- Email: [email protected]
Late fertility in Europe: some comparative and historical data
© Masson, Paris, 2005.
Rev Epidemiol Sante Publique, 2005, 53 : 2S3-2S11
Late fertility in Europe: some comparative and historical data La fécondité tardive en Europe : quelques données historiques et comparatives F. PRIOUX Institut National d’Études Démographiques (Ined), 133, boulevard Davout, 75980 Paris Cedex 20. Email : [email protected] (Reprints: F. Prioux).
The issue: Given the substantial increase in late fertility (at ages 35 and older) in the last few years in Western countries, the paper raises the question of what factors have caused this new trend, and how likely it is to continue. Methods: Late fertility is measured using the sum of age-specific fertility rates at 35 or more; its contribution to the total fertility rate (in % of the total fertility rate) and the proportion of first childbirths in this fertility is then computed. Late fertility among men is assessed using the same method, where the necessary data are available. The data analysed sometimes date back to the 18th century for women (Sweden), and the early 20th century for men (in France). Results: Late fertility is now much lower than it used to be. Despite the current increase — which can be traced back to the early 1980s for most Western countries, and appears to be much more recent and limited in the former socialist countries — current levels are very low compared with those of the past, among men as well as women. Reduced family size was the main factor in that decline, which became more pronounced as a result of women having a first child at younger ages. The reversal of the trend is due only to postponed childbearing, which must be viewed as part of a general trend towards postponement of all demographic events. More and more of the children born in such circumstances are thus first children. Very late fertility (at 40 or more) has increased, though only slightly, for men as well as women, and age 45 seems to be a limit beyond which few are prepared to have children. Social norms concerning the age limit for a person to have a child could account for the differences from one country to another in that respect. Conclusion: Late fertility is likely to continue increasing as long as fertility continues to decrease among young people. However, the increase will continue to be due mostly to the “younger”age groups (under 40 for women, under 45 for men); both biological constraints and social norms probably play a part in the stagnation of very late fertility levels. Fertility at age 35 or more. Europe. Postponement of first childbearing. Female fertility. Male fertility. Social norms.
Position du problème: La fécondité tardive (35 ans ou plus) ayant beaucoup augmenté ces dernières années dans les pays occidentaux, l’article s’interroge sur la cause de cette nouvelle tendance, et sur ses chances de se poursuivre encore. Méthodes: On mesure la fécondité tardive par la somme des taux de fécondité par âge à 35 ans ou plus, et on calcule sa contribution à la fécondité totale (en % de l’indicateur conjoncturel), et la part des premières naissances dans cette fécondité. La fécondité tardive des hommes est évaluée de la même manière, lorsque les données le permettent. Les données analysées pour les femmes remontent parfois jusqu’au XVIIIe siècle (Suède), et pour les hommes au début du XXe siècle (en France). Texte reçu le 22 mars 2004. Acceptation définitive le 10 mai 2005.
2S4
F. PRIOUX
Résultats: La fécondité tardive est aujourd’hui beaucoup plus faible qu’autrefois. Malgré la hausse actuelle, qui remonte au début des années 1980 pour un grand nombre de pays occidentaux, et qui est beaucoup plus récente et timide dans les pays anciennement socialistes, les niveaux actuels sont très faibles comparés aux niveaux anciens, aussi bien pour les hommes que pour les femmes. La réduction de la taille des familles a été le principal moteur de cette baisse, qui a été accentuée par le rajeunissement de l’âge à la première maternité. Le retournement de tendance n’est dû qu’au retard des maternités, retard qui fait partie d’un mouvement général de décalage de tous les calendriers démographiques. De plus en plus souvent, ces naissances tardives sont ainsi des premières naissances. La fécondité très tardive (40 ans ou plus) augmente très peu, et, pour les hommes comme pour les femmes, l’âge de 45 ans semble une limite que peu d’entre eux franchissent. Des normes sociales concernant l’âge limite pour avoir un enfant pourraient expliquer des différences entre pays à cet égard. Conclusion: Il est probable que la fécondité tardive poursuivra sa hausse tant que la fécondité des jeunes continuera à diminuer. Mais, cette hausse portera toujours essentiellement sur les âges les plus « jeunes » (moins de 40 ans pour les femmes, moins de 45 ans pour les hommes); contraintes biologiques et normes sociales jouent probablement toutes deux un rôle dans cette stagnation de la fécondité très tardive. Fécondité après 35 ans. Europe. Retard des premières naissances. Fécondité des femmes. Fécondité des hommes. Normes sociales.
Over the last twenty years, the number and proportion of babies born to older mothers (age 35 or more) have increased almost everywhere in Europe. For example, in France, 134,200 children were born in 2000 to mothers aged 35 or more — 17.3% of the total number of births; in 1980 this number was only 47,000 — 5.6% of the total. In twenty years, the number of children born to older mothers was thus multiplied by 2.9, while the number of women in the corresponding age group was multiplied by only 1.4. Late fertility has thus risen sharply, and France is certainly not the only country in such a situation. That the number of childbirths comparatively late in the mothers’lives should be increasing is a matter of curiosity, or even concern, for the medical world, because the risks for the life and health of both mother and child are heightened as the mother grows older. This, however, does not come as a surprise to demographers, who think that such an evolution is only logical: postponement of childbearing began in the early 1970s, and it seems logical that the decline in early fertility should be followed by a rise in late fertility, otherwise the total number of children would have declined by the same proportion. The question for demographers is the following: to what extent, if fertility among young women continues to decrease, can fertility among older women increase to compensate for the decline.
SOME HISTORICAL LANDMARKS
The recent evolution of fertility at age 35 or more should first be studied within a long-term historical framework, by observing those countries where long-established statistical series are available. By comparing recent and older data, the current levels can be viewed in a relative perspective. Late fertility can be measured by summing the age-specific fertility rates of women aged 35 or more, expressed as an average number of children per woman. The proportion of late fertility in the total fertility of a country in a given year can also be measured by comparing that sum with the total fertility rate (TFR) for the year under study. It should be noted, however, that this does not provide an indicator of fertility timing, since the TFR depends both on the evolution of timing and the intensity (the completed fertility) of fertility within cohorts. This proportion of late fertility is simply a measure of the contribution of women aged 35 or more to the annual fertility of a country. In the mid-18th century, the figure for late fertility was 1.5 children per woman in Sweden, which represented one third of the total fertility of the country. In about 1860-1870, a period when reliable data first became available in several European countries [1], the figures for late fertility were 2.1 children per woman in the
TRENDS IN LATE FERTILITY
Netherlands, 1.8 in Norway, 1.7 in Finland and Sweden, and only 0.9 children per woman in France (fig. 1). As we know, fertility in France began to decline from the late 18th century [3], as a result of early dissemination of traditional methods of birth control. This indicates that birth control was probably practised more intensively among women aged 35 or more, as childbearing at those ages represented only one quarter of the total fertility, as against one third in Finland, 38% in Sweden, and over 40% in Norway and the Netherlands (fig. 2). From the late 19th or early 20th century, there was a widespread, rapid decline in late fertility, For 1 woman
2,5 Netherlands
2,0
2S5
and in the late 1930s, the levels ranged from 0.3 children per woman in France, to 0.7 in the Netherlands. The trend was briefly interrupted by the baby-boom period, when fertility among women aged 35 and older increased again everywhere in Europe, and even quite significantly in Norway and the Netherlands. The decline subsequently resumed, and by the late 1970s, late fertility figures were a mere 0.1 children per woman in the Netherlands, and only slightly higher in the other countries under study; in all five countries, the contribution of late fertility to total fertility rates was under 10%. Late fertility has picked up once again, though it is now very far below its former level. Ranging between 0.25 and 0.28 children per woman, it represents approximately one sixth of the TFR of the countries under study.
Norway
A GENERAL TREND IN WESTERN EUROPE, AFFECTING MOSTLY WOMEN AGED UNDER 40
Sweden Finland
1,5
1,0
France
0,5
0,0 1750
1800
1850
1900
1950
2000
FIG. 1. — Late fertility from the 18th century in Western Europe (the sum of age-specific fertility rates at ages 35 and older, per woman). Sources: [1] and [2].
%
50 Netherlands
40 Norway
Sweden Finland
30 France
20
10
0 1750
1800
All the countries in Western Europe are currently experiencing a new rise in late fertility, though it is more recent in some countries than in others. It is in Southern Europe that the trend has reversed most recently (between 1986 and 1990), and also in Ireland, where a renewed rise in late fertility has been observed since 1994 only. And yet, fertility at age 35 and above was not especially low in Ireland — 0.36 children per woman, more than twice the minimum level in Spain (0.17), and four times the minimum levels in Germany, Belgium and Denmark (0.09). The minimum value generally ranges from 0.1 to 0.15 child per woman.
1850
1900
1950
Year
2000
FIG. 2. — Contribution of late fertility to total fertility rates from the 18th century in Western Europe (the proportion of the aggregate age-specific fertility rates of women aged 35 and older in the total fertility rate). Sources: [1] and [2].
The most significant factor in the decline of late fertility has been the gradual and virtually total disappearance of large families [1] as a result of birth control, at first by means of traditional methods, and later using “modern”, more efficient methods (pill, IUD, sterilization, legal abortion) [4]. As a consequence, the average age of mothers fell to its lowest level in Western Europe in the mid-1970s (and a little later in Southern Europe). This corresponds to a period when family building probably occurred earliest in the whole of history. Marriage — at the time a virtually compulsory precondition — and the
2S6
F. PRIOUX
birth of the first children occurred at particularly young ages, as did completed fertility, which was usually limited to 2 children [5]. Though late fertility has recently started to increase again, this is certainly not due to a new rise in the number of large families. It is mainly due to an increase in age at childbearing [6, 7]. The various aspects of transition to adulthood now all occur later [8]. The time for completion of education, entry into the labour market and formation of a stable couple, are stages in life that people reach at increasingly later ages, and these factors all contribute to postponing the birth of children. Nevertheless, it seems that women cannot postpone childbearing indefinitely. It is mostly between ages 35 and 40 that fertility is rising (fig. 3): in many countries, the fertility of women aged 35-39 is now the same, or even higher than in the early 1970s, and now usually ranges between 0.20 and 0.25 children per woman. Germany, Austria and Greece are exceptions, with fertility remaining comparatively low (0.15 or less in 2000). In Ireland, by contrast, fertility is much higher (0.39). The rise in fertility beyond age 40 is currently still quite small, and fertility at such ages still represents, everywhere except in Ireland, fewer than 0.05 children per woman (fig. 4), or 2 to 3% of total fertility. And childbirths at ages over 45 represent only a very small part of overall fertility beyond age 40 — a bare 3 to 5%, approximately 1‰ of total fertility1 — and the trend does not appear to be rising yet. Age 45 thus seems to represent a boundary beyond which few women are prepared (or able) to go. According to a survey recently conducted in France, the average age from which women consider it desirable not to have any more children is 41 [9]; for more than 4 in 10 women the age is 40, and while 1 in 20 women gave an age over 45, 1 in 6 set the limit at 35 or less2. Social standards thus appear to match behaviour in France. We do not know 1
According to [2], the sum of fertility rates of women aged 45 or more is 1 child per 1,000 women in Germany, Spain, Norway and the Netherlands; it ranges from 1 to 2 children per 1,000 women in France and Sweden, from 2 to 3 in Finland, and 3 per 1,000 in Ireland. 2 Data kindly provided by Laurent Toulemon.
0,5
0,4 Spai n 0,3
Ireland
Netherlands
0,2 France
0,1 Germany 0,0 1960
1970
1980
1990
2000
FIG. 3. — Fertility of women aged 35-39 in several Western European countries since 1960 (sum of age-specific fertility rates of women aged 35-39, per woman). Source: [2].
whether surveys on this theme have been conducted in other European countries3. CONTINUED LOW LEVELS OF LATE FERTILITY IN EASTERN EUROPE
Up to the 1990s, there was a significant contrast between the ages of mothers on either side of the Iron Curtain — in the socialist countries (as defined by their present borders4), mothers were 2 and a half years younger on average than in Western countries [10]. Due to very early childbearing, and families seldom having more than 2 children, late childbirths were rare up to the late 1980s. Moreover, the widespread decline of fertility observed in all the countries of the region since the fall of the communist regimes, has further strengthened this trend. Late fertility was consequently particularly low — under 0.1 child per woman virtually everywhere (except in Poland), or even under 0.05, as in Bulgaria and the former 3 A question on this theme was included in the 1995 Eurobarometer survey, but the results do not appear to have been published. 4 Except for the former German Democratic Republic, which we have included among this group of countries, because fertility behaviours were — and still are — very similar to Eastern ones.
TRENDS IN LATE FERTILITY
0,5
0,4
2S7
per woman Irl
0,4
0,3
0,3
0,2
Ireland
Fl
0,1
0,2
GB
Sw PB GFR
Dk
Spain 0,1
0,0 1970
Netherlands France
1970
1980
Nor Aust Belg Hun
Gr
Pol
Lit Cro Est Latv Slova Slove Rum Czek Rus Bulg GDR
1990
Year 2000
FIG. 5. — Minimum value of late fertility, and the years when that value was reached in Europe (sum of age-specific fertility rates at ages 35 and older, per woman). Source: [2].
Germany 0,0 1960
Spain Por Serb
It
Swed Fr
1980
1990
2000
FIG. 4. — Fertility of women aged 40 and older in several Western European countries since 1960 (sum of age-specific fertility rates of women aged 40-49, per woman). Source: [2].
rapid — the proportion of first children born to older women has increased more than twofold since the 1980s, rising from 15% to over one third.
German Democratic Republic (GDR). It was only in the 1990s5 that the fertility of women aged 35 and older began to rise timidly again in all those countries, much later than in the Western countries (fig. 5). As a consequence, late fertility is still very rare (ranging from 0.06 child per woman in Bulgaria, to 0.15 in Estonia), and represents 5 to 10% of total fertility .
As the proportion of higher order births tends to increase with age, the proportion of primaparous women should always be higher among women aged 35-39 than among those aged 40 or more. This is the case in the Netherlands for example, where first births represent 27% of fertility of the first age group, as against 23% of the second. In Spain and France, however, the proportions tend to be the same in both age groups, and in Sweden, the proportion of primiparous women is even slightly higher among the older group: in 20007, the proportion of first births was 30% at ages 35-39 and 33% at ages 40 and older.
A GROWING PROPORTION OF PRIMIPAROUS WOMEN
In Western Europe, between one-fifth and onethird of women who give birth at age 35 and over are primiparous, whereas half a century ago, the proportion was probably seldom above 10%. In the Netherlands, for which a long data series is available, the proportion of primiparous births at age 35 and over was only 6% in the 1950s and into the mid 1960s. Today, the proportion is 26% (table I). In France, the proportion was still under 9% in the late 1960s, it is now probably over 20%6. But it is in Spain that the rise was most 5 Hungary is an exception, having reached a minimum as early as 1984. Poland is another exception, where the minimum seems to have been reached only in 2000.
6 This is based on data from the 1999 Study of Family History survey, kindly provided by Laurent Toulemon. We do not use the statistics published by the Insee, which were based on birth registration data recorded since 1998, because the proportion of first children, particularly among older women, was considerably overestimated. This is due to the fact that municipal registry offices in France still complete the forms using the “livret de famille” booklet given to each family at the time of marriage, or of the birth of the first child (of the couple, or of the mother if she does not live in a couple). As a result, the first child of a subsequent union is usually categorized as a woman’s first child, though she may have had children from a previous union. 7 The mean figure from 1999 to 2001 was used here, with a view to reducing the risks of errors due to exceedingly low rates at those ages.
2S8
F. PRIOUX
TABLE I. — Change in the proportion of first order births in late fertility and in total fertility (ratio of aggregate first order fertility rates to aggregate fertility rates for all birth orders at corresponding ages). Sources: Spain, Hungary and the Czech Republic: the data were kindly provided by J.P. Sardon (European Demographic Observatory/INED); France: from the 1999 Study of Family History survey; data kindly provided by L. Toulemon (INED); Netherlands and Sweden: the author’s own calculations, based on national statistics. Proportion of first order births (%)
Country
1960
1980
2000
14.6
34.3
France
8.4
a
13.4
19.6b
Hungary
13.4
18.1
17.9
Netherlands
5.9
14.0
26.1
22.2
30.6
14.8
16.0
17.3
Spain Late fertility (aged 35 and older)
Sweden Czech Rep. Spain Total fertility (at all ages)
40.5
53.0
France
34.2
a
41.1
42.2b
Hungary
44.4
47.5
44.2
Netherlands
30.4
42.4
47.9
42.6
44.3
44.8
46.9
Sweden Czech Rep. a
45.3
1965, b 1998.
In a growing number of cases, these very late births appear to be the consequence of a woman’s desire to have a first child before she becomes infertile. At first, the growing proportion of late fertility first births was simply due to a general decrease in the number of higher order births (i.e. 3 or above). As a result, the proportion of first children born to women of all ages (lower part of table I) duly increased, including those born to women aged 35 and older (upper part of the table). Since the 1980s however, the main factor underlying this trend has been an increase in firstbirth fertility of older women: in the Netherlands, for example, the proportion of women who have a first child at age 35 and older (9 children per 100 women in 2002) is currently 5 to 6 times higher than in 1980 (1.5 per 100), and approximately twice as high as in 1950 (5 per 100). Similar trends are also observed in Spain, France and Sweden (fig. 6). Nothing of the kind has been observed in Eastern Europe as yet, where late first-birth fertility remained peculiarly low up to the 1990s. For
example, it was only 1 child per 100 women in the Czech Republic, and 1.5 per 100 women in Hungary. It is currently just beginning to increase (fig. 6). The proportion of primiparous women
per 100 women 10 Netherlands
8
6 Spain France
4 Sweden
Hungary
2
0 1950
Czech Rep.
1960
1970
1980
1990
2000 Year
FIG. 6. — Evolution of late first-birth fertility (sum of agespecific first-birth fertility rates at age 35 or over, per 100 women). Source: see table I.
TRENDS IN LATE FERTILITY
among those giving birth at age 35 and over thus appears to be still under 20% (table I).
1,4
2S9
per woman or man
1,2
A CONCURRENT INCREASE IN LATE FERTILITY AMONG MEN
1,0 Men aged 35 or +
0,8
Statistical data on the ages of fathers are seldom available, if ever. The fact that so little attention is paid to male fertility results partly from a difficulty inherent to nature, as fathers cannot always be identified. However, even in those countries where legal recognition of a child is compulsory, statistics on the ages of fathers are not always available. For example in Sweden, it is only since 1992 that the statistics on the ages of fathers have covered the birth of all children, not just those born in wedlock, as formerly, and it is only since 1999 that male fertility indicators by age group have been computed. Denmark is the only country where age-specific male fertility rates have been published more or less regularly since 1980, but the rates are only computed up to age 49, which makes any specific study of late male fertility impossible. In France, where the Insee has recently reconstructed records [11], a long data series covering the whole of the 20th century is now available (fig. 7). As could be expected, considering that the parents’ages are fairly well correlated, the pattern of late male fertility is quite similar to that of women, though the level is definitely higher, because fathers are usually older than mothers8. At the beginning of the century, the level of late male fertility was 1.3 children per man, 42% of the total male fertility rate at the same date (3.13 children per man). The trend was subsequently quite similar to that of female fertility at the same ages, with each period of war being followed by a rise in late fertility of variable duration. By the time the decline came to an end, in 1978, the level was a bare 0.35 child per man, 19% of the total male fertility rate. At that point, the level of late male fertility began to rise again, until it reached 0.59 in 2001, almost one third of the total 8 However, the method used by the Insee, in which childbirths are classified by age (the age reached in the year) results in slightly overestimating late fertility, as compared with figures 1 to 4, where the data in all the countries concerned the age of the mother at the time of childbirth.
0,6 Women aged 35 or +
0,4 Men aged 40 or +
0,2
Men aged 45 or + Women aged 40 or +
0,0
1900
1920
1940
1960
1980
2000 Year
FIG. 7. — Late male fertility in France since 1900. A comparison with late female fertility (sum of age-specific fertility rates, from ages 35, 40 and 45, per man or woman). Source: [11] and [12].
male fertility (1.86 children per man), or twice as high as female fertility at those ages (0.31). In the early 1980s, the level of fertility among men aged 40 or older was very similar to that of women aged 35 or more. However, the subsequent rise was much slower because male fertility beyond age 45 has stopped increasing since 1990 — it has been 0.08 per man since then, 4% of the total male fertility rate. It thus appears that among men too, postponed parenthood has led to a rise in fertility, particularly before age 40, as the contribution of older men to the rise has been very small up to now. Given the rise in family recomposition, in which men, more than women, can build a new family at later ages if they form a union with a woman younger than themselves, a more substantial rise in male fertility beyond age 45 could have been expected. The limits of male physiological capacities at those ages, as can be inferred from past statistics, have not been reached, far from it: at the beginning of the century, male fertility at age 45 and older was 0.3 children per man, 10% of the total. So there is plenty of scope for a further increase in male fertility at much older ages. The standards concerning the age beyond which it is no longer desirable for a man to have children do not entirely tally with this stagnating male fertility beyond age 45: though the mean
2S10
F. PRIOUX
age mentioned by men is 44.59 [9], more than 1 man in 10 has no opinion on the question, and among those who do express their views, 3 in 10 give an age over 45 (age 50 mostly)10. How do things stand in other countries? In Denmark male fertility at ages between 35 and 50 fell to a level below that of France in the early 1980s, and though it did rise again, the rise seems to have levelled off since 1994. The current level is 0.45 children per man, as against 0.56 in France at the same ages11. In Sweden, in 2001, the level of male fertility after age 35 was also lower than in France (0.49 vs 0.59), particularly for very late fertility (after age 45 or 50). Even though levels are currently tending to rise, particularly among younger men, higher levels of late male fertility could have been expected in those two countries, considering that separation and repartnering are more frequent than in France. TOWARDS AN INCREASING PROPORTION OF OLDER PARENTS
Statistical data on the ages of both parents are also very rare. For many years, only legitimate births were taken into account, and tables on the crossed ages of the parents only became available in national statistics very recently. Moreover, as no rates can be computed in this area, our observation will concern only the proportion of late childbirths in the total numbers. Obviously the proportion depends on behaviours, but it is nevertheless influenced by the age structure of the population of fertile age. In 2000 the proportions of late childbirths from the point of view of both parents were similar in Sweden and in France (table II): approximately 1 in 7 or 8 children had parents both aged 35 or more, and 1 in 15 had a father aged at least 40 and a mother aged at least 35; in Hungary, where late fertility is still very rare, the proportions are 9 In women’s view, the mean age limit for a man to father a child is 45. 10 Data provided by Laurent Toulemon. 11 Though the definition of age used here in France (age attained) entails very slight overestimation of late fertility, as compared with the definition used for Denmark and Sweden (the father’s age at the time of the birth of the child), those differences are very significant.
lower by half, as is probably the case in most former socialist countries. In countries that had a baby-boom period followed by a decline in fertility, the population of fertile age is currently “ageing”, and this entails more late childbirths, in circumstances of unchanged behaviour. The general trend in which motherhood and fatherhood are postponed until later ages simply increases the proportion of late childbirths, from the point of view of either the father, the mother or both parents, in the total numbers of childbirths12. AN OVERALL PERSPECTIVE
As a combined result of reduced family sizes and earlier fertility, first-birth fertility especially, late fertility seemed destined to disappear. In the early 1980s, a few years later in Southern Europe, and in the 1990s in the former socialist countries, the levels were strikingly low. The recent increase in late fertility results not from any resurgence of large families, but from a general postponement of all demographic events. The increase, while still very slight in the countries of former East Europe, is very rapid in Western Europe. On the other hand, the proportion of late fertility first births has increased, at first as a consequence of reduced family size, and later mostly because the proportion of women who have their first child after age 35 is increasing very fast in Western countries. However, there has been only a slight increase in fertility after age 40, as if this age barrier was a difficult hurdle to cross. And yet, comparing the present situation with older data suggests that there is still substantial scope for increased fertility at those ages. In the past, however, late childbirths, which occurred mostly in large families, were the result of hyper-fertile women bearing a last, more or less desired child. Nowadays, women who want a child after age 40 are not necessarily the most fertile ones; in fact they may even be less fertile than the others because they less often have child-bearing experience. Even 12 As mentioned in the introduction to this text, in France the increased numbers of women aged 35 or more have also largely contributed to increasing the numbers of late childbirths since the 1980s.
TRENDS IN LATE FERTILITY
2S11
TABLE II. — The proportion of late childbirths in 2000, from the point of view of the mother, the father and both parents (for a total of 100 births). Sources: National statistics, after due distribution of undeclared ages. Countries France
a
Hungary Sweden a
b
b
Mother aged 35 or more
Father aged 35 or more
Father aged 40 or more
Both parents aged 35 Mother aged 35 or +, or more father 40 or +
17.3%
31.5%
11.6%
13.9%
6.8%
8.0%
20.8%
8.3%
6.2%
3.6%
17.0%
32.8%
11.7%
13.2%
6.2%
age attained in the current year; b age attained at the time of the child’s birth.
without any selection effect, a recent simulation by H. Leridon [13] shows that the chances of conceiving a child are substantially reduced among women aged 40 compared with those aged 35, and that assisted reproduction techniques cannot compensate for the handicap. On the other hand, the levels of very late male fertility (at ages 45 and older) have remained surprisingly unchanged, despite the supposedly favourable effect of family recomposition. In that respect, however, we need to verify whether the proportion of men who form new unions at those ages with much younger women is actually increasing. Moreover, differences remain in Europe as far as very late fertility is concerned, which suggests that there are social norms relating to the ages when a man or a woman should stop having children. Late fertility is thus likely to continue increasing, as long as fertility among young people pursues its decline, and this increase will continue to concern mostly the “younger” groups (under age 40 for women, under 45 for men). Both biological constraints and social norms probably play a part in this stagnation of very late fertility. REFERENCES 1. Festy P. La fécondité des pays occidentaux de 1870 à 1970. Paris, Ined-PUF, Cahier n°85, 1979. 2. Recent Demographic Evolution in Europe, Strasbourg, Council of Europe Publishing, 2003.
3. Henry L, Blayo Y. La population de la France de 1740 à 1860. Population, 30 special issue, November 1975: 71-122. 4. Leridon H. La seconde révolution contraceptive. Paris, Ined-PUF, Cahier n°117, 1987. 5. Prioux F. Fertility and family size in Western Europe. Population, English selection, 2, 1990: 141-61. 6. Frejka T, Sardon JP. Childbearing Trends and Prospects in Low-Fertility Countries. A Cohort Analysis. Dordrecht/Boston/London. Kluwer Academic Publishers, ESPO 13, 2004. 7. Sobotka T. Postponement of Childbearing and Low Fertility in Europe. Amsterdam, Dutch University Press, 2004. 8. Corijn M, Klijzing E. Transition to adulthood in Europe. Boston. Kluwer Academic Publishers, 2001: ESPO 10. 9. Toulemon L, Leridon H. La famille idéale: combien d’enfants, à quel âge ? Insee Première, n°652, 1999. 10. Prioux F. Avoir moins d’enfants, une évolution inéluctable? In : Vivre plus longtemps, avoir moins d’enfants, quelles implications ?, AIDELF-PUF n°10, 2002 : 2948. 11. Daguet F. Un siècle de fécondité française. Insee Résultats : Société n° 8, 2003. 12. Beaumel C, Doisneau L, Vatan M. La situation démographique en 2001. Insee Résultats : Société n°18, 2003. 13. Leridon H. Can assisted reproduction technology compensate for the natural decline in fertility with age? A model assessment, Human Reproduction, 2004; Vol 19, n°7: 1548-53.
Rev Epidemiol Sante Publique, 2005, 53 : 2S3-2S11
Late fertility in Europe: some comparative and historical data La fécondité tardive en Europe : quelques données historiques et comparatives F. PRIOUX Institut National d’Études Démographiques (Ined), 133, boulevard Davout, 75980 Paris Cedex 20. Email : [email protected] (Reprints: F. Prioux).
The issue: Given the substantial increase in late fertility (at ages 35 and older) in the last few years in Western countries, the paper raises the question of what factors have caused this new trend, and how likely it is to continue. Methods: Late fertility is measured using the sum of age-specific fertility rates at 35 or more; its contribution to the total fertility rate (in % of the total fertility rate) and the proportion of first childbirths in this fertility is then computed. Late fertility among men is assessed using the same method, where the necessary data are available. The data analysed sometimes date back to the 18th century for women (Sweden), and the early 20th century for men (in France). Results: Late fertility is now much lower than it used to be. Despite the current increase — which can be traced back to the early 1980s for most Western countries, and appears to be much more recent and limited in the former socialist countries — current levels are very low compared with those of the past, among men as well as women. Reduced family size was the main factor in that decline, which became more pronounced as a result of women having a first child at younger ages. The reversal of the trend is due only to postponed childbearing, which must be viewed as part of a general trend towards postponement of all demographic events. More and more of the children born in such circumstances are thus first children. Very late fertility (at 40 or more) has increased, though only slightly, for men as well as women, and age 45 seems to be a limit beyond which few are prepared to have children. Social norms concerning the age limit for a person to have a child could account for the differences from one country to another in that respect. Conclusion: Late fertility is likely to continue increasing as long as fertility continues to decrease among young people. However, the increase will continue to be due mostly to the “younger”age groups (under 40 for women, under 45 for men); both biological constraints and social norms probably play a part in the stagnation of very late fertility levels. Fertility at age 35 or more. Europe. Postponement of first childbearing. Female fertility. Male fertility. Social norms.
Position du problème: La fécondité tardive (35 ans ou plus) ayant beaucoup augmenté ces dernières années dans les pays occidentaux, l’article s’interroge sur la cause de cette nouvelle tendance, et sur ses chances de se poursuivre encore. Méthodes: On mesure la fécondité tardive par la somme des taux de fécondité par âge à 35 ans ou plus, et on calcule sa contribution à la fécondité totale (en % de l’indicateur conjoncturel), et la part des premières naissances dans cette fécondité. La fécondité tardive des hommes est évaluée de la même manière, lorsque les données le permettent. Les données analysées pour les femmes remontent parfois jusqu’au XVIIIe siècle (Suède), et pour les hommes au début du XXe siècle (en France). Texte reçu le 22 mars 2004. Acceptation définitive le 10 mai 2005.
2S4
F. PRIOUX
Résultats: La fécondité tardive est aujourd’hui beaucoup plus faible qu’autrefois. Malgré la hausse actuelle, qui remonte au début des années 1980 pour un grand nombre de pays occidentaux, et qui est beaucoup plus récente et timide dans les pays anciennement socialistes, les niveaux actuels sont très faibles comparés aux niveaux anciens, aussi bien pour les hommes que pour les femmes. La réduction de la taille des familles a été le principal moteur de cette baisse, qui a été accentuée par le rajeunissement de l’âge à la première maternité. Le retournement de tendance n’est dû qu’au retard des maternités, retard qui fait partie d’un mouvement général de décalage de tous les calendriers démographiques. De plus en plus souvent, ces naissances tardives sont ainsi des premières naissances. La fécondité très tardive (40 ans ou plus) augmente très peu, et, pour les hommes comme pour les femmes, l’âge de 45 ans semble une limite que peu d’entre eux franchissent. Des normes sociales concernant l’âge limite pour avoir un enfant pourraient expliquer des différences entre pays à cet égard. Conclusion: Il est probable que la fécondité tardive poursuivra sa hausse tant que la fécondité des jeunes continuera à diminuer. Mais, cette hausse portera toujours essentiellement sur les âges les plus « jeunes » (moins de 40 ans pour les femmes, moins de 45 ans pour les hommes); contraintes biologiques et normes sociales jouent probablement toutes deux un rôle dans cette stagnation de la fécondité très tardive. Fécondité après 35 ans. Europe. Retard des premières naissances. Fécondité des femmes. Fécondité des hommes. Normes sociales.
Over the last twenty years, the number and proportion of babies born to older mothers (age 35 or more) have increased almost everywhere in Europe. For example, in France, 134,200 children were born in 2000 to mothers aged 35 or more — 17.3% of the total number of births; in 1980 this number was only 47,000 — 5.6% of the total. In twenty years, the number of children born to older mothers was thus multiplied by 2.9, while the number of women in the corresponding age group was multiplied by only 1.4. Late fertility has thus risen sharply, and France is certainly not the only country in such a situation. That the number of childbirths comparatively late in the mothers’lives should be increasing is a matter of curiosity, or even concern, for the medical world, because the risks for the life and health of both mother and child are heightened as the mother grows older. This, however, does not come as a surprise to demographers, who think that such an evolution is only logical: postponement of childbearing began in the early 1970s, and it seems logical that the decline in early fertility should be followed by a rise in late fertility, otherwise the total number of children would have declined by the same proportion. The question for demographers is the following: to what extent, if fertility among young women continues to decrease, can fertility among older women increase to compensate for the decline.
SOME HISTORICAL LANDMARKS
The recent evolution of fertility at age 35 or more should first be studied within a long-term historical framework, by observing those countries where long-established statistical series are available. By comparing recent and older data, the current levels can be viewed in a relative perspective. Late fertility can be measured by summing the age-specific fertility rates of women aged 35 or more, expressed as an average number of children per woman. The proportion of late fertility in the total fertility of a country in a given year can also be measured by comparing that sum with the total fertility rate (TFR) for the year under study. It should be noted, however, that this does not provide an indicator of fertility timing, since the TFR depends both on the evolution of timing and the intensity (the completed fertility) of fertility within cohorts. This proportion of late fertility is simply a measure of the contribution of women aged 35 or more to the annual fertility of a country. In the mid-18th century, the figure for late fertility was 1.5 children per woman in Sweden, which represented one third of the total fertility of the country. In about 1860-1870, a period when reliable data first became available in several European countries [1], the figures for late fertility were 2.1 children per woman in the
TRENDS IN LATE FERTILITY
Netherlands, 1.8 in Norway, 1.7 in Finland and Sweden, and only 0.9 children per woman in France (fig. 1). As we know, fertility in France began to decline from the late 18th century [3], as a result of early dissemination of traditional methods of birth control. This indicates that birth control was probably practised more intensively among women aged 35 or more, as childbearing at those ages represented only one quarter of the total fertility, as against one third in Finland, 38% in Sweden, and over 40% in Norway and the Netherlands (fig. 2). From the late 19th or early 20th century, there was a widespread, rapid decline in late fertility, For 1 woman
2,5 Netherlands
2,0
2S5
and in the late 1930s, the levels ranged from 0.3 children per woman in France, to 0.7 in the Netherlands. The trend was briefly interrupted by the baby-boom period, when fertility among women aged 35 and older increased again everywhere in Europe, and even quite significantly in Norway and the Netherlands. The decline subsequently resumed, and by the late 1970s, late fertility figures were a mere 0.1 children per woman in the Netherlands, and only slightly higher in the other countries under study; in all five countries, the contribution of late fertility to total fertility rates was under 10%. Late fertility has picked up once again, though it is now very far below its former level. Ranging between 0.25 and 0.28 children per woman, it represents approximately one sixth of the TFR of the countries under study.
Norway
A GENERAL TREND IN WESTERN EUROPE, AFFECTING MOSTLY WOMEN AGED UNDER 40
Sweden Finland
1,5
1,0
France
0,5
0,0 1750
1800
1850
1900
1950
2000
FIG. 1. — Late fertility from the 18th century in Western Europe (the sum of age-specific fertility rates at ages 35 and older, per woman). Sources: [1] and [2].
%
50 Netherlands
40 Norway
Sweden Finland
30 France
20
10
0 1750
1800
All the countries in Western Europe are currently experiencing a new rise in late fertility, though it is more recent in some countries than in others. It is in Southern Europe that the trend has reversed most recently (between 1986 and 1990), and also in Ireland, where a renewed rise in late fertility has been observed since 1994 only. And yet, fertility at age 35 and above was not especially low in Ireland — 0.36 children per woman, more than twice the minimum level in Spain (0.17), and four times the minimum levels in Germany, Belgium and Denmark (0.09). The minimum value generally ranges from 0.1 to 0.15 child per woman.
1850
1900
1950
Year
2000
FIG. 2. — Contribution of late fertility to total fertility rates from the 18th century in Western Europe (the proportion of the aggregate age-specific fertility rates of women aged 35 and older in the total fertility rate). Sources: [1] and [2].
The most significant factor in the decline of late fertility has been the gradual and virtually total disappearance of large families [1] as a result of birth control, at first by means of traditional methods, and later using “modern”, more efficient methods (pill, IUD, sterilization, legal abortion) [4]. As a consequence, the average age of mothers fell to its lowest level in Western Europe in the mid-1970s (and a little later in Southern Europe). This corresponds to a period when family building probably occurred earliest in the whole of history. Marriage — at the time a virtually compulsory precondition — and the
2S6
F. PRIOUX
birth of the first children occurred at particularly young ages, as did completed fertility, which was usually limited to 2 children [5]. Though late fertility has recently started to increase again, this is certainly not due to a new rise in the number of large families. It is mainly due to an increase in age at childbearing [6, 7]. The various aspects of transition to adulthood now all occur later [8]. The time for completion of education, entry into the labour market and formation of a stable couple, are stages in life that people reach at increasingly later ages, and these factors all contribute to postponing the birth of children. Nevertheless, it seems that women cannot postpone childbearing indefinitely. It is mostly between ages 35 and 40 that fertility is rising (fig. 3): in many countries, the fertility of women aged 35-39 is now the same, or even higher than in the early 1970s, and now usually ranges between 0.20 and 0.25 children per woman. Germany, Austria and Greece are exceptions, with fertility remaining comparatively low (0.15 or less in 2000). In Ireland, by contrast, fertility is much higher (0.39). The rise in fertility beyond age 40 is currently still quite small, and fertility at such ages still represents, everywhere except in Ireland, fewer than 0.05 children per woman (fig. 4), or 2 to 3% of total fertility. And childbirths at ages over 45 represent only a very small part of overall fertility beyond age 40 — a bare 3 to 5%, approximately 1‰ of total fertility1 — and the trend does not appear to be rising yet. Age 45 thus seems to represent a boundary beyond which few women are prepared (or able) to go. According to a survey recently conducted in France, the average age from which women consider it desirable not to have any more children is 41 [9]; for more than 4 in 10 women the age is 40, and while 1 in 20 women gave an age over 45, 1 in 6 set the limit at 35 or less2. Social standards thus appear to match behaviour in France. We do not know 1
According to [2], the sum of fertility rates of women aged 45 or more is 1 child per 1,000 women in Germany, Spain, Norway and the Netherlands; it ranges from 1 to 2 children per 1,000 women in France and Sweden, from 2 to 3 in Finland, and 3 per 1,000 in Ireland. 2 Data kindly provided by Laurent Toulemon.
0,5
0,4 Spai n 0,3
Ireland
Netherlands
0,2 France
0,1 Germany 0,0 1960
1970
1980
1990
2000
FIG. 3. — Fertility of women aged 35-39 in several Western European countries since 1960 (sum of age-specific fertility rates of women aged 35-39, per woman). Source: [2].
whether surveys on this theme have been conducted in other European countries3. CONTINUED LOW LEVELS OF LATE FERTILITY IN EASTERN EUROPE
Up to the 1990s, there was a significant contrast between the ages of mothers on either side of the Iron Curtain — in the socialist countries (as defined by their present borders4), mothers were 2 and a half years younger on average than in Western countries [10]. Due to very early childbearing, and families seldom having more than 2 children, late childbirths were rare up to the late 1980s. Moreover, the widespread decline of fertility observed in all the countries of the region since the fall of the communist regimes, has further strengthened this trend. Late fertility was consequently particularly low — under 0.1 child per woman virtually everywhere (except in Poland), or even under 0.05, as in Bulgaria and the former 3 A question on this theme was included in the 1995 Eurobarometer survey, but the results do not appear to have been published. 4 Except for the former German Democratic Republic, which we have included among this group of countries, because fertility behaviours were — and still are — very similar to Eastern ones.
TRENDS IN LATE FERTILITY
0,5
0,4
2S7
per woman Irl
0,4
0,3
0,3
0,2
Ireland
Fl
0,1
0,2
GB
Sw PB GFR
Dk
Spain 0,1
0,0 1970
Netherlands France
1970
1980
Nor Aust Belg Hun
Gr
Pol
Lit Cro Est Latv Slova Slove Rum Czek Rus Bulg GDR
1990
Year 2000
FIG. 5. — Minimum value of late fertility, and the years when that value was reached in Europe (sum of age-specific fertility rates at ages 35 and older, per woman). Source: [2].
Germany 0,0 1960
Spain Por Serb
It
Swed Fr
1980
1990
2000
FIG. 4. — Fertility of women aged 40 and older in several Western European countries since 1960 (sum of age-specific fertility rates of women aged 40-49, per woman). Source: [2].
rapid — the proportion of first children born to older women has increased more than twofold since the 1980s, rising from 15% to over one third.
German Democratic Republic (GDR). It was only in the 1990s5 that the fertility of women aged 35 and older began to rise timidly again in all those countries, much later than in the Western countries (fig. 5). As a consequence, late fertility is still very rare (ranging from 0.06 child per woman in Bulgaria, to 0.15 in Estonia), and represents 5 to 10% of total fertility .
As the proportion of higher order births tends to increase with age, the proportion of primaparous women should always be higher among women aged 35-39 than among those aged 40 or more. This is the case in the Netherlands for example, where first births represent 27% of fertility of the first age group, as against 23% of the second. In Spain and France, however, the proportions tend to be the same in both age groups, and in Sweden, the proportion of primiparous women is even slightly higher among the older group: in 20007, the proportion of first births was 30% at ages 35-39 and 33% at ages 40 and older.
A GROWING PROPORTION OF PRIMIPAROUS WOMEN
In Western Europe, between one-fifth and onethird of women who give birth at age 35 and over are primiparous, whereas half a century ago, the proportion was probably seldom above 10%. In the Netherlands, for which a long data series is available, the proportion of primiparous births at age 35 and over was only 6% in the 1950s and into the mid 1960s. Today, the proportion is 26% (table I). In France, the proportion was still under 9% in the late 1960s, it is now probably over 20%6. But it is in Spain that the rise was most 5 Hungary is an exception, having reached a minimum as early as 1984. Poland is another exception, where the minimum seems to have been reached only in 2000.
6 This is based on data from the 1999 Study of Family History survey, kindly provided by Laurent Toulemon. We do not use the statistics published by the Insee, which were based on birth registration data recorded since 1998, because the proportion of first children, particularly among older women, was considerably overestimated. This is due to the fact that municipal registry offices in France still complete the forms using the “livret de famille” booklet given to each family at the time of marriage, or of the birth of the first child (of the couple, or of the mother if she does not live in a couple). As a result, the first child of a subsequent union is usually categorized as a woman’s first child, though she may have had children from a previous union. 7 The mean figure from 1999 to 2001 was used here, with a view to reducing the risks of errors due to exceedingly low rates at those ages.
2S8
F. PRIOUX
TABLE I. — Change in the proportion of first order births in late fertility and in total fertility (ratio of aggregate first order fertility rates to aggregate fertility rates for all birth orders at corresponding ages). Sources: Spain, Hungary and the Czech Republic: the data were kindly provided by J.P. Sardon (European Demographic Observatory/INED); France: from the 1999 Study of Family History survey; data kindly provided by L. Toulemon (INED); Netherlands and Sweden: the author’s own calculations, based on national statistics. Proportion of first order births (%)
Country
1960
1980
2000
14.6
34.3
France
8.4
a
13.4
19.6b
Hungary
13.4
18.1
17.9
Netherlands
5.9
14.0
26.1
22.2
30.6
14.8
16.0
17.3
Spain Late fertility (aged 35 and older)
Sweden Czech Rep. Spain Total fertility (at all ages)
40.5
53.0
France
34.2
a
41.1
42.2b
Hungary
44.4
47.5
44.2
Netherlands
30.4
42.4
47.9
42.6
44.3
44.8
46.9
Sweden Czech Rep. a
45.3
1965, b 1998.
In a growing number of cases, these very late births appear to be the consequence of a woman’s desire to have a first child before she becomes infertile. At first, the growing proportion of late fertility first births was simply due to a general decrease in the number of higher order births (i.e. 3 or above). As a result, the proportion of first children born to women of all ages (lower part of table I) duly increased, including those born to women aged 35 and older (upper part of the table). Since the 1980s however, the main factor underlying this trend has been an increase in firstbirth fertility of older women: in the Netherlands, for example, the proportion of women who have a first child at age 35 and older (9 children per 100 women in 2002) is currently 5 to 6 times higher than in 1980 (1.5 per 100), and approximately twice as high as in 1950 (5 per 100). Similar trends are also observed in Spain, France and Sweden (fig. 6). Nothing of the kind has been observed in Eastern Europe as yet, where late first-birth fertility remained peculiarly low up to the 1990s. For
example, it was only 1 child per 100 women in the Czech Republic, and 1.5 per 100 women in Hungary. It is currently just beginning to increase (fig. 6). The proportion of primiparous women
per 100 women 10 Netherlands
8
6 Spain France
4 Sweden
Hungary
2
0 1950
Czech Rep.
1960
1970
1980
1990
2000 Year
FIG. 6. — Evolution of late first-birth fertility (sum of agespecific first-birth fertility rates at age 35 or over, per 100 women). Source: see table I.
TRENDS IN LATE FERTILITY
among those giving birth at age 35 and over thus appears to be still under 20% (table I).
1,4
2S9
per woman or man
1,2
A CONCURRENT INCREASE IN LATE FERTILITY AMONG MEN
1,0 Men aged 35 or +
0,8
Statistical data on the ages of fathers are seldom available, if ever. The fact that so little attention is paid to male fertility results partly from a difficulty inherent to nature, as fathers cannot always be identified. However, even in those countries where legal recognition of a child is compulsory, statistics on the ages of fathers are not always available. For example in Sweden, it is only since 1992 that the statistics on the ages of fathers have covered the birth of all children, not just those born in wedlock, as formerly, and it is only since 1999 that male fertility indicators by age group have been computed. Denmark is the only country where age-specific male fertility rates have been published more or less regularly since 1980, but the rates are only computed up to age 49, which makes any specific study of late male fertility impossible. In France, where the Insee has recently reconstructed records [11], a long data series covering the whole of the 20th century is now available (fig. 7). As could be expected, considering that the parents’ages are fairly well correlated, the pattern of late male fertility is quite similar to that of women, though the level is definitely higher, because fathers are usually older than mothers8. At the beginning of the century, the level of late male fertility was 1.3 children per man, 42% of the total male fertility rate at the same date (3.13 children per man). The trend was subsequently quite similar to that of female fertility at the same ages, with each period of war being followed by a rise in late fertility of variable duration. By the time the decline came to an end, in 1978, the level was a bare 0.35 child per man, 19% of the total male fertility rate. At that point, the level of late male fertility began to rise again, until it reached 0.59 in 2001, almost one third of the total 8 However, the method used by the Insee, in which childbirths are classified by age (the age reached in the year) results in slightly overestimating late fertility, as compared with figures 1 to 4, where the data in all the countries concerned the age of the mother at the time of childbirth.
0,6 Women aged 35 or +
0,4 Men aged 40 or +
0,2
Men aged 45 or + Women aged 40 or +
0,0
1900
1920
1940
1960
1980
2000 Year
FIG. 7. — Late male fertility in France since 1900. A comparison with late female fertility (sum of age-specific fertility rates, from ages 35, 40 and 45, per man or woman). Source: [11] and [12].
male fertility (1.86 children per man), or twice as high as female fertility at those ages (0.31). In the early 1980s, the level of fertility among men aged 40 or older was very similar to that of women aged 35 or more. However, the subsequent rise was much slower because male fertility beyond age 45 has stopped increasing since 1990 — it has been 0.08 per man since then, 4% of the total male fertility rate. It thus appears that among men too, postponed parenthood has led to a rise in fertility, particularly before age 40, as the contribution of older men to the rise has been very small up to now. Given the rise in family recomposition, in which men, more than women, can build a new family at later ages if they form a union with a woman younger than themselves, a more substantial rise in male fertility beyond age 45 could have been expected. The limits of male physiological capacities at those ages, as can be inferred from past statistics, have not been reached, far from it: at the beginning of the century, male fertility at age 45 and older was 0.3 children per man, 10% of the total. So there is plenty of scope for a further increase in male fertility at much older ages. The standards concerning the age beyond which it is no longer desirable for a man to have children do not entirely tally with this stagnating male fertility beyond age 45: though the mean
2S10
F. PRIOUX
age mentioned by men is 44.59 [9], more than 1 man in 10 has no opinion on the question, and among those who do express their views, 3 in 10 give an age over 45 (age 50 mostly)10. How do things stand in other countries? In Denmark male fertility at ages between 35 and 50 fell to a level below that of France in the early 1980s, and though it did rise again, the rise seems to have levelled off since 1994. The current level is 0.45 children per man, as against 0.56 in France at the same ages11. In Sweden, in 2001, the level of male fertility after age 35 was also lower than in France (0.49 vs 0.59), particularly for very late fertility (after age 45 or 50). Even though levels are currently tending to rise, particularly among younger men, higher levels of late male fertility could have been expected in those two countries, considering that separation and repartnering are more frequent than in France. TOWARDS AN INCREASING PROPORTION OF OLDER PARENTS
Statistical data on the ages of both parents are also very rare. For many years, only legitimate births were taken into account, and tables on the crossed ages of the parents only became available in national statistics very recently. Moreover, as no rates can be computed in this area, our observation will concern only the proportion of late childbirths in the total numbers. Obviously the proportion depends on behaviours, but it is nevertheless influenced by the age structure of the population of fertile age. In 2000 the proportions of late childbirths from the point of view of both parents were similar in Sweden and in France (table II): approximately 1 in 7 or 8 children had parents both aged 35 or more, and 1 in 15 had a father aged at least 40 and a mother aged at least 35; in Hungary, where late fertility is still very rare, the proportions are 9 In women’s view, the mean age limit for a man to father a child is 45. 10 Data provided by Laurent Toulemon. 11 Though the definition of age used here in France (age attained) entails very slight overestimation of late fertility, as compared with the definition used for Denmark and Sweden (the father’s age at the time of the birth of the child), those differences are very significant.
lower by half, as is probably the case in most former socialist countries. In countries that had a baby-boom period followed by a decline in fertility, the population of fertile age is currently “ageing”, and this entails more late childbirths, in circumstances of unchanged behaviour. The general trend in which motherhood and fatherhood are postponed until later ages simply increases the proportion of late childbirths, from the point of view of either the father, the mother or both parents, in the total numbers of childbirths12. AN OVERALL PERSPECTIVE
As a combined result of reduced family sizes and earlier fertility, first-birth fertility especially, late fertility seemed destined to disappear. In the early 1980s, a few years later in Southern Europe, and in the 1990s in the former socialist countries, the levels were strikingly low. The recent increase in late fertility results not from any resurgence of large families, but from a general postponement of all demographic events. The increase, while still very slight in the countries of former East Europe, is very rapid in Western Europe. On the other hand, the proportion of late fertility first births has increased, at first as a consequence of reduced family size, and later mostly because the proportion of women who have their first child after age 35 is increasing very fast in Western countries. However, there has been only a slight increase in fertility after age 40, as if this age barrier was a difficult hurdle to cross. And yet, comparing the present situation with older data suggests that there is still substantial scope for increased fertility at those ages. In the past, however, late childbirths, which occurred mostly in large families, were the result of hyper-fertile women bearing a last, more or less desired child. Nowadays, women who want a child after age 40 are not necessarily the most fertile ones; in fact they may even be less fertile than the others because they less often have child-bearing experience. Even 12 As mentioned in the introduction to this text, in France the increased numbers of women aged 35 or more have also largely contributed to increasing the numbers of late childbirths since the 1980s.
TRENDS IN LATE FERTILITY
2S11
TABLE II. — The proportion of late childbirths in 2000, from the point of view of the mother, the father and both parents (for a total of 100 births). Sources: National statistics, after due distribution of undeclared ages. Countries France
a
Hungary Sweden a
b
b
Mother aged 35 or more
Father aged 35 or more
Father aged 40 or more
Both parents aged 35 Mother aged 35 or +, or more father 40 or +
17.3%
31.5%
11.6%
13.9%
6.8%
8.0%
20.8%
8.3%
6.2%
3.6%
17.0%
32.8%
11.7%
13.2%
6.2%
age attained in the current year; b age attained at the time of the child’s birth.
without any selection effect, a recent simulation by H. Leridon [13] shows that the chances of conceiving a child are substantially reduced among women aged 40 compared with those aged 35, and that assisted reproduction techniques cannot compensate for the handicap. On the other hand, the levels of very late male fertility (at ages 45 and older) have remained surprisingly unchanged, despite the supposedly favourable effect of family recomposition. In that respect, however, we need to verify whether the proportion of men who form new unions at those ages with much younger women is actually increasing. Moreover, differences remain in Europe as far as very late fertility is concerned, which suggests that there are social norms relating to the ages when a man or a woman should stop having children. Late fertility is thus likely to continue increasing, as long as fertility among young people pursues its decline, and this increase will continue to concern mostly the “younger” groups (under age 40 for women, under 45 for men). Both biological constraints and social norms probably play a part in this stagnation of very late fertility. REFERENCES 1. Festy P. La fécondité des pays occidentaux de 1870 à 1970. Paris, Ined-PUF, Cahier n°85, 1979. 2. Recent Demographic Evolution in Europe, Strasbourg, Council of Europe Publishing, 2003.
3. Henry L, Blayo Y. La population de la France de 1740 à 1860. Population, 30 special issue, November 1975: 71-122. 4. Leridon H. La seconde révolution contraceptive. Paris, Ined-PUF, Cahier n°117, 1987. 5. Prioux F. Fertility and family size in Western Europe. Population, English selection, 2, 1990: 141-61. 6. Frejka T, Sardon JP. Childbearing Trends and Prospects in Low-Fertility Countries. A Cohort Analysis. Dordrecht/Boston/London. Kluwer Academic Publishers, ESPO 13, 2004. 7. Sobotka T. Postponement of Childbearing and Low Fertility in Europe. Amsterdam, Dutch University Press, 2004. 8. Corijn M, Klijzing E. Transition to adulthood in Europe. Boston. Kluwer Academic Publishers, 2001: ESPO 10. 9. Toulemon L, Leridon H. La famille idéale: combien d’enfants, à quel âge ? Insee Première, n°652, 1999. 10. Prioux F. Avoir moins d’enfants, une évolution inéluctable? In : Vivre plus longtemps, avoir moins d’enfants, quelles implications ?, AIDELF-PUF n°10, 2002 : 2948. 11. Daguet F. Un siècle de fécondité française. Insee Résultats : Société n° 8, 2003. 12. Beaumel C, Doisneau L, Vatan M. La situation démographique en 2001. Insee Résultats : Société n°18, 2003. 13. Leridon H. Can assisted reproduction technology compensate for the natural decline in fertility with age? A model assessment, Human Reproduction, 2004; Vol 19, n°7: 1548-53.