Home Sociology  &  Culture  Journal Issue Article

Population

2005/4 (Vol. 60)

  • Pages : 200
  • DOI : 10.3917/popu.504.0443
  • Publisher : I.N.E.D

Email alert

Your alert request has been correctly taken into account.
You will receive an email when new issues are put online.

Close

Previous Pages 371 - 414 Next

Overall trends and age structure

The French population in the 2004 census survey

1

In January-February 2004, INSEE (the French National Institute for Statistics and Economic Studies) conducted the first annual survey since France introduced its redesigned, survey-based population census. On the basis of these initial results, INSEE revised the population estimates calculated since the 1999 census (Desplanques, Royer, 2005). At 1 January 2004, the population of metropolitan France (mainland + Corsica) is now estimated at 60.2 million inhabitants rather than 59.9 million, and the total French population (including overseas départements) at 62 million instead of 61.7 million. On the basis of this new assessment, INSEE raised the estimate of net migration in metropolitan France between 1 January 1999 and 1 January 2004 from +275,000 to +410,000. It also made an adjustment of 164,000 persons, or 33,000 per annum, to allow for the “difference in method” between the 1999 census and the 2004 survey [1]  For more details, see box in Desplanques and Royer...[1]. The 1999 census results had entailed a revision in the other direction, with a negative adjustment of 479,000 persons for the 1990-98 period [2]  For a description of adjustment models used after each...[2].

2

Most of the demographic indicators published in this review have been adjusted to take account of this revision of population figures between 1999 and 2004 (see Tables in Appendix).

An exceptional increase in 2004

3

In 2004, the population of metropolitan France grew by an estimated 364,000 to 60.56 million on 1 January 2005 (Richet-Mastain, 2005). This is the steepest increase observed in many years (Table 1) in both absolute and relative terms (6 per 1,000); not since the early 1970s have higher figures been recorded. The exceptional gains of 2004 are mainly due to the sharp fall in the number of deaths (508,500). In 2003—a year marked by the summer heatwave—550,000 were recorded, or 15,000 more than expected. The number of births edged up from 761,500 in 2003 to 767,800 in 2004. This generated an exceptional natural increase of 259,300, i.e. a rate of 4.3 per 1,000, nearly matching the 1987 and 1988 figures (Table 1). In addition to this large natural increase, France recorded net migration of +105,000 in 2004, a very slight rise from the previous year’s +100,000 (INSEE substantially revised the 2003 figure from the estimates published in 2004). Though French population growth is mainly driven by natural increase, the contribution of net migration is now about 30% (2003-04 average) [3]  Note, however, that net immigration is simply an estimate...[3]. According to Council of Europe data (2005) however, France is one of the only countries of Western Europe—with the Netherlands—whose growth is chiefly due to the excess of births over deaths. Even in Ireland, whose rate of natural increase is twice that of France, natural increase accounts for only half of total growth, since net immigration is also high.

The population continues to age slowly

4

As birth rates in the past five years have been higher than in the closing years of the twentieth century, the base of the population pyramid has widened slightly (Figure 1). Nevertheless, the proportion of young people aged under 20 continues to decline by 0.1 percentage point a year (Table 2); the proportion of persons aged 60+ has also been rising by 0.1 percentage point per annum for the past two years. As a result, the share of intermediate groups has remained temporarily stable at 54.3% since 2003. It is only in two years’ time that the first effects of the entry of the first cohort of baby-boomers (born in 1946) into the “60+” age group will be felt in France. This will trigger an inexorable decline in the proportion of persons aged 20-59.

Figure 1 - Age pyramid of the French population on 1 January 2005

Note: Birth-cohort sizes are available on the INED website (http:// www. ined. fr).

Source: INSEE
5

In sum, the shift in the age distribution shift towards older ages is still relatively slow. Thanks to a relatively low proportion of very old persons aged 85+ (the “depleted cohorts” are still very visible in Figure 1) and, above all, thanks to a sustained birth rate over the past twenty years, the proportion of persons aged 65+ is lower in France—at 16.3% in 2003—than in some other Western European countries. Belgium, Spain, Sweden, Greece, and Germany register percentages in the 17-17.5% range, with Italy scoring as high as 18.2% (Sardon, 2004).

Immigration [4]  The author thanks Xavier Thierry (INED) for supplying...[4]

Immigration continued to rise in 2003, but at a slower pace

6

In 2003, the number of new permits issued for residence in metropolitan France rose to 215,000 [5]  Number of persons who received for the first time a...[5] (221,000 including the overseas départements), from 206,000 in 2002, an increase of 4.7% (Table 3). After three consecutive years of growth exceeding 10%, the 2003 figure thus marks a sharp deceleration.

7

While the number of foreigners receiving residence permits continues to rise, the subset of persons authorized to take up residence in the year of their arrival in France shrank from 134,000 in 2002 to 127,000 in 2003 (last section of Table 3). This contrasts with the pattern in previous years, when the volume of new permits reflected a simultaneous rise in the number of current-year entrants and in the number of permits issued to persons having arrived in earlier years. Nevertheless, since 2002, France has seen an increase in the proportion of residence permits issued to foreign nationals who have arrived in preceding years. In 2003, that share reached 41%, a level similar to the one observed in 1998, a year in which large numbers of foreign nationals were “regularized”. But there is no evidence of longer application processing times or an increase in the proportion of long-term permits granted after the issuance of permits valid for less than one year. The gap between the year of entry into France and the year of issuance of residence permits thus corresponds to “de facto” regularizations.

8

The number of entrants from the European Economic Area (EEA) remained stable at 42,000. The year 2003 was the last in which an enumeration was possible on the basis of Interior Ministry data: as from 1 January 2004, France no longer requires EEA immigrants to obtain residence permits. In the ten years in which France was able to measure this component of international migration (1994-2003), the flows proved to be very stable. On the eve of European Union enlargement on 1 May 2004, citizens of the new Member States accounted for some 4,400 entrants, barely more than in 2002 (Table A).

Table A - Legal, long-term immigration, by group of nationalities and for nationalities most represented in 2003
9

Immigration from non-EEA countries totalled 173,000 persons, an increase of 6.4% compared with 16.3% in 2002 (Table 3). Of the total, 92,000 (53%) were admitted mainly because they had relatives in France. Only 8,000 non-EEA nationals were officially admitted as labour immigrants in 2003. Labour immigration is, however, structurally underestimated because family ties have become the predominant factor in decisions to issue residence permits since the suspension of labour immigration in 1974. In reality, an estimated one-third of non-EEA nationals authorized to reside in France find work shortly after arriving in France (Thierry, 2004). On the basis of our earlier calculations, an estimated 60,000 foreign nationals entered France in 2003 for work purposes, a higher figure than the estimate for the late 1990s [6]  Persons admitted for family reasons who enter the labour...[6]. For the first time since 1995, the number of students from outside the EEA is declining: 47,000 residence permits of at least one year’s duration were issued to this category in 2003. The drop was particularly steep for Moroccans (–35%).

10

There was little change in the ranking of the eight main nationalities that constitute half of the total immigration flow (Table A). By decreasing order, they are Algeria (32,600), Morocco (24,900), the United Kingdom (10,800), Tunisia (10,500), China (8,900), Portugal (7,800), Turkey (7,500), and Germany (6,000). The most significant changes in 2003 were the continued growth of North African immigration (+8%), with the notable exception of Morocco, whose numbers are declining, and the stabilization of flows from China and Turkey. Just before it became optional for EEA citizens to hold residence permits, flows from the U.K. and Portugal accelerated sharply, up 15% and 18% respectively [7]  For more detailed statistics, see Thierry (2004) and...[7].

From immigration to net migration

11

For the years 1999-2003—i.e. the last intercensal period—the combined total of residence permits issued each year indicates an immigration flow of 910,000 persons. This aggregate should not be confused with net migration, which takes departures of foreign nationals and flows of French nationals into account. As noted earlier, INSEE estimates net immigration at 410,000 for the same period.

12

The 500,000 gap between the two numbers can be explained by several factors:

  • After obtaining a one-year residence permit, some foreign nationals do not apply for renewal (Thierry, 2004);

  • An unknown proportion of foreign nationals return to their home countries after a longer stay or when they retire;

  • Net migration of French nationals appears to be increasingly negative. This is evidenced by the increase in the number of French persons who have settled in other countries, according to Foreign Ministry statistics (Gentil, 2003). For the 1999-2003 period, the Ministry estimates that the number of French nationals residing outside France rose by 350,000 from 1,774,000 to 2,125,000 (Duchêne-Lacroix, 2005) [8]  These statistics are partly based on estimates, as...[8].

Fertility

A small increase in births in 2004

13

According to the final figures, there were 767,800 births in France in 2004 [9]  The author thanks INSEE’s Demographic Surveys and Studies...[9], up 6,300 (or 0.8%) on 2003 [10]  One-third of the increase can be explained by the fact...[10]. As the mean number of women of childbearing age continues to decline slowly every year, the total fertility rate (TFR) rose by 1.3%. The TFR edged above 1.9 children per woman to 1.904 in 2004, up from just under 1.88 (1.879) in 2003 (Table B). INSEE’s adjustments to the population estimates after the 2004 census survey led to a small reduction in the TFR for 1999-2003 with respect to the values published earlier (Prioux, 2004, Table A). The monthly figures released by INSEE (2005) show that most of the increase in births occurred at year-end. While the first six months of 2004, particularly March-April-May, indicated a shortfall with respect to the previous year, the rise in births in the second half more than offset the January-June deficit. Preliminary estimates for January-June 2005 show no evidence yet that the uptrend is subsiding.

Binational couples sustain the birth rate

14

From 1998 to 2004, the proportion of children born in metropolitan France to at least one foreign parent rose from 14.5% to 18.2%. The largest increase concerned children born to binational couples, i.e. comprising only one foreign parent: their share grew from 7.8% of total births in 1998 to 10.9% in 2004. The proportion of children born to two non-French parents increased much less sharply, from 6.7% to 7.3%. In 2004, 55,700 children were born to two foreign parents (up from 49,200 in 1998), 84,000 to binational couples (up from 57,900 in 1998), and 628,100 to all-French couples (down from 631,000 in 1998 [11]  The number of children born to two French parents peaked...[11]). Among children of binational couples, those with a foreign father and a French mother slightly outnumber those with a French father and a foreign mother, at 57% versus 43%.

15

The increase in births to foreign parents is consistent with rising immigration flows, as well as with the increase in marriages between two foreign nationals, and—most significantly—in binational marriages [12]  A very large majority of these children are born to...[12] observed in recent years (Prioux 2003 and 2004). Unsurprisingly, fathers and mothers of Algerian and Moroccan nationality form the largest group of foreign parents [13]  Unfortunately, the statistics by nationality do not...[13], but the number of parents from sub-Saharan Africa has grown sharply since 1997. Statistics indicate that the arrival of immigrant women in France—very often a consequence of marriage or family reunion—is followed by a period of high fertility (Toulemon, 2004; Tribalat, 2005).

16

Because of naturalization, however, the number of children born to at least one foreign parent is smaller than that of children born to at least one immigrant parent. For the 1991-98 period, the data from the 1999 Family History Survey (EHF) suggest that the latter group accounted for about 17% of births (Toulemon, op. cit.). For the same period, civil registration statistics indicate that the proportion of children born to at least one foreign parent can be estimated at 14.5%.

17

From partial data published by INSEE [14]  Until 1997, when the father’s nationality was unknown...[14], we can reconstruct the series for the proportion of children of foreign nationals born in France since 1980, so as to view the recent pattern in a longer-term context (Figure 2). In 1980, approximately 13% of children were born to couples comprising at least one foreign parent. This proportion has gradually risen, with spells of sharp increases (1980-83, 1987-93) followed by declines (1983-87, 1993-97). The uptrend accelerated steeply between 1997 and 2004, in particular from 2000 to 2004. The discontinuity is due to the often contradictory trends in the two components of this proportion: while births to binational couples rose consistently throughout the period, births to foreign couples declined in two sub-periods (1983-87, 1993-97), followed by two stable spells. Since 2000, the proportion of births to two foreign parents has risen slightly, explaining the faster growth of the overall proportion.

Figure 2 - Change in proportion of children born to at least one foreign parent since 1980 (per 100 live births)
Source: author’s calculations and estimates based on civil records (Beaumel et al., 2005).

A slower increase in age at childbearing

18

As in 2003, the rise in fertility in 2004 was confined to women over 30 years old (Table B). On the other hand, while fertility among the youngest women had started falling again in 2002-03, there was no further decline in 2004. Overall, since the late 1990s, fertility in the 15-27 age group has stabilized, while it continues to rise among the over-27s (Table 4). As a result, the mean age of childbearing is still increasing, but far more slowly than before. Since 2000, it has been gaining about half a month each year, compared with an average of one month a year in the 1990s and one and a half months in the 1980s. Thanks to this inflection in the trend, though the total fertility rate (TFR) still under-estimates completed cohort fertility, the gap is becoming narrower (Table 5).

Table B - Age-specific fertility since 1999 (per 1,000 women)
19

The mean age of childbearing in 2004 was 29.6 years (Table 4). This value should not be confused with age of first childbearing, which is about two years younger. Because of the poor quality of birth-order data collected when births are registered, no accurate calculation of the mean age of first childbearing is available for recent years [15]  The statistics published by INSEE overestimate the...[15].

20

The proportion of children born outside marriage continues to rise, reaching 46.4% in 2004 versus 45.2% in 2003. The contribution of births outside marriage to the TFR is therefore increasing. In 2004, it was 0.89 children per woman, or 46.8% of total fertility (last columns of Table 4).

France in Europe

21

In 2003 and probably 2004 as well, France’s fertility continued to rank second in the European Union behind Ireland (Table 6). The other countries of Western Europe generally fall into two very distinct groups:

  • The northern region, where the TFR ranges between 1.7 and 1.8 children per woman (in increasing order: United Kingdom and Sweden, Netherlands, Denmark and Finland, Norway) and is trending up everywhere, most notably in Sweden and the Netherlands;

  • The central and southern regions, where the TFR ranges from 1.27 (Greece) to 1.44 (Portugal). For the past few years, the indicator has been clearly on the rise in Spain (1.3 in 2003) and Italy (1.29), and, to a lesser extent, in Austria (1.38). In Greece, Germany (1.34), and Switzerland (1.41), fertility has remained stable in recent years. In Portugal, it is still on a mild downtrend.

Only two countries bridge the gap between the two groups: Belgium (1.61) and Luxembourg (1.63).

22

As for the ten new countries that joined the European Union in 2004, their fertility places them in the second group rather than the first. Three of them rank in the upper tier or slightly above the group: Estonia, with 1.37 children per woman in 2002, Malta, with 1.46 in 2003, and Cyprus with 1.50. The seven others are clearly positioned at the lower end (Lithuania, Latvia, and Hungary: from 1.26 to 1.29) or even below the lowest value (Czech Republic, Slovenia, Slovakia, and Poland: from 1.18 to 1.22).

23

French fertility has another key characteristic: the proportion of births outside marriage is among the highest in Western Europe. While the proportion of children born out of wedlock has stopped increasing in Sweden, Norway, and Denmark, it continues to rise in France, which is catching up with Denmark and edging closer to 50%. In Norway, one in two children is already born outside marriage. In Sweden, the proportion has reached 56% [16]  Iceland, which has a long tradition of births outside...[16].

24

France’s sustained high level of fertility compared with some of its neighbours is probably due—at least in part—to this rapid rise in fertility outside marriage. The phenomenon has largely offset the decline in births to married couples ensuing from the fall in the marriage rate. In Western Europe, fertility tends to be higher in countries where the percentage of births outside marriage is high (Figure 3a). But the correlation between the two indicators is not perfect. While the explanation seems satisfactory for France, it is not so for countries such as Austria, Germany, and Spain, whose TFRs are much lower than might be expected from their proportions of births outside marriage. As for the former Eastern-Bloc countries, they do not fit into the pattern at all (Figure 3b). Apart from Serbia and Macedonia, the TFR is very low everywhere and seems totally unrelated to the proportion of births outside marriage. The latter is almost as variable as in the West, ranging from 10% in Croatia to 56% in Estonia.

Figure 3a - TFR (per woman) by percentage of births outside marriage (BOM) in Western Europe, 2003
Source: Council of Europe (2005).
Figure 3b - TFR (per woman) by percentage of births outside marriage (BOM) in former Eastern-Bloc European countries, 2003
Source: Council of Europe (2005).
25

Thus the low fertility of some European countries is not solely due to a stronger social resistance to births outside marriage.

Completed cohort fertility is decreasing

26

When the population revisions after the 2004 census survey are taken into account, some age-specific fertility rates in certain cohorts are reduced, but not enough to have a real impact on cumulative fertility and estimated completed fertility (Table 5).

27

The relatively high level of fertility in recent years has not been sufficient to offset the delay accumulated by the cohorts born after 1960 in completing their fertility. The completed fertility of the 1961-64 birth cohorts has fallen swiftly (by an average 2 points per 100 women per cohort), lowering completed fertility from 212 children per 100 women in the 1960 cohort to 204 in the 1964 cohort. After age 30 [17]  The youngest cohort (1964) is observed until age 4...[17], the fertility of these cohorts was not high enough to make up for the deficit accumulated before age 30, compared with previous cohorts (Table C). Note that this deficit, mostly due to the rise in age of childbearing, was aggravated by the years of low fertility (1992-97) lived through by these cohorts at the ages when fertility is still normally high (28-32 years).

Table C - Cumulative fertility at ages 28-40 for women in the 1960-74 birth cohorts, and differences between successive cohorts (average number of children per 100 women)
28

After the 1964 cohort, the decrease in completed fertility slows down gradually. The decline will peak at one child per 100 women between successive cohorts, and might even cease entirely with the 1970 cohort. This is because the delays accumulated at early ages with respect to previous cohorts are steadily narrowing (Table C), and a full catch-up before age 45 is still possible in the 1970 and 1972 cohorts. Though the projections reported here—based on a continuation of past trends—suggest a mild decrease, a more vigorous recovery of fertility over ages 32 or 34 remains highly likely. In later cohorts (1974), the delay becomes negligible and, absent a new drop in the TFRs, the stabilization of completed fertility should be confirmed.

29

The decline in completed fertility might thus be confined to the cohorts born between 1960 and 1970. This would keep the TFR from falling below two children per woman.

30

However, the decline changes France’s ranking in Western Europe by placing Norway in second position starting with the 1965 cohort (Table 7). Apart from France and Norway, the only country where completed fertility still exceeds two children per women is Ireland, where it has dropped by 23 points per 100 women in five generations. The closest levels are found in the three other Nordic countries (from 1.91 to 1.98 children per woman). The lowest values—between 1.49 and 1.65—are in the Latin countries (Italy and Spain) and Germanic countries (Germany, Austria, and Switzerland). Greece, the Netherlands, Belgium, Luxembourg and Portugal post intermediate levels ranging from 1.75 to 1.82.

31

The ranking is therefore different from that of the TFRs, and the levels are higher. The reason is that the age of childbearing is rising everywhere, entailing a deficit in the TFRs. The faster the increase in the age at childbearing, the larger the deficit.

Abortions

How many induced abortions in 2002?

32

For the first time since 1997, induced abortion statistics were recorded and tabulated by the statistical services of the French Ministry of Health for the year 2002. While the ministry previously collected between 160,000 and 170,000 abortion notifications each year [18]  Except in 1995 due to a shortage of notification forms...[18], it received only slightly above 130,000 in 2002 (Table 8). The data collected by the ministry via the medical statistics database (PMSI, programme médicalisé des systèmes d’information) or the annual statistics of healthcare facilities (SAE, Statistique annuelle des établissements de santé) give much higher abortion statistics however. This suggests that abortion notifications are now submitted on a much less systematic basis, due probably to a lack of motivation on the part of the medical teams and healthcare facilities concerned, given that no statistical analyses were performed for several years.

33

We saw in this same section last year (Prioux, 2004) that, according to the estimates of the ministry’s statistical services based on the PMSI and the SAE, the number of induced abortions has been increasing since 1995 (from 179,600 in 1995 to 205,600 in 2002, according to Vilain, 2004). Yet it would appear that this apparent increase is due to improvements in the statistical coverage of these data sources. A team of INED researchers conducted an in-depth study to understand the reasons for this divergence between the various figures given by the ministry (PMSI and SAE) and the number of abortion notifications, and put forward a new estimate of the number of induced abortions, both in 2002 and in the previous years (Rossier, Pirus, 2005). The results are given in the last column of Table D.

Table D - Number of induced abortions since 1976 according to different sources
34

The number of induced abortions appears to be marginally lower than previously estimated, though the direction of change of the two estimates is similar: after falling in the 1980s, the number of induced abortions has stabilized since the 1990s to an estimated total of around 207,000 per year. It would also appear that the ministry figures based on the SAE and the PMSI, which for many years under-estimated the number of induced abortions, are now somewhat more reliable. Nevertheless, these latest statistics cannot be used to analyse in detail the characteristics either of the women who resort to induced abortion or of the terminated pregnancies (conjugal situation, occupation, number of children, gestational age, etc.), since the information collected is limited in scope [19]  Only the woman’s age, the technique used, the type...[19]. So we must continue to rely on abortion notifications to understand why women opt for induced abortion. For example, they provide the only means to assess the consequences of the two-week increase in the gestational age limit for legal abortion (from 10 to 12 weeks) introduced by the law of 2001.

Rising frequency of induced abortions among women aged 20 to 27

35

The estimated 207,000 induced abortions in 2002 correspond to a total abortion rate of 0.51 abortions per woman (Table 8). This rate, which is the sum of age-specific abortion rates (ratio of the number of induced abortions among women of a given age to the number of women of the same age [20]  The number of induced abortions at each age is estimated...[20]), does not mean that 51% of women have an abortion, since some have several abortions during their lifetime. If we count only first-time induced abortions, we can calculate on the basis of behaviours observed in 2002 that the proportion of women who will have at least one abortion in their lifetime is 38%, consistent with the ongoing downtrend since the 1980s. On the other hand, the second component of this total abortion rate, the sum of repeat abortion rates, appears to be remarkably stable (0.12 to 0.13 abortions per woman), reflecting the very stable frequency of repeat abortions [21]  The second abortion component in 2002 (0.10 per woman)...[21].

36

Though the frequency of induced abortion has remained generally stable since 1997—on the basis of 207,000 abortions in 1997 and in 2002, with a total abortion rate of 0.50 and 0.51 per woman respectively—the distribution of age-specific rates has changed slightly: below age 28, and in particular between age 20 and 27, abortion rates have increased, while beyond this age they have tended to decrease, especially between ages 33 and 38 (Figure 4a). The abortion curve continues to shift towards younger ages as the age of childbearing increases. Moreover, if we observe the ratio, at each age, of the number of abortions to the number of pregnancies that end either with a live birth or with an induced abortion (which gives a reliable indication of the proportion of pregnancies terminated by an abortion), we note that this ratio has decreased quite sharply among women aged between 35 and 45 (Figure 4b). On the other hand, it remains relatively stable below this age, signalling that the previous uptrend in the number of pregnancy terminations at these ages is now at an end. This stabilization can be linked to the stabilization of fertility at these ages. All these changes are closely correlated with changes in fertility timing. Last, the lowest abortion rates are observed among women aged 28 to 32 (around 13 abortions for 100 pregnancies at these ages).

Figure 4a - Age-specific abortion rate in 1997 and 2002, and by age and abortion order in 2002 (for 1,000 women)
Source: author’s calculations based on abortion notifications, Rossier and Pirus (2005) and registration records.
Figure 4b - Abortions per 100 pregnancy outcomes (live births + induced abortions) by age at outcome in 1997 and 2002
Source: author’s calculations based on abortion notifications, Rossier and Pirus (2005) and registration records.

Unions and union dissolutions

A steeper fall in marriages in 2004

37

The provisional figure for marriages celebrated in France in 2004 is 259,400, down 6% from 276,000 the previous year (Table 9). The number of marriages has thus fallen back to a level close to that of the years 1993-95, which saw the lowest figures of the second half of the twentieth century (about 255,000). Since 2000, when 297,200 marriages were recorded, the number has slipped by 38,500 or 12.9%, and the 2004 drop contrasts with the milder fall of earlier years.

38

Final statistics are now available for 2003. As in 2002, the decline mainly concerned marriages of never-married persons. As in 2001 and 2002, the only category to decrease was that of marriages between two French nationals; marriages involving at least one foreign spouse continued to increase, reaching 20% of the total in 2003. The proportion of marriages contracted in order to “legitimize” [22]  Since the reform of filiation in July 2005, the term...[22] children has levelled off at 28% since 2001. After the two sharp rises in 1996-97 and 2000, this proportion has stopped increasing.

The growing success of the PACS (civil union)

39

The number of civil unions known as PACS (pacte civil de solidarité) continued to rise. Almost 40,000 (39,576) were signed in 2004, up from 31,161 in 2003. This represents an increase of 27%, slightly above the 25% recorded in 2003. The fastest increase was recorded in the final quarter of 2004 (Figure 5): one-third of the PACSs for the year were signed between October and December, probably due to the abolition of the three-year waiting period to qualify for joint taxation. The number of new PACSs in the first quarter of 2005 dropped steeply, though this is usually the time of year when the largest number are signed. The atypical pattern shows that the end-2004 spike consisted mostly of PACSs formed a few months ahead of schedule; absent a change in tax legislation, those civil unions would have been signed in the first quarter of 2005 [23]  However, as with marriages, joint taxation applies...[23]. But the PACS remains as popular as ever. Though the second quarter is traditionally a slack period, a record 15,723 PACSs were signed in April-June 2005.

Figure 5 - Quarterly change in number of PACSs registered and dissolved
Source: French Ministry of Justice, SDSED.
40

Absent detailed statistics on partners (age, sex, marital status, nationality, number of children in current union, etc.), it is very hard to assert that the PACS is competing with marriage. The PACS was introduced as a form of legal union for same-sex and cohabiting couples who did not want to commit to marriage. In fact, it is proving about three times as vulnerable as marriage (Prioux 2004). However, a small percentage of dissolutions—10% on average—is due to the fact that the partners marry: in a much smaller number of cases, the PACS probably represents the first step before a genuine commitment, unless such unions merely anticipate already-scheduled marriages!

Nuptiality of the never-married still declining

41

The release of the revised population figures has led to a downward revision of certain nuptiality indicators for the never-married (Table 9). They largely concern total marriage rates. For women, the total rate has fallen from 0.60 first marriages per woman in 2000 to just 0.53 in 2004 (provisional estimate). But these shifts are too small to significantly modify the total cohort rates which give the characteristics of nuptiality by birth cohort (Table 10). Among women, the proportion of ever-married at age 50 continues to fall by an average of one percentage point from one cohort to the next. The proportion declined from 75% in the 1965 cohort to 69% in the 1971 cohort, while mean age at first marriage rose by almost two years from 26.3 to 28.2. Among men, the proportion of ever-married declined by an identical six points between the 1963 cohort (73%) and the 1969 cohort (67%), and the mean age posted a barely smaller increase of 1.7 years from 28.3 to 30.

42

In later cohorts, the proportion of young ever-married men and women (ages 24 and 22 respectively) is still falling, but at a slower pace. It now seems to have bottomed out at around 6%—approximately one-fourth of the value recorded for the male 1963 cohort and the female 1965 cohort.

Increasingly fragile unions

43

The drop in formal marriages is known to be mainly due to the rise in non-marital cohabitation. In the 1970s, about one couple in six started living together without waiting to marry. Today, the proportion is nine out of ten couples (Toulemon, 1996), and more than 95% of partnerships that are not first unions begin outside marriage (Prioux, 2005).

44

Likewise, an ever smaller percentage of non-marital unions end in marriage. Some last for years without being officialized, while others are dissolved. Thanks to the 1999 Family History Survey (Étude de l’histoire familiale, EHF), we can clearly observe this decline in marriage among cohabiting couples. Among women who began their first unions outside marriage around 1975, one-half had married within two years; among women who started their unions a decade later (1985), the proportion of those who married within two years had fallen to one-third; for unions formed around 1995, the percentage dropped to under one-fifth (19%) (Prioux, op. cit.).

45

Unions begun outside marriage tend to be more fragile than direct marriages (Table E). Among women who formed their first partnerships around 1980, 11% of the unions that started outside marriage—whether or not they later led to marriage—were dissolved within five years, versus 5% of direct marriages. The respective percentages of dissolutions within ten years were 22% and 12%. Subsequently, both categories of unions registered a large increase in dissolutions. As a result, the overall frequency of dissolutions in the first five years of union almost doubled from 8% for unions formed in 1980 to 15% for unions formed in 1990. In sum, the growing fragility of unions is due to a combination of the rise in non-marital cohabitation, as non-marital unions remain more vulnerable than marital ones, and the increasing fragility of marriages that are not preceded by cohabitation.

Table E - Percentage of first unions begun by women ca. 1980 and ca. 1990 already dissolved within 5 and 10 years, by type of union formation

Significant rise in divorce in 2003

46

With or without premarital cohabitation, marriage has become shorter-lived and divorce has been increasing. Between 1995 and 2001, however, the number of divorces fell slightly, auguring a stabilization of the dissolution rate. But the increase has resumed in recent years (Table 9). In 2002, it rose by a moderate 3% to 116,000, including direct divorces and conversions of separation into divorce. In 2003, the number surged 8% to a record 125,000. As the number of marriages exposed to the risk is, if anything, trending down, the total divorce rate, which had levelled off at about 38 divorces per 100 marriages between 1995 and 2001, topped 39% in 2002 and rose to 42.5% in 2003. This is the first time that the indicator has crossed the symbolic threshold of 40 divorces per 100 marriages.

47

Figure 6 illustrates the change in the probability of divorce by marriage duration in successive marriage cohorts. It shows an almost continuous increase from one cohort to the next at all durations. However, the shape of the curves has changed considerably, reflecting a major shift in the risk distribution by marriage duration. In the older marriage cohorts (1960 and 1965), the divorce risk remained practically the same between approximately 5 and 20 years of marriage, and then gradually declined. In later cohorts, the risk rose much more sharply for short durations. As a result, the highest risk is now located around the fifth year of marriage. In the 1990 cohort, 2.4% of surviving marriages were dissolved 5 years after the year of union formation, and a barely smaller percentage for proximate durations (4 and 6 years of marriage); from then on, the probability of divorce diminishes with union duration. Nevertheless, the curves do not intermingle. Whatever the increase in risk for the initial marriage durations from one cohort to the next, couples who stay married are more likely to divorce than couples in older cohorts with the same marriage duration. Only couples married after 1990 seem to want to break with the trend toward increasingly frequent, early divorce. However, the recent rise in the divorce rate suggests that this moderation was short-lived.

Figure 6 - Probability of divorce by marriage duration in selected cohorts (per 1,000 surviving marriages)
Source: author’s calculations based on statistics of the Ministry of Justice, SDSED.
48

With the renewed rise in the probability of divorce, the cohort-specific frequency of divorce cannot be accurately predicted until behaviours have become stable again. If the total divorce rate were to stabilize above 40 divorces per 100 marriages, this 40% frequency would certainly be exceeded by couples who married in the 1990s.

An increasing number of men and women live alone

49

The 2004 census survey found that 8.3 million people were living alone, 900,000 more than in the 1999 census. They account for 14% of the total French population (Borrel, Durr, 2005). The fall in mortality and the better health status of older people naturally contribute to the rise in the number of elderly men and women living alone. Half of women living alone—and one-fifth of men living alone—are aged over 65. But the changes in marital behaviour have become the most important factor in the increase in one-person households. Figure 7 compares the age-specific proportions of men and women enumerated as living alone in censuses since 1990. The changes are less significant between 1999 and 2004 because of the shorter interval, but the shifts observed between 1990 and 1999 have persisted, i.e. an increase in the proportion of men living alone between ages 20 and 70, and of women living alone between ages 20 and 65. At young ages, the phenomenon is chiefly due to the rise in the age at first cohabiting partnership and the increase in residential independence. At older ages, the main factor is union dissolution. When a couple with children breaks up, the man is more likely than the woman to go off and live alone if he does not form a new couple immediately. This explains the greater increase in the proportion of men aged 30-50 living alone. The rise is much smaller—practically imperceptible—among women in the same age group, as they are more likely to form a single-parent family after union dissolution. By contrast, the increase is more clearcut between ages 50 and 65, since unless they have formed a new couple, women generally live alone after their children have left home.

Figure 7a - Age-specific percentage of women living alone in 1990, 1999, and 2004
Source: INSEE, population censuses
Figure 7b - Age-specific percentage of men living alone in 1990, 1999, and 2004
Source: INSEE, population censuses
50

By their fifties (age 53 and over), women are more likely to be living alone than men. The likelihood increases with age, at least until age 86-87, when the proportion peaks at nearly 57%. This is the classic effect of greater female longevity and the age difference between partners. Another factor is that women are less likely than men to form a new couple after a union dissolution (Delmeyre, 2005).

51

Living alone does not necessarily mean that a person is not in a couple relationship however. A question on partner relationships was introduced for the first time into the 2004 individual census schedule. A small proportion of persons enumerated as living alone responded that they were living in a couple relationship. The proportion increases with age, especially for men (Table F). Although we know that this situation has not always been freely chosen [24]  The partner’s absence may be due to work-related reasons...[24], the size of these proportions shows that the concept of the couple is changing, and that people do not always need to share the same dwelling to consider themselves as living in a couple. Moreover, this is probably a transitory situation for those who are only beginning their relationship. However, the census data do not indicate which of these couples have deliberately chosen not to share the same dwelling, and to “live apart together” [25]  In Scandinavian countries, such couples are designated...[25] on a long-term basis.

Table F - Percentages of men and women who report living in a couple among those living alone in 2004

Mortality

A sharp decrease in mortality in 2004

52

After an exceptional year in 2003, with 15,000 excess deaths due to the heatwave, 2004 proved exceptional for the opposite reason: a total of 508,500 deaths were registered (provisional figure), 41,500 fewer than in 2003. Not since the early 1960s has the figure been so low, because even though mortality is decreasing, the number of deaths is tending to rise due to population ageing. The crude death rate, which had varied between 9.0 and 9.2 deaths per 1,000 inhabitants for the last ten years, fell to a record low in 2004 (8.4 per 1,000) (Table 1). Life expectancy at birth thus reached 83.8 years for women and 76.7 years for men, a leap of 0.9 and 0.8 years respectively, after falling by 0.1 year for women and increasing by 0.1 year for men in 2003. This lowered the ranking of France in that year, notably for women, who were no longer in first position (Tables 11 and 12). Such large variations in the number of deaths and the mean length of life had not been observed in more than thirty years, though up to the early 1970s they were not unusual, due mainly to influenza epidemics. Since the development of a flu vaccine available free of charge to elderly people, the impact of influenza epidemics has become much more limited and the annual number of deaths much less variable.

53

The mortality decline in 2004 is much larger than expected. If the heatwave had simply hastened by several months the deaths of persons already in poor health—a phenomenon known to epidemiologists as the “harvest effect”—the number of deaths would have not decreased by so much (–15,000 instead of the –25,000 [26]  This is the difference with respect to expected deaths...[26] observed), and life expectancy at birth would have increased by only around 0.6 to 0.7 years, i.e. 0.2 to 0.3 years less than the increase actually observed (Richet-Mastain, 2005; Toulemon, Barbieri, 2005).

54

An analysis of data by département conducted by Toulemon and Barbieri shows that only around 5,000 “missing” deaths in 2004 can be attributed to this harvest effect. According to these authors, though a part of this difference (7 to 10,000 deaths) can be accounted for by the absence of an influenza epidemic in 2004 [27]  During the winter of 2003-04, the influenza epidemic...[27], the remainder can only be explained by a decrease in mortality of elderly people linked to better prevention [28]  Moreover, according to these authors, a small fraction...[28]. Observation of mortality in 2005 will tell us whether this progress is set to continue, while detailed analysis of 2004 data by age and cause of death will provide a clearer picture of the situation.

Mortality of the elderly

55

Under the mortality conditions of 2003 (the last year for which detailed age-specific data is available), 73% of men and 87% of women could expect to live up to the age of 70 at least, and those who reached the age of 70 could expect to live for a further 13.6 and 17.1 years respectively. Ten years earlier, the life expectancy at this age was 12.6 years for men and 16.3 years for women, so the increase is substantial (respectively 1 year and 0.8 year), and only slightly lower than that observed in the previous decade. Increasingly, it is the progress observed among the elderly which explains the increase in the length of life, for women especially: if the period 1993-2003 is divided into two five-year periods, the weight of progress achieved at age 70 or above—i.e. the the proportion of years of life gained thanks to the decrease in mortality at these ages—has risen from one-fifth in the first period to one-third in the second for men, and from 49% to 58% for women. A recent study analysed the main causes of this decrease in mortality among the elderly (Meslé, 2005).

56

Over the decade 1990-2000, as over the previous decade, the progress achieved at these ages is due above all to the decline in mortality from cardiovascular diseases (Table G). All cardiovascular pathologies are now in decline, including mortality from ischaemic heart disease, which followed an upward trend for many years. Nevertheless, progress slowed down between the two decades. Cardiovascular diseases are still the leading cause of death among people aged 70 and over.

Table G - Contribution of the seven major cause-of-death groups to the increase in life expectancy at age 70 from 1980 to 2000 (years and hundredths of years)
57

The second cause is cancers, which are set to become the leading cause of death at these ages for men, since their cancer mortality is much higher than that of women. Moreover, for both sexes, less progress has been achieved in this area than for cardiovascular diseases. However, the trend is increasingly favourable for men. These gains in life expectancy can be attributed above all to the continuous decline is stomach cancer mortality (for both sexes) and, in cancers of the uterus for women, though there has also been a recent stabilization or a slight decrease in cancers of the intestine for both sexes, in prostate, bronchial and lung cancer among men and in breast cancer among women. On the other hand, mortality is continuing to rise for bronchial and lung cancer among women and for cancers of the blood and blood-forming organs in both sexes.

58

Respiratory diseases are the third cause of death at these ages. Mortality from this group of causes is continuing to fall, thanks mainly to the less devastating effect of influenza epidemics, though there is little decrease in mortality from chronic bronchitis (linked mainly to tobacco consumption) among men, and no sign of a decline among women.

59

The next cause is violent death, whose contribution to the decline in mortality at these ages has increased slightly. Thanks to improved prevention and treatment, mortality from falls is decreasing, as is mortality from motor-vehicle accidents and suicide.

60

Finally, though France now lags well behind Japan, which has the world’s highest life expectancy at age 70, its position is still very favourable compared with other industrial countries, notably the USA and the Netherlands (Meslé, 2005).

Infant mortality

61

Unlike mortality of the elderly, changes in infant mortality have an ever smaller impact on the increase in life expectancy at birth. Over ten years, from 1993 to 2003, the infant mortality rate nevertheless fell by almost 40%, from 6.5 deaths per 1,000 newborns to 4 per 1,000 (Table 11). But the level of infant mortality is now so low that this decrease has increased the length of life by only 0.19 years on average for both sexes, and represents just 9% of the progress achieved over the same period (2.11 years).

62

According to provisional data, the infant mortality rate fell below the symbolic threshold of 4 per 1,000 for the first time in 2004. In Europe, only a few Nordic countries (Finland, Norway and Sweden) and, more recently, Spain, have infant mortality rates below this level (Table 13).

Overview

63

Following the 2004 census survey, INSEE has readjusted its estimates of the population of metropolitan France and of net migration since 1999.

64

In 2004, a year marked by a sharp drop in the number of deaths and a small increase in births, natural increase was relatively high. France is one of the few European countries whose population is rising primarily due to natural growth. However, though birth rates have been fairly high in the last five years, the age distribution is continuing to shift slowly towards older ages.

65

Immigration increased again in 2003, though at a slightly slower pace than in 2002. Citizens of the European Economic Area represent a stable figure each year (42,000). The largest flows are from North Africa (mainly Algeria and Morocco), and the main reason for admission is the presence of relatives in France. Family immigration is often followed by a birth in France, and the number of children born to parent(s) of foreign nationality has increased in recent years.

66

The total fertility rate increased slightly to 1.9 children per women in 2004, and the mean age of childbearing continues to rise slowly (29.6 years). Completed fertility drops sharply after the 1960 cohort however, and may fall below 2 children per women from the 1970 cohort. With 2.03 children, the completed fertility of the 1965 cohort will be below that of Norway.

67

According to a new estimate, the number of induced abortions has remained relatively stable, at 207,000 per year, corresponding to a frequency of 0.51 abortions per woman. Rates are increasing among the youngest women (below age 28) and decreasing among the oldest.

68

While the number of civil unions (PACS) continues to increase from one year to the next, the number of marriages has been falling steadily since 2000. The proportion of ever-married men and women at age 50 decreases from one cohort to the next, while the mean age at first marriage is increasing.

69

The number of cohabiting couples who marry is decreasing, and first unions are increasingly fragile: in ten years, the frequency of union dissolution during the first five years of cohabitation has almost doubled (from 8% for women whose first union began in 1980 to 15% of those whose union began in 1990), and the frequency of dissolution has increased, both for women who married directly and for those whose union began outside marriage. The number of divorces (which represent the legal termination of marriages only) is increasing again, after several years of stability.

70

The increase in the age at first stable union and the higher frequency of union dissolutions has led to an increase in the proportion of men and women living alone: from 1999 to 2004, these proportions continued to increase among women aged between 20 and 65 (above all between 50 and 65) and among men aged 20 to 70 (above all 30-50).

71

Mortality was exceptionally low in 2004. The life expectancy of women leapt by 0.9 years, and that of men by 0.8 years. This drop in mortality cannot be explained solely by a “harvest effect” following the exceptional mortality in 2003 due to the August heatwave, or by the absence of an influenza epidemic in 2004.

72

Mortality among the elderly is continuing to decline, and this decrease is largely responsible for the increase in mean length of life, for women in particular. Cardiovascular diseases, the leading cause of death at these ages, are the main determinants of this decreasing mortality among the elderly.

The Statistical Data

Table 1 - Population change (in thousands) and crude rates (per 1,000)(a)
Table 2 - Age distribution of the population of metropolitan France on 1 January (%)
Table 3 - Legal long-term immigration by registration status
Table 4 - Fertility since 1970
Table 5 - Cohort fertility: cumulative fertility to selected ages, estimated completed fertility (mean number of children per 100 women), and mean age of childbearing (in years and tenths of years)
Table 6 - Total fertility rates for western europe (average number of children per woman)
Table 7 - Cohort fertility in Western Europe
Table 8 - Number of abortions and annual indices since 1985
Table 9 - Characteristics of nuptiality and divorce since 1985
Table 10 - Characteristics of nuptiality by birth cohort
Table 11 - Characteristics of overall mortality since 1985
Table 12 - Life expectancy at birth in Western Europe in 2003
Table 13 - Infant mortality in Western Europe (rate per 1,000 live births)
Table 14 - Standardized death rates by sex and groups of causes of death(a)
Appendix Table - Cause-of-death groups and the corresponding items in the International Classification of Diseases (ninth and tenth revisions)

REFERENCES

  • Beaumel C., Richet-Mastain L., Vatan M., 2005, La Situation démographique en 2003. Mouvement de la population, Insee résultats, Société n°41, 45 p. + CD-ROM.
  • Bergouignan C., Blayo C., Parant A., Sardon J.-P., Tribalat M. (eds.), 2005, La population de la France. Évolutions démographiques depuis 1946, CUDEP, Vols. I and II, 884 p.
  • Borrel C., Durr J.-M., 2005, “Enquêtes annuelles de recensement : premiers résultats de la collecte 2004. Principales caractéristiques de la population et des logements”, Insee première, n° 1001.
  • Council of Europe, 2005, Recent Demographic Developments in Europe, 2004, Council of Europe Publishing, 128 p. + CD-ROM.
  • Delmeyre Y., 2005, “Histoire des unions dans les cohortes”, in Bergouignan et al. (eds.), La population de la France. Évolutions démographiques depuis 1946, CUDEP, Vol. I, pp. 143-168.
  • Desplanques G., Royer J.-F., 2005, “Enquêtes annuelles de recensement : premiers résultats de la collecte 2004. 62 millions d’habitants en France au 1er janvier 2004”, Insee première, n° 1000.
  • Duchêne-Lacroix C., 2005, “Les Français établis hors de France : une population méconnue et en transformation”, in Bergouignan et al. (eds.), La population de la France. Évolutions démographiques depuis 1946, CUDEP, Vol. II, pp. 847-858.
  • Gentil B., 2003, “La population française immatriculée à l’étranger est en forte hausse”, Insee première, n° 919.
  • Héran F., Toulemon L., 2005, “What happens when the census population figure does not match the estimates?”, Population and Societies, No. 411, April 2005.
  • Insee, 2005, Bulletin mensuel de statistique, n° 8.
  • Meslé F., 2005, “Espérance de vie et mortalité aux âges élevés”, Retraite et société, n° 45, pp. 90-113.
  • Prioux F., 2003, “Recent demographic developments in France”, Population-E, 58(5), pp. 525-558.
  • Prioux F., 2004, “Recent demographic developments in France”, Population-E, 59(5), pp. 595-634.
  • Prioux F., 2005, “Mariage, vie en couple et rupture d’union sous l’angle de la démographie”, Informations sociales, n° 122, pp. 38-50.
  • Richet-Mastain L., 2005, “Bilan démographique 2004. Nette diminution des décès”, Insee première, n° 1004.
  • Rossier C., Pirus C., 2005, “Combien d’interruptions volontaires de grossesses en France? Une analyse de sources de données divergentes”, Poster paper at the 25th International Population Conference, Tours, 18-23 July 2005.
  • Sardon J.-P., 2004, “Recent demographic trends in the developed countries”, Population-E, 59(2), pp. 263-314.
  • Thierry X., 2004, “Recent immigration trends in France and elements for a comparison with the United Kingdom”, Population-E, 59(5), pp. 635-672.
  • Toulemon L., 1997, “Cohabitation is here to stay”, Population: an English Selection, 9, pp. 11-46.
  • Toulemon L., 2004, “Fertility among immigrant women: new data, a new approach”, Population and Societies, No. 400.
  • Toulemon L., Barbieri M., 2005, “The mortality impact of the August 2003 heat wave in France”, Paper at the 25th International Population Conference, Tours, 18-23 July 2005.
  • Tribalat M., 2005, “Fécondité des immigrées et apport démographique de l’immigration étrangère”, in Bergouignan et al. (eds.), La population de la France. Évolutions démographiques depuis 1946, CUDEP, Vol. II, pp. 727-768.
  • Vilain A., 2004, “Les interruptions volontaires de grossesse en 2002”, Études et résultats, n° 348.

Notes

[*] Institut national d’études démographiques, ParisTranslated by Jonathan Mandelbaum

[1] For more details, see box in Desplanques and Royer (2005).

[2] For a description of adjustment models used after each census, see also Héran and Toulemon (2005).

[3] Note, however, that net immigration is simply an estimate based on incomplete information about immigrant flows.

[4] The author thanks Xavier Thierry (INED) for supplying all the background material for this section.

[5] Number of persons who received for the first time a residence permit valid for a period of one year or more.

[6] Persons admitted for family reasons who enter the labour market within a year are reclassified as persons entering France for work purposes.

[7] For more detailed statistics, see Thierry (2004) and the INED website: http:// www. ined. fr/ englishversion/ figures/ france.

[8] These statistics are partly based on estimates, as the number of persons registered in consulates (57% of the total) rose by only 230,000.

[9] The author thanks INSEE’s Demographic Surveys and Studies Division for supplying the 2004 birth statistics, the 2004 provisional marriage rates, and the tables of INSEE’s 2003 demographic report (Situation démographique de l’année 2003) before publication.

[10] One-third of the increase can be explained by the fact that 2004 was a leap year, since around 2,100 children are born each day on average.

[11] The number of children born to two French parents peaked at 657,600 in 2000.

[12] A very large majority of these children are born to married couples: among binational couples, the proportion is 59%; among foreign national couples, 78.5%.

[13] Unfortunately, the statistics by nationality do not make it possible to distinguish binational couples from the rest.

[14] Until 1997, when the father’s nationality was unknown (a situation that mostly concerned children born outside marriage), INSEE did not adjust the figures for nationality.

[15] The statistics published by INSEE overestimate the number of first births and the age of mothers at first birth, as many children registered as first-born are likely to be children born in new unions to mothers who have had at least one child in a previous union.

[16] Iceland, which has a long tradition of births outside marriage (64% in 2003), is not included in the comparison.

[17] The youngest cohort (1964) is observed until age 40.

[18] Except in 1995 due to a shortage of notification forms in the Paris region.

[19] Only the woman’s age, the technique used, the type of health facility, the date and the département where the abortion took place are recorded.

[20] The number of induced abortions at each age is estimated using the distribution by age of notifications received in 2002. We were able to verify that, although incomplete, the statistics provided by the notifications can be considered as representative of all women who undergo induced abortion.

[21] The second abortion component in 2002 (0.10 per woman) corresponds to a frequency of 10% of women having at least two abortions in their lifetime.

[22] Since the reform of filiation in July 2005, the term has been removed from the French Civil Code.

[23] However, as with marriages, joint taxation applies only to the period after PACS formation, not to total 2004 income. To discourage people from setting up PACSs solely for tax purposes, the law stipulates that partners who dissolve their PACS too quickly—in the same or following year—are ineligible for joint taxation, unless the union is terminated because the partners decide to marry or because one partner has died.

[24] The partner’s absence may be due to work-related reasons (assignment abroad), medical reasons (long-term hospital stay) or other causes (e.g. imprisonment).

[25] In Scandinavian countries, such couples are designated as LAT (living apart together).

[26] This is the difference with respect to expected deaths in 2004 and not with respect to observed deaths in 2003.

[27] During the winter of 2003-04, the influenza epidemic occurred in November and December 2003, while the following winter it arrived later, affecting France from mid-January to mid-March 2005.

[28] Moreover, according to these authors, a small fraction of the mortality decline (that of young adult men) is linked to the increase in tobacco and alcohol prices. The decrease in road deaths may also be a factor, though the largest drop was observed in 2003.

Abstract

English

In 2004, a year marked by a sharp drop in the number of deaths and a small increase in births, natural increase was relatively high. France is one of the few European countries whose population is rising primarily due to natural growth.
Immigration increased again in 2003, though at a slightly slower pace than in 2002. The total fertility rate increased slightly to 1.9 children per women in 2004. Completed fertility drops sharply after the 1960 cohort however, and may fall below 2 children per women from the 1970 cohort. The number of induced abortions has remained relatively stable.
Though the number of civil unions (PACS) is still increasing, the number of marriages has been falling since 2000. The proportion of ever-married men and women at age 50 decreases from one cohort to the next, while the mean age at first marriage is increasing. The frequency of union dissolution (divorces and separation of unmarried couples) has increased considerably. These changes in conjugal behaviour are raising the number of adult men and women who live alone.
Life expectancy registered an exceptional increase in 2004 (+0.9 years for women and +0.8 years for men), and this cannot be explained solely by a “harvest effect” following the exceptional mortality in 2003 due to the August heatwave, or by the absence of an influenza epidemic in 2004. Mortality among the elderly is continuing to decline, and this decrease is largely responsible for the increase in mean length of life, for women in particular.

Outline

  1. Overall trends and age structure
    1. The French population in the 2004 census survey
    2. An exceptional increase in 2004
    3. The population continues to age slowly
  2. Immigration4
    1. Immigration continued to rise in 2003, but at a slower pace
    2. From immigration to net migration
  3. Fertility
    1. A small increase in births in 2004
    2. Binational couples sustain the birth rate
    3. A slower increase in age at childbearing
    4. France in Europe
    5. Completed cohort fertility is decreasing
  4. Abortions
    1. How many induced abortions in 2002?
    2. Rising frequency of induced abortions among women aged 20 to 27
  5. Unions and union dissolutions
    1. A steeper fall in marriages in 2004
    2. The growing success of the PACS (civil union)
    3. Nuptiality of the never-married still declining
    4. Increasingly fragile unions
    5. Significant rise in divorce in 2003
    6. An increasing number of men and women live alone
  6. Mortality
    1. A sharp decrease in mortality in 2004
    2. Mortality of the elderly
    3. Infant mortality
  7. Overview

To cite this article

France Prioux, “ L'évolution démographique récente en France ”, Population 4/2005 (Vol. 60) , p. 443-487
URL : www.cairn.info/revue-population-2005-4-page-443.htm.
DOI : 10.3917/popu.504.0443.


Previous Pages 371 - 414 Next
© 2010-2014 Cairn.info
back to top
Feedback

Message sent

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