1Conjugal separation is an increasingly frequent event and occurs at an ever earlier stage in life (Daguet, 1996). Demographers have examined the differences between men and women in the intensity of and length of time to new union formation, and in the age difference between partners in new unions (Bozon, 1990; Festy, 1991; Villeneuve-Gokalp, 1991). A study based on data from the Etude de l’histoire familiale family history survey of 1999 (Cassan et al., 2001) reveals that a man has a 23% greater chance, all other things being equal, of entering a new union than a woman. The age at conjugal separation is the key variable: before age 35, women and men form new unions much more frequently than after that age. All other things being equal, women who do not have a young child are only slightly more likely to form a new union (7%) than those who have one or more young children. Though the probabilities of entering a new union are relatively well known, we know little about the way in which individuals’ emotional and sexual life is reorganized immediately after a conjugal separation.
2The reorganization of emotional and sexual life after a separation is a preliminary to repartnering, as immediate transition from a first to a second union is uncommon. Very few people refuse to envisage a new romantic relationship, though a larger number say they have no desire to enter a new union (Villeneuve-Gokalp, 1991). Starting a relationship with a new sexual or dating partner does not always lead to union formation, either because the partners are satisfied with a non-cohabiting dating relationship, or because the necessary conditions are not satisfied. Determining the ways in which these new dating relationships are formed sheds light not only on one of the preliminaries to repartnering, but also, more broadly, on the organization of a post-conjugal dating market.
3In this article, we will describe, by age and sex, the factors influencing the pursuit of sexual and romantic relationships after a conjugal separation by examining three aspects in turn. The comparison between men and women is the common thread that links them. Like repartnering, is the pursuit of sexual and dating relationships easier for men than for women? One objective is to identify pertinent indicators for comparing the situations of men and women in terms of access to new partners. A second question concerns the circumstances of separation. How does the presence or absence of parallel relationships before the demise of the couple influence subsequent sexual activity? Is the situation the same for men and for women? Lastly, a third question concerns the way in which individuals adapt to the dating market (Bozon and Héran, 1989). The problem is as follows: while up to age 25 or so, a minority of individuals live with a partner and a large proportion of the people encountered in daily life are potentially available for dating relationships, the situation is rapidly reversed as age increases. After age 25, the vast majority of potential partners are in a union or a steady relationship. Following conjugal separation, individuals looking for a dating partner find themselves in a much “narrower” market. How do they adapt to the situation? Are they willing to date someone who is already in another relationship? In particular, do market constraints affect men and women equally?
4After describing the survey and the study population, we will examine the sexual context in which conjugal separation occurs. We will then present the various trajectories involved in meeting a first post-conjugal sexual partner and the characteristics of this partner. Lastly, we will describe the more complex sexual trajectories of cases where there are several partners after separation.
I – The data: a survey module on sexual activity after conjugal separation
5Few data are available to study sexual activity after conjugal separation. Sexuality surveys are scarce and rarely involve the collection of event history data. Since the emergence of AIDS, the number of surveys has increased slightly, though they tend to focus on public health issues (Spira and Bajos, 1994). This is the case for the survey that we used. It forms part of the series of ANRS-KABP AIDS surveys conducted regularly by the Paris region public health observatory (Observatoire régional de la santé d’Île de France) [1] to monitor changes in knowledge and attitudes relating to HIV-AIDS among the general French population, and in various indicators of sexual behaviour. Interviews are conducted by telephone with random selection of numbers, including unlisted ones.
6We used the 2001 survey on a sample of 3,321 people aged 18-54, which included in that year a specific “Separations” module administered to the sub-sample of persons “who separated finally or temporarily from a person with whom they formed a cohabiting couple” in the five years preceding the survey. The main aim was to examine how the risks of sexuality were managed by this population. It was hypothesized that for people who had previously been in a committed relationship, it might be more difficult to modify sexual behaviour with a new partner to avoid a potential risk of HIV contamination. One of the main findings was that a large proportion of respondents – though variable by age and hence by cohort – had used a condom for first sexual intercourse with a new partner after separation (77% among the 25-34 age group, 55% among the 35-54 age group) but, as in the case of sexual intercourse among young people, condom use was very quickly abandoned (Beltzer et al., 2002).
7Important questions for studying the repartnering process (age difference between partners, or the fact of having had children with the previous partner) were not covered by the survey. But the survey did collect detailed data on sexual trajectories after separation. They included the sexual circumstances of the separation, sexual activity after separation, a description of the characteristics of subsequent partners and of the duration of relationships, along with the means of protection and contraception used. The choice of a five-year reference period prior to the survey reduces recall problems and makes it possible to collect accurate data on the years immediately following separation. However, it does not provide a means to follow the majority of respondents until the formation of a second union, if any.
8In the 2001 ANRS-KABP AIDS survey, among men and women aged 25-54, a total of 12%, i.e. 425 individuals, reported having separated for at least six months in the last five years from a person with whom they had been cohabiting for at least six months. The men and women who had most frequently experienced separation were the youngest: 18.4% of men and 17.7% of women aged aged 25-29 at the time of survey had experienced separation, versus 7.0% and 7.5% respectively among those aged 40-54. The decrease with age in the proportion of respondents who have experienced a separation in the last five years reflects a cohort effect as much as a life-cycle effect. The conjugal history of individuals is today more strongly marked by the experience of separation than in the past, and at an earlier age.
II – The sexual context of separation: presence or absence of extra-conjugal relations
9The population selected for this study includes persons who reported having experienced a separation in the last five years. This separation, by definition, involves decohabitation of the two partners.
10In which sexual context does conjugal breakdown occur? The questionnaire asked respondents about the possible existence of parallel sexual relationships at the time of separation. It did not aim to identify the factors behind separation, and cannot be used to determine whether breakdown was caused by an extra-conjugal relationship. In any case, such a question would be very difficult to answer [2]. The purpose of the questions on this point was simply to shed light on the context in which separation occurred by indicating the proportion of breakdowns in which, according to the respondents, extra-conjugal relationships had preceded the act of separation. Two questions were asked, one concerning the respondent (“Just before this separation, did you have sexual intercourse with anyone other than your partner?”) and another concerning the partner (“Just before this separation from your partner, did he/she have sexual intercourse with anyone other than yourself?”). The question on the partner could be answered in three ways (“yes I’m sure”, “probably, but I’m not sure” and “no”). The answer concerning the partner is of a different nature to the one concerning the respondent: it corresponds to the respondent’s perception of the partner’s behaviour, though this perception may have influenced the respondent’s own behaviour.
11At the time of separation, two respondents in three report the existence of extra-conjugal relationships, concerning either the respondent him/herself, the partner, or both partners. When reporting about him/herself (see Table 1), the proportions of men and women who claim to be personally “unfaithful” are similar, with around 30% saying that they had had sexual intercourse with someone else just before separation. This apparent similarity between men’s and women’s answers tends to disappear when age is taken into account. Young women (ages 25-34) more often claim to be in parallel relationships when they separate than their male counterparts of the same age (34.6% versus 21.6%). The reverse is observed in the 35-54 age group: men personally report twice as many extra-conjugal relationships at the time of separation as women (41.2% versus 21.9%).
Respondent’s extraconjugal relationships before separation, and partner’s perceived extraconjugal relationships (%)
Respondent’s extraconjugal relationships before separation, and partner’s perceived extraconjugal relationships (%)
12However, respondents consistently see their partner as more often involved in extra-conjugal relationships than they admit for themselves: at the time of separation, 54% of women and 44% of men believe that their partner was having a parallel relationship. The partners “infidelity” is generally suspected (answer: “it’s likely, though I’m not sure”) rather than known for a fact (answer: “yes, I’m sure”). Among younger people, a similar proportion of men and women (40% and 42%) claim that their partner was involved in an extra-conjugal relationship. Conversely, in the 35-54 age group, the gap is wider, with many more women than men claiming that their partner had certainly or probably had sexual intercourse with another person before separation (64% of women versus 49% of men). Individuals tend to point the finger at their partners rather than at themselves. And men are more often accused than women. There are real differences in behaviour between men and women, probably quite accurately reflected in individuals’ claims about their own behaviour (notably among the over-35s), though these differences are amplified by representations, clearly expressed in answers concerning the partner’s behaviour, notably in lower SES groups (findings not presented).
13We will examine two situations typical of separation (see Figure 1): the situation in which the respondent reports no extra-conjugal relationships and thinks his/her partner is also faithful, or the reverse situation in which the respondent reports having extra-conjugal relationships and thinks that his/her partner does too. It is mainly among younger couples that separation occurs without extra-conjugal relationships on the part of one or other partner. Among the 35-54 age group however, a much larger proportion of separations involve extra-conjugal relationships, notably according to the answers given by men (graph on the right). These age-specific differences are probably linked to the length of the union. Moreover, men report more often than women that they are in the same situation as their partner, while women more often report a divergence of situations (one is faithful, the other not).
Respondent’s extraconjugal relationships and partner’s perceived infidelity (%)
Respondent’s extraconjugal relationships and partner’s perceived infidelity (%)
14All these situations affect subsequent sexual activity, though their consequences are not directly predictable. Not all separations are linked to the existence of parallel relationships, and the experience of parallel relation-ships does not always lead to the demise of the central relationship [3].
III – The process of meeting the first new partner
15Separation is followed by a period of reorganization of one’s emotional and sexual life, a process which may be practically instantaneous or more drawn out in time. At the time of the survey, i.e. two and a half years on average after separation, 82% of men and 80% of women had had sexual intercourse with at least one new partner. Predictably, the more distant the separation, the fewer the persons who have not yet met a new partner. The process of meeting the first new partner can be described using two indicators: the timing with regard to the previous conjugal relationship (overlapping or sequential), and the duration of sexual inactivity, if any, between conjugal separation and first sexual intercourse with a new partner.
1 – Overlapping or sequential timing of post-conjugal relationships
16Around ten questions were asked about the first partner. The first one was as follows: “Now let’s talk about the person or the first person with whom you had sexual intercourse after this separation. It may be a partner with whom you had already started a relationship before separating from X. When did you first have sexual intercourse with this partner?” Another question provides information on the timing of this relationship: “How long had you known this person before you first had sexual intercourse with him/her?” We thus obtain information on the date of first intercourse and the length of time the two individuals knew each other before first intercourse. Respondents were free to choose the person designated as their first partner.
17On the basis of these two questions (see Table 2), three different patterns in the timing of the transition between former partner and new partner are observed. Most frequently, a new partner is not met until after separation has already occurred: new partners met after separation correspond to three-fifths of cases for men and only two-fifths for women. In the second case, the new partner was met prior to separation, but the first sexual intercourse took place afterwards: new partners met before separation correspond to one-fifth of cases for men and one-third for women. In the third case, first sexual intercourse with the new partner took place before conjugal separation (or at the same time): these sexual relationships initiated before separation concern more than 20% of men and women.
Timing of the first post-conjugal relationship (%)
Timing of the first post-conjugal relationship (%)
18The respondents do not necessarily associate extra-conjugal partners (if any) with first partners after separation. One-third of persons reporting an extra-conjugal partner report that they met their first post-conjugal partner after separation had occurred. Conversely, two-thirds of men and women who say they met their partner before conjugal separation say that that did not have extra-conjugal intercourse before separation (data not presented).
19Contrary to a widely-held opinion, more men meet the first post-conjugal partner after separation, while more women than men start their post-conjugal life with a partner they met before the separation. One-fifth of men and almost a quarter of women report having sexual intercourse with the person reported as the first post-conjugal partner even before the demise of the previous union.
20The differences between men and women are even larger when age is taken into account. Two-thirds of young women meet their future partner before separation, compared with only one-third of men, whereas differences between men and women are smaller for individuals over age 35. If we limit our analysis to individuals of all ages who say they did not have extra-conjugal relationships before separation (70% of men and women who separate), we observe that only 22% of “faithful” men met their future partner before separation, compared with 55% of “faithful” women (data not presented). The post-conjugal love life of men is thus very distinct from that which preceded separation.
21In short, at the time of separation, women more often know the person who will become their new partner (without necessarily having had sexual intercourse with him), while men more often meet the woman reported as their first post-conjugal partner after separation has occurred.
2 – Duration of sexual inactivity between separation and first intercourse with a new partner: differences between men and women
22To determine the time before entry into a new sexual relationship, we created a general indicator, the median duration of sexual inactivity, which starts, by convention, on the date of conjugal separation and ends at the time of first intercourse with the new partner. For persons who had sexual intercourse with their new partner before separation of the previous couple, the duration is zero, since they did not experience a period of sexual inactivity. Given that 80% of respondents had met a new partner at the time of the survey, we used a duration model (Kaplan-Meier model) which can be used to estimate median durations (Table 3), taking account of those who have not yet resumed any sexual activity (right-censored).
Duration of sexual inactivity in months
Duration of sexual inactivity in months
23The duration of sexual inactivity indicator gives a more expected finding than the previous indicator on the succession of relationships over time. Women generally take longer to begin a new sexual relationship after separation (six months compared with three for men). This median duration of sexual inactivity is higher for women, even though more women have a zero score (i.e. they began a relationship with the first post-conjugal partner before separation). This is because those who had not formed a new relationship before separation spend longer looking for a new partner than men in the same situation.
IV – First partners already involved in other relationships
24We will now look at the characteristics of the first post-conjugal partner and the type of relationship that is formed. Only a minority of first partners after separation will eventually form a couple with the respondent. At the survey date, 72% of men’s relationships with first post-conjugal partners had already ended, and for women the proportion was 59%. For those who had a first partner after separation, 16% of men and 25% of women reported being in a couple with this person at the time of the survey. The median duration of the relationship with this partner after first sexual intercourse, including those still ongoing at the time of survey, was 5 months for men and 12 months for women (Kaplan-Meier method). The fact of meeting the first partner before or after separation has no effect on the duration of the relationship (findings not presented). Several characteristics of the first partner suggest that he/she does not necessarily have the profile of a future cohabiting partner.
25One of the questions asked about this partner was as follows: “When you first had sexual intercourse with this partner, did you think that he or she was already in a relationship with one or more other partners?” The choice of answers was: “yes, I’m sure”, “probably, but I’m not sure”, and “no”. For men and women alike, half of all new partners were thought to be in a relationship with someone else when the respondent began the relationship with him/her (Table 4). Persons aged 35-54 more often enter into relationships with someone who already has a partner (61% of first partners for men and 52% for women) than younger persons (43% and 45% respectively).
Proportion of first post-conjugal partners already involved in other relationships
Proportion of first post-conjugal partners already involved in other relationships
26The status of the respondent’s new partner appears to be closely linked to the sexual context of separation (Table 4): when the person and their former partner both had extra-conjugal relationships at the time of separation, the first new partner met is more often perceived to have a relationship with someone else – in 70% of cases – and although the numbers are small, the tendencies for both men and women are similar. Conversely, when there were no extra-conjugal relationships at the time of separation, individuals more frequently enter into relationships with available new partners (in 60% to 70% of cases). Persons belonging to couples with no extra-conjugal relationships before separation are also those who take longest to find a new partner, with 30% having met no-one at the time of the survey (data not presented). In brief, and quite logically, it takes longer to find a first partner who is available than one who is not. The stricter the selection criteria, the longer it takes to find a new partner.
27At the start of the relationship with their first new partner, women more often claim to be in love (deeply or moderately) than men (64% versus 46%), whatever their age (data not presented). Behaviour is thus very similar to that already observed among young people (Bozon, 1996), though around one woman in two and two-thirds of men regard the person they met as simply a casual partner [4]. It is uncommon, among young people especially, to claim to be in love with partners considered as casual. Hence, it is remarkable that at the start of their first post-conjugal relationship, women often report being in love with a partner already involved in another relationship. We can probably conclude that the separation process and the constraints of the dating market tend to orient the individual attitudes of men and women towards greater acceptance of non-committal sexuality (Bozon, 2001).
V – Older women have a lower partner turnover
28Respondents were asked about their number of partners since separation. The data are not suitable for conducting a valid duration analysis. The ratio of the number of reported post-conjugal partners to the time elapsed since separation (at least six months, by construction), gives us an indicator of the “partner turnover” which includes four categories: no partner; just one; fewer than one every six months; at least one every six months [5]. To present the findings, we also distinguished between recent separations (in the last six months to two years) and separations that occurred more than two years previously.
29The differences between men and women are clear, though sometimes unexpected (see Table 5). For example, in the case of separations dating back less than two years, almost one-third of both male and female respondents have had no new partner. Yet at the same time, while the majority of women report having (had) just one partner, the majority of men have already had several (more than one every six months). This suggests that the experiences in the period immediately following separation differ greatly between individual men, and likewise between men and women. As the time elapsed since separation increases, the proportion of those who have not found a partner decreases. Among those who separated between two and five years previously, the proportion of partnerless persons dwindles rapidly among men (3%), while remaining at a non-negligible level among women (15.4%). For the rest, behaviours start to converge and the largest group, for both men and women alike, is now the subset of persons having fewer than one partner every six months, with many women reporting just one partner (40%). A small share of men continue to have many partners, though in most cases their sexual life becomes more stable.
Degree of partner renewal for men and women by time elapsed since separation (col. %)
Degree of partner renewal for men and women by time elapsed since separation (col. %)
30More generally, it is young men and women who have the largest number of partners. For men aged over 35, partner renewal is much more common than for women of the same age group, among whom such behaviour is the least frequent. Only 37% of women aged 35-54 versus 71% of men have had two partners since separation.
31The sexual context of separation (extra-conjugal relationships or otherwise) does not appear to induce a specific attitude among women in terms of partner numbers or renewal: women who have had extra-conjugal relationships do not have more partners after separation than those who have not (data not presented). The situation is apparently different among men, though the sample size is small. When separation occurs in a context where both partners had extra-conjugal relationships, practically all men report having had several subsequent partners, whereas the number is much smaller when separation occurred without extra-conjugal relationships.
VI – Adapting to the new dating context and diverging experiences of men and women after age 35
32When we decided to analyse a public health survey to study the outcomes of conjugal separations from a sociological and demographic viewpoint, we were well aware of the limits of the data we used. The ANRS-KABP AIDS survey does not tell us whether the persons who separate have children, and we do not know the age of the new partners they meet. We do have accurate information on the sexual context of conjugal separation however, and on how the respondents meet their new partners. This enabled us to examine the largely unstudied theme of the pursuit of sexual activity after separation, an essential preliminary, though not sufficient in itself, to the subsequent formation of a new union.
33There are similarities between men and women in the organization of their sexual activity after separation. Similar proportions of men and women aged 25-54 who have separated from a partner within the last five years report having had sexual intercourse with other partners shortly before separation (30% of men and 27% of women) [6]. In addition, around 80% of men and women have had at least one new sexual partner in the first few years after separation. Lastly, in half of all cases, this first sexual partner is perceived to be already involved in other relationships. In short, most individuals pursue their sexual activity after separating from a cohabiting partner, but a large proportion of their first new partners are not available for an exclusive relationship (or not immediately). Note that this experience of new partner non-availability is equally frequent for men and for women. This factor helps to explain why such a small proportion of respondents rapidly enter a new union. Individuals are obliged to adapt to the realities of a dating market very different from that of young people.
34Despite certain similarities in the situations of men and women, their experience after separation differs in many ways, notably when the speed and method of meeting new partners and the age at separation are taken into account. More men than women did not meet their first new partner until the previous relationship was over (61% and 43% respectively), while women, young women especially, have more frequently already met their future new partner before conjugal separation [7]. When separation dates back less than two years, there are practically as many women as men (one-third) who have not yet met a new partner. For more distant separations, things are different, since among those who separated between two and five years previously, the proportions of partnerless persons drops to 3% of men while remaining at almost 15% for women, and 74% of men have had at least two partners, compared with just 45% of women. Women meet a new sexual partner slightly less often than men and, above all, much less often a second partner. This less frequent partner renewal is doubtless one of the key reasons why women take longer to form new unions and do so less frequently. In addition, a comparison of the romantic and sexual trajectories of men and women by age clearly shows that women under 35 pursue their sexual life just as often as men of the same age, and under very similar conditions. This is not the case for women aged over 35, who engage in post-conjugal sexual activity less often than younger women and less often than men of the same age. Age thus represents a specific constraint for the sexual activity of women, and likewise for their chances of forming a new union (Attias-Donfut, 2001).
35The study of sexual trajectories after separation reveals two distinct phenomena. First, the very first sexual or romantic relationships experienced by men and women after separation do not, in the majority of cases, herald the immediate formation of a new couple. They occur quite soon after separation, with partners often involved in other relationships, and frequently classified as casual partners. This capacity to dissociate sexuality and committed relationships, already strongly present among young men (Bozon, 1993; Lagrange et Lhomond, 1997), would appear to be spreading among separated women as well, the youngest women in particular. Conjugal separations bring about a marked change of attitude in the management of romantic relationships.
36A second phenomenon is the difficulty encountered by women over 35 in beginning relationships with new sexual partners, or their absence of desire to do so, compared with men of their age and with younger women. This is generally attributed to the “obstacle” represented by the presence of young children still living with their mother. But we saw in the introduction that this factor is a lesser hindrance to the repartnering of women after separation than is generally believed. And even less so for the pursuit of romantic and sexual relationships. The presence of young children is probably more frequent in the group of women aged under 35, and yet it is these women who most often maintain an active sexual life after separation. Other explanations need to be found therefore. A higher age also corresponds to a longer history of conjugal life. Among women, committed conjugal relationships may leave more subjective traces than among men, reducing their desire to meet a new partner in the immediate aftermath (Bozon, 1990, 1998). Moreover, women aged over 35 tend to be in competition on the dating market with younger women with less conjugal experience. Hence, for women over 35, not only does it become more challenging to form a new union, but the quest for new romantic partners also becomes more difficult and less appealing. This is not the case for men. Age thus has a gender.
Notes
-
[*]
Observatoire régional de santé d’Île-de-France
-
[**]
Institut national d’études démographiques
Translated by Catriona Dutreuilh -
[1]
ANRS-KABP (Agence nationale de Recherche sur le Sida – Knowledge, Attitudes, Beliefs and Practices) AIDS surveys were conducted in 1992, 1994, 1998, 2001 and 2004 (see Grémy, Beltzer et al., 2001).
-
[2]
Firstly, it is difficult to record extra-conjugal relationships, many of which are kept secret, and to measure their consequences. Secondly, when separation occurs in the wake of an extra-conjugal relationship, infidelity may perhaps be the consequence of disagreement or estrangement rather than its cause.
-
[3]
We do not have appropriate data to address the question of the propensity to dissolve a central relationship after a parallel relationship has been formed.
-
[4]
The question was formulated as follows: “What sort of partner did you consider him/her to be?” and the choice of answers was: “a cohabiting partner; a steady partner; a casual partner”.
-
[5]
The first partners after separation with whom the relationship had begun prior to separation (see Table 2) are included in this enumeration.
-
[6]
Clearly, this is not a measure of the tendency to have extra-conjugal relationships within a couple; not all extra-conjugal relationships are followed by separation. Neither does it mean that the extra-conjugal relationship should be viewed as the cause of the separation. It is seen here as a circumstance in the context of separation.
-
[7]
In this context, meeting a first partner before separation simply means making their acquaintance. Only in certain cases did a new sexual relationship begin before the end of the previous one (see Table 2).
