I – General trends and population age structure
Slower natural increase
1On 1 January 2013, the total population of France was estimated at 65.8 million, of which 63.7 million in metropolitan France (mainland France and Corsica). The island of Mayotte became a French département on 31 March 2011. Mayotte has an estimated population of 212,645 according to the latest census dated 21 August 2012 (a census is held in Mayotte every five years). 
2In 2012, the population increased by an estimated 305,000, of which 290,000 in metropolitan France (Bellamy and Beaumel, 2013). As in previous years, natural increase accounts for the majority of overall growth. For metropolitan France, the total rate of increase in 2012 is estimated at 4.6 per 1,000  and the rate of natural increase at 3.6 per 1,000 (Appendix Table A.1),  down on 2011 (4.1 per 1,000) due to a slight dip in births  and an increase in deaths, which totalled 560,000 in metropolitan France, and 571,000 in the country as a whole (Beaumel and Bellamy, 2013b). As 2012 was a leap year, the annual totals cannot be compared directly with those of the previous year. On a daily basis, 1,530 deaths per day were registered in 2012 versus 1,465 in 2011, representing an increase of more than 4%.
Almost half the population is aged below 20 or above 60
3The long-term stability of fertility and birth rates (almost 800,000 annual births, despite slight falls in 2011 and 2012) means that the base of the French population pyramid is still quite broad (Figure 1). While natural increase is still clearly positive, the ageing process is reflected in a rising number of annual deaths (Appendix Table A.1) as the population with the highest risks of dying grows larger.
Figure 1Population pyramid of France on 1 January 2013
Coverage: Whole of France.
4The proportions of under-20s and over-60s are similar, and these two groups together represent almost half of the population (Appendix Table A.2). Ageing due to the relative increase in the older population (summit of the pyramid) will continue in coming decades as the baby-boom cohorts reach advanced ages (Chardon and Blanpain, 2010). The number of multi-generational families (children, parents, grandparents and great-grandparents) is increasing thanks to increased life expectancy. This trend will continue over the coming decades, and families will span an ever wider range of ages.
An intermediate position in Europe
5According to Eurostat data, France occupies an intermediate position in the European Union in terms of population growth and ageing. In 2012, seven countries had a rate of increase greater than or equal to France:  Finland (+4.7 per 1,000), Denmark (+5.1 per 1,000), Austria (+5.4 per 1,000), Sweden (+7.7 per 1,000), Belgium (+7.9 per 1,000), Malta (+8.8 per 1,000), and Luxembourg (+23.0 per 1,000). Growth is negative in nine countries: Latvia (–11.9 per 1,000), Lithuania (–10.6 per 1,000), Bulgaria (–6.2 per 1,000), Estonia (–6.2 per 1,000), Spain (–6.2 per 1,000), Portugal (–5.2 per 1,000), Hungary (–2.6 per 1,000), Romania (–2.4 per 1,000) and Poland (–0.1 per 1,000). Data for Greece, Italy and Cyprus are not available for 2012, but in 2011, growth was negative in Greece (–1.8 per 1,000), positive in Italy (+3.2 per 1,000) and positive in Cyprus (+26.2 per 1,000). The rate of natural increase in Europe is falling due to fertility decline, but negative net migration due to a rise in emigration during the recent economic crisis has also contributed to a lower overall increase. This is notably the case in the Baltic countries and in southern Europe.
6Population ageing in France is slower than in countries which have experienced a rapid and dramatic fertility decline. This trend is clearly illustrated in the very narrow base of the European Union population pyramid (Prioux and Barbieri, 2012). The French population is ageing nonetheless: 9.1% of French residents were aged 75 or over on 1 January 2013, up 50% with respect to 1990 (Appendix Table A.2); and while half were below age 33 in 1960, the median age was 39.8 years in 2010 and 40.5 years on 1 January 2013.
7According to Eurostat projections  used in a summary report on the ageing of the baby-boom generation (Lanzieri, 2011), the median age of the French population will reach 44.5 years in 2060 (Table 1). In the eight countries where median age is projected to exceed 50 years in 2060, persons aged over 65 will represent between 31.5% and 35.7% of the population. According to these same projections, Ireland, the United Kingdom, Sweden, Belgium, Denmark and France will have the lowest median age in 2060 (below 45 years) and over-65s will represent some 25% of the population (Table 1). In some countries, population ageing is dramatic, notably in Romania where median age will be highest in 2060. In fact, the Romanian population pyramid is very singular, marked by abrupt demographic change (Statistical Yearbook of Romania, 2012).
Median age and proportion of persons aged 65 and over on 1 January 1960, 2010 and 2060
Median age and proportion of persons aged 65 and over on 1 January 1960, 2010 and 2060na: not available.
8Generally speaking, and especially when natural increase is slow or even negative in some countries, migration can play an important role in balancing the age structure thanks to the immigration of relatively young adults. Analysing the role of migration in the demographic ageing process in Europe, and using the Eurostat projections as a basis, Ambrosetti and Giudici (2013) imagined projection scenarios with or without migration, and showed the potential of migration to rejuvenate the population age structure and slow the ageing process in Europe. Beyond purely demographic aspects, the positive economic impacts of migration are also considerable.
II – Immigration from non-EEA countries, as reflected in the granting of long-term residence permits
9Flows of foreigners  arriving legally in France to establish residence in the country can be estimated from the statistics on long-term residence permits and long-term visas (one year or more) valid as residence permits. These statistics only concern countries whose nationals require a residence permit or visa to live in France, so they exclude all migration within Europe. They are based on data from the system used by the French Ministry of the Interior to track the status of foreigners residing in France (AGDREF), and were compiled at INED.
10To ensure consistency of comparisons over time, the statistics presented below are established for a constant geographical area. They therefore exclude residence permits granted previously to immigrants from countries whose nationals no longer need a residence permit. 
11The residence permits considered here have two important characteristics which explain the differences between the figures given below and those published elsewhere. First, the permits counted are valid for a period of 364 days or more, so all short-term permits are excluded. Second, among the permits of 364 days or more granted to a given immigrant, only the first is taken into account to avoid counting the same person more than once. Moreover, flows are characterized on the basis of the permits themselves: validity start date, period of validity, age at validity start date. These methodological choices enable us to focus on permanent migration and to count the inflow of foreigners with long-term migrant status. This means that migrants who are granted two successive seven-month permits then leave the country, for example, are not counted. So the results presented here do not correspond to the standard estimates of immigrant numbers determined in line with international recommendations and based on actual or planned duration of stay; they describe the activity of government services using specific information such as residence permit date of issue, period of validity, etc. Other sources, including the annual census surveys,  provide information on inflows.
12The data described below include all first residence permits with a validity of 364 days or more, granted both to minors and adults, while the estimates in Appendix Table A.3 are also based on the AGDREF database, but exclusively for immigrants who arrived as adults in metropolitan France, with minors being estimated from other sources (family reunification and asylum data: Office des migrations internationales (OMI), Agence nationale de l’accueil des étrangers et des migrations (ANAEM), Office français de l’immigration et de l’intégration (OFII)). The Ministry of the Interior also publishes figures for the total number of permits granted (including short-term permits), while INSEE estimates the migrant’s actual date of arrival in France and duration of stay.
13Among migrants receiving a first residence permit of 364 days or more, adults aged 18 or above in the year when their residence permit becomes valid are counted separately from minors, who are generally not legally required to apply for a permit. Table 2 gives the numbers of first permits awards between 2006 and 2011. The number of permits granted to foreign adults remains very stable, at around 164,000 per year. There is a slight downtrend since 2009, but the differences are small. The permits are issued in France while the long-term visas valid as residence permits are issued in French consulates abroad. The latter accounted for more than 40% of permits granted to adults.
Number of first permits of more than 364 days granted to non-EEA nationals (constant geographical area) by year of validity start date
Number of first permits of more than 364 days granted to non-EEA nationals (constant geographical area) by year of validity start dateCoverage: Permits granted in France and abroad to citizens of countries not listed in note 8. Permits granted in year n and recorded in the data extracted in July of the year n + 2, except for 2009, when extraction took place in July 2012.
14The age distribution of adult permit holders remained very stable over the period (Table 3). It is concentrated in the younger age groups, with immigrants aged 18-34 accounting for 72% of adult holders, while the over-45s represent less than 12%. On average, the majority of residence permit holders are women (Table 4). There is no clear trend over the period. On a longer timescale, however, this feminization of inflows is a new phenomenon (Beauchemin et al., 2013). African nationals represent a large majority, although the proportion of immigrants from other continents increases slightly over the period (Table 5).
Distribution of adult holders of a first residence permit of 364 days or more by age group, year and validity start date (%)
Distribution of adult holders of a first residence permit of 364 days or more by age group, year and validity start date (%)Coverage: Permits granted to adult citizens of countries not listed in note 8.
Distribution of adult holders of a first residence permit of 364 days or more by sex, year and validity start date (%)
Distribution of adult holders of a first residence permit of 364 days or more by sex, year and validity start date (%)Coverage: Permits granted to adult citizens of countries not listed in note 8.
Distribution of adult holders of a first residence permit of 364 days or more by continent of origin, year and validity start date (%)
Distribution of adult holders of a first residence permit of 364 days or more by continent of origin, year and validity start date (%)Coverage: Permits granted to adult citizens of countries not listed in note 8, by nationality of origin. Turkey is included in Asia. The total does not necessarily sum to 100 due to rounding and missing values.
15Tables 3, 4 and 5 show the typical profile of long-term residence permit holders in recent years. They are typically young (around age 25), most often a woman from an African country.
16The sex ratios of residence permit holders vary considerably by nationality (Table 6). The proportion of men is much lower among immigrants from America, Asia and above all Europe, but they represent a small majority among those from Africa. Migration from Asia and Europe is becoming increasingly feminized over time.
17The age structure of permit holders also varies by continent of origin. Asian immigrants are younger than the average, while those from America and Europe are older. Here too, there are contrasting dynamics: African and Asian immigrants are proportionally younger than before, while those from Africa and Europe are older (Table 6).
Proportions of men and of young adults among holders of a first residence permit of 364 days or more by continent of origin, year and validity start date (%)
Proportions of men and of young adults among holders of a first residence permit of 364 days or more by continent of origin, year and validity start date (%)Coverage: Permits granted to adult citizens of countries not listed in note 8.
III – Births and fertility
Fewer births, stable fertility
18Since 2010, the year when births reached a record high of 832,800 (of which 802,200 in metropolitan France), their number has fallen slightly. In 2012, they totalled 821,000, of which 790,000 in metropolitan France (Beaumel and Bellamy, 2013a), 2,400 fewer than in 2011 (Appendix Table A.1). As 2012 was a leap year, the extra day added a “bonus” of 2,500 births, so the decrease in births is more pronounced when determined on a daily basis.
19This fall is the consequence of stable fertility (Appendix Table A.4) – the total fertility rate (TFR) barely changed over the two years, standing at 2.01 children per woman (1.99 in metropolitan France) – combined with a decline in the proportion of women of reproductive age (Bellamy and Beaumel, 2013a). Fertility in France remains high in comparison with other European countries  (Appendix Tables A.6 and A.7).
20This levelling of fertility is observed at all ages, with only moderate upward or downward movements within different age groups. Age-specific rates increased slightly above age 30, and fell very slightly at younger ages (Table 7, Figures 2 and 3). The largest increase in fertility is at ages 35-39, although the relative contribution of this age group to fertility is quite small. Observed over the long term, recent trends in age-specific fertility, in terms of contribution to overall fertility, appear to be stabilizing (Figure 2). The age groups that contribute most to fertility are 30-34 (33.1%), followed by 25-29 (31.4%) and 35-39 (16.7%), while the over-39s (4.2%) and last, the under-20s (1.7%) are the age groups which contribute least.
Fertility by age group since 2006 (per 1,000 women)*
Fertility by age group since 2006 (per 1,000 women)** Due to rounding, the total may differ slightly from the sum, and the variations may not correspond to apparent differences.
Coverage: Metropolitan France.
Figure 2Percentage contribution of each age group to the total fertility rate since 1960
Coverage: Metropolitan France.
Figure 3Age-specific fertility rates between ages 20 and 39 since 1985 (per 1,000 women)
Coverage: Metropolitan France.
21While the total fertility rate is unlikely to vary substantially in terms of intensity, the changes in birth timing may stabilize in years to come. The trend towards later childbearing should continue and eventually stabilize, as it is determined largely by the increase in age at first birth. Note that in 2012, the mean age at childbearing was 30 years (Bellamy and Beaumel, 2013), and the mean age at first birth was slightly above 28 (Davie, 2012). The distribution of women by number of children will probably not change much (Masson, 2013). The proportion of childless women should also remain steady, as the proportion of men and women wishing to remain childless is quite stable (Debest and Mazuy, 2013), a minority of couples are sterile, and only a small proportion of the couples who have difficulty conceiving and who resort to assisted reproductive technologies (ART) succeed in having children – when their fertility intentions are seen in a broader biographical context (La Rochebrochard, 2011).
22France seems to have evolved from a “model” of early fertility to one of later fertility, concentrated between ages 25 and 35 (Figures 2 and 3). Fertility at ages 34-39 continues to increase (Figure 3). With the two-child norm and a tendency towards closer average birth spacing (Davie, 2012), the current pattern of fertility timing in France is very likely to stabilize, despite the growing diversity of family situations.
Levelling of completed fertility at 2 children per woman
23The long-term stability of the total fertility rate (but with large annual fluctuations) is reflected in the longitudinal data (Appendix Table A.5). The fertility decline is quite sharp for all the cohorts born between 1940 and 1970 (Figure 4). For the cohorts born in the 1970s and after, the mean number of children has stabilized at around 2 children per woman. If there is no unexpected break in the trend, completed fertility should increase for the cohorts born between 1970 and 1980, whose mean age at childbearing has risen from 27.5 to 30 years.
Figure 4Incomplete fertility at ages 24, 29, 34 and 39 and completed fertility*
* of women born after 1930 * Completed fertility of recent cohorts is estimated.
Coverage: Metropolitan France.
Increase in births outside marriage
24The proportion of non-marital births is still increasing. A total of 356,000 children were born outside marriage in 2012, representing 57% of total births.
25The proportion was below 9% in the early twentieth century, and changed little in peacetime until the 1960s (Figure 5). It fell to its lowest level in the 1960s, when around 6% of births occurred outside marriage. Since then, the trend has been upward, with non-marital births accounting for 30% of total births in 1990 and 43% in 2000 (Daguet, 2002b). New forms of union are emerging, as witnessed by the growing popularity of consensual unions, civil partnerships (introduced in 1999, and called pacte civil de solidarité or PACS), and unions where the partners “live apart together” or live together only part of the time. More than half of all births today occur outside marriage, and they account for 55% of the total fertility rate (Appendix Table A.4).
Figure 5Percentage of births outside marriage since 1901
Coverage: Metropolitan France.
26Since July 2006, French law no longer distinguishes between children born to married or unmarried parents. In the past, so-called “illegitimate” children had no filiation until the act of recognition was registered at the registry office. Today, the birth certificate certifies recognition by the mother. If the child is recognized by both parents, they are entitled to shared parental authority, each holding equal rights to the child. Other family reforms have been introduced, notably for the choice of surname (see Box), with few practical effects as yet, and, most recently, the legalization of same-sex marriage, voted in 2013, which will have practical implications for the legal recognition of women who have a child with a same-sex partner but are not the child’s biological mother (Descoutures, 2010). Many other issues are still under debate, such as the anonymity of egg and sperm donors for assisted reproduction (at present, gamete donation is anonymous and unpaid) or the legal status of third parties (notably step-parents), which as yet remains unclear. Parenthood covers an increasingly diverse ranges of situations which are progressively being legitimized (or otherwise) under law (Charrier and Clavandier, 2013, chapter 8).
Box. The family name given to children born in 2012
Nine percent of children born in 2012 bear both parents’ names“When a child’s filiation is established with respect to its two parents, at the latest on the day of birth registration or afterwards but simultaneously, the said parents choose the child’s family name: either that of the father or of the mother, or the names of both given in the order of their choice, but with no more than one family name for each. In the absence of a joint declaration mentioning the choice of the child’s family name, the registrar shall take the name of the parent with respect to whom filiation is first established and the name of the father if filiation is established simultaneously with respect to both parents. In case of disagreement between the parents, communicated by one of them to the registrar, on the day of birth registration, or after the birth when filiation is simultaneously established, the child shall take the names of both parents, limited to one family name for each, placed in alphabetical order.” Article 311-21 inserted by Law no. 2002-304, modified by Law no. 2013-404 of 17 May 2013 - art. 11.
Almost ten years after the new law was introduced, it is interesting to see how parents make use of the many choices available for naming their child. A large majority of parents (83%) choose to give the father’s name (Table 8), and the mother’s name is chosen for just 6.6% of children. In practically all cases, these are children not recognized by their biological father at birth, though a small number of fathers and mothers do deliberately opt for the mother’s name. Some cases correspond to children born within a lesbian union and who are recognized by the biological mother only (the law authorizing the registration of two persons of the same sex on a child’s birth record came into force in 2013).
Distribution of family name choices for children born in 2012
Distribution of family name choices for children born in 2012Coverage: Live births registered in whole of France excluding Mayotte.
27There is broad adherence to the norms of parenthood, notably the concepts of “planned parenthood”, of agreement within the couple and of birth control associated with widespread contraceptive use (Bajos and Ferrand, 2006; Mazuy, 2009; Bajos et al., 2012). In parallel, social and family situations preceding childbearing have become more diverse: parents’ age at first birth and birth spacing (both linked to educational level; Davie and Mazuy, 2010), assisted reproductive technologies to achieve a pregnancy or to shorten time to conception,  parents’ conjugal status, family recomposition, absence of a partner, same-sex partnerships, etc. are some of the many factors that come into play.
28Around 7% of births are registered to lone mothers. This proportion has remained stable since 2003 according to the latest perinatal surveys (Blondel and Kermarrec, 2011). Although these quantitative data give no information on the actual situations of these mothers, it is tempting to draw a parallel with the proportion of children registered under the mother’s name (6.6%) (Table 8). Lone parent situations at the time of the child’s birth vary by mother’s age. Some of the youngest mothers may not (yet) be living with the father at the time of the birth. Some of the older lone mothers may be women who have recently separated or found a new partner, or who are in other atypical situations (a stable non-cohabiting union for example). Such situations are difficult to detect through surveys or censuses that gather data using the question “Do you live with a partner”, which may be interpreted in rather restrictive way by respondents.
29In France today, the social differences in fertility mainly concern birth timing. Highly educated women have their first child at a later age (30 years on average versus 25 for the least educated) and their births are more closely spaced (Davie and Mazuy, 2010).
IV – Induced abortion
30The number of induced abortions remained stable in 2011. According to statistics based on medical procedures, 222,452 induced abortions were performed in 2011,  of which 209,291 in metropolitan France (Appendix Table A.8).
31There are few variations in abortion indicators.  The overall abortion rate has remained steady at around 14-15 per 1,000 women aged 15-49 since 1990. The total abortion rate remains stable at 0.53 per women (Appendix Table A.8). Taking account of repeat abortions, it is estimated that 35% of women have an induced abortion at least once in their life (Prioux and Barbieri, 2012; Bajos et al., 2013). This is an overall indicator, with some geographical disparities, notably in the overseas départements, and in Languedoc-Roussillon, Île-de-France, Corsica and Provence-Alpes-Côte d’Azur, where abortion rates are higher (Vilain et al., 2013). More detailed studies of fertility and of the social and family situation of the women who undergo abortion would provide a clearer picture of the reasons behind these regional differences.
32Abortion is most frequent at ages 20-24, when fertility is still quite low. The abortion rates of women under age 18 are between 1.8 per thousand at age 14 and 14.9 per thousand at age 17 (Figure 6) and have remained steady since 2006 (Prioux and Barbieri, 2012). The higher abortion rate in the 20-24 age group mirrors the growing concentration of pregnancies and births at ages 25-35, which is associated with a stigmatization of fertility and births among very young and older women; there is increasing social pressure to have children neither too early nor too late (Bessin and Levilain, 2012). Repeat abortions are becoming more frequent and mean age at abortion is decreasing, notably for a first abortion; it fell from 28.0 years in the early 1980s to 26.6 years in 2007. These indicators translate the effects of a longer interval between sexual debut and first birth, and bear witness to the growing diversity of sexual and conjugal trajectories, which may affect the likelihood of an unplanned pregnancy and the decision to abort. Although contraceptive coverage is wide (Bajos et al., 2012), recourse to abortion may imply that the contraceptive methods available to a woman do not correspond to her practical needs at a particular moment in her life (Bajos et al., 2012).
Figure 6Age-specific abortion rate per 1,000 women in 2011
Coverage: Whole of France.
V – Civil partnerships (PACS), marriage and divorce
PACS statistics affected by a new registration procedure
33As of 2011, the PACS statistics supplied by the Ministry of Justice are incomplete because a law was passed on 28 March 2011 empowering notaries to register civil partnerships. The statistics (Table 9) do not yet include data provided by notaries, but will do so in the near future. The apparent fall in the number of PACS unions thus corresponds to a smaller share of registrations at magistrates courts: the decline is only observed in the three last quarters of 2011, following a 15% increase in the first quarter (Prioux and Barbieri, 2012). We can thus assume that a large number of PACS unions are registered with a notary.
34The number of PACS dissolutions continues to increase each year. As notaries are also empowered to register dissolutions, the figure given is a low one, although it is reasonable to assume that couples who registered a PACS at the magistrates court will also register a dissolution there. Separations account for six in ten PACS dissolutions registered at magistrates’ courts, while almost four in ten are dissolved because the couple decides to marry. A very small but stable fraction (below 1%) are dissolved following a death.
Marriage no longer in decline?
35According to provisional INSEE data, 241,000 weddings were celebrated in 2012, of which 235,000 in metropolitan France (Bellamy and Beaumel, 2013). If these provisional figures are confirmed, this would suggest that the number of marriages is levelling off (Appendix Table A.9) after a steady decline since 2000 (excepting 2005).
Numbers of PACS unions registered and dissolved since 2007*
Numbers of PACS unions registered and dissolved since 2007** Only PACS unions registered in a magistrates’ court.
Coverage: Whole of France.
36The seasonal distribution of marriages always follows a similar pattern, with fewer than 5% of annual marriages per month in the winter, and a very high proportion in summer (Figure 7). The Paris region stands out, however, with few weddings in July and August (5% in August), contrary to small towns and villages where summer marriages are much more frequent.
Figure 7Monthly distribution of marriages in 2011, by size of urban unit (%)
Note: Monthly distribution based on a constant number of days per month (30.4). Rural municipalities and units with below 5,000 population are grouped in the first category.
Coverage: Whole of France.
Later and fewer first marriages
37In 2011, 71% of marriages were first unions between two never-married partners. This proportion remained stable with respect to 2010 and represented a total of 168,583 marriages.
38While 85-95% of men and women born in the early 1950s were ever-married at age 35, the proportion has fallen to 50% for men and 60% for women in the cohorts born in the 1970s (Figure 8). An estimated 65% of women in the 1975 cohort will have married at least once by age 50, and 63% of men in the 1973 cohort. First marriages in these cohorts were contracted at age 28.9 for women and age 30.6 for men (Appendix Table A.10).
Figure 8Proportion of ever-married men and women at different ages, by cohort (%)
Note: The dotted lines correspond to an estimate based on stable marriage rates in 2010.
39The difference in age at first marriage between never-married men and women (1.7 years) reflects the age difference between spouses (the man is generally older than the woman). The mean relative age gap between spouses (first and subsequent unions) is 2.7 years for married couples and 2.1 years for heterosexual PACS unions. Age at marriage and age at entry into a PACS union are very similar however.
VI – Mortality
Stagnation of progress in life expectancy in 2012
40According to INSEE, for the first time since the mid-1970s, the number of deaths in 2012 reached the threshold of 560,000 in metropolitan France (569,800 for the whole of France and 11,000 in the overseas départements). The large increase in deaths in 2012 is due mainly to the cold weather early in the year which favoured several epidemics – influenza and respiratory diseases in particular – while the previous winter had been relatively epidemic-free (Bellamy and Beaumel, 2013). The year 2012 is not exceptional, since epidemics of infectious diseases are a regular winter occurrence. In fact, until flu vaccine became widely available in the 1970s, they were the leading cause of death in the winter months (Meslé, 2010). They remain a major threat for older adults and, as the population increases in age, are producing a parallel increase in the number of deaths from epidemics.
41The other mortality indicators show a similar pattern. In 2012, the crude death rate stood at 8.7 deaths per 1,000 population, versus 8.4 in 2011. For the whole country, life expectancy at birth is estimated at 78.4 years for men and 84.8 years for women (78.5 and 84.9 years in metropolitan France in 2012, versus 78.4 and 85.0 in 2011). Despite a drop of more than two months in mean length of life for women (one month in metropolitan France – see Appendix Table A.11), France remains high in the European rankings for female life expectancy at birth (just behind Italy and Spain). For men, whose life expectancy remained stable with respect to 2011 (+0.1 year in metropolitan France), France is still in the top third of the European rankings (Appendix Table A.12). As mortality has evolved differently for men and women, the gender gap in life expectancy has narrowed from 6.6 years in 2011 to 6.4 years in 2012. This represents an acceleration of the convergence between male and female mortality levels observed since the 1980s, when the gender gap was more than 8 years.
A narrowing of the gender gap in mortality before age 65
42Figure 9 illustrates the progress achieved at each age for both sexes over the two previous decades. It represents the ratio between the probabilities of dying available for the most recent period and those observed at the same ages 20 years previously. Random variations, which may be large at ages where the risk of dying is very low, are minimized by using multi-year life tables established by INSEE for metropolitan France (1988-1990 and 2008-2010), and by smoothing the ratios with a moving average calculated over three years of age. In these tables, life expectancy at birth is estimated at 72.5 years for men and 80.7 years for women in 1988-1990, and at 77.9 and 84.5 years, respectively, in 2008-2010.
43At all ages, the mortality levels observed are well below those of the baseline period (baseline 100 in Figure 9). It is the risks of dying below age 20 that have decreased most over the period, with a fall of more than 50% on average for both sexes. This is especially the case for infant mortality, which fell from 7.3 deaths per 1,000 births in 1990 to 3.5 per 1,000 in 2010 (and 3.3 per 1,000 in 2012 according to provisional INSEE estimates). So France remains in a favourable position in the European rankings for infant mortality (Appendix Table A.13). However, as child mortality had already fallen to low levels by the end of the 1980s, the resulting gains in life expectancy at birth are low: 0.6 years (out of a total of 5.3 years) for men and 0.4 years (out of 3.8 years) for women, representing 11% in both cases. 
Figure 9Decrease in mortality at each age from 1988-1990 to 2008-2010 (ratio of age-specific probabilities smoothed over 3 years of age)
44At all other ages, mortality fell by 25% or more between the two periods for both sexes, with the exception of two stages in the life cycle: around ages 40-50 and at extreme old age (above 85 years for men, above 90 for women). In these two age groups, the decrease was below 20%.
45Progress has been systematically slower for women up to around age 65, and this accounts for the narrowing of the gender gap in life expectancy at birth. This gap has only decreased by 18 months, however, since mortality beyond age 67 has fallen more sharply for women than for men. For women, the progress in life expectancy at birth is thus due largely to lower mortality at advanced ages, while for men the mortality decline before age 65 makes a larger contribution. In fact, the mortality decrease between ages 10 and 65 accounts for more than 45% of the total increase in life expectancy at birth for men between 1988-1990 and 2008-2010, and just 25% for women. Conversely, more than 35% of life expectancy gains for women are due to mortality beyond age 80, and less than 15% for men. Mortality in childhood and at ages 65-80 contribute similarly for both sexes to gains in life expectancy at birth (30% and 35%).
46Figure 10 shows the male-to-female ratio of probabilities of dying by year of age in the 1988-1990 and 2008-2010 life tables. It shows a reduction in male excess mortality, at ages 40-65 especially. The gender gap in mortality has also decreased at the age where it was highest at the start of the period. Male excess mortality peaks at age 22 for both periods, but the male-to-female ratio of probabilities decreases slightly, from 3.5 in 1988-1990 to 3.1 in 2008-2010; in other words, male mortality was 3.5 times higher than that of females in 1988-1990, and 3.1 times higher in 2008-2010. After around age 75, the gender gap in life expectancy continues to widen due to faster progress for women at advanced ages.
Figure 10Male excess mortality at each age in 1988-1990 and 2008-2010 (ratio of male to female probabilities, smoothed over 3 years of age)
Decline in cardiovascular diseases is still the main driver of progress
47For both men and women, the decline in cardiovascular diseases accounts for the largest share of gains in life expectancy at birth between 1988-1990 and 2008-2010 (Table 10). Substantial progress has been achieved in ischaemic heart diseases, whose standardized rate has fallen from 96 deaths per 100,000 for men and 42 per 100,000 for women in 1990 to 48 and 17 per 100,000, respectively, in 2010 (Appendix Table A.14). Deaths from cerebrovascular diseases have also declined sharply, with rates of 30 and 22 per 100,000 in 2010 versus 71 and 52 in 1990. The gain attributable to the decline in all diseases of the circulatory system is comparable for both sexes, standing at 1.8 years for men and 1.9 for women, although it represents more than half of the total gain for women (51%) versus just a third for men (34%).
Contributions of groups of causes of death to differences in life expectancy at birth between 1988-1990 and 2008-2010 by sex (in years)
Contributions of groups of causes of death to differences in life expectancy at birth between 1988-1990 and 2008-2010 by sex (in years)Note: See Appendix Table A.16 for the definition of groups of causes of death.
48For men, the decline in cancer mortality also contributes strongly to life expectancy gains, explaining 24% of the total (1.3 years) versus just 12% for women. This progress is due mainly to a decline in lung and prostate cancers, whose standardized rates have fallen, respectively, from 70 to 58 per 100,000, and from 32 to 20 per 100,000. For women, breast cancer mortality has declined sharply, from 29 to 23 per 100,000. Conversely, lung cancer mortality has increased considerably, with a doubling of the standardized rate (from 8 to 16 per 100,000) between 1990 and 2010, eroding the gains achieved for other types of cancer.
49The decrease in deaths from external causes has also contributed to the increase in life expectancy at birth for both sexes, though slightly more for men (1.0 years representing 20% of the total) than for women (0.6 years, 16% of the total). The fall in road deaths has been especially beneficial for men in terms of years of life gained, with a standardized rate that fell from 26 to 10 per 100,000 between 1990 and 2010; for women it dropped from 9 to 3 per 100,000 over the same period. While the rate has been divided by 3 for both sexes, female mortality was already very low, so the decrease is much more visible for men. All other types of deaths from external causes have also decreased, including suicide, whose standardized rate has fallen by a quarter for men and by almost a third for women.
50The other groups of diseases (notably infectious diseases, respiratory diseases and digestive diseases) make a much smaller contribution to the progress observed over the period, even though the decrease in terms of standardized rates was large. This is the case, for example, for all respiratory diseases, which have declined by almost 50% for both sexes if influenza is included, but whose contribution to gains in life expectancy at birth is just 7% for both men and women.
The major role of cancers in reducing the gender gap in life expectancy
51Analysing the contribution of causes of death to the gains for men and women helps to explain why the gender gap in life expectancy at birth has narrowed. To this end, we began by calculating the contribution of each age group and each category of causes of death to the gender difference in life expectancy in 1988-1990 and 2008-2010 (Appendix Table A.15). The difference between contributions for each age group and each group of causes is shown in Figure 11 (which represents the difference between the first and second parts of Appendix Table A.15). The age groups and groups of causes which helped to narrow the gender gap have positive values, while those that widened it have negative values.
Figure 11Contributions of age groups and cause-of-death groups to the narrowing of the gender gap in life expectancy at birth between 1988-1990 and 2008-2010 (1.5 years in total)
Note: See Appendix Table A.16 for the definition of groups of causes of death.
Coverage: Metropolitan France.
52Our results reveal the progress (to the advantage of males) in the first year of life. While the decrease in infant mortality was similar for both sexes in relative terms, the higher rate for males (hence the larger decrease in absolute terms) helped to narrow the gender gap in life expectancy: the progress achieved in this one year of life account for 8.5% of the narrowing of the gender gap.
53The gender gap has also narrowed in all age groups and for all causes up to age 80 (except for “other diseases”) because female gains have been smaller than male gains for external causes (especially among young people and adults up to age 35) as well as for cancers (from age 40) and cardiovascular diseases (notably at ages 55-75). At ages 75-79, women have progressed less than men for cardiovascular diseases and certain infectious diseases, but more than men for cancers and “other diseases”. Last, from age 80, gains are systematically larger for women, especially for cancers and cardiovascular diseases, the two leading causes of death.
Mortality differences between the départements
54There are marked differences in mortality across the départements of metropolitan France, with a gap of 6 years between extreme values for men and 3.4 years for women. As in the past, mortality is higher in the Nord, Alsace and Brittany regions. It is quite low in Paris and in the départements to the south-west of the Paris region (Île-de-France), as well as in the Rhône-Alpes region, in the Midi-Pyrénées (mainly for men) and, for women, in the northern départements of the Poitou-Charentes region and in Pays de la Loire.
55As shown in the second article of this issue, which provides a detailed analysis of the ages at death and causes of death that account for these differences between départements (Barbieri, 2013), the geography of life expectancy at birth is now shaped by that of mortality above age 30. Among young people, children especially, the sharp mortality decline is associated with a notable reduction in geographical disparities. At working ages, man-made diseases (smoking, alcoholism and, for young adults, transport accidents and suicides) are the main causes of inequalities between départements. At ages 30-60, differences in cancer mortality explain most of the excess mortality in the most disadvantaged départements, while cardiovascular diseases play a major role beyond age 60, for women especially.
56On 1 January 2013, the population of France was 65.8 million, of which 63.7 million in metropolitan France, recording an increase that was due mainly to sustained natural growth (despite a slight slowdown in 2012). The relatively broad base of the population pyramid implies that population ageing is not occurring as fast as in other European countries affected by a severe and persistent fertility decline. Median age in France is forecast by Eurostat to reach 44.5 years in 2060 versus 39.8 years in 2010.
57The annual number of first residence permits (excluding short-term permits) granted to foreign adults over recent years (2006 to 2011) is stable, and totalled 160,142 in 2011. More than half the adults concerned are women (51.7% in 2011) and most are relatively young (more than 71.6% are below age 35).
58Fertility is stable (two children per women) but as the proportion of women of childbearing age is decreasing, the number of births fell slightly in 2012 (821,000 births in 2012 versus 823,000 in 2011). The timing of fertility is now changing less rapidly, and age-specific fertility rates could stabilize in coming years. The relative share of fertility at ages 35-39 is still increasing, however, while below age 20 and above age 40 the shares remain quite small. Overall, it is women in the 25-35 age group who contribute most to fertility. Births outside marriage are increasingly frequent, and 57% of children born in 2012 were registered to unmarried parents.
59The number of induced abortions remained stable in 2012, with little change in the total abortion rate. Age-specific rates were also stable. The increase in repeat abortions reflects the growing complexity and diversity of the sexually active periods of life.
60After years of steady decline, provisional figures suggest that the number of marriages increased slightly in 2012. Marriage is still highly seasonal, with a clear preference for weddings in June, July and August (except in Paris). Regarding PACS civil partnerships, in 2011 it became possible to register a PACS with a notary, so statistics on this type of union are incomplete and recent trends are difficult to determine. Once statistics become available from notaries, it will be possible to see whether or not the growing popularity of this type of union in the 2000s was confirmed in 2011 and 2012. For now, the sharp drop in the number of PACS unions published by the Ministry of Justice corresponds to a decrease in PACS registrations at magistrates’ courts. PACS dissolutions are growing in number, mostly due to separation by mutual consent.
61The number of deaths increased between 2011 and 2012 to a total of 569,868, of which 559,227 in metropolitan France. Progress in life expectancy slowed in 2012 due to a series of winter epidemics. However, this slight pause does not affect the general mortality patterns observed over the last two decades, with regard either to changes in the structure by age or cause of death, or to the narrowing of the gender gap in life expectancy.
AcknowledgementsThe authors wish to thank Elodie Baril and Arnaud Bringé from the INED Statistical Methods department for their help in preparing the database.
Age distribution of the population on 1 January (%)*
Age distribution of the population on 1 January (%)** Provisional.
Population: Metropolitan France.
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*
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** European Union member states + Iceland, Liechtenstein and Norway; enlargement from 14 to 24 countries from 2004; from 24 to 26 from 2007 with the entry of Bulgaria and Romania.
Pursuant to the Act of 26 November 2003, foreign nationals of the 14 old EU member states are no longer required to hold a residence permit. A provisional estimate of 40,000 admissions of these EU nationals from 2004 to 2007 was introduced to correct the resulting under-estimation. From 2005, figures are estimated from annual census survey data.
Cohort fertility: cumulative fertility up to selected ages, estimated completed fertility (mean number of children per 100 women), and mean age at childbearing (in years and tenths of years)*, **
Cohort fertility: cumulative fertility up to selected ages, estimated completed fertility (mean number of children per 100 women), and mean age at childbearing (in years and tenths of years)*, *** For the 1930-62 cohorts, observed completed fertility and mean age at childbearing; for later cohorts, unobserved rates are assumed equal to rates observed at the same age in 2011.
** For the 1930-62 cohorts, observed completed fertility and mean age at childbearing; for later cohorts, unobserved rates have been estimated by extrapolating the trend of the last 15 years.
Population: Metropolitan France.
Cohort fertility in Europe(1), (2)(1) Two estimates are proposed. One is based on rates that remain unchanged with respect to the last observation year, the other on a continuation of the trend at each age over the last 15 observed years.
(2) The series of published rates (2002-2010) cannot be used to calculate and estimate completed fertility.
Number of induced abortions and annual indices since 1976*, (1), (2), (3), (4)* Total abortions recorded by SAE and the CNAM-TS health insurance fund.
(1) Statistics from notifications including elective and therapeutic abortions.
(2) Administrative statistics based on recorded medical procedures. Data from 2010 includes data from the CNAM-TS and takes account of abortions covered by specific health insurance funds (MSA and RSI). Source: DREES and CNAM-TS from 2010.
(3) INED estimate (elective abortions). From 2002, the hospital statistics are considered exhaustive.
(4) Based on INED statistics up to 2001, and on hospital statistics from 2002.
Coverage: Metropolitan France.
Characteristics of nuptiality and divorce since 1985*, (1), (2), (3), (4)* Provisional.
(1) Ratios of number of first marriages to number of persons of same age, summed to age 49.
(2) Ratios of number of first marriages to (estimated) number of never-married persons at the same age, combined to age 49.
(3) Direct divorces and separations converted into divorces.
(4) Only PACS signed in court. Since April 2011 notaries have been empowered to register PACS unions. These PACS are not yet included in the statistics of the Ministry of Justice.
Note: The numbers of PACS reported in Population up to 2012 concerned the whole of France (including overseas départements). The numbers given in this table concern metropolitan France only (mainland France and Corsica).
Coverage: Metropolitan France.
Standardized death rates (per 100,000) by sex and groups of causes of death(1)(1) Standardized rate calculated from mortality rates by five-year age group (in completed years) and from standard European population (according to the structure proposed by the WHO). Thanks to a new analysis of INSERM data, the age groups now have the same definition for all years. The contents of the cause-of-death groups are defined in Table A.15 (item numbers refer to ICD-9 for 1980 to 1999 and ICD-10 from 2000).
Coverage: Metropolitan France.
Contributions of age groups and cause-of-death groups to gender differences in life expectancy in 1988-1990 and in 2008-2010 (in years)
Contributions of age groups and cause-of-death groups to gender differences in life expectancy in 1988-1990 and in 2008-2010 (in years)
Cause-of-death categories and the corresponding codes in the International Classification of Diseases (ninth and tenth revisions)
More detailed statistical data concerning Mayotte are currently being validated before they can be routinely produced and published (Bellamy and Beaumel, 2013). For this reason, the population of Mayotte is excluded from most of the updated statistical data in this year’s article.
The provisional estimate of net migration given by INSEE is +62,000. Whether or not the final estimate is positive, most of the total increase will still be due to natural increase.
Appendix Tables A.1 to A.16 are given at the end of the article. They are updated annually if new data becomes available. Their numbers do not always correspond to the order in which they are cited in the text.
The final figures are practically identical to the provisional figures published by INSEE and INED at the beginning of the year in their annual demographic reports (Bellamy and Beaumel, 2013; Pison, 2013).
Eurostat website consulted in July (update dated 18 July 2013) http://appsso.eurostat.ec.europa.eu/nui/show.do?dataset=demo_gind&lang=en.
http://epp.eurostat.ec.europa.eu/statistics_explained/index.php/Population_projections. For France, see Chardon and Blanpain (2010) and the animated population pyramids on the INSEE website.
Born abroad to non-French parents.
Member countries of the European Union on 30 June 2013, Vatican City State, Iceland, Liechtenstein, Norway, the principalities of Andorra and Monaco, the Republic of San Marina and Switzerland are excluded.
The census bulletin includes questions on the year of arrival in France and the place of residence one year previously which can be used to estimate “net entries”: persons present at the time of the census who were not living in France on 1 January of the previous year. INSEE conducts methodological studies on the estimation of entries based on these two variables.
Appendix Tables A.6 and A.7 present recent cohort data on fertility in Europe (see also Prioux and Barbieri, 2012).
According to data from the latest perinatal survey, 0.49% of pregnancies in 2010 were obtained using assisted reproductive technology (ART) (Blondel and Kermarrec, 2011), in half of cases by means of ovarian stimulation, and in the other half by ART with manipulation of gametes. Note that in most cases, the parents’ own gametes are used (Agence de biomédecine, 2011; Rozée and Mazuy, 2012).
From 2010, the data include the procedures covered by specific health insurance funds for the self-employed and farmers: the Régime social des indépendants (RSI) and the Mutuelle sociale agricole (MSA), so the data given in Appendix Table A.8 show a slight increase. Excluding terminations covered by these special regimes, 207,860 induced abortions were performed in metropolitan France in 2011, and 220,962 in the whole of France (Vilain et al., 2013).
The abortion rates were calculated from the detailed age distribution of induced abortions recorded in the PMSI (supplied to us by DREES), adjusted to cover all abortions (including RSI and MSA). The 2011 abortion notifications are not available at the time of writing. A history of the various data sources can be found in Rossier et al. (2009).
The contribution of groups of ages and of causes of death to differences in life expectancy at birth between the two periods or between the sexes is calculated in this article using the method proposed by Andreev et al. (2002).