Population  2009/3


2009/3 (Vol. 64)


doi 10.3917/popu.903.0445

publisher I.N.E.D

E-mail Alerts

Please provide your e-mail address to receive a Table Of Content alert every time a new issue is published.

you're reading

Recent Demographic Developments in France: Tenth Anniversary of the PACS Civil Partnership, and Over a Million Contracting Parties

byFrance Prioux[*]By the same author
andMagali Mazuy[*]By the same author
Page 393-442

I - Overall population trends and age structure

The population of France at the 2006 census


Based on the final results of the first five-year cycle of annual census surveys, the French statistics office (Institut national de la statistique et des études économiques, INSEE) has made a further upward revision of French population growth between the two most recent censuses (Pla, 2009). Between 1 January 1999 and 1 January 2006, the total population (metropolitan France + overseas départements) increased by 3.1 million, from 60.1 to 63.2 million inhabitants. The natural increase (excess of births over deaths) in this period was 1.78 million, while overall net migration is estimated at 0.62 million. An adjustment of 0.66 million, equivalent to approximately 95,000 extra people for each year between 1999 and 2005, was thus made to restore continuity between the two censuses. Provisional adjustments had already been made following the first annual census surveys (Desplanques and Royer, 2005; Richet-Mastain, 2006 and 2007), but this time they are final adjustments for the period 1999-2005. Each year, the results from a new census based on five consecutive annual surveys (2007 census for the period 2005-2009, 2008 census for 2006-2010, and so on) are used to revise INSEE’s population estimates for the years after 2006, which at present are thus provisional. Future adjustments should be smaller, however, because successive intercensal intervals have shortened considerably, and because no further changes are expected in the methodology. [1]  The large adjustment between the censuses of 1999 and...[1]


The adjustment made this year also concerns the population age distribution. The provisional population estimates from 2000 to 2006 were based on the totals by age enumerated in 1999, but the new age distributions also incorporate the totals by age from the 2006 census. [2]  See the article by G. Desplanques (2008), which discusses...[2] This leads to a slight adjustment of the structure by broad age group (Table 2) [3]  Tables 1 to 16, updated annually, are given in the...[3] and of the various demographic indicators (fertility, marriage, and mortality rates) based on these totals by age.

Stable growth in 2007 and 2008


The population of France on 1 January 2009 is estimated at 64.3 million, [4]  The population of the overseas départements (DOM) no...[4] of whom 62.45 million in metropolitan France (Pla, 2009). In 2008, the population of metropolitan France rose by an estimated 337,000, a figure practically identical to the previous year’s increase of 335,000 (Table 1). This is because the 10,000 extra births and the estimated net gains of 5,000 from international migration were almost entirely cancelled out by an additional 13,000 deaths.


At 4.2 per 1,000, the rate of natural increase in France remains among the highest in the European Union (EU). Only in Ireland (10.4 per 1,000) and Malta (5.2 per 1,000) is the rate higher. Together with those two countries, France is one of the few countries where the natural increase equals or exceeds 3 per 1,000, the others being Luxembourg (4.1 per 1,000), United Kingdom (3.5 per 1,000), and the Netherlands (3.0 per 1,000) (Marcu, 2009). Despite a fairly generalized increase in births in 2008 – Germany was the only EU country to register a decline in births – they were still outnumbered by deaths (natural decrease) in eight countries: Bulgaria (–4.3 per 1,000), Hungary (–3.1 per 1,000), Latvia (–3.1 per 1,000), Lithuania (–2.6 per 1,000), Germany (–2.1 per 1,000), Romania (–1.5 per 1,000), Estonia (–0.5 per 1,000), and Italy (–0.1 per 1,000). Of these eight countries, only in Italy did the population continue to grow, thanks to substantial net gains from immigration. In the other seven countries, net migration was either negative (Bulgaria, Latvia, and Lithuania) or too small to offset the natural decrease (Germany, Estonia, Hungary, and Romania). Of the 21 EU countries with net immigration, the net migration rate in France (+1.2 per 1,000) was low compared to the estimated rates for most of the original EU-15 member states (except Germany, where net migration was practically zero in 2008) and to those of three new members which had large net inflows in 2008 (Czech Republic, Slovenia, and Malta: 6-10 per 1,000).

A progressively ageing population structure


The final results of the 2006 census have led to a revision of the population age distributions since 2000 [5]  INSEE now calculates two population age distributions...[5] and a slight modification in the structure by broad age groups (Pla, 2009) (Table 2). A small downward revision in the percentage aged 20-59 has benefited mainly the 60 and over age group. The relative share of this age group has risen sharply since 2006, as the first baby-boom cohorts turn sixty. The percentages in the older groups (ages 65 and over, and 75 and over) are rising much more slowly; the 1946 birth cohort will not reach age 65 before 2011.


The upward trend in births since 1994 has produced a slight broadening at the base of the population pyramid, particularly in the years 2006-2008 (Figure 1). Despite this, however, the percentage of under-20s is still falling by 0.1 point every year, reaching 24.5% on 1 January 2009 in metropolitan France and 24.8% for France as a whole (Pla, 2009; Table 2). Including the population of the DOMs – slightly younger than that of metropolitan France – produces a “younger” distribution by broad age groups, with the percentage of persons aged 65 and over reaching 16.5% (versus 16.7% for metropolitan France).

Population pyramid of France on 1 January 2009

Population: Metropolitan France (provisional estimate).

Source: INSEE.

Beyond age seventy, the population pyramid becomes less symmetrical due to the imbalance of the sexes. At age 85, women outnumber men by more than two to one, and at age 95 by three to one.


By comparison with other EU countries, [6]  Eurostat website accessed on 04/09/2009.[6] France’s population has a relatively young age structure, at least as concerns the share of under-20s. On 1 January 2008, only Ireland reported a higher percentage (27% versus 24.9% for France as a whole), while the average for all 27 EU countries was considerably lower (21.7%). For the population aged 65 and over, on the other hand, their proportion in France is close to the median, and 14 countries have lower values. Despite this, the total share of the over-65s in the EU countries (17%) is still higher than in France (16.3%), mainly due to the rather high share of this age group in Germany and Italy (20.1% and 20% respectively) whose combined populations make up more than a quarter (28.4%) of the EU total.

II - Foreign immigration [7]  The authors thank Xavier Thierry (INED) for supplying...[7]

A sharp decrease in 2007


The analysis of immigration flows to France is based on statistics of residence permits valid for one year or more issued each year and is therefore limited to foreigners from the countries still subject to a residence permit requirement for settlement in France. [8]  Although under-age children can be admitted without...[8] Since 2004, citizens of the European Economic Area (EEA), i.e. the EU Member States plus Iceland, Liechtenstein, Norway, have not been subject to this requirement. An exception concerns foreigners from certain new EU member states [9]  All citizens of new Member States are subject to this...[9] if they are intending to work in France, though this only applies during the early years of EU membership. [10]  Since 1 July 2008, foreign nationals from countries...[10] The enlargement of the European Union in 2004 (10 new Member States) and in 2007 (accession of Bulgaria and Romania) has thus progressively reduced the scope of the immigration statistics, making it harder to evaluate the total number of foreign entrants, which is thus partly based on an estimation (Table 3). Assuming stable flows from countries that belonged to the EEA before 2004, the number of “arrivals” [11]  Some foreigners do not obtain their first one-year...[11] fell substantially in 2007 (to 192,500, around 11,500 fewer than in 2006), continuing the gradual decline that began in 2004, after the number of permits issued peaked at 215,400 in 2003.

For detailed analysis of recent migration flows to France, only third country (non-EEA) nationals can now be counted. However, to obtain an accurate measure of trends, the same population base must be used for each year. For foreigners coming from third countries as defined since 2007, i.e. crossing the frontiers of the present EU-27, 144,658 residence permits were issued in 2007, 13,062 fewer than in 2006 (–8.3%), following a decline of 3,624 (–2.2%) in 2006 (Table A).

Residence permits issued to third-country nationals by reasons for admission

Family immigration is declining but remains the largest category


The breakdown of residence permits by reasons for admission reveals a clear drop in the number of foreigners admitted on grounds of family ties in France (down by 10,172, or 11.3%). This fall is due equally to fewer admissions of spouses of French citizens and fewer foreigners obtaining a permit on “personal and family life” grounds. The sharp rise in “personal and family life” permits in 2006 reflected the exceptional legalization in that year of undocumented foreigners with children enrolled in school in France, who were issued with this type of permit. Permits issued to spouses of French nationals remained stable in 2006 but declined substantially (–13%) in 2007, due probably to the lengthier procedures for contracting or recognizing binational marriages (between a French and a foreign spouse), a change introduced under new legislation in 2003 for marriages celebrated in France, and in 2006 for those celebrated in French consulates abroad. The number of binational marriages celebrated in France has in fact been falling since 2003 (32,889 in 2008 versus 47,579 in 2003), while the number celebrated abroad and transcribed in the French registers, after growing relatively quickly in recent years, fell for the first time in 2007 (47,869 versus 50,350 in 2006). [12]  The figures for foreign marriages transcribed to French...[12] Despite this, marriage to a French national remains, along with studying in France, the main reason for legal immigration, each of these reasons accounting for one-quarter of all residence permits issued in 2007 (Table A).


Overall, family immigration still represents the largest category of admissions (55% of permits in 2007), despite a slight decline in its share since 2006 (57%) mainly in favour of student and worker admissions, whose numbers did not fall in 2007. By contrast, there has been a considerable fall in legalizations and in permits issued to retired people. Permits issued to refugees and stateless persons and to ill foreigners have also fallen, though not as abruptly. After returning to equilibrium in 2001-2003, the ratio of women to men has risen again in recent years, standing at 112 women for 100 men in 2007 (107 in 2006). Family immigration remains a largely female phenomenon (136 women for 100 men) although less markedly so for immigrant spouses of French nationals. Labour immigration involves mostly men (44 women for 100 men), while among university students the sexes are in balance. For all reasons for admission combined, the average age of immigrants is 30.6 years for men and 29.7 years for women.

Among the different nationalities admitted for residence in France, Algerians still ranked first in 2007 (24,041), despite a substantial fall in their number (–16%) relative to 2006 (Figure 2). Moroccan entrants remained in second place (numbering 19,017 in 2007, only slightly fewer than in 2006), and Chinese nationals in third place with 10,040 entrants (+2%) ahead of Tunisians (8,832, down 3%) and Turks, whose number fell appreciably (7,170, down 14%). None of the other nationalities topped 4,000 in 2007, and permits issued to Cameroon and US nationals, which numbered 4,228 and 4,011 in 2006, both fell sharply in 2007 (to 3,695 and 3,444, respectively). [13]  More detailed data on nationalities are available on...[13] Despite this context of generalized decline, numbers for some nationalities continue to grow. Those registering the largest increase were Romanians (3,336 admissions in 2007, up 30% on 2006), Poles (2,937, up 34%), Brazilians (2,706, up 13%), and Indians (2,004, up 11%).

Permits issued since 1994 to the nationalities most represented in 2007

Scope: Metropolitan France.

Source: AGDREF, collated by INED (X. Thierry).

III - Fertility

An increase in births and in fertility


More than 828,400 births were registered in France (796,000 of them in metropolitan France) in 2008, about 10,000 more than in 2007 (Beaumel et al., 2009a). Births in the DOMs totalled 32,400 in 2008, slightly down on the figure for 2007. The 10,000 extra births thus came solely from metropolitan France, where more than three-quarters of départements recorded higher birth rates.


This increase in births is due to the increase in fertility. The average number of children per woman reached an estimated 2.02 in France as whole and 2.0 in metropolitan France (Table 4). If these figures are confirmed, [14]  These figures are currently provisional estimates that...[14] this will be the first time since 1974 that the total fertility rate (TFR) in metropolitan France has reached the symbolic threshold of two children per woman. After falling back slightly in 2007 (from 1.98 children per woman in 2006 to 1.96 in 2007), the upward trend in fertility has resumed, thanks mainly to higher fertility among women aged 30-39.

Children born outside and within marriage: equal in number and in status


The proportion of births to unmarried couples continues to climb. The 50% threshold for children born outside marriage was crossed in 2006, and the proportion reached 52.5% in 2008. The proportion of births outside marriage remains below 50% in only seventeen départements, many of which are highly urbanized: the Île-de-France region (minus Seine-et-Marne), Alsace and Moselle, part of Franche-Comté (Doubs and Territoire de Belfort), and five départements of the Rhône-Alpes region (Ain, Loire, Rhône, Isère, and Haute-Savoie). By contrast, percentages over 60% are observed for some départements with a more rural character, a notable instance being the Creuse where two in three mothers are unmarried. The proportion of births outside marriage is particularly high in the DOMs (74.4%), notably in French Guiana (87.9%).


In the past, children born outside marriage were referred to as “illegitimate” or as “natural children”, to distinguish them from “legitimate” children born within marriage. But the notion of legitimacy ceased to have a legal basis in France under the ordinance of 4 July 2005 ratified by the Act of 16 January 2009. Henceforth, if the mother’s identity is mentioned on the birth certificate, the maternal filiation is established automatically when the birth is registered, independently of the marital status of the parents. Consequently mothers no longer need to undertake a formal recognition procedure. Prior to this change, however, mothers who did not follow this procedure could still have their maternity established on the basis of “possession of status”. Recognition by the father is still necessary when the parents are not married. Some children are not recognized by their father, though by counting the children born to married parents (for whom paternity is established automatically at birth) and the recognitions registered before and after birth, it is estimated that only 2-3% of children born today do not have paternal filiation (estimate based on research by Munoz-Pérez and Prioux, 2005).


Some children are born with no filiation, with neither parent mentioned on their birth certificate, since under French law a mother has the right to remain anonymous when giving birth (Civil Code article 341-1) or when registering the birth (Civil Code article 57) (Munoz-Pérez, 2000). Such children number around 600 each year and represent less than 0.1% of all births. In a minority of cases, the mother changes her mind shortly afterwards and decides to keep her child. In the other cases, the children become orphans in state care and are placed with a family for adoption (Halifax, 2009). Because they are adopted in the first months of life, a filiation is quickly established for these children (Munoz-Perez, 2000). Since the new law of 22 January 2002 reforming access to birth origins, mothers who give birth anonymously can keep their identity secret but leave written information to which the children have a right of access (Act of 22 January 2002 modifying the law on access to origins).

Women aged 25-35 account for two-thirds of births


Mean age at childbearing was close to 30 in 2008 (Table 4). The distribution of fertility by age has changed considerably over the last thirty years (Figure 3). At present, it is women aged 25-29 and 30-34 years who account for two-thirds of the TFR, whereas until the late 1970s it was women aged 20-29 years who did so. Most of the remaining one-third of total fertility occurs after 35 years, with women under age 25 contributing only a small share of births. Childbearing under age 20 is rare, accounting for less than 2% of overall fertility in 2008 versus nearly 5% in 1960. Childbearing after age 40 is also uncommon, amounting to less than 4% of total fertility, close to the level recorded fifty years ago. Combining this with the contribution from women aged 35-39 (16%) gives a total of roughly 20%, or one in five, of births in France today that can be described as “late”. Childbearing after age 40 has increased little relative to its level in the 1960s and in the early twentieth century (Toulemon, 2005; Prioux, 2005). It is the “nature” rather than the level of late parenthood that is changing. Late births are now less often additions to large families, but are more frequently only children or children born into reconstituted families (Bessin et al., 2005).

Contribution of each age group* to the total fertility rate since 1960 (%)

age reached in the year.

Population: Metropolitan France.

Source: INSEE.

The increase in the total fertility rate since 1995 is due mainly to the increase in fertility among women aged over 27, together with a levelling off since the late 1990s in rates at younger ages, which had been declining since the mid-1960s. The increase in fertility from age 28 since the late 1970s has almost compensated for the fall in fertility rates at younger ages that slowed sharply in the early 1990s (Table 4). The recent uptrend in period fertility is not, however, accompanied by an increase in cohort fertility levels. Completed fertility is falling slightly (Table 5). From a little over 2.1 children among women born between 1950 and 1960, completed fertility falls back to 2 children per woman in the 1969 birth cohort, after which it stops falling and is predicted to stabilize or even recover slightly, depending on the projection scenario adopted. [15]  Fertility rates are projected under two scenarios....[15] The TFR currently stands very close to cohort fertility levels, and the mean age at childbearing in 2008 (29.9 years) is similar to that of women born between 1972 and 1974 (depending on the projection).

Persistent but evolving geographical disparities


Relatively large geographical variations in fertility exist between départements in metropolitan France (Table 16). Fertility levels in 2006-2007 were lowest in Corsica (1.54 children per woman in Haute-Corse and 1.56 in Corse-du-Sud) and highest in Val-d’Oise (2.32) and Seine-Saint-Denis (2.31). The zones of high and low fertility have shifted over time, with the disappearance of France’s “high-fertility crescent” [16]  The crescent-shaped belt of high fertility extending...[16] (Daguet, 2005; Prioux, 2006). A large zone of high fertility now extends without interruption from Vendée to Aisne, encompassing almost all of the Île-de-France region, except for Paris – where fertility is particularly low – and two départements of the outer Paris suburbs (Figure 4). Conversely, the centre and the south-west (excepting Tarn-et-Garonne), are characterized by low fertility, with indicators commonly under 1.8 children per woman and in some cases down to 1.7 or even lower (in Vienne, Haute-Garonne, Gironde, and Cantal). Fertility in the DOMs is above 2 children per woman, except in Martinique where it has fallen below metropolitan France (1.9 against 2.0).

Total fertility rate in the départements in 2006-2007 (mean number of children per woman)
Source: Calculated from INSEE data (Table 16 and map in Appendix).

Fertility in northern France thus remains slightly higher than in the south. The urban-rural variable, with the migration of young working adults, probably explains part of the diversity. Young adults abandon the regions that attract them least in terms of employment, while adults who already have jobs and children tend to remain in the more rural areas. In Île-de-France, one of the lowest fertility rates is recorded for Paris intra muros, whereas fertility is high in the Seine-et-Marne and in the outer suburban ring. The constraints of size and cost associated with housing in central Paris are important in explaining why young couples who want to have a family move out to the suburbs.

Fertility among the highest in the European Union


With similar fertility levels (Table 6), France and Ireland are still the two most fertile countries in the European Union, followed by those of northern Europe. Fertility remains very low in the southern European countries, as it does in central and eastern Europe where in many countries the total fertility rate stands at 1.4 or 1.5 children per woman. Fertility is currently lowest in Poland (1.23 in 2008) and Poland is also one of the few countries where the period indicator is continuing to decline. In direct contrast, the largest fertility increases in recent years have occurred in the Czech Republic, Lithuania, and Slovenia (at least +0.2 children per woman since 2005).


A slightly different picture emerges from a comparison of European fertility levels in the 1970 birth cohort (Table 7 [17]  These estimates are liable to be revised upward since...[17]). Ireland and France, followed by northern Europe, still lead the field, but most of the former Eastern bloc countries – and in particular Estonia, Hungary, Poland, the Czech Republic, and Slovakia – show relatively high completed family size compared with their respective period indicators. This is because the trend towards to later maternity is rapid in these countries, thus exerting a corresponding downward pressure on the annual fertility rates. [18]  Assuming completed fertility does not change, a steady...[18] The countries where completed fertility is lowest are Italy, Spain, and Germany (between 1.42 and 1.47 children per woman) and the downward trend is continuing.

Despite the general trend to later maternity, mean ages at childbearing in the countries of Eastern Europe are still relatively low (between 25 and 27 depending on the country) compared with those in the EU-15 countries, where mean ages are between 28 and 30 (Table 7).

IV - Abortions

A stable number of abortions


The number of induced abortions recorded in hospital statistics for metropolitan France fell back slightly (–2.1%) in 2005 but not in 2006 (Vilain, 2008), when they totalled 209,700, an increase of 3,400 or 1.6% on 2005 (Table 8). For 1,000 women aged 15-49, the number of induced abortions thus climbed back slightly (14.5 per 1,000 in 2006 against 14.2 in 2005), as did the total abortion rate (0.52 per woman in 2006 and 0.51 in 2005). This indicator, which is constructed in the same way as the total fertility rate, does not mean that 52% of women have induced abortions, since some women use the procedure several times in their life. As 25-30% of induced abortions are performed on women who have already aborted, the estimated percentage of women undergoing abortion at least once in their life is 38% (Rossier et al., 2009).


In addition to hospital statistics, the notification forms completed for each abortion can be used to analyse abortion in greater detail. Recording of data from the notification forms was suspended for several years but resumed in 2005 [19]  The complete data sets for 2006 and 2007 are currently...[19] so that detailed information on women presenting for abortion and on terminations is again available. There is less information than before, however, as the notification form has been greatly simplified. [20]  See Rossier et al. (2000) for a description of the...[20]


The Act of 4 July 2001 extended the legal limit for abortion from 10 to 12 weeks of gestation, i.e. from 12 to 14 weeks of amenorrhea. This produced only a transient lengthening of the average gestational age at termination (Rossier et al., 2009) because the share of surgical terminations, the only possible procedure at the highest gestational ages, has declined sharply in favour of medical terminations, generally performed around the sixth week of amenorrhea. Abortion at slightly higher than average gestational ages by some women (the youngest, those without a cohabiting partner or who are unemployed) probably reflects their greater difficulty in obtaining an abortion.

Although the proportion of women who present for abortion at least once in their life has been stable since 1990, the share of repeat abortions has gradually increased. The “learning effect” of a first termination seems to have vanished, since the abortion rate is now the same whatever the number of previous abortions (Rossier et al., 2009).

A slight increase in abortions at young ages


Abortion rates at the youngest ages have continued to increase (Vilain, 2008). This may be the sign that unplanned pregnancies are becoming more frequent due to less careful contraceptive practice and higher levels of sexual activity among young people, reflecting the fall in the median age at first intercourse (Bozon, 2008). To observe trends in pregnancy, the numbers of induced abortions and births must be combined to estimate age-specific conception rates. [21]  Spontaneous abortions (miscarriages) are not recorded...[21] We do this by moving back births by age (reached in the year) by two-thirds of a year, so they can be counted at the start of pregnancy and added to the number of induced abortions at each age (reached in the year) (Rossier et al., 2009).


Table B presents abortion rates by age group (with greater detail for the under-20s), conception rates by age at conception estimated in this way, and the ratio between these two values, which gives an estimate of the proportion of terminated pregnancies by age. This ratio suggests that the rise in abortion rates among women under 18 arises mainly from the growing propensity to terminate a pregnancy at these ages. In 2005, four in five pregnancies were terminated at ages 14-15 and two in three at ages 16-17, compared with two in three and slightly over one in two, respectively, in 1990. The increase in the conception rates at these ages is small and is limited to the 2000s.

Estimation of induced abortion rates, conception rates by age group (per 1,000 women), and the ratio between induced abortions and conceptions, 1990-2005

Conception rates are fairly stable between ages 18 and 30 and even fell at ages 20-24 in the early 1990s, but they rise steadily between ages 30 and 45, the ages at which the propensity to terminate a pregnancy decreases. The overall frequency of conceptions has thus increased, which explains the slight rise in the total abortion rate, whereas the propensity to terminate a pregnancy has on the whole fallen (from 22% to 21%). These changes can be related to those in fertility: the increase in conceptions is reflected in a higher total fertility rate, while the change in age-specific conception rates and termination rates contributes to fertility postponement.

V - Marriage, PACS, and Divorce

A further decline in marriage


Following two years of stability, the number of marriages in France resumed its downward course in 2008, with a total of 265,400, almost 8,300 fewer than in 2007 (Beaumel et al., 2009b). The decline was specific to metropolitan France since in the DOMs slightly more marriages were celebrated in 2008 than in 2007 (6,665 versus 6,475).


In metropolitan France, the annual number of marriages fell below 260,000 for the first time since 1995 (Table 9). The 3.2% decline in marriages in 2008 concerned binational marriages especially, which fell by 8%, and notably marriages between a French woman and a foreign man, down by 12.3%, and marriages between two foreign nationals, down by 5.7%. Marriages between French persons registered a more moderate fall of 2.3%. Thus the share of binational marriages in all marriages has continued to fall, standing at only 12.7% in 2008 versus 16.8% in 2003. [22]  The fall in binational marriages since 2003 is probably...[22]


The proportion of first marriages in total marriages is gradually shrinking for men and women alike, whereas that of remarriages of divorcees is growing. Four out of five persons marrying in 2008 were never-married (79.2% of men and 80.4% of women), a little under one in five were divorcees (19.4% and 18.2% respectively), and a mere 1.4% were widow(er)s. By way of comparison, in 2000, marriages of never-married persons accounted for 81.5% of the total among men and 82.4% among women. The majority of marriages are thus still first marriages, and their relative decline has slowed since the 1980s, due to the continual increase in the “stock” of never-married persons resulting from the fall in nuptiality and the shift to later marriage observed since the 1970s.


The decline in marriages in 2008 is reflected in a further decrease in total first marriage rates for both men and women (Table 9). The total rate stood at 49 marriages per 100 men (and 51 per 100 women) in 2008, while the overall probability was 56% for men and 58% for women. This is the first time that the annual probability of marrying for never-married women (overall probability) has fallen below 60%.

In fact, the total first-marriage rate by cohort points to a continued decline in cohort nuptiality (Table 10). Although nuptiality has not yet fallen as low as the period total rates in any of the cohorts for which we have estimated the total marriage frequency before age 50, less than two-thirds of men born in the early 1970s and two-thirds of women born around 1975 will ever marry. Meanwhile, mean age at first marriage continues to rise in successive cohorts. It is estimated at 29.1 years for women born in 1975, and 30.7 years for men born in 1973.

The PACS civil partnership: over a million contracting parties in ten years


Since the PACS (pacte civil de solidarité – civil partnership between same- or different-sex partners) was instituted ten years ago (Act of 15 November 1999), French couples can sign a contract defining the organization of their joint life and giving access to certain advantages formerly only available to married couples. However, these advantages are not as extensive as those of marriage, notably with respect to inheritance and filiation.


The popularity of the PACS has increased year on year. More than 6,000 PACS contracts were concluded between 15 November and 31 December 1999, 30,000 in 2003, more than 100,000 in 2007 and nearly 150,000 in 2008. This type of civil union has met with substantial success over the ten years of its existence, totalling close to 600,000 PACS and at least one million contracting parties (because the same person may have formed and dissolved several PACS unions in this period). The PACS dissolution rate is reasonably stable, and for heterosexual unions is gradually catching up with the divorce rate (Carrasco, 2007).


Since homosexual marriage is not authorized under French law, the PACS enables same-sex couples to obtain legal recognition for their union and to benefit from a more favourable tax regime. But the PACS is also chosen by an increasing number of different-sex partners. Of some 146,000 PACS registered in 2008, 94.4% involved heterosexual couples. [23]  A total of 4,780 PACS were between two men (3.3%) and...[23] This proportion varies between départements, ranging from 91% in Alpes-Maritimes to 98% in Mayenne. In Paris, however, the proportion of PACS registered by heterosexual couples is distinctly lower, standing at 82.7% versus 17.3% registered by same-sex couples, of which 13.5% are between two men and 3.8% between two women.

Heterosexual couples account for the great majority of PACS unions simply because of the very large stock of unmarried couples available to form a PACS. But although the share of same-sex couples fell slightly from 6.1% in 2007 to 5.6% in 2008, the frequency of PACS unions increased among these couples. Their number rose by 32% between 2007 and 2008, from 6,217 to 8,203, with a slightly larger increase for PACS between two women (+ 36%) than for those between two men (+ 29%). However, without estimates for the number of couples who are neither married nor in a PACS union, and particularly the number of same-sex couples, it is impossible to measure the appeal of the PACS for the different categories of couples potentially concerned by this type of union.

Just over one PACS for two marriages in 2008


The number of heterosexual PACS unions is increasing fast and the number of marriages is tending to fall. In 2008, just over one PACS was registered for every two marriages, [24]  265,404 marriages and 146,030 PACS, of which 137,820...[24] which means that slightly more than one in three registered unions was a PACS. Making the comparison between PACS and marriage does not mean that these two modes of legalizing a union are mutually exclusive, since a proportion of PACS unions lead on to marriage. [25]  9,610 PACS were dissolved due to marriage in 2008 and...[25] The comparison is nonetheless instructive since some PACS partners chose this contract precisely because it represents an alternative model to marriage (Rault, 2009).


The ratio between PACS and marriage in 2007-2008 shows considerable variability between the French départements, from a minimum of 10 heterosexual PACS per 100 marriages in Martinique, to a maximum of 66 in Haute-Vienne (Figure 4). The lowest values, between 10 and 21 PACS per 100 marriages, are observed in the four DOMs, followed by Seine-Saint-Denis, with 26 per 100. In all the other départements the ratio stands at or above 35 PACS per 100 marriages. At the other end of the scale, ratios above 60 PACS per 100 marriages are found in only five départements (Haute-Vienne, Pyrénées-Atlantiques, Hautes-Pyrénées, Gers, and Puy-de-Dôme). Apart from two zones of adjacent départements, one in the south-west, the other to the east of the Paris region, the départements where the ratio is over 50 PACS per 100 marriages are evenly distributed across France.

Number of heterosexual PACS per 100 marriages, 2007-2008
Source: Calculations based on data from the French Ministry of Justice and INSEE (Table 16 and map in Appendix).

In addition to a preference for the PACS, other factors that may explain these differences at département level include the respective proportions of the population already in marital and PACS unions, and the relative size of specific sub-populations with differing propensities to form marital or PACS unions (students, older couples, rural dwellers, foreigners, etc.). Because the status of PACS partners is not identical to that of married partners (notably for foreign nationals [26]  Unlike marriage, the PACS confers no entitlement to...[26] and for persons whose partner has died [27]  In the event of death, the remaining PACS partner does...[27]), certain sub-populations prefer to choose marriage rather than the PACS to obtain legal recognition for their union.


While PACS partners tended to have quite specific profiles in the years after the PACS first came into force (gay and lesbian couples, couples seeking alternative forms of union, civil servant couples for whom a PACS union makes it easier for both partners to obtain posts in the same geographical area), its rapid spread will probably lead to a weakening of the regional contrasts. On the other hand, the variations between départements linked to the proportion of young people in the population, which became evident from the first years of the PACS (Belliot, 2005), may persist for several more years yet, since the stocks of potential PACS partners are numerically smaller in the départements with older populations. The same is true in the départements with relatively large numbers of foreign nationals, unless the legislation in this area is changed. Some of the recent increase in PACS unions can probably be attributed to improvements in the tax advantages – notably in 2005, when the income tax regime for PACS partners was brought into line with that for married couples – but it can be assumed that as the PACS gains in popularity and becomes more integrated into the French legal system it will be chosen by ever larger numbers of young couples, irrespective of their sexual orientation, fiscal motives or career mobility strategies.

The PACS is associated with a varied range of “social practices”. While some contracting partners view it as an alternative to marriage, others see it as a step towards marriage at a later date or, in the case of same-sex couples, as a substitute for marriage. The signing of the PACS contract can be the occasion for a “celebration” (publicized or not; possibly with a ceremony to celebrate and announce the union). The symbolism of these practices and the form they take can be identical for same-sex and different-sex couples (Rault, 2009).

A majority of divorces by mutual consent


The number of divorces fell slightly in 2008, continuing the downward trend observed since 2005. In all, 132,594 divorces were granted, [28]  Direct divorces and conversions of separations. We...[28] some 1,900 fewer than in 2007 and a fall of 1.4%.


The number of divorces in metropolitan France was just under 130,000 for the first time since 2003 (Table 9). After peaking at 52.3 divorces per 100 marriages in 2005, the first year the new Divorce Act of 26 May 2004 came into force, [29]  The 2005 peak in divorce arose primarily from the simplification...[29] the total divorce rate has fallen by a few tenths of one percentage point, from 45.5 in 2007 to 45.1. The total rate could therefore stabilize at around 45, substantially higher than its level of 38-39 per 100 marriages in the period 1995-2000.


Can the higher frequency of divorce be attributed to the new legislation that unquestionably makes obtaining a divorce easier, even when contested by one of the spouses? [30]  Divorce for irretrievable marriage breakdown (which...[30] A recent report by the Ministry of Justice examines the petitions for divorce filed since 1996 and their outcomes in the two years that followed (Chaussebourg et al., 2009). It finds that the increase in divorce petitions occurred a few years before the divorce law reform of 2005 came into force, while their number remained reasonably stable between 2003 and 2006 and then fell back slightly in 2007 and 2008 (Figure 6a). The increase in petitions concerned mainly mutual consent divorce [31]  Until 2004 the procedure was known as “joint petit...[31] though also to a lesser extent “contested” divorce, i.e. where the spouses disagree over the effects of the divorce (alimony, child maintenance payments and child custody) or over the actual principle of the divorce. Contested divorces still account for a large majority of total divorces, since they represented 59-60% of petitions filed in 2006, 2007, and 2008, versus 66% in 1996.

Figure 6Divorce petitions filed and divorces granted by type of proceeding, since 1996
Divorce petitions
Divorces granted
Sources: Chaussebourg et al. (2009), Lermenier and Timbart (2009).

A very different picture emerges when we consider divorces granted (Figure 6b). This time, the abrupt increase, within the space of a few years (2003-2005), is limited to mutual consent divorce. The number of contested divorces, by contrast, has tended to fall slightly, and since 2005 they have been outnumbered by uncontested divorces. The large disparity between the numbers of contested petitions filed and of contested divorces granted can be explained largely by the fact that a proportion of the petitions filed do not end in a divorce. This is less often the case with uncontested petitions: the proportion of petitions filed between 1998 and 2004 for which a divorce was granted within two years was 80-84% for mutual consent proceedings, against only 47-52% for contested proceedings (Chaussebourg et al., 2009, p. 46). The other proceedings are either still going through the courts (roughly 20%) or have been settled outside the divorce courts (around 30%). [32]  In the judicial nomenclature of the decisions that...[32] Under the new legislation, the procedure for mutual consent divorce was simplified and shortened. Abolition of the compulsory six-month waiting time has resulted in most divorces now being granted at the first hearing, two or three months after the petition is filed, compared with between seven and nine months previously. The frequency of mutual consent divorces granted within two years has thus increased: more than 92% of petitions filed in 2005 had already ended in a divorce before the end of 2007. For contested divorces the opposite has occurred and proceedings have lengthened, with 32% still going through the courts at the end of 2007. In addition, the new legislation makes it possible to switch between types of divorce procedure, so some contested divorces have been changed into mutual consent divorces in the course of proceedings.


Taken together, these factors explain why mutual consent divorces represent the majority of divorces granted even though contested divorces are most numerous when the petitions are filed. So the new legislation does not seem to account for the increase in petitions for divorce. But it does help to speed up the outcome when there is agreement between the spouses, as well as encouraging the choice of an uncontested proceeding. In addition, the most conflictual of the contested proceedings are becoming less common and fault-based divorces have declined steadily (Lermenier and Timbart, 2009); they accounted for 38% of divorces granted in 2001, but only 13% of those granted in 2008.

Last, uncontested divorces are increasingly numerous, making up 23% of the total in 2008. Strong growth is also observed in divorces for “irretrievable marriage breakdown”, which represented 9% of the divorces granted in 2008. While in the first procedure the spouses agree over the divorce (though perhaps disagreeing over the post-divorce arrangements), in the second, it is sufficient to prove that the parties have lived apart for at least two years.

Divorce is more frequent in Paris and along the Mediterranean coast


The recent frequency of divorces across France can be compared by dividing the average number of divorces in 2006, 2007, and 2008 by the number of people at risk of divorce in each département. [33]  Taking the annual averages of newly divorced persons...[33] Relatively large differences between départements are observed (Table 16). The lowest value for the indicator, 7.5 divorces per 1,000 married persons, is in Lozère, and the highest, at 20.5 per 1,000, or almost triple, is recorded in Paris, which thus has a particularly high risk of divorce, since the next highest levels – 17.2 per 1,000 in Guadeloupe and 16.5 per 1,000 in Bouches-du-Rhône – are appreciably lower. [34]  Census declarations may introduce a small bias into...[34]

Despite the general increase in divorce and a clear convergence in behaviour between the départements with the highest and lowest values, the current regional disparities (Figure 7) are broadly similar to those observed some thirty years ago (see the map of divorces in 1974-1975 in Munoz-Pérez, 1981 [35]  This study uses the ratio between the average number...[35]). For example, as in the mid 1970s, divorce is at relatively low levels in the four départements of the southern Massif Central (Cantal, Haute-Loire, Lozère, and Aveyron) as well as in Brittany and in the four neighbouring départements (Manche, Mayenne, Vendée, and Deux-Sèvres). Conversely, of the ten départements where divorce was previously most frequent, six (Alpes-Maritimes, Bouches-du-Rhône, Rhône, Haute-Garonne, Territoire de Belfort, and Vaucluse) still occupy the same position today. A very high rate is also registered for the Île-de-France region, presumably due to the higher incidence of divorce in the population of Paris. [36]  Paris could not be treated separately in that study....[36]

Divorce rates (per 1,000)* in the départements in 2006-2007

Newly divorced persons per 1,000 married men and women aged under 70 in 2006.

Source: Calculations based on data from the French Ministry of Justice and INSEE (Table 16 and map in Appendix).

The most salient changes since the 1970s are threefold. First, the formation of an unbroken zone of high divorce in the départements bordering the Mediterranean. Second, the virtual disappearance of a zone of high divorce to the north east of the Paris basin (Oise, Aisne, and Marne). And last, the eastward extension of the zone of low divorce centered on Brittany (notably into Maine-et-Loire, Loire-Atlantique, and Sarthe).


The relationships identified in the 1970s between the frequency of divorce and, first, urbanization levels, and second, the percentage of children attending private (typically Catholic) elementary and pre-elementary schools (Munoz-Pérez, 1981, pp. 102-108) are still relevant today. Attitudes towards divorce currently seem less influenced by urbanization levels, except perhaps in the most rural and the most urban départements. However, the two zones with the lowest levels of divorce correspond almost exactly to those where private education was the most strongly implanted in the 1970s – the southern Massif Central, and Brittany and its adjacent area. It is known that the strength of religious values is still a factor influencing family behaviour in France (Régnier-Loilier and Prioux, 2009).

VI - Mortality

No improvement in female life expectancy in 2008


The number of deaths in 2008 is estimated at 543,500 (of which 534,000 in metropolitan France), an increase of 12,300 on 2007. It corresponds to a rate of 8.5 per 1,000 inhabitants. This increase in deaths is reflected in a small reduction in life expectancy at birth for women – estimated at 84.33 years in 2008, 0.06 years less than in 2007 (84.39 years), while for men it rose by a mere 0.14 years, from 77.38 to 77.52. [37]  Provisional data supplied by INSEE (Demographic surveys...[37] The stagnation in female mortality has affected the overseas départements but also metropolitan France, where mean life expectancy for women fell from 84.43 to 84.37 years, while for men it rose from 77.43 to 77.59 years (Table 11). The gender gap in mean life expectancy has thus narrowed to 6.8 years, falling below seven years for the first time. After decreasing steadily from the early 1990s – the difference was still eight years in 1995 – the female mortality advantage stabilized at around seven years as of 2003.


This female advantage is due to the male excess mortality observed at all ages and in particular at ages 20-25 (when the probability of dying is three times higher for men than for women) and at ages 50-70 (probability at least 2.2 times higher). The gender gap in life expectancy is 6.6 years at age 20, and 4.9 years at age 60, when life expectancy stands at 26.9 years for women and 22.0 for men (Table 11). Male deaths outnumber female deaths at all ages, but after age 80 – and although men’s probability of dying is still 1.5 times higher – female deaths increase in number due to the dissymmetry in the population pyramid (Figure 1).

Mortality in the first year of life (infant mortality) has been stable since 2005 (Table 11). The infant mortality rate was halved between 1986 and 2003, falling from 8 deaths per 1,000 births to 4 per 1,000, but now seems to have bottomed out at 3.6 per 1,000 births in metropolitan France (Table 13) and 3.8 per 1,000 if the DOMs are included (Pla, 2009). Infant mortality in metropolitan France is generally lower in the centre and west than in the north-east, and is high throughout the DOMs, where the average level is double that of metropolitan France.

Different causes of death at each age


Cancers are the main cause of mortality in France (Table 14). The standardized mortality rates (i.e. controlling for differences in age structure) show that cancer became the dominant cause of death for men in the late 1980s and for women in the early 2000s, ahead of cardiovascular diseases. In 2006, cancer mortality represented about one-third of the standardized rate for all ages (34.8% for men and 32.4% for women), while the share of cardiovascular diseases is currently down to one-quarter (24.4% and 25.6% respectively). Next come “other diseases”, many of which are associated with the oldest ages [38]  This category includes diseases of the respiratory...[38] (17.8% and 21.8% respectively), while “injury and poisoning” (accidents, suicides, etc.) rank fourth (9.3% for men and 7.0% for women).


Cause-of-death patterns for the period through childhood and adolescence (0-14 years) are highly specific. The most important causes are congenital malformations and early childhood diseases, followed by accidental deaths. Beyond the first year of life, however, mortality is extremely low; it reaches a minimum at ages 9-10, when the risk of dying is below 1 per 10,000. At ages 15-24, injury and poisoning, mostly from road accidents but also from suicide, is the most important cause, responsible for 70% of male deaths and 53% of female deaths (Table C). This group of causes accounts for much of the high male excess mortality at these ages.

Standardized mortality rates by broad age groups in 2006* (per 100,000) and distribution by causes of death (%)

At ages 25-44, injury and poisoning are still the leading causes of death for men (48%): in this age group, the majority of these deaths are suicides but they also include deaths from road accidents and other accidental deaths. Injury and poisoning contribute less (27%) to female mortality at this age, when cancer becomes the single most important cause of death (40%). The dominance of cancer is even more marked (57%) among women aged 45-64 years. At these ages cancer is also the main cause of death among men (48%), although cardiovascular diseases are also a major killer (18%). At ages 65-79, cancer remains predominant (45% for men, 46% for women), ahead of cardiovascular diseases (24% and 25% respectively). Cardiovascular diseases finally become the principal cause of death among persons aged 80 or over (41% of female and 37% of male deaths), while “other diseases”, essentially conditions of the oldest ages, overtake cancers among women.


The increasing weight of cancer mortality at some ages, particularly for women, is due to the slow progress in combating this disease compared to advances in the prevention and treatment of cardiovascular disease (Meslé, 2006). In addition, while recent years have seen a favourable trend in male mortality for all cancer sites, this is not the case for women, who face increased mortality from cancer of the lung and the larynx, particularly at ages 45-64. Male cancer mortality nonetheless remains much higher than that of women (after age 45 at least), notably for cancer of the lung and the larynx.

According to the most recent report on cancer in France (INCa, 2009), the geography of cancer mortality (all sites) for 2002-2004 was markedly less favourable in départements located in the north of France, especially for men.

Mortality in the European Union: the east-west divide


Based on female life expectancy at birth, the European Union countries where mortality is lowest are France, Spain, and Italy, where women have mean length of life of 84.2 to 84.3 years. This was bettered in Switzerland, with 84.4 years (Table 12). In Eastern Europe, mortality is generally higher and female life expectancy is often below 80 years. The highest mortality levels, with female life expectancies of between 76.5 and 76.9 years, are observed in Latvia, Bulgaria, and Romania. These high-mortality countries are also those where the gender gap is widest, with more than 10 years’ difference between female and male life expectancy in Estonia (11.6 years), Latvia (10.7), and Lithuania (12.4). In six of the new EU member countries – Bulgaria, Estonia, Hungary, Latvia, Lithuania, and Romania – male life expectancy at birth is below 70 years. The gender gap is smallest in Denmark, Greece, Ireland, the Netherlands, United Kingdom, and Iceland (3.8 years). In terms of overall mortality, France compares favourably with its European neighbour countries, but it counts among the high-mortality countries for certain causes of premature death, notably suicide and alcohol-related deaths (Eurostat, 2009).


The observations relative to infant mortality follow a broadly similar pattern. Comparatively low rates, close to that in France, are observed in the countries of Western and Southern Europe (Table 13): Italy (3.7 deaths per 1,000 live births), Spain (3.5), Greece (3.5), Portugal (3.3), Netherlands (3.8), Austria (3.7), Germany (3.5), Belgium (3.4). Infant mortality levels are generally higher in Eastern Europe, notably in Romania (11.0) and Bulgaria (8.6), probably due to the less favourable economic and health conditions in those countries. Note, however, that infant mortality in France is higher than in many European countries, seven of which – Finland, Luxembourg, Czech Republic, Slovenia, Sweden, Iceland, and Norway – record infant mortality below 3 deaths per 1,000 live births. For individual countries, therefore, no direct relationship exists between levels of infant mortality and levels of life expectancy at birth.



Now that the final results of the first five-year cycle of annual census surveys are available, INSEE has made further upward adjustments to French population growth for 1999-2005.


The total population of France on 1 January 2009 is estimated at 64.3 million inhabitants, of which 62.45 million in metropolitan France where it increased by an estimated 337,000 in 2008, almost identical to the previous year’s increase of 335,000. At 4.2 per 1,000, the rate of natural increase remains among the highest in the European Union.


The number of residence permits issued to foreign nationals from outside the European Economic Area (EEA) fell by 8% in 2007, following a decline of 2% in 2006. Family immigration fell slightly though it remains the single most important reason for legal immigration to France (55%).


Fertility rose in 2008. The total fertility rate exceeds an estimated 2 children per woman (2.02) and, according to provisional figures, stands at 2.0 children per woman in metropolitan France, a level not reached since 1974, and thanks to which France still ranks first among the European countries. The fertility increase is due to births to women aged 30 or over. Women aged 25-34 account for two-thirds of total fertility and the mean age at childbearing is close to 30 years. Despite this, completed fertility falls off slightly after the 1960 birth cohort. It is expected to stabilize at around 2 children per woman from the 1970 cohort or thereabouts.


The number of induced abortions in metropolitan France remains stable at between 205,000 and 210,000. However, the frequency of abortion is rising slightly among the youngest women, a consequence of the increasing propensity to terminate pregnancies at these ages. There is no decline in the proportion of repeat abortions.


The number of marriages fell slightly in 2008 while that of civil unions (PACS) continued to increase, with nearly 150,000 PACS registered in the year. It is estimated that over one million men and women have signed such a contract since 1999, though they may or may not still be in a PACS union. The increase concerns same-sex couples as well as different-sex couples. The geography of the PACS continues to exhibit relatively sharp contrasts. A comparison of PACS and marriages shows that marriage still predominates heavily in some départements while in others the number of PACS unions is catching up with marriage. For France as a whole, just over one heterosexual PACS was concluded for two marriages in 2008, equivalent to just over one PACS for every three unions registered.


The number of divorces fell slightly in 2008, by 1,900, continuing the downward trend observed since new legislation came into effect in 2005. The total divorce rate stands at 45.1 divorces per 100 marriages. Mutual consent divorce has become the most widely used procedure since 2005. The highest levels of divorce are in Paris and the départements bordering the Mediterranean, while the lowest are in the southern Massif Central and in Brittany and its adjacent départements.

Male life expectancy at birth improved slightly in 2008, while female life expectancy stagnated. Cancers and cardiovascular diseases are the leading causes of death. Men still have a much higher risk of dying from a cancer of the bronchus or the lung, but female cancer mortality has declined little in recent years due to an appreciable increase in mortality from lung and laryngeal cancer related to the spread of smoking among women. Female life expectancy in France nonetheless remains among the highest in the European Union.

Statistical Data

Population change (in thousands) and crude rates (per 1,000)(a)
Age distribution of the population of France on 1 January (%)
Legal long-term immigration of foreign nationals (adults and minors) from the European Economic Area (EEA) and from countries without freedom of movement rights in Europe
Fertility since 1970
Cohort fertility: cumulative fertility up to selected ages, estimated completed fertility (mean number of children per 100 women), and mean age of childbearing (in years and tenths of years)
Total fertility rates in Europe (total number of children per woman)
Cohort fertility in Europe
Number of induced abortions and annual indices since 1976
Characteristics of nuptiality and divorce since 1985
Characteristics of nuptiality by birth cohort
Characteristics of overall mortality since 1985
Life expectancy at birth in Europe in 2007
Infant mortality in Europe (rate per 1,000 live births)
Standardized death rates (per 100,000) by sex and groups of causes of death(a)
Cause-of-death groups and the corresponding items in the international classification of diseases (ninth and tenth revisions)
The départements of metropolitan France
Département indicators


  • Beaumel C., Pla A., Vatan L., 2009a, Statistiques d’état civil sur les naissances en 2008, Insee Résultats, Société, 97.
  • Beaumel C., Pla A., Vatan L., 2009b, Statistiques d’état civil sur les mariages en 2008, Insee Résultats, Société, 100.
  • Belliot N., 2005, “Cinq années d’application du pacte civil de solidarité en France”, in Bergouignan C., Blayo C., Parant A., Sardon J.-P., Tribalat M., (eds.), La population de la France. Évolutions démographiques depuis 1946, Pessac, CUDEP, vol. 1, pp. 253-270.
  • Bessin M., Levilain H., Régnier-Loilier A., 2005, “Avoir ses enfants sur le tard. Une exploration statistique de la parenté tardive à partir d’EHF 1999”, in Lefèvre C., Filhon A. (eds.), Histoires de familles, histoires familiales, Paris, INED, Cahier 156, pp. 283-307.
  • Bozon M., 2008, “Premiers rapports sexuels, premières relations : des passages attendus”, in Bajos N., Bozon M. (eds.), Enquête sur la sexualité en France. Pratique, genre et santé, Paris, La Découverte, chapter 7, pp. 117-148.
  • Carrasco V., 2007, “Le pacte civil de solidarité : une forme d’union qui se banalise”, Infostat Justice, 97, October.
  • Chaussebourg L., Carrasco V., Lermenier A., 2009, Le divorce, Ministère de la Justice, Sous-direction de la statistique et des études, 100 p.
  • Daguet F., 2005, Données de démographie régionale, 1954 à 1999, Insee Résultats, Société 49, 48 p. + CD Rom.
  • Desplanques G., 2008, “Strengths and uncertainties of the French annual census surveys”, Population, English Edition, 63(3), pp. 415-440.
  • Desplanques G., Royer J.-F., 2005, “Enquêtes annuelles, de recensement. Premiers résultats de la collecte 2004”, Insee Première, 1000, January.
  • Eurostat, 2009, Health Statistics – Atlas on Mortality in the European Union, Eurostat Statistical books, 205 p.
  • Halifax J., 2009, Situation des pupilles de l’État au 31 décembre 2007, ONED, Paris, January.
  • Héran F., Toulemon L., 2005, “What happens when the census figures do not match the estimates?”, Population and Societies, 411, April.
  • INCa, 2009, La situation du cancer en France en 2009, Institut national du cancer, Collection Rapports & synthèses, 209 p.
  • Lermenier A., Timbart O., 2009, “Les divorces prononcés de 1996 à 2007”, Infostat Justice, 104. www.justice.gouv.fr/art_pix/1_stat_infostat104divorces_20090218.pdf
  • Marcu M., 2009, “The EU-27 population continues to grow”, Data in focus, Eurostat, 31/2009.
  • Meslé F., 2006, “Recent improvements in life expectancy in France: men are starting to catch up”. Population, English Edition, 61(4), pp. 365-388.
  • Munoz-Pérez B., 1981, “La répartition géographique des divorces de 1970 à 1975”, in Baillon D., Costecalde N., Godin G., Munoz-Pérez B., Le divorce en France, Ministère de la Justice/Insee, Collections de l’Insee, D85-86, vol. 1, pp. 99-108.
  • Munoz-Pérez F., 2000, “Les enfants nés sans filiation en France, 1965-1994”, Population, 55(4-5), pp. 663-690.
  • Munoz-Pérez F., Prioux F., 2005, “Filiation des enfants nés hors mariage en France depuis 1950”, in Bergouignan C., Blayo C., Parant A., Sardon J.-P., Tribalat M., (eds.), La population de la France : évolutions démographiques depuis 1946, Pessac, CUDEP, vol. 1, pp. 333-354.
  • Pla A., 2009, “Bilan démographique 2008 : plus d’enfants, de plus en plus tard”, Insee Première, 1220, January.
  • Prioux F., 2005, “Late fertility in Europe: Some comparative and historical data”, Revue d’épidémiologie et de santé publique, vol. 53, N° HS2, November, pp. 211-223.
  • Prioux F., 2006, “Recent demographic developments in France”, Population, English Edition, 61(4), pp. 323-364.
  • Prioux F., 2008, “Recent demographic developments in France: life expectancy still rising”, Population, English Edition, 63(3), pp. 375-414.
  • Rapport au Parlement, 2008, Les orientations de la politique de l’immigration, Comité interministériel de contrôle de l’immigration, Paris, La Documentation française, 235 p.
  • Rault W., 2009, L’invention du Pacs : pratiques symboliques d’une nouvelle forme d’union, Paris, Presse de la fondation nationale des sciences politiques, 275 p.
  • Régnier-Loilier A., Prioux F., 2009, “La pratique religieuse a-t-elle une influence sur les comportements familiaux en France ?”, in Régnier-Loilier A. (eds.), Portraits de familles. L’enquête Étude des relations familiales et intergénérationnelles, INED, Grandes enquêtes, pp. 397-424.
  • Richet-Mastain L., 2006, “Bilan démographique 2005. En France, la fécondité des femmes augmente toujours”, Insee Première, 1059, January.
  • Richet-Mastain L., 2007, “Bilan démographique 2006 : un excédent naturel record”, Insee Première, 1118, January.
  • Rossier C., Toulemon L., Prioux F., 2009, “Abortion trends in France, 1990-2005”, Population, English Edition, 64(3), pp. 443-476.
  • Toulemon L., 2005, “Who are the late mothers?”, Revue d’épidémiologie et de santé publique, vol. 53, HS2, November, pp. 13-24.
  • Vilain A., 2008, “Les interruptions volontaires de grossesses en 2006”, Paris, DREES, Études et résultats, 659.


[*] Institut national d’études démographiques.Correspondence: France Prioux, Institut national d’études démographiques, 133 boulevard Davout, 75980 Paris Cedex 20, tel.: +33 (0)1 56 06 21 44, e-mail: prioux@ined.fr

[1] The large adjustment between the censuses of 1999 and 2006 is explained by the length of the intercensal interval and by differences in methodology (with lower under-enumeration thanks to more highly trained census agents and to the use of the identified buildings register, and probably also more double-counting). On the question of adjustments after each census in France, see Héran and Toulemon (2005).

[2] See the article by G. Desplanques (2008), which discusses the possible causes of these differences and the necessary adjustments to the population age distributions.

[3] Tables 1 to 16, updated annually, are given in the Appendix. Their number does not always correspond to the order in which they are referred to in the text.

[4] The population of the overseas départements (DOM) no longer includes the inhabitants of the islands of Saint-Martin and Saint-Barthélemy, which have had overseas collectivity status since 15 July 2007. Hence they are no longer integral parts of Guadeloupe.

[5] INSEE now calculates two population age distributions using different methodologies. One is based on the five-year average of people enumerated at a given age, the other uses the average of the total enumerated in each cohort: www.insee.fr/fr/methodes/sources/pdf/methodologie_Estimations_de_population.pdf

[6] Eurostat website accessed on 04/09/2009.

[7] The authors thank Xavier Thierry (INED) for supplying the background material for this section.

[8] Although under-age children can be admitted without a residence permit they are included in these statistics (and accounted for 7% of total admissions in 2007).

[9] All citizens of new Member States are subject to this obligation, except for citizens of Malta and Cyprus.

[10] Since 1 July 2008, foreign nationals from countries that have been Member States since 2004 have been exempted from this requirement.

[11] Some foreigners do not obtain their first one-year residence permit until after several years of residence in France.

[12] The figures for foreign marriages transcribed to French records come from the report to the French Parliament (Rapport au Parlement, 2008).

[13] More detailed data on nationalities are available on the INED website at www.ined.fr/en/pop_figures/france/immigration_flow/

[14] These figures are currently provisional estimates that will be revised by INSEE in 2010.

[15] Fertility rates are projected under two scenarios. The first applies the rates from the last observed year (in this case 2008) to each age for the years of the projection; the second extrapolates the trend of the fifteen previous years.

[16] The crescent-shaped belt of high fertility extending from Brittany and Pays-de-la-Loire to Lorraine, taking in the Nord but skirting around the Île-de-France region.

[17] These estimates are liable to be revised upward since the method used (keeping the rates from the last year of observation) usually under-estimates fertility at older ages.

[18] Assuming completed fertility does not change, a steady increase in age at childbearing produces a proportional shortfall in annual period fertility: an increase of 0.1 years in the mean age per cohort gives a TFR 10% below completed fertility, an increase of 0.2 years per cohort gives a shortfall of 20%, and so on.

[19] The complete data sets for 2006 and 2007 are currently being processed.

[20] See Rossier et al. (2000) for a description of the differences between the old and new notification forms.

[21] Spontaneous abortions (miscarriages) are not recorded and are not considered here.

[22] The fall in binational marriages since 2003 is probably a consequence of the new regulations governing marriages of foreign nationals in France (see above, section on foreign immigration).

[23] A total of 4,780 PACS were between two men (3.3%) and 3,423 between two women (2.3%).

[24] 265,404 marriages and 146,030 PACS, of which 137,820 were between a man and woman.

[25] 9,610 PACS were dissolved due to marriage in 2008 and 10,781 in 2007.

[26] Unlike marriage, the PACS confers no entitlement to family reunion or to acquisition of French nationality (which in any case is not granted automatically or immediately after marriage).

[27] In the event of death, the remaining PACS partner does not receive a survivor’s pension.

[28] Direct divorces and conversions of separations. We would like to thank the Ministry of Justice (SDSE-BDSE) for supplying the 2008 data before their publication.

[29] The 2005 peak in divorce arose primarily from the simplification of proceedings for divorce by mutual consent (Prioux, 2008; Lermenier and Timbart, 2009).

[30] Divorce for irretrievable marriage breakdown (which replaced divorce for breakdown of conjugal life) can be granted after two years’ separation instead of six previously.

[31] Until 2004 the procedure was known as “joint petition”.

[32] In the judicial nomenclature of the decisions that terminate the proceedings, the three main ones, other than divorce, are petitioner withdrawal, cancellation or expiry of the petition, to which are added dismissal of the petition and the other cases where the court’s competence ceases.

[33] Taking the annual averages of newly divorced persons in 2006-2008 and dividing by the number of men and women aged 15-69 reporting as married in the 2006 census, we calculate an indicator that can be likened to an annual risk of divorce per 1,000 married persons in 2006. Since divorce is infrequent among people aged 70 or over, we include only the married population aged under 70 to limit under-estimation of divorce in the départements with large numbers of older residents. The data came from the INSEE website.

[34] Census declarations may introduce a small bias into the results and artificially accentuate the disparities between Paris and the other départements if we suppose, for example, that some Parisians with a second home in the country are enumerated in a different département. In the event of divorce, however, they probably apply to the court in Paris.

[35] This study uses the ratio between the average number of divorces in 1974 and 1975 and the total number of married women in the 1975 census. The correlation coefficient between the values for 1974-1975 and those for 2006-2008 is 0.8.

[36] Paris could not be treated separately in that study. A change made to the territorial jurisdictions of several courts in the Paris region in the early 1970s obliged the study author to aggregate the départements of the Île-de-France region. Working on the seven départements of the Paris region together in 2006-2008, gives a divorce rate of 16.9 divorces per 1,000 married persons.

[37] Provisional data supplied by INSEE (Demographic surveys and studies division).

[38] This category includes diseases of the respiratory system (bronchitis, emphysema, etc.) and mental disorders (including senile dementia).



The population of metropolitan France (mainland and Corsica) on 1 January 2009 was estimated at 63.2 million. Natural growth was again strongly positive in 2008, and almost identical to that of 2007, with the increase in births partly offset by a rise in deaths. The number of foreigners admitted for residence fell slightly more sharply in 2007 than in the two previous years. The estimated total fertility rate was 2 children per woman in 2008, a level close to the completed fertility of the 1970 cohort. The number of abortions remained stable, but their frequency increased slightly among the youngest women. The number of PACS civil partnerships signed in 2008 increased yet again, both for same-sex couples and for heterosexual couples. Marriages fell slightly, and the probability of marrying for single people has never been lower. Most of the divorces pronounced in 2008 were by mutual consent. Male life expectancy at birth increased slightly (+0.14 years) and that of women remained stable (–0.06 years).


  1. Overall population trends and age structure
    1. The population of France at the 2006 census
    2. Stable growth in 2007 and 2008
    3. A progressively ageing population structure
  2. Foreign immigration7
    1. A sharp decrease in 2007
    2. Family immigration is declining but remains the largest category
  3. Fertility
    1. An increase in births and in fertility
    2. Children born outside and within marriage: equal in number and in status
    3. Women aged 25-35 account for two-thirds of births
    4. Persistent but evolving geographical disparities
    5. Fertility among the highest in the European Union
  4. Abortions
    1. A stable number of abortions
    2. A slight increase in abortions at young ages
  5. Marriage, PACS, and Divorce
    1. A further decline in marriage
    2. The PACS civil partnership: over a million contracting parties in ten years
    3. Just over one PACS for two marriages in 2008
    4. A majority of divorces by mutual consent
    5. Divorce is more frequent in Paris and along the Mediterranean coast
  6. Mortality
    1. No improvement in female life expectancy in 2008
    2. Different causes of death at each age
    3. Mortality in the European Union: the east-west divide
  7. Overview

To cite this article

France Prioux and Magali Mazuy "L'évolution démographique récente en France : dix ans pour le pacs, plus d'un million de contractants", Population 3/2009 (Vol. 64).
URL  www.cairn.info/

Back to Top


Why is this article not available in English?

Cairn International Edition is a service dedicated to helping a non–French–speaking readership to browse, read, and discover work published in French journals. You will find English full–text translations, in addition to French version already available on Cairn regular edition. Full text translations only exist for a selection of articles.

If you are interested in having this article translated into English, please enter your email address and you will receive an email alert when this article has been translated.

Message sent

Your email address has been saved.
We will notify you when this article becomes available in English.