth in 2021 but a larger increase than in 2020
12 On 1 January 2022, the population of France was 67.8 million, of whom 65.6 million in metropolitan France (mainland France and Corsica) and 2.2 million in the overseas departments and regions (3.2% of the population) (Papon, 2022a).  An increase of 187,000  was recorded in 2021 (172,000 in 2020), representing a growth rate of 2.77 per 1,000 (2.55 per 1,000 in 2020).  Growth thus recovered slightly in 2021, or at least made up for the slowdown in 2020 when it fell to its lowest level since the Second World War. The downtrend in French population growth that began 20 years ago is continuing, however.
13 The COVID-19 health crisis is ongoing. The year 2021 was marked by several new transmission waves, a third lockdown (though less strict than that of the spring of 2020), and major vaccination campaigns, initially targeting the most vulnerable and then extended to the younger population from the spring. While the impact of COVID-19 was much smaller than in 2020, it continued to affect population dynamics in 2021, producing excess mortality but also an upturn in births after the dip observed between November 2020 and early 2021 (Brée et al., 2021; Breton et al., 2021; Papon, 2022a). The quarterly number of births fell in the 3rd and 4th quarters of 2020 and in the 1st quarter of 2021 but then rose in the 3rd and 4th quarters of 2021 and the 1st quarter of 2022—9 months after the end of the first lockdown in April 2021 (Figure 1A). This upturn has fully made up for the shortfall of births in early 2021. The quarterly death figures spiked in the 2nd and 4th quarters of 2020 and rose again in 2021 and early 2022 but much less sharply (Figure 1B). In the 1st quarter of 2021, as in the 4th quarter of 2020, natural growth (difference between births and deaths) was negative, as in the 1st quarter of 2022 (provisional data).
14 The crude death rate in 2021 (9.8 deaths per 1,000 population) was very similar to that of 2020 (9.9 per 1,000) and well above that observed before the COVID-19 pandemic (9.1 per 1,000 in 2019). The crude birth rate was also stable. Despite a slight increase (11.0 per 1,000 in 2021 vs. 10.9 per 1,000 in 2020), it remained below the pre-pandemic level (11.2 per 1,000 in 2019).
Table 1. Indicators of population change (metropolitan France and whole of France, 1990, 2000, 2010, and since 2015)
Years Population on 1 January (thousands) Births (thousands) Deaths (thousands) Natural increase (thousands) Net migration (thousands) Adjustment for migration (thousands) Total change after adjustment (thousands) Crude birth rate (per thousand) Crude mortality rate (per thousand) Total increase before adjustment (per thousand) Total increase after adjustment (per thousand) Metro. France Whole of France Metro. France Whole of France Metro. France Whole of France Metro. France Whole of France Metro. France Whole of France Metro. France Whole of France Metro. France Whole of France Metro. France Whole of France Metro. France Whole of France Metro. France Whole of France Metro. France Whole of France 1990 2000 2010 2015 2016 2017 2018 2019* 2020* 2021* 2022* 56,577 57,996 58,858 60,508 62,765 64,613 64,301 66,422 64,469 66,603 64,639 66,774 64,844 66,992 65,097 67,258 65,284 67,454 65,448 67,626 65,628 67,813 762 793 775 807 802 833 760 799 745 784 730 770 720 759 714 753 697 735 702 743 n/a n/a 526 534 531 541 540 551 582 594 581 594 594 606 597 610 599 613 655 669 644 662 n/a n/a 236 259 244 267 262 282 179 205 164 190 137 163 123 149 115 140 42 66 58 81 n/a n/a 80 77 70 72 43 39 53 40 88 65 167 154 103 87 155 140 155 140 155 140 n/a n/a –53 –52 94 94 0 0 –63 –64 –82 –84 –98 –100 27 30 –82 –84 –34 –34 –33 –34 n/a n/a 316 336 314 339 305 320 232 246 252 255 304 317 226 236 270 280 197 206 213 221 n/a n/a 13.0 14.0 13.0 13.0 13.0 13.0 11.8 12.0 11.5 11.8 11.3 11.5 11.1 11.3 11.0 11.2 10.7 10.9 10.7 11.0 n/a n/a 9.3 9.2 9.0 8.9 8.6 8.5 9.0 8.9 9.0 8.9 9.2 9.1 9.2 9.1 9.2 9.1 10.0 9.9 9.8 9.8 n/a n/a 5.6 5.8 5.3 5.6 4.9 5.0 3.6 3.7 3.9 3.8 4.7 4.7 3.5 3.5 4.1 4.2 3.0 3.1 3.3 3.3 n/a n/a 4.7 4.9 6.9 7.2 4.9 5.0 2.6 2.7 2.6 2.6 3.2 3.3 3.9 4.0 2.9 2.9 2.5 2.6 2.8 2.8 n/a n/a
Table 1. Indicators of population change (metropolitan France and whole of France, 1990, 2000, 2010, and since 2015)Note: * Estimated data; n/a = data not available. The numbers of births and deaths in 2021 are those published by INSEE on 28 July 2022 (they are slightly different from those published in the demographic report in January 2022).
Coverage: Whole of France, metropolitan France.
15 While natural growth in 2021 was positive (+81,000) and greater than in 2020 (+66,000), it remained much lower than the levels observed in previous decades. For the second straight year, natural growth contributed less to population increase than net migration, both before and after adjustment  (106,000 and 140,000, respectively). Immigration is an ever stronger component of the demographic dynamics of France (Lê, 2021).
Figure 1. Quarterly numbers of births and deaths in France, 2010–2022
Figure 1. Quarterly numbers of births and deaths in France, 2010–2022Note: Dotted lines indicate estimated data.
Coverage: Whole of France (including Mayotte).
2 – France is one of nine European countries with positive natural growth and net migration
16 While its population growth is slowing, France is still an exception. In the 27 countries of the European Union (EU-27), the population fell by 170,000 in 2021  due mainly to strongly negative natural growth (–1.2 million). Only nine of the 27 countries, including France, had positive natural growth.  While France was the country with the largest natural growth in absolute terms, it ranks 5th relative to its population (behind Ireland, Luxembourg, Sweden, and Cyprus). These nine countries also have positive net migration and represent a demographic driving force in the EU. Among the 18 countries with negative natural growth, half recorded an overall increase in population due to positive net migration. This was the case in Germany (+82,000 in 2021) and Spain (+40,000), for example. In Italy (–253,000) and Poland (–185,000), positive net migration was not sufficient to offset negative natural growth. In five countries—Greece, Latvia, Croatia, Romania, and Slovakia—net migration and natural growth were both negative.
3 – Less rapid ageing in France than in Europe as a whole, but France has the highest dependency ratio
17 The French population pyramid, which shows the population age–sex structure, is continuing to narrow at the base (i.e. at the youngest ages) due to a steady decrease in births over the last 10 years, and is widening at the top as the baby-boom cohorts grow older (Figure 2). People born in 1945 reached age 75 in 2020, and the share of over-75s has started increasing more sharply over the last 2 years, reaching 9.8% in metropolitan France on 1 January 2022 and 9.9% in France as a whole (Appendix 2). This age group will soon account for more than 10% of the population. 
Figure 2. Population pyramid of France on 1 January 2022
Figure 2. Population pyramid of France on 1 January 2022Coverage: Whole of France (including Mayotte).
18 In the 20 years from 2001 to 2021, median age increased by 4.5 years, from 37.6 to 42.1 years.  This increase is smaller than that observed in EU-27 as a whole (+5.7 years, from 38.4 to 44.1 years). Population ageing as measured by this indicator is especially severe in the countries of Eastern Europe (Romania, +8.6 years; Lithuania, +7.9 years; Slovakia, +7.1 years) and Southern Europe (Portugal, +7.9 years; Italy, +7.2 years; Spain, +7.1 years). Sweden and, to a lesser extent, Belgium and Luxembourg, have a more stable age structure, with respective increases of just 1.2, 2.5, and 2.8 years. 
19 Despite a steady decrease in births over the last 10 years, the French population pyramid is among the youngest in Europe (23.9% of under-20s, in second place behind Ireland, with 26.4%).  And as the proportion of over-65s is close to the European average, this means that France was the EU country with the highest dependency ratio (0.8) in 2021,  with almost one person of generally non–working age (under 20 or over 65) for each working-age person (ages 20–64).
4 – Strong regional disparities across France
20 Patterns of change are very variable across the country. To observe them, we use a regional breakdown corresponding to the former French regions,  equivalent to Level 2 of the European Nomenclature of Territorial Units for Statistics (NUTS-2) (Appendix Figure A.1). 
21 Over the 5 years between 1 January 2016 and 1 January 2021, the largest mean annual population increases were observed in Mayotte (+36.7 per 1,000) and French Guiana (+15.3 per 1,000), the two overseas territories that still have a total fertility rate well above replacement level.  These two outlier departments with relatively small populations (each representing 0.4% of the national total) are followed by the regions of southern and western France (Figure 3; Corsica, +9.6 per 1,000; Languedoc-Roussillon, +7.1 per 1,000; Aquitaine, +6.9 per 1,000; Midi-Pyrénées, +6.8 per 1,000; Pays de la Loire, +6.0 per 1,000; Rhône-Alpes, +5.6 per 1,000; Brittany, +4.8 per 1,000; Provence-Alpes-Côte d’Azur, +3.8 per 1,000; Île-de-France [Paris region], +3.7 per 1,000; Alsace, +3.6 per 1,000). At the other extreme, two other overseas departments, Martinique (–11.1 per 1,000) and Guadeloupe (–10.3 per 1,000), have seen a population decline. This was also the case in the regions of the east (Lorraine, Champagne-Ardenne, Franche-Comté) and the Centre (Limousin, Burgundy, Centre), located in the increasingly pronounced ‘empty diagonal’ (Oliveau and Doignon, 2016; Breton et al., 2017) but also in the regions of northern France (Nord-Pas-de-Calais, Picardie, Basse- and Haute-Normandie). In all, out of the 27 former French regions, 12 experienced population shrinkage between 2016 and early 2021, in which year they represented 29.2% of the population. Declines were recorded in only five regions between 2011 and 2016, when these regions accounted for just 9.3% of the population. The regions where growth became negative between 2016 and 2021 are the former Nord regions (Nord-Pas-de-Calais and Picardie), the two regions of Normandy, and Franche-Comté. Be it positive or negative between 2016 and 2021, the rate of population change slowed in all regions between the 2011–2016 and 2016–2021 periods, except in Alsace, where it remained practically stable (from +3.4 per 1,000 to +3.6 per 1,000).
Figure 3. Mean annual population increase in the French regions (NUTS-2) between 1 January 2016 and 1 January 2021
Figure 3. Mean annual population increase in the French regions (NUTS-2) between 1 January 2016 and 1 January 2021
22 While the COVID-19 epidemic—which had a major impact on population structure at the end of the period—did not affect all regions equally (Le Minez and Roux, 2021), the hierarchy of regions in terms of population change remained very similar between the start of the period (from 1 January 2015 to 1 January 2020) and the end (from 1 January 2020 to 1 January 2021).  The regions where the population declined between 2020 and 2021 are the same as those where it also fell between 1 January 2015 and 1 January 2020.
23 Total population growth in the regions is due mainly to net migration.  Four groups can be defined in terms of their positive or negative natural growth and net migration (Figure 4). In all, five former regions have both negative net migration and negative natural growth. They are all located in the northern half of France. Île-de-France follows a singular pattern, with an increasing population (Figure 3) despite negative net migration (Figure 4).
Figure 4. Natural population growth and net migration in the French regions (NUTS-2) between 1 January 2016 and 1 January 2021
Figure 4. Natural population growth and net migration in the French regions (NUTS-2) between 1 January 2016 and 1 January 2021
24 The variability of population growth dynamics in the French regions is unique in Europe, both in terms of total growth (Appendix Figure A.2) and the contribution of its components (Appendix Figure A.3). Regarding population growth, Spain is the only other country with strong regional contrasts, like those observed in France. In other countries, the situation is clearly more homogeneous, with either population growth, as in Sweden, Germany, Austria, Slovenia, and the Benelux countries, or decline, as in Portugal, Italy, Croatia, Greece, Bulgaria, Romania, Hungary, Poland, Finland, and the Baltic countries. Concerning the two components (natural growth and net migration), only Poland and Slovakia, like France, have regions in all the four possible combinations. In Germany, Sweden, and the Benelux countries, all regions have positive net migration combined with either positive or negative natural growth. In Italy, as in all countries of south-eastern Europe (Croatia, Hungary, Romania, Bulgaria, Greece), natural growth is negative in all regions. This is also the case in Spain, excepting the regions of Madrid, Murcia, and the Balearics, and in Portugal (excluding Lisbon).
25 The age structures of the former French regions are also very diverse and, again, atypical. The regions with a relatively high proportion of over-60s (+30%) are Languedoc-Roussillon, due largely to retirement migration, but also the regions of central France where populations are declining because younger people are moving away (Figure 5; Brutel, 2022). Conversely, the share of over-60s is relatively small in the overseas regions (excepting the two Caribbean islands), in northern France, and Île-de-France, regions where the population is young and births are numerous, but which also attract students and young workers (mainly in Île-de-France).
Figure 5. Proportion of over-60s in France on 1 January 2021 by region
Figure 5. Proportion of over-60s in France on 1 January 2021 by region
26 At the European level, contrasts between regions are large in some countries, as is the case in France. In Spain, for example, the eastern regions are younger, while in Germany the eastern regions have the oldest populations. In other countries such as Italy, Greece, or Finland, the proportion of over-60s is above 30% in all or most regions. None of the countries of Eastern Europe (Appendix Figure A.4) is in this situation. Studies of age structure and population ageing in the NUTS-2 regions have shown an effect of urbanization, but no rural–urban divergence in ageing is observed between 2003 and 2013 (Kashnitsky et al., 2021). The convergence of age structures in urban and rural areas will depend partly on the demographic dynamics of working-age populations (migration and mortality), in the countries of Eastern Europe especially (Kashnitsky et al., 2017).
5 – One in five French people live in a rural municipality
27 Since November 2020, municipalities qualified as sparsely or very sparsely populated in the municipality density grid are defined by INSEE as rural. (Beck et al., 2022).  Under this definition, 1 in 5 French people live in a rural municipality (20% of women and 22% of men). This proportion varies with age. It is highest below age 18  and lowest at ages 18–25 (Figure 6), when young adults leave home to attend university or to work in towns and cities. The share of women is slightly lower in rural municipalities, and the mean age of the population is slightly younger (Figure 6). The relative proportion of families with small children is higher due to migration at working ages and the movement of dependent older adults to urban settings where care is more easily available. Mobility at municipality level can be measured by estimating the number of people already living in the municipality the previous year (Figure 7, dotted curves). Mobility is high at early adult ages (when people move for reasons of education or employment), but also at advanced ages, above 80, when dependent older adults may need to move to a town to receive care or enter a care home.
Figure 6. Population structure by age and sex in France on 1 January 2019 by place of residence and recent mobility
Figure 6. Population structure by age and sex in France on 1 January 2019 by place of residence and recent mobilityCoverage: Whole of France (excluding Mayotte).
Figure 7. Share of French population living in rural and urban municipalities at each age on 1 January 2019
Figure 7. Share of French population living in rural and urban municipalities at each age on 1 January 2019Coverage: Whole of France (excluding Mayotte).
II – Immigration from third countries
28 This section describes recent immigration trends  based on residence permit statistics. It concerns third-country nationals who are required to hold a residence permit to reside in France, so immigrants from member countries of the European Economic Area  (EEA) and Switzerland are excluded (see Box 1 for an estimation based on other sources). To ensure consistency of comparisons over time, the statistics are established for constant geographical areas. We do not count people of nationalities that were formerly required to hold a residence permit but are now exempted.  Due to the delays in registering certain types of residence permit, migration flows cannot be accurately determined beyond 2020. And the effect of the departure of the United Kingdom from the European Union cannot be analysed until statistics for 2021 become available. 
29 Inflows of third-country nationals settling legally in France are estimated via the numbers of residence permits and long-term visas valid as residence permits granted by prefectures in France  and by embassies and consulates abroad. The relevant data are drawn from the system used by the French Ministry of the Interior to track the status of foreigners residing in France (AGDREF) and transmitted annually to INED. The method developed by d’Albis and Boubtane (2015) is used to construct these flows. It applies the basic principle whereby people are counted in the flows of the year in which they receive their first residence permit valid for 1 year or more.  This is generally the same as the year of entry, although in some cases it may be later (for example, if the person previously held a more short-term residence permit or if they are a minor). The term inflows used here thus corresponds specifically to the flow of people who obtain permanent migrant status as legal, long-term residents in France. It is not an estimation of physical entries into French territory, whose numbers are not directly recorded in France. Likewise, departures are not measured here.
1 – An 18% decrease in inflows from third countries in 2020
30 Table 2 shows inflow data, as defined previously, for the years 2015 to 2020. Entries in 2020 totalled 211,255, far fewer than in 2019. The measures enforced to combat the COVID-19 pandemic included restrictions on international mobility.  Total inflows fell by 17.8% between 2019 and 2020, the largest recorded drop since the data series was first created (Appendix Table A.3).
31 The decrease was stronger for holders of residence permits valid for less than 10 years (–18.1%) than for those with a permit valid for 10 years or more (–15.9%). Among the individuals counted, the share of immigrants receiving a residence permit valid for 10 years or more (11.8% in 2020) increased very slightly. Residence permits valid for 10 years or more (typically resident cards) are rarely granted as first permits but generally after one or more permits of less than 10 years.
Box 1. Estimating inflows from the countries of the European Economic Area (EEA) and Switzerland
Table 2. Number of first residence permits valid for more than 1 year by first year of validity and period of validity
Period of validity 2015 2016 2017 2018 2019 2020 Less than 10 years 187,626 193,163 208,772 222,155 227,409 186,242 10 years or more 22,414 25,191 28,969 27,319 29,728 25,013 Total 210,040 218,354 237,741 249,474 257,137 211,255 Share of permits of 10 years or more in the total (%) 10.7 11.5 12.2 11.0 11.6 11.8
Table 2. Number of first residence permits valid for more than 1 year by first year of validity and period of validityCoverage: Permits granted in France and in embassies and consulates abroad to foreign nationals, excluding citizens of the EEA and Switzerland (constant geographical area from 2015 to 2020). Permits granted in year n and included in the AGDREF data extracted in July of the year n + 2. Permits of less than 10 years are valid for between 364 and 3,649 days; permits of 10 years or more are valid for more than 3,649 days.
32 Because of the restrictive measures taken to combat the pandemic, 2020 was marked not only by a decrease in the number of residence permits issued but also by a change in their distribution over the year. This can be illustrated by comparing the distribution of permits granted each month in 2019 and in 2020 (Figure 8). Due to the closure of prefectures, only a small share of the annual total was granted during the first lockdown that began on 17 March 2020: only 1.8% of the 2020 total was granted in April compared with 6.6% the previous year. This shortfall was partly offset by an increase in the numbers issued in the last quarter of the year.
Figure 8. Share of annual residence permits granted each month in 2019 and 2020
Figure 8. Share of annual residence permits granted each month in 2019 and 2020Coverage: See Table 2.
2 – Other measures of inflows
33 Inflows of foreigners can also be estimated from other statistical sources. INSEE uses population censuses and specifically a question on the place of residence in the year preceding the census survey. According to Eurostat, which disseminates INSEE data, 180,252 third-country nationals entered France in 2019,  76,885 fewer than the number estimated based on residence permits (Table 2). This low estimate is perhaps due partly to the non-exhaustive coverage of foreign students in the census survey, even if they remain in France for more than 1 year.
34 Counting students in migration flows is a complex problem. Under certain statistical conventions, they should be counted if they stay for at least 1 year in the host country; this is the approach applied here. Under others, based on the rules of the National Account System, they should not be counted unless they plan to stay in the host country after completing their education. This is the convention applied by the Organisation for Economic Co-operation and Development (OECD). Flows based on the OECD convention can be estimated using the method proposed by d’Albis and Boubtane (2021b) by applying the rule of the first residence permit valid for more than 1 year to all permits except those granted for educational reasons. Hence, while people who hold a student permit throughout their stay in France are not counted (even if they stay for more than 1 year), those who receive a permit for another reason after holding a student permit are counted in the year of their first change of status. Figure 9 shows inflows since 2000 measured with these two methods. If we ignore first residence permits granted for educational reasons, annual inflows are necessarily smaller (184,210 people in 2020), but the numbers calculated using both methods have been evolving in parallel for the last 2 decades.
Figure 9. Annual inflows based on number of residence permits, including or excluding those granted for educational reasons, 2000–2020
Figure 9. Annual inflows based on number of residence permits, including or excluding those granted for educational reasons, 2000–2020Coverage: See Table 2.
3 – Two-thirds of permits granted to young adults
35 Whatever the reason for admission, new immigrants counted in total inflows are young. In 2020, people aged 18–34 represented 66% of all arrivals (Table 3) and 73% of all immigrants who were adults when they received their first residence permit. The year 2020 was marked by a sharp decrease in the share of minors, whose number totalled 19,704 (9.3% of admissions), down from 31,491 the previous year. However, this figure for minors should be treated with caution: by definition, only minors with a residence permit  are counted in inflows. Foreign minors do not have to hold a residence permit but may need to obtain one if they wish to travel outside France, for example. This drop in the number of minors is likely a consequence of the border closures linked to the COVID-19 pandemic. Note also that minors born in France to foreign parents are, by definition, not counted as immigrants. The first row of Table 3 thus only includes minors born abroad who hold a residence permit.
Table 3. Distribution (%) of first residence permits valid for 1 year or more by first year of validity and age group
Age group 2015 2016 2017 2018 2019 2020 0–17 (a) 10.2 10.3 10.3 10.8 12.2 9.3 18–34 62.5 63.1 64.5 64.0 62.5 66.0 35–64 25.5 24.9 23.7 23.5 23.6 23.2 65+ 1.7 1.8 1.6 1.6 1.7 1.5 Total 100 100 100 100 100 100
Table 3. Distribution (%) of first residence permits valid for 1 year or more by first year of validity and age group(a) Foreign minors are not required to hold a residence permit. The first row includes minors born abroad who hold a residence permit.
Coverage: See Table 2.
36 Figure 10 gives a more detailed representation of inflows by age and sex in 2020. The spike at ages 18 and 19 observed each year corresponds to people who arrived as minors and who apply for a residence permit when they reach age 18, often to enter the labour market. This spike is very pronounced for men  and represented more than 15,000 permits at ages 18 and 19. In all, 23.2% of men and 18.2% of women with a new residence permit are below age 20. Students also contribute to the large share of young people in the age distribution of inflows. Apart from the male spike at ages 18–19, the age distributions of women and men are very similar. Their mean age at arrival was 29.0 years and 27.9 years, respectively, in 2020 compared with 29.3 and 29.1 years in 2015, a notable age decrease for men.
Figure 10. Age and sex distribution of first residence permits valid for at least 1 year granted in 2020
Figure 10. Age and sex distribution of first residence permits valid for at least 1 year granted in 2020Coverage: See Table 2.
4 – Almost two-thirds of permits granted to Africans
37 The majority of third-country nationals  who enter France are from Africa, and their share has increased considerably in recent years. Africans accounted for 64.6% of total inflows in 2020 (Table 4), the highest percentage since this data series was first recorded (d’Albis and Boubtane, 2015). The share of nationals from other continents is correspondingly lower. Inflows of Asian nationals are almost 3 times smaller than inflows of Africans.
Table 4. Breakdown and distribution (%) of first residence permits valid for 1 year or more by first year of validity and continent of origin
Continent of origin 2015 2016 2017 2018 2019 2020 Africa 122,294 126,171 138,921 147,611 157,886 136,544 Americas 21,835 20,437 21,146 22,961 21,977 16,033 Asia(a) 51,301 55,920 61,315 63,224 61,735 46,976 Europe(b) 13,266 14,541 15,036 14,423 14,251 10,712 Oceania 809 803 924 927 805 632 Total(c) 210,040 218,354 237,741 249,474 257,137 211,255 Africa (%) 58.2 57.8 58.4 59.2 61.4 64.6 Americas (%) 10.4 9.4 8.9 9.2 8.5 7.6 Asia(a) (%) 24.4 25.6 25.8 25.3 24.0 22.2 Europe(b) (%) 6.3 6.7 6.3 5.8 5.5 5.1 Oceania (%) 0.4 0.4 0.4 0.4 0.3 0.3 Total(c) 100 100 100 100 100 100
Table 4. Breakdown and distribution (%) of first residence permits valid for 1 year or more by first year of validity and continent of origin(a) Türkiye is included in Asia.
(b) Europe includes all European countries outside the EEA and Switzerland.
(c) The total is not equal to the sum due to missing values (origin of person unknown).
Coverage: See Table 2.
38 Since 2017, women no longer account for the majority of inflows, and in 2020 they represented only 47.3% of total entries versus 51.6% in 2015 (Table 5). In 2020, while there were fewer women than men among inflows from Africa (46.4%) and Asia (44.7%), they formed a large majority among inflows from the Americas (56.6%) and Europe outside the EEA and Switzerland (56.9%). The rising share of inflows from Africa partly explains the decreasing share of women in overall inflows.
Table 5. Share of women (%) among recipients of a first residence permit by first year of validity and continent of origin
Continent of origin 2015 2016 2017 2018 2019 2020 Africa 49.3 48.3 46.7 47.5 46.4 46.4 Americas 56.7 57.3 58.1 57.7 57.6 56.6 Asia 53.0 51.3 48.5 47.9 46.5 44.7 Europe (excluding EEA and Switzerland) 60.0 58.6 58.1 58.3 57.7 56.9 Oceania 52.7 53.5 54.8 52.5 48.8 50.0 Overall 51.6 50.6 48.9 49.2 48.0 47.3
Table 5. Share of women (%) among recipients of a first residence permit by first year of validity and continent of originCoverage: See Table 4 and note to Table 6.
5 – A sharper fall in inflows from third countries for employment reasons
39 Inflows can be broken down by reason for admission stated on the first residence permit valid from more than 1 year (Table 6). Family reasons accounted for the largest share of admissions in 2020, representing 41.1% of the total. The other main reasons were education (28.3% in 2020), humanitarian (12.8%), and employment (11%). Admissions for all reasons decreased in 2020 but to very different extents. In 2020, 86,830 admissions were for family reasons, 18.2% fewer than in 2019, of which 44% concerned people with family ties to a French citizen. Admissions for educational reasons  in 2020 fell by just 6.8%, to 59,879, probably due to the easing of travel restrictions in the summer of that year. This smaller drop explains the increase in students among 2020 inflows, with admissions for educational reasons reaching a level unequalled since the data series was first established (i.e. since 2000; d’Albis and Boubtane, 2015). There were 27,128 admissions for humanitarian reasons in 2020, a drop of 19.7%. They mainly concern two types of situations: (a) people admitted as refugees, stateless persons, or beneficiaries of territorial asylum or subsidiary protection (23,922 people in 2020) ; or (b) foreigners who are ill (2,967 people). The number of permits granted to refugees has fallen back from the 2017 peak that was due largely to the war in Syria, when 30,180 permits were granted in response to the increase in asylum applications since 2014 (d’Albis and Boubtane, 2018a). Last, with a 25% drop between 2019 and 2020, admissions for employment reasons were severely affected by the restrictions associated with the COVID-19 pandemic. They totalled 23,166 in 2020, broken down into three categories: employees or self-employed (14,502), ‘passeport talent’  holders (5,192), and seasonal or temporary workers with a residence permit valid for more than 1 year (3,472).
Table 6. Breakdown and distribution (%) of first residence permits valid for 1 year or more by first year of validity and reason for admission
Reason for admission 2015 2016 2017 2018 2019 2020 Family 93,069 89,169 89,849 93,969 106,213 86,830 Education 53,069 56,507 63,809 65,786 64,262 59,879 Humanitarian 19,490 25,866 34,233 31,883 33,800 27,128 including refugee(a) 13,265 19,581 30,180 27,887 29,403 23,922 Employment 16,287 18,151 21,812 27,518 30,888 23,166 Other 28,125 28,661 28,038 30,318 21,974 14,252 Total 210,040 218,354 237,741 249,474 257,137 211,255 Family (%) 44.3 40.8 37.8 37.7 41.3 41.1 Education (%) 25.3 25.9 26.8 26.4 25.0 28.3 Humanitarian (%) 9.3 11.8 14.4 12.8 13.1 12.8 including refugee(a) (%) 6.3 9.0 12.7 11.2 11.4 11.3 Employment (%) 7.8 8.3 9.2 11.0 12.0 11.0 Other (%) 13.4 13.1 11.8 12.2 8.5 6.7 Total 100 100 100 100 100 100
Table 6. Breakdown and distribution (%) of first residence permits valid for 1 year or more by first year of validity and reason for admission(a) The refugee row covers permits granted on the following grounds: refugee and stateless, territorial asylum, and subsidiary protection.
Note: The table is updated each year to take account of new information on reasons for admission communicated by the Ministry of the Interior.
Coverage: See Table 2.
40 Reasons for admission vary by sex. Women are still over-represented among immigrants admitted for family reasons and under-represented among those admitted for humanitarian and, above all, employment reasons (Table 7). Among students, women are only slightly outnumbered by men.
Table 7. Share of women (%) among recipients of first residence permits valid for 1 year or more by first year of validity and reason for admission
Reason for admission 2015 2016 2017 2018 2019 2020 Family 61.1 61.2 61.3 62.4 60.0 59.5 Education 49.0 49.4 48.1 49.2 48.1 48.8 Humanitarian 41.3 38.9 34.8 34.6 35.3 33.0 Employment 30.2 21.5 16.9 22.9 22.9 19.6 Overall 51.6 50.6 48.9 49.2 48.0 47.3
Table 7. Share of women (%) among recipients of first residence permits valid for 1 year or more by first year of validity and reason for admissionCoverage: See Table 2.
41 Reasons for admission are distributed very differently across continents of origin (Table 8). The main reason for admission from all continents is still family-related, except for people from Asia, among whom this reason is under-represented (28.1% of these admissions in 2020 compared with 41.1% on average). It is over-represented, on the other hand, among Europeans from outside the EEA and Switzerland (50.5%), from Africa (44.5%), and from the Americas (44.4%). Educational reasons are under-represented among third-country Europeans (11.5% compared with 28.3% on average) and over-represented among those from the Americas (30.0%), and Africans (29.6%), while Asians are close to the average (28.5%). Humanitarian reasons account for a large share of permits granted to Asians (25.4% vs. 12.8% on average), notably due to the impact of the Syrian war, and to Europeans (14.7%), and are under-represented among Africans (9.3%) and, above all, those from the Americas (4.3%), for whom employment-related reasons are over-represented (13.3% vs. 11.0% on average), while Europeans are under-represented (8.8%).
42 Among Africans admitted to France in 2020, the share admitted for educational reasons increased sharply. Despite the pandemic, the number of admissions for educational reasons even increased, rising from 39,986 in 2019 to 40,367 in 2020. Among inflows from the Americas, the share of admissions for educational reasons also rose, although their number decreased. Among Asians, the share admitted for educational and humanitarian reasons increased, with admissions for educational reasons outnumbering those for family reasons (13,369 and 13,199). Last, migration from third-country Europe observed here was marked by a further increase in the share of admissions for family reasons that began in 2019, and a decrease in humanitarian admissions (these were unusually high in 2016 and 2017 due to an inflow of Albanians) (Table 8).
Table 8. Breakdown and distribution (%) of first residence permits valid for 1 year or more by first year of validity, continent of origin, and reason for admission
Continent of origin and reason for admission 2015 2016 2017 2018 2019 2020 Africa Family 64,852 60,895 61,468 64,204 71,432 60,766 Education 27,858 31,321 38,288 38,891 39,986 40,367 Humanitarian 8,209 11,329 14,357 13,583 15,563 12,732 Employment 7,026 8,440 10,939 15,307 18,933 14,445 Africa (%) Family 53.0 48.3 44.2 43.5 45.2 44.5 Education 22.8 24.8 27.6 26.3 25.3 29.6 Humanitarian 6.7 9.0 10.3 9.2 9.9 9.3 Employment 5.7 6.7 7.9 10.4 12.0 10.6 Americas Family 8,228 7,518 7,653 8,279 9,343 7,115 Education 6,504 6,113 6,427 6,806 6,024 4,810 Humanitarian 389 416 709 789 980 683 Employment 2,922 2,514 2,717 3,075 3,055 2,135 Americas (%) Family 37.7 36.8 36.2 36.1 42.5 44.4 Education 29.8 29.9 30.4 29.6 27.4 30.0 Humanitarian 1.8 2.0 3.4 3.4 4.5 4.3 Employment 13.4 12.3 12.8 13.4 13.9 13.3 Asia Family 14,342 14,791 14,990 15,536 18,192 13,199 Education 16,834 17,261 17,291 18,464 16,749 13,369 Humanitarian 7,882 10,399 15,257 14,366 14,468 11,921 Employment 5,273 6,107 6,901 7,779 7,468 5,439 Asia (%) Family 28.0 26.5 24.4 24.6 29.5 28.1 Education 32.8 30.9 28.2 29.2 27.1 28.5 Humanitarian 15.4 18.6 24.9 22.7 23.4 25.4 Employment 10.3 10.9 11.3 12.3 12.1 11.6 Europe (excluding EEA and Switzerland) Family 5,268 5,581 5,359 5,554 6,801 5,413 Education 1,743 1,685 1,674 1,495 1,409 1,235 Humanitarian 2,775 3,527 3,700 2,947 2,533 1,578 Employment 868 909 1,004 1,127 1,190 948 Europe (%) Family 39.7 38.4 35.6 38.5 47.7 50.5 Education 13.1 11.6 11.1 10.4 9.9 11.5 Humanitarian 20.9 24.3 24.6 20.4 17.8 14.7 Employment 6.5 6.3 6.7 7.8 8.4 8.8
Table 8. Breakdown and distribution (%) of first residence permits valid for 1 year or more by first year of validity, continent of origin, and reason for admissionCoverage: See Table 2 and notes to Table 4.
6 – Asylum seekers admitted for residence account for almost 19% of inflows from third countries
43 Asylum seekers may be admitted for residence in France (i.e. receive a residence permit of 1 year or more) in several ways. In the first case, if their application is accepted, they obtain a permit on humanitarian grounds and are counted as ‘refugees’ (Table 6, Row 5). The second case concerns those whose application is rejected, some of whom may be admitted for residence on different grounds, most often family reasons. The rates of admission for residence by date of submission of the asylum application are given in d’Albis and Boubtane (2018a). The perspective here is different. Table 9 shows the annual number of people having submitted an asylum application to the French Office for the Protection of Refugees and Stateless Persons (OFPRA) and who are admitted for residence, for any reason. They totalled 39,497 people in 2020 and represented 18.7% of overall inflows. The number of asylum seekers admitted for residence fell by 16.6% with respect to 2019, but their proportion in total inflows remained stable. Asylum seekers are not admitted solely on humanitarian grounds, so the total number of admissions is higher than the number admitted for this reason alone; 25.7% of asylum seekers who entered France in 2020 were admitted for family reasons.
Table 9. Numbers of people admitted for residence after applying for asylum, by first year of validity of first residence permit valid for 1 year or more, and distribution (%) by sex, continent of origin, and reason for admission
2015 2016 2017 2018 2019 2020 Numbers 27,507 35,262 46,174 44,470 47,353 39,497 Share of women (%) 41.1 39.7 36.6 36.5 36.7 35.3 Continent of origin (%) Africa 35.9 36.9 38.2 38.8 42.2 44.6 Americas 4.1 3.0 2.8 2.8 3.0 2.5 Asia 44.0 43.9 45.5 46.0 43.4 42.6 Europe 14.9 15.3 12.9 11.9 10.7 9.6 Reason for admission (%) Family 27.1 24.2 21.2 24.0 24.8 25.7 Education 0.5 0.3 0.3 0.3 0.2 0.3 Humanitarian 57.8 62.9 68.9 65.3 65.2 63.5 Employment 6.2 6.3 5.1 5.8 5.5 6.3
Table 9. Numbers of people admitted for residence after applying for asylum, by first year of validity of first residence permit valid for 1 year or more, and distribution (%) by sex, continent of origin, and reason for admissionCoverage: Permits granted in France and abroad to foreign nationals who applied for asylum between 1985 and the first year of validity of the first residence permit valid for 1 year or more. Permits granted in year n and recorded in the data extracted in July of the year n + 2. See notes to Table 4.
44 The share of women among asylum seekers admitted for residence (33.5% in 2020) is lower than among the general population of residence permit holders. This proportion has been falling sharply since 2015, when it stood at 41.1% Over recent years, the proportion of Africans has increased steadily, and they now account for the largest share of asylum seekers admitted for residence, at 44.6%, up from 35.9% in 2015. Conversely, the share of Asians is falling, down from 44.0% in 2015 to 42.6% in 2020.
7 – Inflows centred on Île-de-France
45 The inflows presented above concern the whole of France, but spatial disparities are very large (Breton et al., 2017; d’Albis and Boubtane, 2018b). Practically all holders of first residence permits of 1 year or more (96.7% in 2019 and 97.2% in 2020) live in metropolitan France. Table 10 shows the number of permits granted and the inflows to each region of metropolitan France as a share of total inflows. The listed regions correspond to the NUTS-2 regions that existed until 2015. New arrivals are highly concentrated in Île-de-France, which accounted for 40.1% of total inflows to metropolitan France in 2019 and 38.4% in 2020. It is followed by the regions that include France’s two other largest cities, Lyon and Marseille: Rhône-Alpes (9.5% of inflows in 2020) and Provence-Alpes-Côte d’Azur (7.9%). These three regions alone account for more than half of all inflows to metropolitan France. Flows to Île-de-France and Provence-Alpes-Côte d’Azur have been trending downward for the last 20 years, however.
46 Entries of third-country nationals to the overseas departments and regions of France are shown in Table 11. It gives the numbers and share of the total for the five overseas departments and regions, and for three overseas collectivities (Saint-Barthélemy, Saint Martin, and Saint Pierre and Miquelon). Mayotte accounts for more than half of inflows and French Guiana for more than a quarter. Inflows in 2019 represented 1.7% of its estimated population in 2017, a proportion 5 times higher than that of total inflows to France with respect to the French population.
Table 10. Breakdown and distribution (%) of first residence permits valid for 1 year or more by first year of validity and region of residence, metropolitan France
2000 2019 2020 Numbers Share of total (%) Numbers Share of total (%) Numbers Share of total (%) Alsace 4,837 3.6 7,707 3.1 6,175 3.0 Aquitaine 3,572 2.7 8,890 3.6 7,086 3.5 Auvergne 1,274 1.0 3,578 1.4 2,718 1.3 Basse-Normandie 1,004 0.8 2,387 1.0 2,592 1.3 Brittany 2,115 1.6 6,409 2.6 5,811 2.8 Burgundy 1,601 1.2 3,719 1.5 3,411 1.7 Centre 3,440 2.6 7,159 2.9 6,400 3.1 Champagne-Ardenne 1,645 1.2 3,655 1.5 3,197 1.6 Corsica 875 0.7 838 0.3 515 0.3 Franche-Comté 1,809 1.4 2,811 1.1 2,241 1.1 Haute-Normandie 1,810 1.4 4,869 2.0 4,541 2.2 Île-de-France (Paris region) 58,806 44.3 99,734 40.1 78,815 38.4 Languedoc-Roussillon 6,176 4.7 8,037 3.2 6,458 3.1 Limousin 662 0.5 2,025 0.8 1,628 0.8 Lorraine 3,216 2.4 6,149 2.5 5,046 2.5 Midi-Pyrénées 4,518 3.4 10,166 4.1 8,494 4.1 Nord-Pas-de-Calais 4,378 3.3 9,467 3.8 8,833 4.3 Pays de la Loire 3,003 2.3 10,214 4.1 8,666 4.2 Picardie 2,232 1.7 4,702 1.9 4,210 2.0 Poitou-Charentes 1,272 1.0 3,908 1.6 2,696 1.3 Provence-Alpes-Côte d’Azur 12,097 9.1 19,105 7.7 16,314 7.9 Rhône-Alpes 12,415 9.4 23,067 9.3 19,525 9.5 Total 132,757 100 248,596 100 205,372 100
Table 10. Breakdown and distribution (%) of first residence permits valid for 1 year or more by first year of validity and region of residence, metropolitan FranceCoverage: See Table 2.
Table 11. Breakdown and distribution (%) of first residence permits valid for one year or more by first year of validity and region of residence, French overseas departments and territories
2019 2020 Numbers Share of total (%) Numbers Share of total (%) Guadeloupe 631 7.4 262 4.5 French Guiana 2,153 25.2 1,531 26.0 Réunion 865 10.1 743 12.6 Martinique 316 3.7 237 4.0 Mayotte 4,396 51.5 2,955 50.2 Overseas collectivities 180 2.1 155 2.6 Total 8,541 100 5,883 100
Table 11. Breakdown and distribution (%) of first residence permits valid for one year or more by first year of validity and region of residence, French overseas departments and territoriesCoverage: See Table 2.
III – Births and fertility
1 – An increase in births after an atypical year in 2020
47 According to the most recent INSEE estimates, 742,500 children were born in 2021,  7,300 more than in 2020 but fewer than in 2019 (753,400), the last year before the COVID-19 pandemic. This slight increase in births between 2020 and 2021 is due mainly to a catch-up effect following the slump observed between November 2020 and March 2021 (Brée et al., 2021). Contrary to certain predictions, including those formulated in the 2021 demographic report (Breton et al., 2021), this deficit of births has now been fully offset. The second lockdown in late 2020 (30 October to 15 December) had no apparent impact (no reduction in births in July and August 2021), and it is too early to assess the repercussions of the third lockdown in 2021 (3 March to 3 April).  However, the effects of these second and third lockdowns are more difficult to measure as they partly coincide with the period of recovery in births following the end of the first lockdown. That said, the population has likely become used to these periods of uncertainty linked to COVID-19 and reassured by the arrival of effective vaccines.
48 The year 2021 was marked by large seasonal fluctuations (a peak in births in October and not in July) and by an unprecedented amplitude, with a difference of more than 7,000 births between the months with the most and with the fewest births (February), compared with 3,500 to 4,000 in a normal year (Figure 11). The seasonality of births in the first months of 2022 appears to be returning to the pattern observed before the health crisis.
49 This was the seventh consecutive year that the number of births fell below 800,000 and the second with fewer than 750,000. A similar situation was observed between 1989 and 1999, with a very sharp dip in 1993 and 1994 (Appendix A.1), around 25–30 years ago.  Under a stable fertility regime (timing and intensity of fertility), the current dip in births was relatively predictable, given the slightly smaller size of the mean cohort  of women of reproductive age contributing to fertility in the year. If fertility behaviours remain unchanged, the number of births should remain below 800,000 for at least another 3 or 4 years.
50 In 2021, the increase in births is linked mainly to a change in behaviour (increase in age-specific fertility rates).
Figure 11. Monthly variation in births in France, January 2016 to June 2022
Figure 11. Monthly variation in births in France, January 2016 to June 2022Note: Dotted lines indicate estimated data.
Coverage: Whole of France (including Mayotte since 2014).
2 – A slight fertility rebound after 6 years of decline
51 The total fertility rate (TFR)  in 2021 was 1.83 children per woman for the whole of France and 1.80 for mainland France and Corsica, up slightly on the previous year in both cases but still below the rates of the years preceding the health crisis (Table 12). This slight rebound is due entirely to the fertility of women aged 30 and over, primarily the 30–34 age group (+20 per 1,000) and the 35–39 age group (+12 per 1,000). The steady increase in fertility after age 40 observed since the 1980s in France (Daguet, 2022), levelled off in 2021, doubtless due to the health crisis and the restriction of medically assisted reproduction services (Brée et al., 2021). Fertility before age 30 continued to decline. The age-specific fertility curve is increasingly symmetrical on either side of the peak childbearing age (Figure 12), which is gradually shifting to the right, from 29 years in 2000 to 30 years in 2007 and 31 years in 2019.
52 The mean age at childbearing is still rising (30.9 years in 2021). This is due mainly to the increase in age at first birth, which also pushes back the mean age at childbearing, and the mean age at second and third birth (Figure 13). These increases are almost parallel, but with a slightly steeper slope for the increase in mean age at first birth. 
Table 12. Fertility by age group in France, 2016–2021
Sum of age-specific rates (per 1,000 women) Absolute variation* Age reached in the year 2016 2017 2018 (p) 2019 (p) 2020 (p) 2021 (p) 2016 2017 – 2017 2018 – (p) 2018 2019 – (p) 2019 2020 – (p) 2020 2021 – (p) Under 20 years 32 30 29 28 26 24 –2 –1 –1 –2 –2 20–24 232 224 215 212 201 187 –8 –9 –3 –11 –14 25–29 575 559 545 539 525 518 –16 –14 –7 –14 –7 30–34 645 636 634 633 624 644 –9 –3 –1 –9 20 35–39 345 345 347 350 347 359 0 2 3 –3 12 40–44 89 92 94 96 95 95 3 2 2 –1 0 45+ 6 6 7 7 7 7 1 0 1 0 0 Total (TFR*) 1,924 1,895 1,870 1,864 1,824 1,834 –29 –25 –6 –40 10 Mean age 30.5 30.5 30.6 30.7 30.8 30.9 0.0 0.1 0.1 0.1 0.1
Table 12. Fertility by age group in France, 2016–2021* TFR: total fertility rate (sum of age-specific fertility rates) expressed as a mean number of children per 1,000 women. Due to rounding, the total may differ slightly from the sum, and the variations may not correspond exactly to apparent differences.
(p) Provisional data.
Coverage: Whole of France (including Mayotte since 2014).
53 Women born in 1986 turned 35 in 2021. Assuming that fertility behaviours beyond age 35 will be the same as those observed the previous year (constant age-specific fertility rates), this birth cohort will not be fully replaced if its completed fertility is 2.01 children. Completed fertility may thus decrease from the 1982 to 1986 birth cohorts (Appendix 5),  but more slowly than the TFR, whose level and trends depend on changes in completed fertility across cohorts and the increase in age at childbearing. The completed fertility of the 1986 birth cohort may thus be equivalent to that of the 1968 cohort, but higher than that of the 1972 cohort (1.99 children per woman). France is still one of the few EU-27 countries, alongside Ireland and Sweden, with near-replacement levels of fertility, until the 1980s birth cohorts at least (Breton et al., 2019).
54 Despite the acceleration of fertility decline due to the pandemic, France was still the EU-27 country with the highest TFR in 2020. But its lead is now very small, with respect to Romania especially, where period fertility has been increasing each year since 2011 to reach 1.80 in 2020.  As in practically all Eastern European countries, fertility in Romania fell very sharply in the decade following the collapse of the Eastern Bloc and then rebounded somewhat, but not to its previous level. While the period fertility of France and that of Romania have been converging in recent years, and were similar in 2020, their age profiles are very different; fertility at young ages remains very high in Romania (Figure 12). The pattern of age-specific fertility in Romania hints at the existence of two populations with different behaviours: one that has started to postpone births and to adopt behaviours akin to those of Western Europe, and a second that maintains the tradition of early fertility characteristic of Eastern European countries. This polarization is similar to that observed in England and Spain in the 2000s (Chandola et al., 1999; Kostaki and Paraskevi, 2007).
Figure 12. Age-specific fertility in the two EU-27 countries with the highest period fertility in 2020, France and Romania, 2000 and 2020
Figure 12. Age-specific fertility in the two EU-27 countries with the highest period fertility in 2020, France and Romania, 2000 and 2020Note: The rates for 2020 are 1-year moving averages (15 and 49 years), 3-year moving averages (16 and 48 years) and 5-year moving averages for the other ages.
Figure 13. Mean age at childbearing by birth order of children, France, 2014–2020
Figure 13. Mean age at childbearing by birth order of children, France, 2014–2020Coverage: France.
3 – Fertility variations across French regions
55 According to the provisional estimates transmitted to Eurostat by INSEE, fertility declined in all the former French regions (NUTS-2) between 2019 and 2020, except Guadeloupe. The largest decreases were observed in Mayotte (–9.8%) and Martinique (–5.5%), but also in Auvergne (–4.0%), Lorraine (–3.7%), Poitou-Charentes, and Corsica (–3.5%). They reflect the economic and health impact of the COVID-19 pandemic (Brée et al., 2021). This decline did not change the hierarchy of regions observed before the health crisis (Desplanques, 2011; Aerts, 2013; Breton et al., 2017). The differences in fertility profiles and across regions remain small:  in 2019–2020, three-quarters of the NUTS-2 regions had a TFR between 1.66 and 1.93 (1.63 and 1.86 for metropolitan France) and a mean age at childbearing between 30.1 and 31.0 years (30.1 and 20.9 years in metropolitan France) (Table 13). The regions can be divided into four groups by age-specific fertility rate (Figure 14):
- Regions of metropolitan France with relatively early fertility timing (peak fertility between ages 25 and 29) and a TFR generally between 1.7 and 1.8, except in Lorraine (1.6) and Centre-Val de Loire (1.9). More than half of the regions of metropolitan France are in this group (12), and most are located in the north, north-east, north-west, and centre of France (Figure 14, Group 1).
- Regions of metropolitan France with relatively late fertility timing (peak fertility between ages 30 and 34). This group includes eight regions that can be divided into two subgroups, the first comprising the south-western regions and Alsace, where the TFR is between 1.6 and 1.7, and the second comprising the western and south-eastern regions, with TFRs between 1.8 and 1.9 (Figure 14, Group 2).
- The overseas regions form a third group, with very high levels of fertility before age 25 in Martinique, Guadeloupe, and Réunion, and at all ages for French Guiana and Mayotte, not shown here (Figure 14, Group 3).
- ‘Atypical’ regions of metropolitan France, including Île-de-France, which has high fertility above age 30 (mean age 32.1 years), and Corsica, where rates are lower in all age groups (TFR = 1.4) (Figure 14, Group 4).
56 While differences across the regions of metropolitan France are small, they are persistent and correspond to differences in the socio-economic structure of their populations. In northern and eastern France, for example, entry into adulthood occurs at a younger age, and the share of low-educated women is above the national average (Breton, 2010). Sociocultural factors linked to family systems (as defined by Le Bras and Todd in 2012) may also play a role, with a high prevalence of ‘stem families’ in the south-west, and of nuclear families in the north-west.  This observation is consistent with findings at the departmental level described in the 2017 demographic report (Breton et al., 2017).
Table 13. Fertility indicators in the French regions (NUTS-2), 2010 and 2020
Region Total fertility rate Mean age Proportion of births outside marriage 2009–2010 2019–2020 2009–2010 2019–2020 2010 2020 France 2.02 1.84 29.9 30.8 54.9 62.2 Alsace 1.86 1.67 29.9 30.8 46.4 53.5 Aquitaine 1.84 1.64 30.1 31.0 60.6 68.9 Auvergne 1.88 1.70 29.7 30.4 60.1 71.0 Basse-Normandie 2.01 1.75 29.4 30.2 62.3 72.7 Burgundy 1.94 1.75 29.5 30.2 57.4 66.7 Brittany 2.02 1.77 30.0 30.7 60.4 71.6 Centre–Val de Loire 2.07 1.87 29.5 30.3 56.6 65.7 Champagne-Ardenne 1.99 1.77 29.2 29.8 60.2 69.2 Corsica 1.63 1.39 30.0 30.9 58.1 61.0 Franche-Comté 2.06 1.78 29.4 30.2 54.0 62.9 Haute-Normandie 2.08 1.86 29.2 30.1 60.3 67.7 Île-de-France 2.04 1.91 31.1 32.0 46.0 48.9 Languedoc-Roussillon 1.97 1.77 29.8 30.6 58.1 65.1 Limousin 1.82 1.69 29.5 30.2 62.4 72.3 Lorraine 1.83 1.61 29.5 30.2 54.7 63.0 Midi-Pyrénées 1.85 1.66 30.3 31.1 58.7 66.4 Nord-Pas-de-Calais 2.09 1.84 29.1 30.0 57.3 66.1 Pays de la Loire 2.12 1.84 29.7 30.5 58.1 69.6 Picardie 2.09 1.83 29.1 30.0 60.3 68.2 Poitou-Charentes 1.93 1.69 29.4 30.1 64.1 73.7 Provence-Alpes-Côte d’Azur 2.03 1.94 30.1 30.8 54.0 60.1 Rhône-Alpes 2.05 1.85 30.2 31.0 49.2 56.5 Guadeloupe 2.13 2.33 29.4 30.2 77.2 84.1 French Guiana 3.43 3.71 28.0 29.0 87.5 89.3 Réunion 2.37 2.38 28.4 29.2 71.9 78.2 Martinique 2.05 1.95 29.3 30.1 76.6 80.9 Mayotte — 4.39 — 28.9 — 91.8
Table 13. Fertility indicators in the French regions (NUTS-2), 2010 and 2020Note: The fertility rates are a mean of the rates from two years (2009–2010 and 2019–2020). The highest one-third of indicators are in bold, the lowest one-third are in italics.
Figure 14. Models of age-specific fertility in the French regions (NUTS-2), 2019–2020
Figure 14. Models of age-specific fertility in the French regions (NUTS-2), 2019–2020Note: The fertility rates are a mean of the rates in 2019 and 2020.
57 As is the case at the national level, the TFR has been falling each year since 2010 in practically all regions (Table 13). The decline has been sharpest in the Poitou-Charentes and Nord-Pas-de-Calais regions (–12.4%) and in Provence-Alpes-Côte d’Azur (–4.4%). The TFR is increasing in just four regions— Réunion (+0.4%), Mayotte (+6.0%), French Guiana (+8.2%), and Guadeloupe (+9.4%)—all located overseas. These increases are partly attributable to the particular characteristics of migration flows in these regions, with the departure of more highly educated populations and the arrival of less educated migrants.
58 At the European level, recent geographical studies show a relative convergence of fertility intensity between 1960 and 2015 within NUTS-2 regions and a weakening of the effects of living in a particular country (Buelens, 2022). Conversely, these studies reveal a persistence of cross-country disparities and particularities in fertility timing, beyond subnational differences, linked mainly to the presence or absence of large cities: age at childbearing is later in the most urbanized countries (Buelens, 2021). Other studies at the NUTS-2 level have shown that the effects of the 2008 economic crisis varied across regions but that the impact at national level was a more important factor than local characteristics (Matysiak et al., 2021). France was among the countries where fertility was least affected by this crisis (Papon, 2021). While France has ranked 1st in Europe for fertility for many years, in 2019–2020, it was the country where subnational variance was greatest among those with at least eight NUTS-2 regions.  This dispersion is largely due to the presence of Mayotte and French Guiana. Even when these two territories are excluded, France is still among the countries with the highest level of dispersion, alongside Southern countries (Spain, Italy, and Greece) and certain Eastern countries (Romania and Hungary). Dispersion is twice as large as in Germany, Belgium, Sweden, and the Czech Republic (Appendix Figures A.5 and A.6).
4 – Higher rural fertility
59 Beyond the regional dimension, fertility in France varies substantially at the more local pseudo-canton level (Breton, 2010; Daguet, 2021; INSEE, 2021).  All these spatial analyses of differential fertility confirm the powerful effect of social and economic structures on these disparities, often proxied by education, and more specifically by level of qualification, but also by living standard (Reynaud, 2022). French studies of fertility differentials by socio-economic variables often use indirect methods based on census microdata, such as the own-children method.  These methods generally underestimate fertility (under-reporting of very young children, no way of linking certain children to their mother in a household, etc.) but can be used to assess the effect of certain characteristics by looking at relative differences. 
60 Urban fertility is slightly lower than in rural areas (those outside urban units), where estimated cumulative fertility at age 35 is 1.14 times higher than in urban areas. This is partly linked to the larger share of highly educated women in urban areas (Table 14) who have fewer children. Cumulative fertility at age 35 decreases more sharply with educational level in urban than in rural areas, where educational differences are less pronounced. The urban–rural differential is especially large for the most highly educated women (4 or more years of higher education): the ratio of the cumulative fertility at age 35 of highly educated rural women to that of all women in France is 0.92, and to urban women it is 0.69. These same ratios are 1.05 and 0.89 for women with 2 or 3 years of higher education. The differences are much smaller, or even negligible, for low-educated women (Figure 15, green curves).
Table 14. Distribution and structure of female fertility at age 35 by education, labour market status, and place of residence, France, 2019
Rural (20.8%) Urban (79.2%) % of total % of the category Fertility rate % of total % of the category Fertility rate No qualification Employed 2.4 48.1 1.35 3.4 35.1 1.41 Unemployed 1.0 19.8 1.26 2.3 23.9 1.14 Inactive 1.6 32.2 1.57 4.0 41.0 1.63 Overall 5.0 100.0 1.38 9.7 100.0 1.41 Lower secondary Employed 17.4 72.9 1.25 11.9 58.8 1.18 Unemployed 3.8 15.9 1.18 4.3 21.4 1.06 Inactive 2.7 11.2 1.61 4.0 19.8 1.55 Overall 23.9 100.0 1.24 20.2 100.0 1.18 Upper secondary Employed 21.7 83.4 1.08 15.0 71.0 1.01 Unemployed 2.5 9.5 1.07 3.2 15.4 0.93 Inactive 1.8 7.1 1.44 2.9 13.7 1.44 Overall 26.1 100.0 1.10 21.1 100.0 1.01 2–3 years higher ed. Employed 36.1 90.7 0.94 29.7 82.6 0.86 Unemployed 2.2 5.6 1.05 3.4 9.5 0.85 Inactive 1.5 3.7 1.34 2.9 7.9 1.26 Overall 39.8 100.0 1.05 36.0 100.0 0.88 4 or more years higher ed. Employed 9.1 89.0 0.90 19.5 85.9 0.67 Unemployed 0.7 6.6 0.92 1.9 8.2 0.69 Inactive 0.5 4.4 1.12 1.3 5.9 1.04 Overall 10.2 100.0 0.93 22.7 100.0 0.71 Total 100 1.11 100 0.97
Table 14. Distribution and structure of female fertility at age 35 by education, labour market status, and place of residence, France, 2019Note: Fertility rate = ratio between estimated cumulative fertility at age 35 of the category concerned to that of all women aged 35.
Coverage: Women aged 35 at time of census in the whole of France (excluding Mayotte).
Figure 15. Cumulative fertility at age 35 by educational level, employment status, and place of residence, France, 2019
Figure 15. Cumulative fertility at age 35 by educational level, employment status, and place of residence, France, 2019Note: The reference category is all women.
Coverage: Women aged 35 at time of census in the whole of France (excluding Mayotte).
61 Beyond urban–rural residence and educational level, women’s fertility also depends on employment status; the weaker the attachment to the labour market, the higher the level of fertility (Figure 15, black curves).
62 Whatever their employment status or education, rural women, excepting the least educated, have higher fertility. The direction of causation is difficult to determine, however. For an equivalent level of education and labour market status, rural living may favour fertility (more affordable housing and proximity to family, friends, and carers). Alternatively, couples wanting more children may prefer to settle in a rural area.
5 – Fewer births outside marriage and more traditional child names in eastern France and Île-de-France
63 The uptrend in non-marital births in France is continuing, with the share reaching 63.5% in 2021 (up from 62.2% in 2020). In 2020, this proportion was the highest in Europe, ahead of Bulgaria (59.6%), Portugal (59.7%), and Sweden (55.2%). This is almost double the share observed in Italy (33.8%), Germany (33.1%), and Romania (32.1%), and almost 5 times higher than that of Greece (13.8%). The European rankings are the same as those of 2018 (Breton et al., 2019).
64 At the regional level, the differences observed in 2020 were quite large, with the lowest shares in Île-de-France (48.9%) and Alsace (53.5%) and the highest in Poitou-Charentes (73.7%), Basse-Normandie (72.7%), and Limousin (72.1%) for metropolitan France, and in all the overseas regions (between 91.8% in Mayotte and 78.2% in Réunion) (Table 13). These proportions are rising in all regions. However, this increase is not correlated with the initial level of 2009–2010 (r² = 0.1), and the regions of eastern France, with their more ‘traditional’ family behaviours, are still outliers (Figure 16). The increase is especially strong in the three north-western regions (Pays de la Loire, Brittany, and Basse-Normandie) and in Auvergne, but it is slower in the overseas regions where the proportion of births outside marriage is high, and likewise in Île-de-France, where the proportion is low.
Figure 16. Proportion of births outside marriage in the French NUTS-2 regions in 2010 and 2020
Figure 16. Proportion of births outside marriage in the French NUTS-2 regions in 2010 and 2020
65 The use of double family names increased again slightly in 2020, with 12% of children bearing the name of both their mother and father (vs. 11.8% in 2019), generally the father’s name followed by the mother’s (Table 15). To gauge the prevalence of this practice, it is preferable to calculate this proportion, whatever the order of names, with respect to children who bear their father’s name only. This is because the proportion of children with the mother’s name only is correlated with that of births not recognized by the father and with non-marital births (Breton et al., 2021). In 2020, this proportion was 12.9% (vs. 12.6% in 2019). Among children who bear their father’s name, the proportion with double names exceeds 15% in Aquitaine, Corsica, Midi-Pyrénées, Réunion, Languedoc-Roussillon, and Poitou-Charentes, and is below 11% in Haute-Normandie, Lorraine, Alsace, Nord-Pas-de-Calais, Franche-Comté, and Mayotte (Table 15).
Table 15. Family name attribution in the NUTS-2 regions, France, 2020
Region Attribution of family name Father’s name only Mother’s name only Father’s name followed by mother’s Mother’s name followed by father’s Double name prevalence index France 81.1 6.9 9.3 2.7 12.9 Alsace 84.8 5.5 7.3 2.5 10.3 Aquitaine 77.5 4.8 14.1 3.5 18.6 Auvergne 83.5 4.8 9.4 2.3 12.3 Basse-Normandie 83.3 5.2 9.1 2.4 12.1 Burgundy 82.7 4.9 9.7 2.6 13.0 Brittany 82.0 4.5 10.8 2.7 14.1 Centre-Val de Loire 80.6 5.7 10.8 2.9 14.5 Champagne-Ardenne 79.9 9.6 8.4 2.2 11.7 Corsica 79.6 3.8 11.8 4.8 17.2 Franche-Comté 86.0 5.6 6.6 1.8 8.9 Haute-Normandie 84.1 5.9 8.1 1.9 10.7 Île de France 82.8 4.7 9.4 3.2 13.2 Languedoc-Roussillon 78.1 7.0 12.2 2.6 16.0 Limousin 80.0 6.8 10.5 2.7 14.2 Lorraine 84.8 5.4 7.6 2.2 10.4 Midi-Pyrénées 78.9 5.2 12.8 3.2 16.8 Nord-Pas-de-Calais 84.4 6.8 7.0 1.8 9.4 Pays de la Loire 82.5 5.2 9.7 2.5 13.0 Picardie 80.7 7.9 9.2 2.1 12.3 Poitou-Charentes 78.8 6.4 11.9 2.9 15.8 Provence-Alpes-Côte d’Azur 82.9 5.5 9.2 2.4 12.3 Rhône-Alpes 85.5 3.5 8.3 2.6 11.4 Guadeloupe 30.0 65.7 3.3 1.0 12.5 French Guiana 39.3 55.5 3.7 1.5 11.6 Réunion 61.4 26.4 8.5 3.7 16.5 Martinique 36.4 58.5 3.0 2.1 12.2 Mayotte 90.6 8.8 0.4 0.1 0.6
Table 15. Family name attribution in the NUTS-2 regions, France, 2020Note: The fertility rates are a mean of the rates from two years (2009–2010 and 2019–2020). The highest one-third of indicators are in bold, and the lowest one-third are in italics.
IV – Induced abortions
1 – More than three-quarters of induced abortions are medical abortions
66 In 2021, 224,200 induced abortions were performed in France,  including 209,000 in metropolitan France (Table 16; Appendix Table A.6). Abortions slightly increased between 2016 and 2019 before decreasing in 2020, a year heavily marked by the COVID-19 crisis and particularly the first lockdown, which were followed by decreases in pregnancies, abortions and births, especially among the youngest age groups (Breton et al., 2021; Vilain et al., 2021). In 2021, the number of abortions remained lower than in the pre-COVID year 2019.
67 Over three-quarters of abortions performed in 2021 (77%) were medical abortions (42% in hospital and 35% in non-hospital settings) (Table 16). The availability of this option played an important role during the COVID-19 crisis. Following the mobilization of actors in the sector, who observed and anticipated difficulties with access to abortion at the local level—as issues of sexual and reproductive health are often overlooked in the context of crises, even though they represent major public health issues as well as key sexual and reproductive rights (Mazuy et al., 2020)—the time limit for medical abortions was extended by 2 weeks. This extension has since been made permanent, and the legal limit for surgical abortions has also been extended by 2 weeks. 
68 Dividing the number of abortions by the number of women of reproductive age, we estimate that 15 abortions were performed per 1,000 women in 2021. However, this rate is not geographically uniform, and the local supply of services plays a determinant role in these variations. This section presents an analysis at the department level, which is better suited (because of how care is organized) to revealing these variations than the level of NUTS-2 regions. The rate was above the average in most departments in the Île-de-France, Provence-Alpes-Côte d’Azur, Languedoc-Roussillon regions, and in the overseas departments; it was below the average in many departments in the Burgundy, Franche-Comté, Alsace, Brittany, Normandy, Pays de la Loire, Limousin, and Auvergne regions (Figure 17).
Box 2. Data sources
The sources of these data, with the dates when reporting began, are as follows:
Since 2005: number of reimbursements (physicians’ fees or prescription drugs) for medical abortions provided outside a hospital (CNAM), based on data from the general health insurance scheme (régime général) and the inter-scheme data marts (DCIR and DCIRS).
Since 2009: health centres and family planning or education centres.
Since 2010: data from the Mutualité sociale agricole and the Sécurité sociale des indépendants health insurance schemes.
Since 2014: the Programme de médicalisation des systèmes d’information (medical systems database, or PMSI) for hospital abortions.
Medical data have become increasingly complete with the progressive integration of the various health insurance schemes and healthcare actors in France.
The instructions for coding abortions in the PMSI have evolved as it has grown. Until February 2019, abortions were counted on the basis of diagnosis-related group (groupe homogène de malades) code 14Z08Z. Since March 2019, principal diagnoses with an extension code O04 (ICD-10 coding) have been used, which may have caused slight irregularities in the data series, especially for the year 2019 (Vilain et al., 2020, 2022).
Table 16. Distribution of abortions by site and method (%)
Year Hospitals and clinics Non-hospital Total Numbers* Medical abortion Surgical abortion Medical abortion 2015 42.3 38.1 19.6 100 220,300 2016 44.9 34.9 20.2 100 216,000 2017 45.4 31.8 22.8 100 217,800 2018 44.4 30.2 25.4 100 225,500 2019 43.6 29.3 27.1 100 233,300 2020 44.3 24.4 31.3 100 222,300 2021 42.5 22.6 34.9 100 224,200
Table 16. Distribution of abortions by site and method (%)* Numbers including teleconsultations.
Note: Annual numbers are rounded and have been updated; they may thus be slightly higher than the estimates presented in previous years.
Coverage: Whole of France.
69 The distribution of abortion between types of healthcare structures is also geographically differentiated. Women living in the Paris, Rhône-Alpes, and Provence-Alpes-Côte d’Azur regions and in certain departments in the southern parts of the Pyrénées and Aquitaine regions are more likely to have a medical abortion in non-hospital settings: more or less half of abortions (Figure 18). In contrast, women living in Corsica and in the Centre, Pays de la Loire, Nord-Pas-de-Calais, Alsace and Lorraine, and Midi-Pyrénées regions and in the north-eastern part of the Aquitaine region are much less likely to have a medical abortion in non-hospital settings (less than 10% in each department).
70 Surgical abortions are the most common (ahead of medical abortion in hospitals/clinics as well as abortions in non-hospital settings) in two departments: Indre-et-Loire (53 %) and Loire-Atlantique (45%).
Figure 17. Abortion rate, per 1,000 women aged 15–49 years, in 2021
Figure 17. Abortion rate, per 1,000 women aged 15–49 years, in 2021Coverage: Whole of France.
Figure 18. Proportion of abortions performed in non-hospital settings in 2021 (%)
Figure 18. Proportion of abortions performed in non-hospital settings in 2021 (%)Coverage: Whole of France.
2 – More than 1 in 10 abortions performed by midwives working in non-hospital settings
71 Authorized to provide medical abortions in non-hospital settings since 2016, midwives performed nearly 25,000 abortions in 2021. This represents more than 10% of all abortions and nearly a third of abortions provided in non-hospital settings.
72 Here again, the availability of this service through midwives is unevenly distributed at the local and department levels. Consquently, no clear tendencies emerge at the regional level. Midwives performed more than 75% of the abortions in non-hospital settings in the departments of Allier, Ardennes, Meuse, Mayenne, Côtes d’Armor, Pas-de-Calais, Landes, Puy-de-Dôme, and Lot-et-Garonne. At the other end of the scale, in Indre, Haute-Saône, Yonne, and Loiret, the corresponding proportion was below 10%.
73 While the scope of services that midwives are authorized to provide has expanded (gynaecological care since 2009, medical abortions since 2016, and, recently, surgical abortions), local specificities—in access to public services, the socio-economic characteristics of the population, environment (rural or urban), sexual and reproductive health norms, medical cultures, etc.—lead to major variations. In September 2022, the High Council for Gender Equality, with the support of many other organizations and actors, called for the enshrinement of the right to abortion in the French constitution in order to ensure its recognition as a fundamental human right,  guaranteeing access at any time throughout France.
3 – Surgical abortions most often performed under general anaesthesia
74 In 2021, nearly a quarter of abortions were surgical. These were almost exclusively performed in public hospitals, which carried out 137,000 abortions, versus fewer than 10,000 in private clinics or hospitals (i.e. less than 7% of all hospital abortions and less than 5% of all abortions). Public hospitals are thus the main providers of surgical abortions.
75 Most (80%) are performed under general anaesthesia. Some are performed under local anaesthesia, with large variations between institutions (Figure 19). Local anaesthesia is more prevalent in geographical areas where the proportion of surgical abortions performed in hospitals and clinics is above average (Bracq, 2022). This is most likely due to the greater presence of specialized units and personnel within these hospitals, and thus the greater availability of personnel able to use this technique. In all other départements, the likelihood of being offered a surgical abortion with local anaesthesia is virtually nil, with hospitals and clinics mainly using general anaesthesia. This means that in these areas, persons who have a surgical abortion do not have a choice about the type of anaesthesia used. This method is dominant almost everywhere in France.
Figure 19. Proportion of general anaesthesia among surgical abortions in France in 2021 (%)
Figure 19. Proportion of general anaesthesia among surgical abortions in France in 2021 (%)Coverage: Whole of France.
76 The diversity of available methods and sites for abortions is increasing, but the national context (resources allocated to public hospitals and to sexual and reproductive healthcare) and the local environment (number of healthcare structures and professionals, medical culture, interprofessional networks, urban vs. rural) drive variations in practices among departments. This highlights the importance of improving the visibility of the different options available to individuals seeking an abortion, or simply information about abortions. Legal provisions do set out a general framework, but access, methods, and thus, importantly, availability and choice remain insufficient in certain respects.
V – Marriages, civil unions, and their dissolution
1 – A clear rebound in marriages in 2021 but not enough to make up for 2020
77 After falling to a historically low level in 2020 (155,000 marriages) due to the restrictions associated with the COVID-19 pandemic, marriages strongly rebounded in 2021 (+42%), to 220,000 marriages  in France as a whole (Papon, 2022a). A large number of additional marriages might have been expected to compensate for those cancelled or delayed in 2020. But the number of marriages did not fully return to its pre-pandemic level (Table 17, Figure 20), with the number in 2021 remaining slightly below that of 2019 (more than 224,000 marriages). No doubt the still restrictive conditions on marriage celebrations in 2021  played a role, especially among the youngest couples, whose number of marriages fell the most in 2020 (Papon, 2022b). This postponement may be spread out over multiple years, although it is also likely that the marriage plans of some couples have been or will be cancelled or redirected towards a civil union (pacte civil de solidarité [PACS]). 
Table 17. Numbers of marriages and PACS unions by partners’ sex, 2013–2021
Marriages PACS union Year Between a man and a woman Between men Between women Total Percentage same-sex Between a man and a woman Between men Between women Total Percentage same-sex 2013 231,225 4,307 3,060 238,592 3.1 162,714 3,354 2,734 168,802 3.6 2014 230,770 5,666 4,856 241,292 4.4 167,487 3,519 2,745 173,751 3.6 2015 228,565 4,085 3,666 236,316 3.3 181,949 3,933 3,085 188,967 3.7 2016 225,612 3,672 3,441 232,725 3.1 184,444 3,863 3,251 191,558 3.7 2017 226,671 3,637 3,607 233,915 3.1 188,233 4,084 3,252 195,569 3.8 2018 228,349 3,268 3,118 234,735 2.7 200,282 8,5 89 208,871 4.1 2019 218,468 3,061 3,211 224,740 2.8 188,014 8,3 56 196,370 4.3 2020 149,983 2,241 2,357 154,581 3.0 165,911 7,9 83 173,894 4.6 2021 214,000 6,0 00 220,000 2.7 n/a n/ a n/a
Table 17. Numbers of marriages and PACS unions by partners’ sex, 2013–2021Note: Provisional data for marriages in 2021 and PACS unions from 2017 to 2020; n/a = data not available.
Coverage: Whole of France (including Mayotte since 2014).
78 In 2020, for the first time, in the exceptional conditions created by the COVID-19 crisis, civil unions outnumbered marriages. As we previously hypothesized (Breton et al., 2021), notably because it remained possible to register a PACS union remotely through a notary throughout the pandemic, the number of PACS unions in 2020 decreased much less than the number of marriages (–31%). In 2020, 173,894 PACS unions were registered, or around 23,000 fewer than in 2019, a relative decrease of 11.4%. In addition to the more flexible conditions for registering PACS unions, this relative persistence may be explained by the lesser association of this form of union with a tradition of celebrating with a large group of family and friends, and by its viability for couples wishing to quickly make their union official.
Figure 20. Number of marriages and PACS unions by partners’ sex, 2000–2021
Figure 20. Number of marriages and PACS unions by partners’ sex, 2000–2021Note: Provisional data for marriages in 2021 and PACS unions from 2017 to 2020.
Coverage: Whole of France (including Mayotte since 2014).
79 As in 2019, the decline in 2020 was significantly smaller for same-sex couples (–27% for marriages and –4.5% for PACS unions) than for different-sex couples (–31% and –12%). The proportion of same-sex unions among all unions continued to rise: 3% of marriages and 4.6% of PACS unions in 2020 (Table 17). For PACS unions, this is the highest level since 2013. However, the provisional figures for 2021 show a smaller rebound in same-sex marriages (6,000), and the relative share of these unions among all marriages (2.7%) is at the lowest level since marriage was made available to all couples in 2013, matching the previous low in 2018. Finally, in 2020, for the second year in a row, more couples of women than couples of men married (51.3% of same-sex marriages).
2 – Regionally varied decreases in the number of marriages
80 The size of the decrease in marriages in 2020 differed between regions (Figure 21). It was above 35% in the regions of northern France, Corsica, and Poitou-Charentes, while it was lower (below 30%) in the southern regions and in Alsace. Restrictions on gatherings during ceremonies at town halls and registry offices were applied in a fairly uniform fashion throughout metropolitan France. This unequal distribution may thus be linked to regional differences in celebration practices, with marriages more likely to be postponed where their festive and convivial character is more heavily emphasized, particularly in northern France. The decrease was also larger in regions where marriages are more heavily concentrated in summer (probably due to the desire for favourable weather) and in regions where couples who marry are more likely to include at least one native of the region (in northern France especially), although no explanation is apparent.
Figure 21. Regional decreases (NUTS-2) in the number of marriages celebrated in France, 2019–2020 (in %)
Figure 21. Regional decreases (NUTS-2) in the number of marriages celebrated in France, 2019–2020 (in %)Coverage: Marriages by region of residence; whole of France.
81 Marriages are not always celebrated in the couple’s area of residence.  Younger, different-sex and never-married couples in particular are more likely to marry elsewhere (Papon, 2020). For France as a whole, 7.4% of marriages in 2018–2019  were registered outside the two partners’ region of residence. The proportion was above the mean for only one region, Île-de-France: 12.5% of couples residing in this region married in another, probably because they chose to celebrate where they have family or residential attachments. The proportion in all other regions is below 7%. It is particularly low in the overseas departments and in Corsica (for obvious reasons of distance) as well as in Nord-Pas-de-Calais (below 3%). Regions close to Île-de-France (Burgundy, Basse-Normandie, Centre, Poitou-Charentes), Brittany and the Massif Central do not include a large city. They are mainly areas of emigration for young people, and 10%–15% of marriages celebrated there are between individuals who live elsewhere.
82 Patterns in the choice of where to marry differ by region of birth. In 2019, 61% of marriages took place in a region where at least one of the partners was born.  This proportion is above 70% in the north-eastern regions and as high as 83% in Nord-Pas-de-Calais. The situation is much less common in French Guiana (30%), Corsica (33%), Île-de-France (48%) and the southern regions. 
Table 18. Marriages by region of registration, of couple’s residence, and of partners’ birth, 2018–2020
Region Proportion of marriages registered in another region (in %) Proportion of couples not resident in the region of marriage* (in %) Share (in %) of marriages where… Change in resident marriages between 2019 and 2020 At least one of the partners was born in the region of marriage** Both partners were born in the region of marriage** Alsace 4.3 5.0 68.8 45.4 –26.9 Aquitaine 5.0 9.8 56.5 26.5 –30.3 Auvergne 4.6 11.2 63.8 33.7 –32.4 Basse-Normandie 3.2 13.6 69.4 43.3 –35.5 Burgundy 5.4 15.7 64.2 32.5 –34.4 Brittany 4.1 11.2 66.0 35.0 –30.1 Centre 6.1 11.5 59.9 28.6 –27.9 Champagne-Ardenne 5.7 10.3 73.4 45.0 –38.1 Corsica 3.0 13.8 33.6 19.7 –37.0 Franche-Comté 5.3 9.1 71.3 42.9 –34.1 Haute-Normandie 4.3 7.8 70.3 45.3 –37.4 Île-de-France 12.5 3.1 48.5 30.6 –29.6 Languedoc-Roussillon 4.4 9.8 50.0 20.8 –30.1 Limousin 6.8 13.1 61.4 27.7 –31.0 Lorraine 3.9 6.0 76.4 54.4 –34.7 Midi-Pyrénées 6.8 8.1 54.2 25.0 –24.9 Nord-Pas-de-Calais 2.8 5.4 82.9 66.1 –36.0 Pays de la Loire 5.3 8.5 68.1 38.0 –32.3 Picardie 5.5 10.2 65.5 36.8 –36.3 Poitou-Charentes 4.7 13.1 61.1 30.3 –36.2 Provence-Alpes-Côte d’Azur 4.6 6.3 54.8 28.8 –27.8 Rhône-Alpes 6.0 5.1 63.7 39.1 –28.8 Guadeloupe 3.1 5.4 71.5 50.2 –27.1 French Guiana 5.2 2.4 30.4 18.5 –44.2 Réunion 2.2 7.8 81.0 60.2 –33.3 Martinique 2.5 8.3 79.9 56.2 –27.3 Mayotte 4.2 4.1 –46.2 Whole of France 7.4 7.4 61.3 36.3 –31.2
Table 18. Marriages by region of registration, of couple’s residence, and of partners’ birth, 2018–2020* marriages in 2018 and 2019; ** marriages in 2019.
Interpretation: 3.1% of marriages of couples living in Guadeloupe in 2018–2019 took place in another region.
Interpretation: 5.4% of marriages in Guadeloupe in 2018-2019 involved a couple who live in another region.
Interpretation: In 71.5% of marriages in Guadeloupe in 2019, at least one of the partners was born in the region.
3 – In 2020, total marriage rates at a historic low and marriages happening later
83 In 2020, marriages (and first marriages) became less frequent at all ages, for both men and women, regardless of the partner’s sex (Figure 22). But it is in the age range where marriages are most frequent (between ages 25 and 40) that they fell the most. Consequently, the average age at marriage increased significantly (+0.3 years) and identically for both women and men (Appendix Table A.12). It reached 35.7 years for women and 38.2 years for men in 2020,  resulting from the slight delay in the average timing of first different-sex marriages (+0.2 years, or 33.2 years for women and 35.2 years for men) and the much larger shift for remarriages (from +0.5 to +1 year depending on sex and divorced/widowed status). After continuously decreasing since 2013, the average age of the partners in same-sex marriages went up in 2020 for the first time, to 43.0 years for men (+0.9 years) and 37.2 years for women (+0.1 year).
Figure 22. Total marriage and first marriage rates by age group (per 1,000 individuals) by partners’ sex, 2019 and 2020
Figure 22. Total marriage and first marriage rates by age group (per 1,000 individuals) by partners’ sex, 2019 and 2020Coverage: Whole of France.
84 In 2020, consistent with the historic collapse in marriage rates, the total marriage rate fell to 377 marriages per 1,000 men and 372 per 1,000 women. And the total first marriage rate  fell to its lowest ever level, at 301 marriages per 1,000 men and 299 per 1,000 women (from 440 and 437 respectively in 2019, a decrease of 31.6%). The decrease was slightly lower for remarriages of different-sex couples (down less than 30%) and for same-sex marriages (–27%).
4 – More marriages in eastern France, more PACS in the west
85 Here we examine regional differences in marriage and PACS based on total rates, calculated from annual union registration data. These indicators reflect regional dynamics in unions, without interference from dissolutions, which lead to the exclusion of some unions from instantaneous observations (when divorced people have repartnered). The last 2 years for which detailed regional data are available are 2018–2019  for marriages and 2015–2016 for PACS unions.
86 Regional disparities (Figure 23A) in different-sex unions are relatively similar for men and women, with more frequent marriages in the eastern regions of France, from Alsace to the Mediterranean basin. Marriage is less frequent in more rural regions such as Brittany, Limousin, Auvergne, and Midi-Pyrénées. The regional geography of the PACS (Figure 23B) is different, confirming analyses carried out at the department level (Breton et al., 2017; Robert-Bobée, 2018): it is in western France, from Basse-Normandie to Midi-Pyrénées, that PACS rates are highest. Finally, although PACS unions are more common in urbanized departments, it is in Île-de-France that they are least common among different-sex couples. One explanation may relate to the characteristics of the population of the region, who are more likely to be immigrants and less likely to own their home (Breton et al., 2017; Robert-Bobée, 2018).
87 The contribution of first marriages to overall marriage rates can be estimated by dividing the total first marriage rate by the total marriage rate. The regional analysis shows that this ratio is higher in western France (in particular due to the lower frequency of divorces and remarriages) and lower in southern regions, in Guadeloupe, in French Guiana, and in eastern France.
Figure 23. Total marriage and PACS rates by sex (per 1,000 people), different-sex unions
Figure 23. Total marriage and PACS rates by sex (per 1,000 people), different-sex unions
5 – Unions between men are concentrated in Île-de-France, while those between women are more evenly distributed
88 For same-sex unions, the total marriage and PACS rates are strongly and positively correlated. In other words, for both men and women, marriages are more frequent where civil unions are as well. These rates are thus due more to the regional presence of same-sex couples than to any specific preference for registering a union in one or the other of these ways. Finally, regional variations in same-sex unions, which are quite different from those in different-sex unions, are considerably greater for men than for women, almost certainly reflecting a greater concentration of couples of men in certain regions (Figure 24).
89 Geographically, the rate of same-sex unions is by far the lowest in the overseas regions, singularly so in Guadeloupe and Martinique, as well as in Corsica. Greater difficulty living as a same-sex couple, reinforced by an insular context (Gordien, 2018), may encourage emigration to metropolitan France as well as a lower propensity to form a union among gay and lesbian people in these regions.
90 Unions between two men, both PACS and marriages, are most common in Île-de-France, in the Mediterranean regions, and in the southwest. Nearly 30% of marriages between two men in 2019 were between partners living in Île-de-France, versus only 21% of those between partners of different sexes, and only 17% of those between two women.  Île-de-France is the only region in metropolitan France where two men are more likely to marry than are two women—indeed, twice as likely (24 same-sex marriages per 1,000 men vs. 12 per 1,000 women). In all other regions, women are more likely than men to marry a person of the same sex, while this is only the case in two regions for PACS unions (Picardie and Poitou-Charentes).
Figure 24. Total marriage and PACS rates by sex (per 1,000 people), same-sex unions
Figure 24. Total marriage and PACS rates by sex (per 1,000 people), same-sex unions
91 The frequency of marriage between two women is more homogeneous across regions. It is somewhat higher in the southern regions close to the Atlantic Ocean and the Mediterranean, as well as in the northeast around the Paris region. PACS unions are also much more common in the Mediterranean basin and in all regions along the Atlantic seaboard.
92 Although these geographical variations tell us nothing about how likely a same-sex couple is to register their union, they nevertheless reveal what regions these couples prefer to live in and thus the underlying geographical mobility. This is confirmed by same-sex couples tending to live further from their place of birth at the time of marriage (Papon, 2020).
6 – A decrease in divorces in 2020
93 In 2020, 57,437 divorces  were recorded in France, nearly 9,000 fewer than in 2019 (–13%). These figures are currently incomplete, as only statistics on divorces that are settled in court  have been compiled since 2017. Divorces registered by notaries, without the involvement of a judge, have not yet been integrated into the published data. The observed decrease is clearly related to the particular context of 2020, which limited access to legal services and lengthened divorce proceedings due to the various lockdowns. Dissolutions on all grounds (Table 19) are affected by the decrease, including accepted divorces (divorces acceptés) (–15%) and divorces due to irretrievable marriage breakdown (pour altération définitive du lien conjugal) (–7%), the most common grounds and which have been increasing since the 2017 reform of divorce legislation in France. Fault-based divorces have also continued the decline that began in 2010 (–17%), while divorces by mutual consent are now almost never referred to the courts, with couples in this situation now working with a notary. It is also possible that the context of the pandemic encouraged more couples to turn to notaries, who are more accessible, rather than to judges to organize the dissolution of their union. We will have to await the publication of notarial data to confirm this point. The decrease in divorces in 2020 was largest in the overseas regions and in the southwest (20% or more).
Table 19. Annual divorces by grounds, 2010–2020
Year Absolute change between 2019 and 2020 Relative change between 2019 and 2021 (%) Grounds for divorce 2010 2016 2017 2018 2019 2020 By mutual consent 72,433 71,933 33,457 283 91 65 –26 –28.6 Accepted divorce 32,603 29,854 30,404 36,373 40,048 33,937 –6,111 –15.3 Irretrievable marriage breakdown 14,107 17,010 17,790 17,635 18,411 17,105 –1,306 –7.1 Fault divorce 13,117 8,036 7,665 6,989 6,662 5,516 –1,146 –17.2 Indeterminate direct divorce 468 731 935 747 590 599 9 1.5 Conversion of separation into divorce 993 479 362 290 279 215 –64 –22.9 Total* 133,721 128,043 90,613 62,317 66,081 57,437 –8,644 –13.1
Table 19. Annual divorces by grounds, 2010–2020* Divorces and conversions of separation.
Since 2017, only divorces settled through a court have been counted and appear in this table.
Coverage: Whole of France.
7 – More divorces in the south and east
94 Here we analyse regional variations in the frequency of divorces on the basis of divorces that occurred before the 2017 reform (latest comprehensive data). The analysis is thus not influenced by differential behaviours linked to the new options for divorce proceedings. We use the total divorce rate  for the period 2015–2016. This is the proportion of marriages that, under the conditions of the period, would end in divorce, as established each year for France up to 2016.  It has the advantage of being clearer and easier to interpret than the divorce rate. 
95 Looking at marriages counted according to the couple’s place of residence, large regional variations can be seen for the years 2015 and 2016. The frequency of divorce increases with movement from the north-west to the south-east. It is highest in the southern regions (Figure 25) along with Alsace (more than 0.53 divorces per marriage) and lowest in Île-de-France (0.4 divorces per marriage) and in north-western France. These observations match fairly closely with those that emerge from the analysis of the divorce rate (per 1,000 married people under the age of 70), except for a few regions such as Martinique, French Guiana, and Île-de-France, where the estimated divorce rate is lower than the corresponding total divorce rate, whereas it is higher for Corsica (0.6 divorces per marriage). These differences are likely due to migration after marriage: the movement of married couples towards southern regions, particularly from Île-de-France, leads the total divorce rate to overestimate the frequency of divorce in the regions of arrival and to underestimate it in the regions of departure. 
Figure 25. Total divorce rate in 2015–2016 in French regions (per marriage)
Figure 25. Total divorce rate in 2015–2016 in French regions (per marriage)
8 – A PACS union, then a marriage: more common in the north-west and in Alsace
96 In 2017, the latest year for which figures are available, 82,345 PACS unions were dissolved.  Nearly half (48%) of these dissolutions are linked to the marriage of at least one, and most often both partners; the other half (48%) are linked to their separation by agreement; the remaining 4% are for other reasons (death or at the request of one of the two partners). In 2016, the frequency of PACS dissolutions  was 49% for all reasons combined (separation, death, marriage). Excluding dissolutions by marriage, under the conditions of 2016, approximately 1 in 4 PACS would ultimately be dissolved.
97 The chances of ending a PACS to get married differ quite widely between regions (Table 20): dissolutions through marriage are high, representing the majority of PACS dissolutions, in north-western France (57% in Pays de la Loire, 52% in Brittany and Basse-Normandie) and Alsace (53%). They are less common in the overseas departments (around 30%), in Corsica (30%), and in the south (43% in Provence-Alpes-Côte d’Azur and 42% in Languedoc-Roussillon). Given regional differences in marriage rates, marriages were more often preceded by a PACS in Alsace, Brittany, and Pays de la Loire (more than 22% of marriages). This is less common in the overseas departments (less than 5% of marriages), Corsica (6%), the north (15% in Île-de-France, 14% in Picardie, 16% in Nord-Pas-de-Calais), and on the Mediterranean coast (15% in Provence-Alpes-Côte d’Azur and 14% in Languedoc-Roussillon).
98 The total PACS dissolution rate  in 2015–2016 (Table 20), excluding PACS unions ended by marriage (Figure 26), is used to capture the frequency of permanent union dissolution, most often by agreement, and not couples’ propensity to end their civil union by getting married. With this indicator, regional variations in the dissolution of civil unions are more overt: PACS unions are more often dissolved in the south-eastern regions and the overseas departments (above 28%) and in the south-west and north-east (from 26% to 28%) but less often in the northwest (below 22%) and in central France and Basse-Normandie (24%). These different behaviours may be associated with the interregional mobility of working-age populations (Fabre and Dejonghe, 2015) and the tendency of different regions to attract them or drive them away. This can encourage couples in common-law unions to register their union in order to facilitate movement between regions (particularly for civil servants), which brings an increased risk of dissolution.
Table 20. PACS union dissolutions in 2015–2016 by region
Region Total PACS dissolution rate in 2016 (per PACS) PACS dissolved in 2015–2016 Proportion of marriages in 2015–2016 that followed a PACS dissolution** By mutual consent between the partners (%) By marriage (%) Other grounds* (%) Alsace 0.56 43.6 52.8 3.6 22.5 Aquitaine 0.52 50.0 46.3 3.7 19.4 Auvergne 0.47 50.5 45.7 3.8 18.4 Basse-Normandie 0.49 44.4 52.5 3.1 19.4 Burgundy 0.48 49.2 46.6 4.1 18.8 Brittany 0.45 44.9 51.7 3.4 22.2 Centre 0.47 46.5 50.0 3.5 17.9 Champagne-Ardenne 0.49 47.4 49.6 2.9 20.1 Corsica 0.42 66.0 30.0 4.0 5.9 Franche-Comte 0.51 48.0 48.6 3.5 19.7 Haute-Normandie 0.51 45.6 50.6 3.7 18.3 Ile-de-France 0.48 47.2 48.5 4.3 14.7 Languedoc-Roussillon 0.49 54.5 42.4 3.1 14.3 Limousin 0.50 48.7 46.8 4.5 19.6 Lorraine 0.53 49.0 47.3 3.7 19.4 Midi-Pyrénées 0.47 51.1 45.7 3.2 19.2 Nord-Pas-de-Calais 0.48 50.7 46.0 3.3 16.2 Pays de la Loire 0.48 40.3 56.9 2.8 22.3 Picardie 0.44 52.6 43.4 4.0 13.8 Poitou-Charentes 0.48 49.3 47.5 3.2 20.0 Provence-Alpes- Cote d’Azur 0.53 52.7 43.0 4.2 14.8 Rhone-Alpes 0.49 46.8 50.1 3.1 16.9 Guadeloupe 0.48 62.1 28.1 9.8 2.8 French Guiana 0.52 70.9 22.6 6.5 3.5 Réunion 0.47 64.8 31.4 3.8 4.6 Martinique 0.31 80.9 10.0 9.1 0.5 Mayotte 0.45 52.2 30.4 17.4 2.0 Whole of France 0.49 48.3 48.0 3.6 17.0
Table 20. PACS union dissolutions in 2015–2016 by region* The other grounds for dissolution are: unilateral request by one of the partners, death or on the initiative of the guardian.
** Ratio of PACS dissolved by marriage in 2015–2016 to the number of resident marriages in 2015–2016.
Figure 26. Total rate of PACS union dissolutions except by marriage in 2015–2016
Figure 26. Total rate of PACS union dissolutions except by marriage in 2015–2016
99 In summary, these indicators highlight distinct union dynamics and models in different parts of the country. They show a contrast, for example, between Brittany, where PACS unions are very widespread and often followed by marriage, and where unions are more stable (fewer divorces and PACS breakdowns), and Provence-Alpes-Côte d’Azur, where marriage is more common, including for same-sex couples, and unions are more fragile.
VI – Mortality
1 – Life expectancy increasing again, but still below pre-COVID levels
100 According to INSEE’s still provisional figures, there were an estimated 639,000 deaths in 2021 in metropolitan France and 18,000 in the overseas departments, making 657,000 in the whole of France (INSEE, 2022a). The corresponding crude death rate is 9.7 deaths per 1,000 inhabitants (9.8 in metropolitan France). Estimated life expectancy at birth in 2021 was 79.3 years for men and 85.4 years for women (79.4 and 85.5 in metropolitan France). After losing 7 months of life expectancy for men and 5 months for women between 2019 and 2020 due to the COVID-19 pandemic, the country regained 2.4 and 3.6 months, respectively, in 2021. In 2019, life expectancy for the whole of France was 79.7 years for men and 85.6 years for women (79.8 and 85.6 in metropolitan France). The country thus did not fully return to its pre-COVID level in 2021, particularly for men (–0.4 years compared to 2019 and –0.1 years for women).
101 Based on the provisional figures provided by CépiDc (INSERM), the body responsible for collecting, coding, and publishing deaths by medical cause, INSEE estimates that the number of deaths directly attributable to COVID-19 stands at ‘between 130,000 and 145,000’ for the 2 years combined. Given the age structure of the population and change in mortality rates during the recent period, estimated excess mortality for 2020 is 7.5% and 6.3% for 2021 compared to the expected level. The improvement observed in 2021 seems to have mainly benefited women, whose excess mortality decreased from 6.4% in 2020 to 4.3% in 2021, while it remained high for men (8.6% and 8.3%) (Blanpain, 2022).
2 – The pandemic did not change France’s position in Europe
102 In 2021, France remained close to the European average for life expectancy at birth and infant mortality, with no notable change compared to 2019.
103 Women’s life expectancy in France continues to be close to the top of the ranking (85.4 years), just behind Switzerland (85.9) and Spain (86.2), and ahead of Italy (85.1). Italy moved down one place in the European ranking due to excess mortality during the pandemic, losing 0.6 years of life expectancy at birth among women between 2019 and 2021. Life expectancy increased during the period in Sweden (+0.2 years), Belgium (+0.1) and Switzerland (+0.1). France’s position for male mortality is much less favourable and deteriorated slightly during the pandemic, dropping a place in the European ranking (from 10th to 11th) behind Denmark.  Denmark is one of the few European countries in which male life expectancy at birth increased between 2019 and 2021 (+0.1 years), along with Luxembourg (+0.5) and Norway (+0.4). Eastern European countries suffered by far the most from the pandemic, with losses in life expectancy at birth reaching up to 3.5 years in Bulgaria. 
104 France ranks particularly well in mortality at advanced ages, but its position in infant mortality has gradually deteriorated since the early 2000s, due to the stagnation of the probability of dying before age 1 and even a slight increase in this probability in the periods 2005–2009 and 2011–2019. This contrasts with the continuous progress recorded in many other European countries. However, French infant mortality seems to have resumed its decline in 2019 and 2020. According to provisional figures from INSEE, the rate fell from 3.6 per 1,000 births in 2019 to 3.4 in 2020 and 2021. This is higher than the rate in Denmark, Greece, Germany, Ireland, and Austria, where it is between 3.0 and 3.2 per 1,000 births, and especially in Spain, Finland, Iceland, Italy, Norway, Portugal, and Sweden, where less than 3 out of 1,000 children die during their first year (less than 2 in Finland and Norway). It is comparable to the rate in Belgium and Switzerland, and lower than that of many Eastern European countries as well as Luxembourg and the Netherlands.
3 – Stalling decline in infant mortality due to deaths during the first months of life
105 The stagnation of infant mortality in France since the mid-2000s contrasts with the very rapid decline in the probability of dying before age 1 observed during the last 2 decades of the 20th century. The infant mortality rate fell from 10 deaths per 1,000 births in 1980 to 4.4 in 2000, a decrease of more than 50% in metropolitan France. The rate continued to decrease between 2000 and 2005, reaching 3.6 per 1,000. It has since fluctuated slightly around that level.
106 Infant mortality is broken down into three periods: the early neonatal period, constituted by the first week of life; the late neonatal period, which covers the following 3 weeks; and the post-neonatal period, which represents the remainder of the first year of life (after the first 4 weeks of life and until the child’s first birthday). These different components of infant mortality have changed at very different rates in recent decades. Since the 1950s,  first-month mortality has always progressed more slowly than post-neonatal mortality, except in the 1980s and 1990s. In that period, mortality during the first 4 weeks after birth decreased very rapidly, with the probability of dying falling from 5.8 per 1,000 in 1980 to 2.9 in 1995, while post-neonatal mortality stagnated at around 3.6 to 3.7 per 1,000 between the mid-1980s and the early 1990s. Post-neonatal mortality then dropped significantly in the 1990s, from 3.8 per 1,000 in 1991 to 1.6 per 1,000 in 2000. Change in mortality between 1 and 12 months of age reflects change in the prevalence of sudden infant death syndrome (SIDS), the leading cause of death at the time in this age group. Rates of SIDS had increased sharply during the 1980s following recommendations given to parents to lay their babies on their stomachs to sleep. Changes in recommendations, along with high-profile public awareness campaigns on putting babies to sleep on their backs or sides starting in 1994, led to a rapid drop in post-neonatal mortality in the second half of the 1990s (Sénécal et al., 1998; Meslé, 2000).
107 Unlike the previous period, the stagnation of infant mortality since the mid-2000s is entirely attributable to the unfavourable trend in first-month mortality. Post-neonatal mortality continues to decline steadily, although more slowly than before, with the probability of dying between the 1st and 12th months falling from 1.6 to 0.9 per 1,000 between 2000 and 2020, according to still provisional figures from INSEE. In contrast, neonatal mortality, which reached 2.3 per 1,000 in 2005, increased to 2.7 per 1,000 in 2017 before stabilizing at 2.5 per 1,000 in 2020. This unfavourable trend has affected mortality both in the first week and in the following 3 weeks (Figure 27). The reasons for this development are difficult to determine in so far as, unlike in other countries, the data needed for an analysis of the factors responsible are not available in France. A number of countries record information on gestational age and birth weight on the death certificate, but not France, even though this information is decisive in understanding the role of the conditions of pregnancy in the stagnation of early infant mortality. Similarly, the mother’s characteristics, regarding both social situation (level of education or income, for example) and behaviour (smoking or alcohol consumption) or health status (obesity, diabetes, hypertension), are not routinely available, making it very difficult to analyse the influence of these risk factors on mortality in young children (Blondel et al., 2019; Claris, 2022; Trinh et al., 2022). There are considerable geographical disparities in levels and trends in infant mortality, as in general mortality.
Figure 27. Absolute and relative change in infant mortality and its components in France, 1980–2020
Figure 27. Absolute and relative change in infant mortality and its components in France, 1980–2020Coverage: Metropolitan France.
4 – The persistence of geographic inequalities in mortality
108 INSEE publishes mortality indicators (infant mortality and life expectancy at birth by sex) for the old (NUTS-2) and new regions of metropolitan France since 1990 and for the overseas departments since 1999 (2014 for Mayotte).  The current coverage of these series extends to 2020, the latest year available. In addition, CépiDc (INSERM) publishes standardized mortality rates by group of causes of death and department and region, but only for the years 1990 to 2017.
109 In 2020, a year of high excess mortality due to COVID-19, life expectancy at birth for the whole of France was 79.1 years for men and 85.1 years for women, but these averages conceal important regional differences (Figure 28). Across the 27 NUTS-2 regions, the gap between the extreme values of life expectancy at birth is considerable: 8.3 years for men (from 72.5 years in Mayotte to 80.8 years in Corsica) and 12.6 years for women (from 73.9 years in Mayotte to 86.5 years in Corsica). However, mortality is much higher in Mayotte than in any other French region. Regional life expectancy is second lowest in Guadeloupe, where it is nonetheless almost 10 years higher than in Mayotte for women (83.2 years) and 4.3 years higher for men (76.8 years). In metropolitan France, the gap between the regions at the bottom and the top of the ranking has improved for men since 1990, narrowing from 4.8 to 4.1 years, but increased for women, going from 2.8 to 3.0 years. These changes have come along with changes in the regional map of life expectancies at birth.
110 Figure 28 presents life expectancy at birth for the two sexes in France in 2020. For metropolitan France, the maps show a fragmentation of the crescent of high mortality which, bypassing Île-de-France, traditionally extended along the western, northern, and eastern borders of the country, from Brittany to Alsace. According to the latest available data, the lowest life expectancies are still mainly found in the regions to the north and east of Île-de-France (from Haute-Normandie to Lorraine), along with the overseas departments. Mortality is relatively low in Île-de-France and in the regions located south of Brittany and to the west of a line that can be traced from Tours to Nîmes (Pays de la Loire, Poitou-Charentes, Aquitaine, Auvergne, and Midi-Pyrénées); and in Corsica, the region with the fastest progress in life expectancy at birth for the past 30 years, for both women and men. Finally, the mortality situation in the south-eastern quarter of the country (the Rhône-Alpes, Provence-Alpes-Côte d’Azur, and Languedoc-Roussillon regions) is more favourable than in the centre of the country (Centre, Burgundy, Franche-Comté, and Auvergne regions), where levels of life expectancy at birth are close to those of Brittany and Basse-Normandie.
Figure 28. Life expectancy at birth by sex and region (NUTS-2), 2020
Figure 28. Life expectancy at birth by sex and region (NUTS-2), 2020
111 Regional inequalities in mortality are particularly large for infant mortality. On one side, there are the overseas departments, where more than 4 out of 1,000 children died during their first year (7.4 in Mayotte, between 4 and 5 in other overseas territories) in 2018–2020; on the other are the regions of metropolitan France, where the rate is no more than 3 per 1,000 (in Île-de-France), and less than 2 per 1,000 in Burgundy, Poitou-Charentes, and Provence-Alpes Côte d’Azur (Figure 29).
112 But the biggest single factor in regional mortality differences is mortality linked to cardiovascular diseases because of its particularly large contribution to general mortality. In 2017 (the last year for which data are available), the standardized mortality rate  for this group of causes was 2.5 times higher in Mayotte (482 deaths per 100,000 inhabitants) than in Île-de-France (192 per 100,000) for men and 2.8 times higher for women (333 deaths per 100,000 inhabitants in Mayotte vs. 120 in Île-de-France). Aside from Mayotte, the regions with particularly high cardiovascular mortality rates are, in decreasing order, Nord-Pas-de-Calais, Brittany, Champagne-Ardenne, Poitou-Charentes and French Guiana for men (with standardized rates above 300 per 100,000 in all these regions) and Languedoc-Roussillon, Rhône-Alpes, French Guiana, Réunion, Nord-Pas-de-Calais, and Picardie for women. Regional variations in cancer mortality are also considerable, with a twofold difference in rates between the highest and lowest rates for men (215 per 100,000 in Mayotte vs. 425 in Lorraine) and a 1.5-fold difference for women (from 142 per 100,000 in Guadeloupe to 213 in Nord-Pas-de-Calais). Unlike cardiovascular mortality, cancer mortality is relatively low in the overseas departments, a phenomenon that may be due to differences in behaviour between these regions and metropolitan France. A recent study (INSEE, 2021), confirming some of the results of another analysis (Barbieri, 2013), highlights the role of individual behaviours (smoking, alcohol consumption, and driving) in explaining regional differences in mortality. However, given the particularly high proportion of deaths from ill-defined causes in certain departments (such as Mayotte, Guadeloupe, and Guyana), caution is needed in interpreting the apparent differences.
Figure 29. Infant mortality rate per 1,000 births (average for 2018–2020) by region (NUTS-2)
Figure 29. Infant mortality rate per 1,000 births (average for 2018–2020) by region (NUTS-2)
113 But 2020 was not an ordinary year. The COVID-19 epidemic muddied the waters to some extent: the regions with the highest general mortality in 2019 are not the same ones that suffered the highest excess mortality due to COVID-19 in 2020 and 2021, at least in metropolitan France.  Figure 30 presents the differences in life expectancy at birth between 2019 and 2021 for men and women at the regional level. Men lost more years of life than women (see below), and there were also larger regional disparities in men’s loss of life expectancy (although the scale on the two maps is not equivalent). Excess mortality in 2020 and 2021 was particularly high in Île-de-France, Lorraine, and the south-east (Languedoc-Roussillon and Provence-Alpes-Côte d’Azur)— regions that, except for Lorraine, previously had above-average life expectancies. In these four regions, the loss of life expectancy at birth was more than a year for men (up to 1.3 years in Provence-Alpes-Côte d’Azur), and between 0.7 years in Languedoc-Roussillon and 0.9 years in Lorraine for women. Excess mortality linked to COVID-19 was also relatively high in other northern regions, in particular in Basse-Normandie, Nord-Pas-de-Calais, and Picardie, for both sexes, and Alsace for men. The only regions where no loss of life expectancy was recorded between 2019 and 2021—and only for women—are Burgundy and Corsica. Decreases were also relatively low for women in the south-western regions (Poitou-Charentes, Aquitaine, Limousin, and Midi-Pyrénées) and, for men, in Brittany, Pays de la Loire, Midi-Pyrénées, and Franche-Comté. Analysis of detailed mortality data by cause of death in 2020 and 2021 will be able to identify the proportion of these regional disparities that are due to direct effects of COVID-19, and the proportion due to the indirect effects of the epidemic (through other causes of death).
Figure 30. Differences in years of life expectancy at birth for each sex between 2019 and 2021 by region (NUTS-2), metropolitan France
Figure 30. Differences in years of life expectancy at birth for each sex between 2019 and 2021 by region (NUTS-2), metropolitan France
AcknowledgementsThe authors would like to thank Eric Wiest, an engineer at the Strasbourg university data platform, and Denys Dukhovnov, a student at the University of California, Berkeley, for designing the maps; Clémence Bracq (ENS); Elodie Baril and Arnaud Bringé at the INED Statistical Methods Department for their help in constructing the databases and for the initial analyses; and Ekrame Boubtane, associate professor at CERDI (École d’Économie, Université Clermont Auvergne) for her help with the section on immigration.
114 The appendix tables are available on INED’s website and on the Archined open archive:
Figure A.1. Map of French regions (NUTS-2), codes and names of regions
Code 2010 NUTS level 2 FR10 Île-de-France FR21 Champagne-Ardenne FR22 Picardie FR23 Haute-Normandie FR24 Centre FR25 Basse-Normandie FR26 Burgundy FR30 Nord-Pas-de-Calais FR41 Lorraine FR42 Alsace FR43 Franche-Comté FR51 Pays de la Loire FR52 Brittany FR53 Poitou-Charentes FR61 Aquitaine FR62 Midi-Pyrénées FR63 Limousin FR71 Rhône-Alpes FR72 Auvergne FR81 Languedoc-Roussillon FR82 Provence-Alpes-Côte d’Azur FR83 Corsica FR91 Guadeloupe FR92 Martinique FR93 French Guiana FR94 Réunion FR96 Mayotte
Figure A.1. Map of French regions (NUTS-2), codes and names of regions
Figure A.2. Mean annual population increase in NUTS-2 regions of EU-27 between 1 January 2016 and 1 January 2021
Figure A.2. Mean annual population increase in NUTS-2 regions of EU-27 between 1 January 2016 and 1 January 2021Note: The scale used is that of France applied to all NUTS-2 regions of Europe.
Figure A.3. Components of population increase in NUTS-2 regions of EU-27 between 1 January 2016 and 1 January 2021
Figure A.4. Proportion of population aged 60 and above in the NUTS-2 regions of EU-27 on 1 January 2021
The database of appendices to this article can be accessed at: http://hdl.handle.net/20.500.12204/AYTs6koXkOqZPUwFsEWi
This figure is the difference between the population counts on 1 January 2022 and 1 January 2021 before adjustment for migration. Taking account of INSEE’s adjustment for migration in 2021, the total increase is around 221,000 (Table 1).
These rates are different from those without adjustment for migration traditionally published by INSEE and shown in Table 1 (Papon, 2022a).
While net migration is difficult to estimate—hence the regular INSEE adjustments (Breton et al., 2020)—the problem is not specific to France and is shared by many countries (Beauchemin et al., 2021).
Data available on the Eurostat website: https://ec.europa.eu/eurostat/en/data/database
Alongside Ireland (+25,300), Sweden (+22,300), the Netherlands (+8,500), Denmark (+6,300), Belgium (+6,000), Cyprus (+3,000), Luxembourg (+2,200), and Malta (+200).
This increase is a follow-on from the increase in over-60s from 2006 and in over-65s from 2011.
Rising from 38.9 to 43.6 years for women and from 36.2 to 40.2 years for men.
Sweden has moved from being Europe’s third oldest country in 2001 (France was 16th) to its fifth youngest in 2021 (France is ninth). The median age has remained stable over the period, partly due to a steady number of births since 2010 and even an increase since 2001 (+25% vs. –8% in EU-27).
Data available on the Eurostat website: https://ec.europa.eu/eurostat/en/data/database.
The dependency ratio measures the share of the non-working-age population (children and older adults) with respect to the working-age population (adults), i.e. the ratio of the number (or proportion) of people aged 0–19 and 65+ to the number (or proportion) of people aged 20–64.
The number of administrative regions of metropolitan France was reduced from 21 to 12 in January 2015. The territorial collectivity of Corsica and the five overseas departments and regions maintained their existing status under various specific arrangements. While the former regions no longer exist on an administrative or political level in France, they continue to exist for Eurostat as NUTS-2 territories (Seys, 2017).
EU-27 is divided into 272 NUTS-2 regions, of which 27 are in France (Eurostat, 2020). The names and geographical location of these regions are shown in Appendix Figure A.1.
Replacement level is between 2.05 and 2.10 children per woman. This is the mean number of children needed to ensure that 100 women give birth to 100 girls who will survive until mean childbearing age.
Spearman rank coefficient of 0.96.
The linear correlation coefficient between total growth over the period 1/1/2016 to 1/1/2021 and net migration rate is 0.62 (r²) for the regions of metropolitan France, while the coefficient between total growth and natural growth rate is 0.1.
‘The municipality density grid ranks [French] municipalities by their population and the distribution of inhabitants over their land area. Density increases as population size and concentration increases…. The definition used by the European Union is based on a breakdown of land area into squares of 1 km2 and the aggregation of squares of equivalent density into ‘clusters’. Each municipality is then placed in a category according to the shares of its population living in the different types of clusters’ (https://www.insee.fr/fr/information/6439600).
It is also highest at ages 60–65, possibly due to a cohort effect that disappears at older ages and to the effect of rural–urban migration linked to ageing and old-age dependency.
Immigrants are defined as persons born outside France to non-French parents, whether or not they subsequently acquire French nationality.
Austria, Belgium, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Liechtenstein, Lithuania, Luxembourg, Malta, the Netherlands, Norway, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden, United Kingdom.
The nationalities considered may vary from one demographic report to the next in response to legislative changes in rights of residence. Appendix 3 takes account of changes in scope.
The United Kingdom has been counted as a third country since January 2021.
Excepting New Caledonia, Polynesia, and Wallis and Futuna.
The Ministry of the Interior also publishes a complementary series of migration flow statistics based on a count of all first residence permits issued to adults. Its scope is different as it includes residence permits valid for less than 1 year which will not necessarily be renewed by a longer-term permit. The inflows calculated by the Ministry thus include temporary migration.
Under a prime-ministerial order of 18 March 2020, third-party nationals were prohibited from entering France. In coordination with other EU member countries, depending on country of origin and reason for entry, this ban was progressively lifted from 1 July 2020.
Variable MIGR_IMM1CTZ; only available for the 2013–2019 period on the Eurostat website at the following address: http://ec.europa.eu/eurostat/en/data/database (page accessed on 12 September 2022).
In most cases, this permit is called a document de circulation pour étranger mineur. It was instituted by a decree published on 24 December 1991.
This difference may be linked to the fact that girls more frequently apply for a residence permit before age 18 than boys, although the reason for this remains unknown.
These inflows do not include nationals from the EEA or from Switzerland.
All registered students with a residence permit did actually come to France as these permits are validated by the police after arrival.
Admissions for humanitarian reasons only include people whose asylum application has been processed and approved, so this figure does not include all asylum seekers.
The ‘passeport talent’ residence permit is granted to people who are highly qualified, who want to set up a business or invest in France, or who work in the arts. It aims to promote qualified migration (d’Albis and Boubtane, 2021a).
This figure was obtained by summing the monthly birth estimates published on 28 July 2022. It is slightly higher (+0.5%) than the figure of 738,000 births published in early January. It will be recalibrated in early 2023.
The second and third lockdowns were less stringent (reopening of schools and resumption of activity in numerous economic sectors).
The dip in births in the early 1990s was linked to a long-term increase in age at childbearing and a probable postponement of births due to the economic crisis and the high levels of youth unemployment at that time.
This number is obtained by combining the components of fertility timing in the year with the age structure in that same year. For more details, see the 2021 demographic report (Breton et al., 2021, p. 599). The detailed breakdown is not given here, but the figure fell from 405,395 to 404,772.
Calculated as the sum of age-specific fertility rates over the period considered.
The trend can be modelled by birth order with a linear regression (mean age by year). The slopes of the regression lines are 0.109 for first births, 0.073 for second births, and 0.057 for third births, with a parallel trend by birth order.
This scenario may seem somewhat pessimistic as fertility rates above age 35 fell between 2019 and 2020 due to the COVID-19 pandemic. But if the variations between 2018 and 2019 are applied, rather than those observed between 2019 and 2020, the result is identical. The correlation between observed completed fertility and cumulative fertility at age 35 is very strong (r² = 0.99 for the 1960–1970 cohorts).
Estimate published by Eurostat and which contradicts the decrease initially announced in 2021.
Excluding French Guiana and Mayotte, where fertility is still above 3.5 children per woman.
The stem family model is an extension of the household society system defined by ethnologists. It refers to a system of social organization built around a unique heir. The term ‘nuclear family’ refers to a system in which parent–child relations are liberal, and relations between siblings are egalitarian or relatively egalitarian.
Variance is measured by the coefficient of variation (standard deviation/mean). The countries with at least eight NUTS-2 regions are Austria, Belgium, Czech Republic, France, Germany, Greece, Italy, Hungary, the Netherlands, Poland, Romania, Spain, and Sweden.
Grouping of one or more entire municipalities.
See ‘Box 1. Using census data to estimate cumulative fertility at age x’ in the 2020 article on demographic trends (Breton et al., 2020, p. 476). This method provides good estimates for mothers up to age 35. One of the earliest studies of this kind in France dates back to the mid-1990s (Desplanques, 1993).
Based on the assumption, used here, that the degree of underestimation is similar in the different categories.
On the estimation of the number of abortions, see Box 2.
In 2020, in the context of the COVID-19 crisis, a series of legal measures expanded access to abortion: a 2-week extension of the deadline for medical abortions in non-hospital settings (to the duration that was already in effect in hospitals), a measure with major implications (de Zordo et al., 2020); and the authorization for health professionals to provide them via telemedecine, as a special temporary measure. See decision no. 2020.0092/DC/SA3P/SBPP of the Haute autorité de santé (National Authority for Health) of 9 April 2020, on the adoption of rapid responses to COVID-19, medical abortion during the 8th and 9th week of amenorrhoea outside hospitals; the decree of 14 April 2020 supplementing the order of 23 March 2020; and the decree of 7 November 2020 amending the decree of 10 July 2020. A 2-week extension of the legal time limit for surgical abortions had been rejected, but it was adopted in March 2022 (Law no. 2022-295 of 2 March 2022), along with several other measures, including midwives being authorized (after specific training) to perform surgical abortions (an authorization that had initially been granted on an experimental basis for 3 years by Article 70 of the 2021 Social Security Financing Act), and making permanent the temporary measures enacted during the COVID-19 crisis (Marguet, 2022).
On 24 November 2022, this bill was approved in the National Assembly (lower house of parliament in France). It must be approved by the Senate and then in a nationwide referendum (Article 89 of the French Constitution).
The national lockdown in April was preceded by local lockdowns which began in mid-March 2020. It was followed by a curfew until 21 June, the reopening of establishments open to the public in June but with occupancy restrictions, the implementation of a health pass for entry to restaurants in August, and the restriction of access to free PCR tests in October.
This form of official civil union has been available to all couples in France since 1999.
The distinction between where the partners live and where their marriage was registered is available through data from marriage certificates, which record the place of marriage, each partner’s place of birth, and the couple’s place of residence.
The data for the year 2020 were not analysed, given the extraordinary conditions of the period.
In the case of marriages between a man and a woman, the marriage is slightly more likely to be registered in the woman’s region of birth.
The same patterns can be seen in the proportion of marriages in which both partners were born in the region of marriage (Table 18).
These are mean ages at marriage, all orders combined, calculated from marriage rates at all ages, including above the age of 50 years. They thus differ from the most commonly presented indicator, which is calculated on the basis of marriages before age 50.
The total marriage (or PACS) rate represents the number of marriages (PACS) that would be observed in a cohort of 1,000 people who, at each age, experience the age-specific marriage (PACS) rates observed in a given year. This indicator can be calculated for a specific union order and type of union (marriage, PACS). By distinguishing marriages based on the partner’s sex, two ‘components’, different-sex and same-sex marriages, can be calculated for both men and women. This indicator presents the advantage of eliminating the effects of different age structures, in this case between regions, for comparative purposes.
The data on marriage rates in 2020 were not included due to the exceptional circumstances.
On 1 January 2021, 18.2% of the French population lived in Île-de-France.
Direct divorces and separations converted into divorces.
Law no. 2016-1547 of 18 November 2016. Since 1 January 2017, divorces can be concluded without the involvement of a judge where the spouses agree on the divorce and its effects, and if no child of the couple asks to be heard by a judge.
This indicator represents the ratio of the average number of divorces in each region in 2012 and 2013 to the weighted average of marriages of residents in each region in the preceding years. To simplify, the weighting here consists in the arithmetic mean of marriages in the years corresponding to the marriage durations where divorces are the most frequent. In the present case, the denominator is thus the average of marriages for the years 2001 to 2010, and the durations considered are from 4 to 13 years.
This indicator is obtained for France by taking the sum of duration-specific divorce rates up to 2016. It cannot be calculated for 2017 or subsequent years due to incomplete data.
The divorce rate is the ratio of the mean number of divorces during a period in each region to the number of married people under the age of 70 in the middle of the period (expressed as a number per 1,000). To validate this choice, we verified that these two indicators are strongly correlated for the period 2012–2013. The choice of 2012–2013 is linked to the available data, both for divorces and for the size of the married populations in the NUTS2 regions (the official administrative regions of France up to the year 2014). The divorce rate is thus the ratio of the average number of divorces in 2012 and 2013 in each region to the number of married people aged below 70 on 1 January 2013.
A comparison with census data would be useful, but it is impossible because conjugal status (reported on census surveys since 2015) only covers individuals’ current situation, and not any previous unions. The size of the divorced population would thus be largely underestimated, since conjugal status (being in a couple, whether married, PACS or consensual union) now takes precedence over marital status.
The dissemination of these data has been slowed by the transfer of the registration of PACS declarations and dissolutions from the courts to municipal registry offices in late 2017, combined with the establishment in 2011 of the legal option of completing these formalities through a notary. Source: Ministry of Justice/SG/SEM/SDSE/Statistical analysis of the Répertoire général civil and the notarial database.
Representing the cumulative sum of duration-specific PACS dissolution rates: in the conditions of the year 2016, 49% of PACS would be dissolved before reaching a duration of 17 years (as the first PACS were registered in 1999).
Obtained by taking the ratio of PACS dissolutions in 2015–2016 to a weighted mean of PACS registered annually between 1999 and 2016. The weighting coefficients are the duration-specific PACS dissolution rates calculated in 2016 for the whole of France.
After excluding the United Kingdom and Iceland, for which the necessary indicators are not available for 2021.
More than 2 years for men in Hungary, Estonia, Latvia, Poland, Czech Republic, Romania, and Slovakia, and for women in Lithuania, Poland, Romania, and Slovakia.
The figures in the following paragraphs only concern infant mortality in metropolitan France, as INSEE has only published the corresponding indicators for the overseas departments since 1994.
INSEE, Vital statistics and population estimates (Table P3D: General population indicators by department and region).
Using the standardized rate allows age structures that vary in time and space to be taken into account. Unlike the crude mortality rate, which is higher in an older than a younger population under equal mortality conditions, the standardized mortality rate eliminates the effects of age structure, reflecting only differences in the risk of death.
The data used to compare levels of mortality by sex and age in 2021 to those of 2020 at the regional level are drawn from the French Human Mortality Database (https://frdata.org/fr), consulted on 22 August 2022. Unfortunately, they do not include data for the overseas departments. This is the only source of regional mortality statistics for 2021 that is already available.