1Within ten years, between 1975 and 1985, the number of marriages celebrated each year in France fell by 30 %, and the number of illegitimate births was multiplied by 2.5. These developments signal a profound change in attitudes and behaviour relating to marriage, an institution that, for long, had been a pre-condition for childbearing. By disregarding marriage, we obtain a novel conception of the couple, married or unmarried, that started emerging during the late 1960s.
2It soon became apparent that though the young were becoming less keen on marrying, they did not turn their backs on cohabitation. It was to distinguish these new patterns of behaviour from what was traditionally termed "concubinage" and to stress the often temporary character of these new living arrangements that Louis Roussel suggested the term "Juvenile Cohabitation". His analysis of the first INED survey on this subject [1] showed a spectacular increase in pre-marital cohabitation : 17 % of couples married in 1968-69 had already lived together before their marriage compared with 44 % of those married in 1976-77. Later analysis suggested that though these rates had been somewhat overestimated, the change was so fast that they merely anticipated reality by a year or two [2]… The frequency of cohabitation at the time of the survey was more modest (10 % of the total of 18-29 year-olds) but among men and women aged between 22 and 23 years, it already amounted to 15 %. What was significant was that one-third of the cohabiting partners below 30 years of age expressed the wish "to continue to live together without giving a thought to marriage", and 14 % even refused to consider the idea. A new type of union was, perhaps, about to emerge.
3Pierre-Alain Audirac has recently summarized the available statistics in two articles [3] by making the best use of two major sources : the censuses of 1962, 1968, 1975 and 1982, and the yearly employment surveys (1982 to 1985). These sources tend to underestimate the incidence of extra-marital cohabitation. They also show a fall in the proportion of men and women below the age of 30 cohabiting (married or unmarried) which is not, however, confirmed by the survey results that we shall present later on. However, this work revealed a considerable increase in the number of unmarried couples : between 1962 and 1985, there was a tenfold rise in the number of such couples in which the male partner was between 35 and 60 years old. A wealth of information was also provided about the characteristics of these couples.
A new survey
4A random representative sample of households with a constant sampling fraction (taken from the census, for example) would not have been very efficient, as the households with the most interesting situations for the survey are a small minority. To improve on this result, it was necessary to have advance information on the structure of the households.
5INSEE’s annual Employment Survey fulfilled this condition and INSEE agreed to allow us to re-interview a sub-sample of "the outgoing third" (the 21,000 households questioned for the last time about their working conditions) from the March 1985 survey. We, therefore, used the information available on the composition of households in the data file of the Employment Survey, to select slightly more than 5,000 households for our own survey, with sampling rate varying from one-fifth to one. After appropriate weighting, this sample is representative of :
- the adult population between 21 and 44 years of age on 31-12-1985, living in a private household ;
- the child population between 0 and 18 on the same date [4] ;
- single parent families (with at least one unmarried child below 25 years of age).
6The interview took place in November-December 1985 ; 4,433 households were interviewed, 87 % of the targeted sample. Non-response was due to refusal (5 %), inability to contact those who had moved away to an unknown address (3 %), those temporarily absent (3 %), and other causes (2 %).
7There were 4,091 adults between 21 and 44 years of age, and by using appropriate weights we were able to reconstruct the exact composition of this population by sex and age as on 31-12-1985. The same is true for the 4,554 children between 0 and 18 years of age.
What is a couple ?
8When a social institution is changing, the words used to describe it can be misleading. So far, words like "cohabitation", "concubinage", "consensual union", "common-law unions" or "unmarried couples", have been used indiscriminately. However, etymologically, they are not synonymous. The word cohabitation merely implies that two persons are living under the same roof ; hence a "cohabiting couple" may or may not be married, but the term is usually reserved for unmarried couples.
9This leads us to the question whether there exist "non-cohabiting couples", i.e. individuals who feel committed to a stable union with their partner but who do not live under the same roof. Such situations may be voluntary, or due to an outside constraint (for example, separation for career reasons). In our survey, between 2 and 3 % of currently married couples as well as between 7 and 8 % of unmarried couples reported that they had always lived separately… The characteristics of these "non-cohabiting couples" will be examined in another paper but, as they were ‘couples’ in their own estimation, we shall treat them as such hereafter.
I – Situations at the time of survey
10The survey was conducted in November-December 1985 : we shall consider that the situation described here was that on January 1, 1986.
11In Table 1 we present de facto matrimonial situations for different sex and age groups. 19 % of women aged between 21 and 24 years and 17 % of men aged between 25 and 29 years lived as unmarried couples, a proportion respectively of one-third and one-quarter of the total number of couples in these cohorts. 10 % of the men and women aged 21-44 years lived as unmarried couples and 66 % of them as married couples.
De facto marital situations on 1-1-1986 by sex and age
De facto marital situations on 1-1-1986 by sex and age
12Unmarried cohabitation is more common among the younger respondents. The age difference between married and unmarried couples is five years (Table 2) ; this also explains why three-quarters of cohabiting persons are never-married (Table 3). The ever-married (mostly divorced) only become a majority among those aged 35 years or older (men as well as women), but the proportions of extra-marital cohabitation rates are very low at these ages (4 to 7 %). In this regard, cohabitation still deserves its qualifier "juvenile", if it were defined only by age and marital status.
13We can verify this still further by comparing the respective marital status of each spouse : in two-thirds of couples both partners are never-married ; the remaining third can be roughly divided into three groups :
- a never-married man with a previously married woman ;
- a never-married woman with a previously married man ;
- two previously married persons.
14Naturally, the proportion of unmarried couples consisting of two never-married partners depends largely on the ages of the respondents :
Mean age of respondents by sex and marital status (de facto) (exact average age on 1-1-1986)
Mean age of respondents by sex and marital status (de facto) (exact average age on 1-1-1986)
- 91.1 % among those aged 21-24,
- 73.8 % among those aged 25-29,
- 52.8 % among those aged 30-34,
- 31.1 % among those aged 35-39,
- 28.4 % among those aged 40-44,
- and 65.6 % for all those aged 21-44.
16These initial results confirm the extent of cohabitation. It was possible to check individual replies given in our survey against those given in the Employment Survey at the same date (March 1985) : it was evident that a number of persons living as couples were not spotted as such in the Employment Survey interview, and that a significant fraction of couples were wrongly classified in that Survey as "married". Altogether, the number of unmarried cohabiting persons estimated by INSEE should be increased by 30 % if a strict definition of cohabitation were applied, and by 40 % if "non-cohabiting" couples, were to be included in this category, as was done in our survey (that is to say 2.5 to 2.7 million cohabiting persons in all cohorts, according to the definition given) [5].
17Finally, we must point out that there are great variations by socio-cultural characteristics and place of residence (in the case of cohorts aged 21-29 living in the Paris area, for instance, one-quarter were living as unmarried couples compared with one-seventh in France as a whole) ; we shall discuss this subject in another article, in which we shall describe how these new forms of behaviour have progressively spread to different social groups.
(Legal) marital status of cohabiting partners and that of persons not living as a couple
(Legal) marital status of cohabiting partners and that of persons not living as a couple
II – Biographical data : union history and fertility
18A detailed study in which matrimonial careers will be traced beginning with the first union and looking at its outcome will be undertaken later [6]. Here we need only consider certain aggregate historical data : the number of previous unions, the length of the current union, pre-marital cohabitation, number of children already born.
19In Table 3 we have already provided information on previous marriages : 20 % of the men and 26 % of the women living as unmarried couples had been married previously ; the proportion exceeded 50 % among those aged 35 years and over. Moreover, as we shall see, these cohabiting persons are not averse to the idea of re-marriage : quite the contrary, since 44 % of them wish to remarry rapidly compared with 38 % of never-married cohabiting persons [7]. These data suggest that the current union was frequently not the respondent’s first union and could have been preceded by others, formal or informal.
The number of unions
20In Table 4, we find the number of unions reported by the respondent in answer to questions about periods spent in union (including, where applicable, the current union), by sex and age. The first (and well-known) finding is that women enter into a union at younger ages than men (the difference is approximately two years) : among those aged between 21 and 24 years, 60 % of the women have already had a partner as against 35 % of the men. Only between 5 and 6 % of respondents between the ages of 35 and 44 years had never lived with a partner.
21Second unions are not very frequent. In the age group 30-34 years and above, the proportion who had had at least two partners peaks at 10-12 % ; even if those who had never had a partner were excluded, the average number of partners never exceeds 1.17 and the results are similar for men and women.
22Naturally, a marriage preceded by cohabitation with the same partner is treated here as one and the same union. The figures in Table 5, which are based on matrimonial situation at the time of survey, confirm that ever-married persons have rarely had previous partners other than the spouse : the average number of unions peaks at 1.08-1.09 for women and 1.07-1.12 for men. For unmarried couples, the average is noticeably higher : it varies, for those aged 30 years and over, between 1.52 and 1.84 according to age and sex. But the difference between them and the ever-married is largely due to the fact mentioned above : that among those aged 35 years and over, a marriage to another person preceded current cohabitation in more than half the cases. If we take into account only the cohabiting never-married, the average number of unions no longer exceeds 1.50 : at age 30-39, the mean value is 1.34 instead of 1.71 for all cohabiting persons. It is thus obvious that the choice of extra-marital cohabitation does not necessarily lead to a higher degree of conjugal instability.
Total number of previous (and current) unions by age and sex(a)
Total number of previous (and current) unions by age and sex(a)
(a) No minimum duration of union required (in practice : one month).Average number of unions by sex, age and marital status at the time of survey
Average number of unions by sex, age and marital status at the time of survey
23As regards individuals not living with a partner at the time of survey, their experience, on average, was much more limited : at ages 35 and over, 60 % of the men and 33 % of the women had never had a partner.
Duration of the current union, total time spent in union
24A longitudinal analysis of the union history is necessary to fix the exact length of each episode, particularly for the phases of unmarried cohabitation and for the periods of life when individuals lived on their own. The average length of a current cohabitating union already exceeds 3 years for those aged 25-29 years, and 5 years for those over 30 years old (it is only slightly longer for never-married cohabiting individuals than for those who were divorced). This is not an accurate estimate of the average duration of this type of union, but it gives an idea of the degree of their stability [8]. Naturally, the duration increases with age, primarily because of progressive selection : for example, a union begun at the age of 24 is more likely to be observed among older cohorts if it lasts longer.
25However, this increase remains limited because many de facto unions end in marriage : we shall examine this transition later on. For the time being, it should be noted that the total length of time spent in union is noticeably higher than the duration of the present union, especially for divorced persons (who spent about the same time in a union as currently married persons), and also for the never-married (because of earlier unions) (Table 6).
26As regards individuals who currently lived on their own, the never-married spent less than one year in union and those divorced spent less time in union than the currently married.
27Note that the differences between mean ages at the beginning of the first union are much smaller than those between durations of unions. Divorced persons began their first unions earlier than those currently married, which confirms that the risk of divorce varies inversely with age at marriage. Conversely, never-married individuals (especially women) began living with a partner at a later age than currently married ones (even when the latter’s first union was a premarital cohabitation) : the difference exceeds two years for female cohabiting partners aged between 25 and 29 years, and three years for those aged between 30 and 34 years. Here, we can see the outlines of a very specific matrimonial career, marked by a distinct delay in the first union, and the choice of a particular kind of union, which may be less stable, because those never-married who were not living as a couple at the time of survey, and who had previously lived in a union, began that union at approximately the same age as non-married cohabiting persons.
Total duration spent in union (in years) and average age at the beginning of the first union, according to marital status at the time of survey
Total duration spent in union (in years) and average age at the beginning of the first union, according to marital status at the time of survey
() 20 to 50 cases.(()) Less than 20 cases.
28This type of behavior was most common amongst those with the highest academic qualifications : university graduates, for example, commit themselves to a union two years later than those holding only a primary school-leaving certificate.
Pre-marital cohabitation
29One of the first aspects noticed in the evolution of behaviour relating to marriage was the very rapid increase in pre-marital cohabitation. This is confirmed by the figures in Table 7 : 8 % of the marriages celebrated in 1960-1969 (and continuing on 1-1-1986) had been preceded by a period of cohabitation, compared with 57 % of those celebrated in 1980-1985 [9]. These proportions are not precisely comparable, because of differences in ages at marriage : currently-married persons who married between 1960 and 1969 could only have done so before the age of 29, those who married between 1980 and 1985 could have been as old as 44 years at marriage.
30When reading Table 7 along the diagonal (from left to right), we can control for age at marriage : the figures show that pre-marital cohabitation has made considerable progress at all ages, increasing from less than 10 % to 43 % for those married before their 25th birthday (mean age : 21), and from less than 10 % to 59 % for those married around the age of 25 years. The rate exceeds 60 % in 1980-85 for those married between the ages of 30 and 35 years.
31The average duration of cohabitation is close to 2 years for the entire sample of married persons. It has increased over time but much less than the proportion cohabitating. For recent marriages of those aged between 25 and 34 years at the end of 1985, the average duration of cohabitation is approximately two years ; it is between three and three and a half years for those married at ages exceeding 35 years.
Pre-marital cohabitation
Pre-marital cohabitation
32Without going into detail about differences by socio-cultural characteristics, it should be pointed out that if the average overall proportion of pre-marital cohabitation among those married recently (1980-85) is 57 %, it exceeds 67 % in Greater Paris and 75 % in Paris proper ; depending on academic qualifications, the proportion for CEP diploma holders (the ‘Primary School Certificate’ can be obtained after five years of schooling), as well as for graduates, is higher than 70 % : in fact, members of these two groups do not show a greater propensity to extra-marital cohabitation than their neighbours on the scale of academic qualifications, but those who do get married are more likely to have cohabited before marriage ; the average duration of cohabitation exceeds three years for those with a CEP.
33Observance of religious beliefs leads to a similar paradox : whereas, at any given time, there is only a small proportion of cohabiting persons among regular practising believers (or even irregular ones : 2.7 and 4.4 % respectively for those aged between 21 and 44 years), those recently married often went through a cohabiting phase : the proportion is one out of two among believers, and the duration of cohabitation is identical with the overall average. It should be understood that in this group, "the husband and wife-to-be" very probably account for most of the cohabiting partners at any given time, but their number is very low because the time spent in this phase is rather brief (2 years). It should also be stressed that pre-marital cohabitation is a very recent phenomenon in this group : the proportion was less than 10 % before 1980.
Fertility and marital status
34The average parities (by age) observed in the survey are quite close to the vital statistics estimates (Figure 1). Among the five cohorts analysed, the result for the oldest agrees perfectly with the vital statistics, the survey results are a little lower for those aged between 30 and 34 years, a little higher for those between 20 and 24 or 25 and 29 years, and noticeably higher for those aged 35-39 years. We shall be mainly concerned here with the differentials found in the survey by marital history.
Average parities of women by age and marital status
Average parities of women by age and marital status
35At each age, average parity is largely determined by the currently-married. The differences between the sexes essentially reflect the age difference between spouses, and thus are of little significance by themselves. The mean parity of cohabiting partners is, on an average, half that of the married, but matrimonial history plays an important role : men and women who have been married previously (hence are divorced or separated) have more children than currently-married couples (Figure 1, Table 8). By contrast, fertility of never-married cohabiting partners is only one-quarter of the overall rate for married couples. But this overall result is not the most significant one : between the ages of 30 and 34 years, the average number of children of men in this group is 1.0 (compared with 1.7 for currently married). For women, the corresponding figures are 0.7 and 1.8 respectively ; hence the fertility of "confirmed" cohabiting partners is far from negligible, especially since, as may be seen in Table 6, the mean time they spent in union is noticeably lower than that of married persons. An "average fertility rate" can be computed by dividing mean parity (Table 8) by duration spent in union (Table 6). The result (Table 9) confirms the slightly higher fertility of divorced persons, which is, therefore, not wholly due to early marriage, and shows that the fertility of never-married cohabiting partners is approximately two-thirds of that of currently married couples [10].
36The fertility of persons not committed to a union is very low if they are never-married (almost nil in the case of men), but rather high for the divorced : among them, the number of children is close to that for the currently married of the same age and sex.
37On the whole, these preliminary results confirm that divorce does not reduce fertility. The interruption in childbearing for those who do not take another partner, is compensated by the slightly higher rate for those who commit themselves to a new union outside marriage (re-married divorced persons are included, here, in the currently-married category because they are too few to be analysed separately). Moreover, we observe that the fertility rate of never-married cohabiting persons is far from negligible, as could have been foreseen from the substantial increase in illegitimate births [11].
III – Changes between 1968 and 1985
38As was mentioned at the beginning of this article, marriage behaviour has changed very rapidly during recent years. In particular, the fall in the number of marriages has been so rapid that one wonders whether it has been fully compensated by increased extra-marital cohabitation, or whether there may not be a certain disinclination to living as a couple, married or otherwise.
39From the detailed union history provided by each respondent, the individual’s earlier matrimonial status can be determined for any previous date, thus making it possible to analyse changes in the proportions of men and women living in union at any given date, for various age-groups compatible with the sampling frame. Actually, we can trace those aged :
- 18-20 from January 1968 to January 1983 ;
- 21-24 from January 1968 to January 1986 (or, to be more precise, up to the survey date : November or December 1985) ;
- 25-29 from January 1971 to January 1986) ;
- 30-34 from January 1976 to January 1986 ;
- 35-39 years from January 1981 to January 1986.
Mean number of children ever born by age and marital status
Mean number of children ever born by age and marital status
() 20 to 50 cases.(()) Less than 20 cases.
Average number of children per year spent in union (per 1 000)
Average number of children per year spent in union (per 1 000)
40The age and status observed relate to January of each year. The sample size for each age and sex group never falls below 220 and is usually in the 300-500 range.
41Considering all inherent limitations of a retrospective reconstitution based on persons present in the area at a given date (within the same cohorts, some individuals were not present at a previous date D, others had left the area or died between date D and the survey), we can assume that marriages (and their dates) were reported correctly, as well as the break-up of these unions (with or without a legal divorce). Other unions may pose a problem but it is worth noticing that many respondents even reported previous (or current) periods of "life as a couple" without real cohabitation (meaning co-residence).
Proportion living in union : 1968-1986
42In Figures 2a and 2b, we show the change in the proportions living as a couple, for each sex and age group (upper curves). Between 1976 and 1981, the figures vary little and no clear trend can be discerned. However, the proportion of currently-married persons among those aged between 21 and 24 started declining from 1973 onwards, among those aged 25-29 from 1976, and among those aged 30-34 from 1979. Initially, the decline in marriage rates and the rise in the number of divorces was fully compensated by the spread of extra-marital cohabitation. After 1981 or 1982, such compensation no longer occurred and the proportions in union fell at all ages (above 20 years). This was especially true for those aged 21-24 years, though the decline was a little more gradual than in the proportion of currently-married couples ; among the latter, change has been spectacular : the probability of an individual of either sex being married between the ages of 21 and 24 years was 50 % lower at the beginning of 1986 than in 1972, and no change in trend is noted.
43Figures 3a and 3b show changes in the difference between the two proportions plotted in Figure 2 : they illustrate extra-marital cohabitation rates. The growth has been spectacular at ages 21-24 years where the proportion increased from 2 % in 1968 to 22 % in 1984 for women, and reached 17 % in 1983 for men. Owing to the usual age difference between spouses, this change became more rapid at ages 18-20 years for women and 25-29 years for men. These diagrams also show that a reversal in trend occurred in 1983-84 : the proportions for both men and women aged between 21 and 24 years have fallen noticeably since then, and several other rates also seem to show the same shift [12].
Proportions of women living in union and currently-married women, by age (1968-1986)
Proportions of women living in union and currently-married women, by age (1968-1986)
Proportions of men living in union and currently-married men, by age (1968-1986)
Proportions of men living in union and currently-married men, by age (1968-1986)
Proportion of women living in consensual unions by age (1968-1986)
Proportion of women living in consensual unions by age (1968-1986)
Proportion of men living in consensual unions by age (1968-1986)
Proportion of men living in consensual unions by age (1968-1986)
44We seem to have entered a new phase, when extra-marital cohabitation rates could stabilize during the next few years.
A relation with the age when the young leave the parental home
45An increase in the number of young persons aged between 21 and 24 years not living in a union has been noted since 1980-1981. This is probably not due to deliberate choice. Other results of our survey have shown that the number of young persons in this cohort who have left their parental home has been diminishing for some years. The shift occurred between 1975 and 1980, and is mainly responsible for the decline in the proportion of young men and women not living in a union at that age (see Tables 10 and 11 and Figure 4 in the original article).
IV – Marriage intentions and opinions of cohabiting partners
46What does "cohabitation" mean for unmarried couples ? Is it simply a period of transition which "normally" ends in marriage or is it a durable way of life meant to replace marriage ? Until the mid-1970s, it was but a stage in the history of couples, those who remained together longer ended up by marrying : in 1985, only 1 % of the couples who had been together for more than ten years had not legalized their union. Today, more and more couples do not marry and the fact that 30 % of them have been living together for more than five years suggests that cohabitation is no longer a mere prelude to marriage but a type of union in its own right which is increasingly replacing marriage (some of these 30 % will ultimately marry, but it is possible that they will be few in number since the average duration of pre-marital cohabitation is two years).
47How do cohabiting partners envisage the future of their relationship ? From the very beginning of their union, do they or do they not already have well-defined plans ; and is their opinion likely to change while cohabiting ? The replies to two questions asked in the survey provide partial answers. We asked cohabiting partners whether "personally" they wanted to get married, or whether they preferred to live together without considering marriage, or whether they rejected marriage. The attitudes reported have no absolute value as predictions (except perhaps for the 5 % who had already planned their wedding, or knew when they wanted to marry) : the attitude may not be shared by their partner and might only lead to action under different circumstances, or it may change in time. But the intentions of cohabiting partners express, perhaps as much as their behaviour, their attitude towards the institution of matrimony.
48We also asked the same question about any plans the partners might have made at the beginning of their union : these are no more than memories and, intentionally or not, respondents may have distorted them to project a coherent image of themselves or to justify their behaviour. However, a comparison between attitudes expressed at two different moments of time during the union allows us to discover whether a successful experiment [13] encourages couples to modify their attitudes or whether it reinforces them.
49Two-thirds of cohabiting partners either rejected marriage or did not consider it at the beginning of their union. Table 12 makes it possible to compare the attitudes expressed at the beginning of the union (column 1) and those expressed after having cohabited, on average, for nearly four years (column 2 : attitudes expressed at the time of the survey). Before going on to examine the reasons which might have prompted couples to change their plans, we consider their present intentions.
The present intentions of cohabiting partners : fewer than half envisage marriage
50Only 26 % of cohabiting partners wished to legalize their union quickly by marrying, 17 % planned to do so after a "trial period", half of them (50 %) "did not think" about it and 6 % rejected the very idea (Table 12, column 2).
51We have already noted that cohabitation is particularly frequent in younger cohorts and that it most frequently ends in marriage. In fact, many of them plan to marry : half of the under-25 age group envisage marriage, 37 % plan to marry quickly and 17 % after a trial period. Among those aged between 25 and 30 years, 48 % of cohabiting partners intend to marry, but half of them prefer to wait.
52Among those over 30 years old, couples "settle" more frequently for consensual union, without really rejecting marriage : even among those aged 35 years and over, only 15 % reject the possibility.
Marriage intentions of cohabiting partners : 1. at the beginning of their union 2. at the time of survey
Marriage intentions of cohabiting partners : 1. at the beginning of their union 2. at the time of survey
* Column 1 : age at the beginning of the union.Column 2 : age at the time of survey.
53In those social classes in which cohabitation is most widespread, it is no longer considered as a trial period, but as a substitute : 74 % of senior executives do not plan to marry ; this is also true of those with academic qualifications higher than BEPC (a certificate obtained after 4 years of secondary school attendance), and for 64 % of Parisians. Contrariwise, marriage is generally more popular among farmers (66 % plan to marry) and unskilled workers (54 %). But there is no perfect parallel between the frequency of cohabitation and "indifference" to marriage : middle level executives intend to marry as frequently (43 %) as do employees (43 %) and skilled workers (44 %). One of the most surprising findings is that in rural communes, the proportion of cohabiting partners who wish to marry is as low as among Parisians [14].
54The Church exerts greater pressure than society : few practising believers cohabit outside marriage, and those who do regard themselves only as anticipating their marriage.
55The number of couples who wish to marry is greater among those who have been living together for only a short time : 50 % of the couples whose union began less than five years previously consider marriage, compared with 26 % of those who have been together for more than five years ; this result should be compared with another : 93 % of cohabiting couples who finally legalized their union did so after less than five years. The differences related to the duration of the union reinforce those due to the age of the cohabiting partners, the duration of cohabitation being shorter where partners are younger.
56Previous marital experience has unexpected consequences ; when the specific effects of age and duration of the union are controlled, it would appear that more divorced persons [15] than never-married persons wish to marry, with the exception of divorced persons over 30 years old who have been cohabiting for less than five years (Table 13). Even the presence of a child from an earlier union has no dissuasive effect : 32 % of cohabiting partners over 30 years old who already have a child hope to marry compared with 30 % of those who are childless.
57These observations should not lead us to conclude that an individual’s family history does not modify his or her conception of the future. On the contrary, other survey results show that among persons living alone, fewer of the divorced than of the never-married wished to share their life with another, or to re-marry : 71 % of the never-married less than 30 years old who were not living in a union wished to marry some day, but only 47 % of divorced persons in the same age group. For those aged between 33 and 44 years, 54 % of the never-married, 31 % of the divorced without children, and 28 % of the divorced with children, hoped to marry. In fact, divorced persons do not object so much to the institution of marriage as to living again as a couple, since when choosing between consensual union and a marriage, they more often opt for the latter than do the never-married. One of the consequences of a divorce is that it does not leave those who have experienced marriage indifferent to it : more divorced persons than never-married either want it or reject it, and fewer of them are indifferent to it.
Proportions of cohabiting partners who hoped to marry quickly or after a trial period, by their matrimonial history
Proportions of cohabiting partners who hoped to marry quickly or after a trial period, by their matrimonial history
Changes in intention after a period of cohabitation
58Couples’ plans at the beginning of their life together either materialize or fail, i.e. they either marry or separate. To be able to trace changes accurately, we need to know the intentions of all cohabiting partners at the beginning of their life together, regardless of whether they have since married or separated, or are still living together without being married. We do not know the initial hopes of those who have separated, before the date of survey. As far as currently-married persons are concerned, nearly two-thirds reported that they already intended to marry when they started to live together. Most of the others admitted to not having thought about marriage, almost none rejected marriage. The very fact that they had married probably encouraged some to report that they had always hoped to do so rather than the contrary. In spite of this limitation, it is obvious that married persons have a more positive attitude to marriage than those who cohabit out of wedlock, especially among those who started living together after the age of 30.
59Cohabiting partners show greater stability in their attitude towards the institution of marriage than married persons : 73 % had not changed their minds since the beginning of their union or had remained consistent in their attitudes (after the "trial" period, they intended to marry) (Table 14, columns a + b + e). However, this stability is not as strong as it appears to be when we compare columns 1 and 2 from Table 12, because of the compensating effects of different changes of plan. In fact, those who began by being indifferent or hostile to marriage and turned to favour it (19 %) (column d), were more numerous than those who originally intended to marry and subsequently thought better of it (8 %) (column c).
60There are two decisive moments in the life of a couple. For those who did not intend to marry at the beginning of their union, the first months are the most crucial : 41 % of those who had lived together for less than 6 months, and who had no plans for marriage at the beginning, were already envisaging the possibility (columns d/d + e). This proportion then falls and remains close to 25 %, but there is a new rise in the numbers of those favouring marriage between the second and third year of cohabitation (d/d + e = 36 %). The trial period lasts a maximum of five years for those who treat cohabitation as a test. If at the end of this period, partners have neither married nor separated, they take the view that legalizing their union would not bring them anything new, and so give up their plans for marriage (column b).
61No demographic, social, religious or geographical characteristic can help explain why cohabiting partners change their minds and plan to marry after having lived together, sometimes for a number of years. Only the desire for a child seems to be a sufficient motive. Table 14 shows that people often consider marriage more favourably when they want to have a child, or when they already have one. Those who do not wish to have any more children, because they already have some from a previous union or because they consider themselves too old, made up their minds at the beginning of their union and never considered marriage (70 %). By contrast, other couples wait until they know each other better before planning to have children : without rejecting the possibility of marriage, they begin their life together without firm commitments "to try it out" or "without considering marriage". It would seem that the decision to have a child is taken simultaneously with or even before the decision to marry, even though the marriage usually precedes the birth. Eventually, 53 % of the women and 62 % of the men who want children, wish to marry (35 % of the women and 46 % of the men want to marry "quickly", 18 % and 16 % after a "trial"). By contrast, only 20 % of the women and 31 % of the men who did not want children think that they will marry some day, and 26 % and 12 % among them reject marriage (Table 15).
Changes in couples’ marriage intentions since the beginning of their union
Changes in couples’ marriage intentions since the beginning of their union
62Couples who decide to have a child without getting married (28 % already have a child) seem to be those with most reservations about marriage. At the beginning of their cohabitation, 73 % did not consider marriage (Table 14) ; the birth of a child often changes their attitudes and they become more favourably inclined to marriage (22 %), but the proportions who look at it unfavourably (12 %) are also higher than in other groups.
63To gain a better understanding of the attitude of consensual partners about the institution of marriage, we have tried to analyse the relation between the desire to have children and the partners’ marriage. Do both have a common cause, such as the desire to strengthen their links, or does one lead to the other and, if so, which is cause and which effect ?
64We showed consensual partners a list of the advantages and disadvantages of marriage and asked them for their opinions. The only motive that would lead more than half of them to marry was the good of children. Among those who wanted children, two out of three declared that this would be the decisive reason for them. More of the men less than 30 years old were willing to marry for this reason than women. But men aged 30 and over considered this reason to be less important. The only consensual partners who thought that it was unnecessary to marry in order to have children were those who already had a child… : only 34 % of them considered this to be an important reason and, as we have seen, 37 % in fact expected to marry.
6533 % of the cohabiting partners felt that the principal disadvantage of marriage was that it made an eventual break-up more difficult. Only one out of five cohabiting partners was bothered by any of the other disadvantages which also figured on the list (see Tables 16 and 17 in the original article).
66The analysis of cohabiting partners’ intentions showed that most did not think of marriage, that very few rejected it and that those who wished to marry were the ones who desired children. Their reaction to a series of positive and negative propositions concerning marriage leave an identical impression : although they have no good reason for refusing to marry, they see no point in solemnizing their union. Only 20 years ago, it was difficult to escape marriage, those who did so had to have very strong reasons ; today, there is no single model for childless couples and the trend is reversing : it is the couples who marry who have to justify their choice. Later, however, the desire to have children becomes a decisive factor for half of the cohabiting partners, and 85 % of cohabiting partners already have a child or wish to have one with their current partner or to have one later with someone else.
67A child is thus often the immediate cause of marriage. This has been confirmed by the opinion of cohabiting partners on the "best way to live as a couple" (Table 18) : 73 % feel that it is best for a couple who do not want children to live together without being married, 11 % consider marriage to be the best solution, 10 % opt for separate residences. On the other hand, only 44 % feel that even if a couple have a child, there is no need to get married and only 1 % advise parents to maintain separate residences. In fact, the couples’ intentions and opinions are often contradictory depending on their circumstances. Those who wish to get married do not always consider this solution to be the ideal, even for couples with children (14 % think otherwise). Conversely, among those who do not intend to get married themselves, one out of ten believe that even a childless couple should marry.
Only one-tenth of couples do not want children
6828 % of unmarried couples already have a child of their union, 38 % would like to have one, others feel that it is too early to think of it (14 %), some have never discussed the subject (2 %) or disagree on this point (4 %) (Table 19). Finally, only 10 % of cohabiting couples mutually agree to remain childless : many of them only met after the age of 30 and were already over 35 years old at the time of survey, and most had children from an earlier union. It should be noted that of this 10 %, 78 % did not want any children (85 % of the men and 73 % of the women), 5 % were unable to have any more, and 15 % (men 12 %, women 18 %) did not wish to have children with their current partner, but also expressed the intention to have a(nother) child some day. As for couples who have no definite plans (spouses disagree, do not discuss the subject or think that it is too early…), most of them are likely to have a child : 72 % of the surveyed persons expressed a personal desire to have one.
"When two persons have a steady, sexual relationship and : 1) do not want any children, 2) would like to have a child together, their best form of action would be…" (cohabiting person’s answer)
"When two persons have a steady, sexual relationship and : 1) do not want any children, 2) would like to have a child together, their best form of action would be…" (cohabiting person’s answer)
69In total, two-thirds of cohabiting partners are determined to have a child : the percentage is lower than for currently-married persons (90 % already have a child and 6 % would like one), but higher than for single persons with a friend of the opposite sex and who wish to live with him/her (55 % of such couples already intend to have a child). However, the role of cohabitation cannot be reduced to an intermediate stage in the formation of a couple, somewhere between celibacy and marriage. This may be the case for those already married, or for the majority of the young people who have been living together for a short time. But when we compare the marriage plans of cohabiting partners with those of single persons involved in a sexual relationship, we find a fraction of cohabiting couples who consider consensual union as a lasting alternative to marriage. When a couple desire to have a child (or already have one), only 33 % of the cohabiting partners want to marry quickly, 16 % want to wait a little to make sure, but these percentages rise to 51 % and 36 % respectively among those who have yet to begin their life together. Where couples do not want children, only one-quarter of the cohabiting partners plan to marry, compared with two-thirds of those who are still living alone (Table 15). In other words, a number of those who want to live in extra-marital cohabitation do not have immediate plans for marrying, regardless of their fertility intentions. They do not reject marriage and if plans regarding children were "activated", the same could hold for marriage plans. Thus, cohabitation appears as a new step with many potentials, the outcome of which cannot be forecast.
"Are your partner and you planning to have a child together ?" (unmarried couples)
"Are your partner and you planning to have a child together ?" (unmarried couples)
Overview
70The survey confirmed our original hypothesis that the very object of our study (the couple) was difficult to define. Not only do today’s couples consist increasingly of two persons who are not married to each other (13 % of those aged 21-44 years living as couples), but a couple do not necessarily "cohabit" or "co-reside" : between 7 and 8 % of the unmarried couples and between 2 and 3 % of married couples had two separate residences at the time of survey. This might be for career reasons or a voluntary decision with or without the intention of living together one day. Moreover, one-quarter of the men and one-third of the women who were not living with a partner admitted, when questioned, to having a steady, sexual relationship with a partner, which could in some cases be a preliminary to living together or might constitute a specific type of union which would allow a great degree of freedom to both.
71During the last ten to fifteen years, the fall in marriage rates and the rise in divorce rates have lowered the proportion of persons living as married couples. The decline is especially noticeable at ages 21-24 where it has fallen by half between 1972 and 1986. However, until 1982-83, this decline was completely compensated by the increase in extramarital cohabitation, so that the probability of living as a couple varied very little. The only exception was the 21-24 years group : from 1980, the propensity to live as a couple has declined in this group, but this change is fully explained by delay in leaving the parental home. In other words, these young people do not tend to live alone (out of union) outside the parental home, but rather live out of union at home. Changes in the labour market do play a role here, as is suggested by the high frequency of unemployment among men who still live with their parents.
72Thus, a very rapid rise in extra-marital cohabitation was observed from 1975 until 1982-83, but the rate remained constant thereafter and even fell at ages 21-24. On January 1, 1986, 66 % of the respondents (aged between 21 and 44 years) were living as married couples and 10 % as unmarried couples. The latter proportion exceeds 19 % for women between 21-24 years and is close to 17 % for men between 25-29 years ; owing to the rapid decrease in these rates for older cohorts, cohabiting partners surveyed were, on an average, significantly younger than married ones. Among them, one-quarter had been married before, but this was true of more than half of those aged 35 and over. Among those not living in union, one-tenth of the men and nearly one-third of the women had already been married.
73Therefore young cohabiting partners are mostly never-married, and older ones are mostly divorced or separated. As a result, there is a big difference in the number of years already spent in union and even more in the number of earlier unions : however, the last figure hardly ever exceeds 1 in the case of the never-married (1.2 on an average), and 2 in the case of divorced persons (2.1 on an average), which shows that choice of a non-marital union does not necessarily imply conjugal instability. Moreover, never-married cohabiting partners had generally started their first union later than those currently-married or divorced cohabiting partners.
74For those in consensual union at the time of survey, the average duration of cohabitation did not generally exceed three years for those aged 25-29 years and 5 years for those aged 30-39. This is obviously due to a number of extra-marital unions later being legalized. If we were to tackle the problem from the other end, we would notice that the rise in pre-marital cohabitation has been very rapid : around 25, the age at which most people marry, the pre-marital cohabitation rate rose from less than 10 % (before 1970) to nearly 60 % (after 1980). At the same time, the average duration of this phase seems to have lengthened to two years in recent years.
75Finally, these unions do not remain entirely infertile. At each age, the mean parity of ever-married cohabiting partners is slightly higher than that of currently-married persons, but most of their children were born in an earlier marriage. The picture is different for never-married cohabiting partners : for example, the mean parity of women aged 30-34 in this situation only amounts to 38 % of that of married women of the same age. However, this difference is partly due to the smaller number of years spent in union : if this were corrected, the relative level of fertility would reach 63 %.
76When questioned about their future plans and intentions they had had at the outset of their union, the cohabiting partners seemed to be rather empirical in their thinking… At the time of survey, only one-quarter envisaged marrying rapidly, half of them had not given thought to the matter. Rather logically, the longer the cohabitation, the weaker is the inclination to marry, since those who are most determined to marry do so during the first years of their life together. As a result, initial plans (made at the outset of their union) appear quite different, depending on whether the union has already been legalized by marriage or not : two-thirds of married persons expected to marry rapidly right from the beginning of their life together, seven out of ten currently cohabiting partners did not have that expectation and most of them did not change their minds. Those currently married have no trouble in finding reasons to justify their choice, whereas cohabiting partners find it difficult to name "the major or decisive" factor that would lead them to marry. The other partner’s strong desire to marry is often quoted as a reason, but the "children’s good" is the most decisive : thus, marriage plans are often linked to plans to raise a family ; it is as though one activates the other, but it is neither easy to determine which is planned first, nor is it easy to say which of them materializes first…
77The overall impression conveyed by this preliminary analysis can be summed up thus : non-marital cohabitation most often constitutes a somewhat "neutral" phase, where neither marriage nor its rejection, nor the desire for children nor its rejection figure. This does not imply instability which is not welcome : consensual unions do not mean free love. It is simply a desire to leave one’s options open, at least symbolically : even if couples decide to live together at an age when it is (or was) more "normal" to marry, they do not see the need to be tied down by such a formality, so long as outside constraints do not eventually make it necessary to get married.
Notes
-
[*]
Translated by Nita Lery.
-
[**]
A series of articles has recently been published in Population, giving detailed descriptions of the changes observed in several developed countries.
-
[***]
INED.
-
[1]
Louis Roussel, "La Cohabitation juvénile en France", Population, 33 (1), 1978, p. 15-42.
-
[2]
INED, "Douzième rapport sur la situation démographique de la France", Population 38 (4-5), 1983, p. 699.
-
[3]
P.A. Audirac, "Cohabitation and marriage : qui vit avec qui ? " Economie et Statistique, n° 145, juin 1982, p. 41-59.
P.A. Audirac, "La cohabitation : un million de couples non-mariés". Economie et Statistique, n° 185, février 1986, p. 13-33. -
[4]
A first presentation of the family situations of these children appeared in January 1988 in Population et Sociétés : "Entre père et mère", by H. Leridon and C. Villeneuve-Gokalp.
-
[5]
The original paper contains an Annex in which the need for an upward adjustment of the estimates is explained. This has not been included in this English version : interested readers should refer to the original article.
-
[6]
cf. H. Leridon, "L’Analyse des biographies matrimoniales dans l’enquête sur les situations familiales", INED, Dossiers et Recherches, n° 19, 1988 and forthcoming in Population Studies, 1990.
-
[7]
See Table 13. Moreover, an analysis of the periods spent alone after a separation showed that a year after the break, 69 % of the men and 45 % of the women wished to live again as a couple (or were already in such a situation).
-
[8]
See ref. given in fn. 6 for a detailed cohort analysis.
-
[9]
An analysis based on all first marriages solemnized during the years under observation (regardless of marital status at the time of survey) yielded results very similar to those in Table 7.
-
[10]
Provided, of course, that the periods lived in consensual union, particularly in the case of childless unions, were properly reported. We should also stress that the birth of a child during a period of cohabitation may be linked with plans to marry, in which case the fertility rate would not be typical of "cohabitation without marriage".
-
[11]
Far more detailed analysis of fertility rates for unmarried couples can be found in Leridon, op. cit. in fn. (6).
-
[12]
The trends remain the same when we exclude the "non-cohabiting couples".
-
[13]
This part deals with all unmarried couples whether they share a dwelling or not. But it does not include those who wanted to separate, nor those who did not wish to live together. Here we are interested in finding out whether the cohabiting partners wanted to legalize their union or not. Break-ups will not be dealt with here.
-
[14]
We verified that the age distribution of cohabiting partners from rural areas was the same as that of Parisians. But both distributions differ from those in the rest of France in that in both there is a smaller proportion of cohabiting partners under 30 years of age ; the figure for Paris is 19 % whereas that for the rest of France is 30 %, and the percentage is only 16 % for the rural communes.
-
[15]
For the sake of convenience, this group includes some individuals who have not yet been legally divorced. Among cohabiting partners, 76.9 % had never married, 21.2 % were legally divorced, 1.5 % were still married to a previous spouse and 0.4 % were widowed.
