1 While there have been increasing numbers of studies in recent years on lesbian, gay, bisexual, and trans (LGBT) populations, comparatively few analyses are specifically devoted to people who identify as bisexual, as opposed to homosexual. Studies on bisexuality have identified two main factors that explain this relative invisibility [1, 2]. The first is the lack of social recognition of bisexuality with respect not only to heterosexuality but also to homosexuality. Bisexuality is often considered as a transitional identification on the way towards homosexuality or, on the contrary, as a circumstantial sexuality that does not challenge the individual’s heterosexuality.

2 Its invisibility is also linked to the difficulty of capturing it in statistical surveys. In France, analyses often focus on people with homo-/bisexual practices, without systematically differentiating between homosexual and bisexual trajectories [3]. This choice is due to methodological constraints and in particular to the relatively small number of individuals concerned. Independently of these constraints, however, statistical surveys often analyse bisexuality with reference to homosexuality and thus tend to treat bisexuality as a secondary sexuality [4]. But recent studies in the United States have shown that, while the proportions of those who define themselves as bisexual or homosexual vary between surveys, there may be as many individuals who define themselves as bisexual as homosexual, or even more [5].

3 Thanks to its large sample size and its attention to sexual minorities, the VIRAGE survey (Box 1) provides a wealth of information on the sociodemographic characteristics of these populations as well as their conjugal and sexual trajectories. It can be used to study the experiences of people who identify as bisexual and to compare them to those of people who identify as homosexual or heterosexual. Who are the individuals who describe themselves as bisexual, and what underlies this identification? What are the specificities of people who identify as bisexual in relation to those who identify as homosexual or heterosexual?

Identification as bisexual varies with gender

4 Three indicators from the VIRAGE survey can be used to define an individual’s sexuality: attractions, practices, and identification (Box 2). These three facets of sexual experience are relatively independent. Sexual attraction is not necessarily reflected in practices, and sexual practices with people of the same sex do not necessarily lead individuals to define themselves as homosexual or bisexual.

5 People who reported attraction to both sexes over their lifetime made up 2.7% of the sample (3.4% of female respondents and 2.0% of male respondents), while those who reported sexual practices with people of both sexes made up only 1.9% of the sample (2.2% of female respondents and 1.6% of male respondents). Most of the respondents who reported attractions to both sexes defined themselves as heterosexual: this was true of 65% of women in this category and 45% of men. Two-thirds of men who reported sexual practices with both sexes defined themselves as homosexual or bisexual, rather than heterosexual, while more than half of women who had engaged in sexual practices with both sexes defined themselves as heterosexual (Figure 1).

6 Attractions to, and sexual practices with, people of both sexes are thus not specific to those who identify as bisexual. The majority of those who describe themselves as homosexual (61% of homosexual women and 49% of homosexual men) have had partners of both sexes over their lifetime. A significant proportion of them reported being equally attracted to both sexes (17% of homosexual women and 9% of homosexual men; see Table). A behavioural definition of bisexuality, based on sexual practices with people of both sexes, thus does not take into account the differences in the value that individuals attribute to their sexual experiences.

Figure 1

Sexual identification of people who have engaged in sexual practices with both sexes over their lifetime, by respondent’s sex (%)

Sexual identification of people who have engaged in sexual practices with both sexes over their lifetime, by respondent’s sex (%)

Coverage: All people having engaged in sexual practices with persons of both sexes over their lifetime.
Note: Out of the 15,556 women and 11,712 men surveyed, 369 women and 238 men reported having engaged in such practices.
.

7 Identification as bisexual may thus be understood as a specific way in which individuals give a place to certain attractions and sexual practices in their self-definition. This identification can vary over the individual’s lifetime [6], but it can also be maintained by individuals who have sexual relations exclusively with members of one sex and even be reinforced as the individual gets older [7]. Moreover, this identification varies by gender. Slightly more women (0.9%) than men (0.6%) described themselves as bisexual in the survey. Women were more likely to report practices with partners of both sexes, reproducing a result found in other surveys [8]. Women who report such practices are less likely to identify as bisexual than men (Figure 1).

8 These gender variations can be understood in terms of gendered differences in sexual socialization and representations of sexuality. Expressing desire for, and engaging in sexual practices with, people of both sexes are relatively more accepted for women; these experiences may be considered an element of a feminine sexuality, particularly when they do not call into question a woman’s attraction to men [2]. However, it seems easier for men who report attractions and practices with both sexes to identify as bisexual, whereas women tend instead to identify as heterosexual regardless of their desires and their practices. Describing oneself as bisexual involves an affirmation of desires that may be more readily accepted for men.

Table

Sexual characteristics by sexual identification (%)

		Women			Men
		Bisexual	Homosexual	Heterosexual	Bisexual	Homosexual	Heterosexual
Sexual	Only to the same sex	1.4	48.2	< 0.1	4.1	51.7	< 0.1
attractions	Mainly to the same sex	4.6	30.7	< 0.1	7.5	36.5	< 0.1
	Equally to both sexes	34.1	17.5	0.2	29.2	8.9	0.1
	Mainly to the opposite sex	58.5	0.4	2	46.8	1.9	0.8
	Only to the opposite sex	0.7	3.2	96.2	12	1	97.8
	No response	0.7	0	1.6	0.4	0	1.3
Sex of sexual	Only same sex	0.3	35.2	0.1	4.4	45.4	0.4
partners over	Both sexes	76.5	61.4	1.2	88.7	49.5	0.5
lifetime*	Only other sex	16.7	1.1	88.1	3.5	0.0	91.6
	No response	6.5	2.3	10.6	3.4	5.1	7.5
Sex of first	Same sex	11.2	39.4	0.5	21.4	65.4	0.9
sexual partner*	Other sex	88.8	60.4	97.3	77.2	34.6	97.5
	No response	0.0	0.2	2.2	1.4	0	1.6
Mean age at sexual debut (years)*		16.5	18.0	18.3**	17.4	17.7	17.5**
Mean age at first sexual relations with a same-sex partner (years)*		22.1	21.3	22.4**	22.8	18.7	19.0**
Median number of same-sex partners***		2	4	0	3	20	0
Median number of other-sex partners***		7	1	3	7	0	5

Sexual characteristics by sexual identification (%)

Coverage: All individuals identifying as bisexual, homosexual, or heterosexual.
Note: Out of the 15,556 women surveyed, 135 identified as bisexual, 94 as homosexual, 15,297 as heterosexual, and 30 did not know or did not want to provide their identification. Out of the 11,712 men surveyed, 94 identified as bisexual, 180 as homosexual, 11,402 as heterosexual, and 36 did not know or did not want to provide their identification.
* Only persons who had already had their sexual debut.
** Certain persons who identify as heterosexual have had same-sex sexual partners.
*** Half of the group has had fewer partners than the median, and half has had more.

A sexuality predominantly oriented towards the other sex

9 A majority of those who described themselves as bisexual had had sexual practices with people of both sexes, but not all. The proportion was 89% for men and 76% for women (Table). A majority reported being more attracted to the other sex than to the same sex (58% of women and 47% of men), while only around a third (34% of women and 29% of men) reported being equally attracted to both sexes. What is more, 12% of bisexual men reported being attracted only to the other sex. Identification as bisexual thus does not necessarily imply undifferentiated attraction to both sexes. Bisexual identification most often accompanies a sexuality predominantly oriented towards the other sex, which these individuals’ sexual trajectories confirm.

10 While a majority of bisexual people have had partners of both sexes over their lifetime, these partners tend to be mainly of the other sex. Regarding sexual debut, 89% of bisexual women had their first sexual relations with a man, while 77% of bisexual men had their first sexual relations with a woman. The mean age at which both bisexual women and men first had sexual relations with a person of the same sex was 22 years. This does not distinguish bisexual women from homosexual women, whose first sexual experience with a woman took place at age 21. For men, however, this does distinguish bisexuals from homosexuals, who first had sex with a man at the age of 19 on average. Looking at the median number of sexual partners (which is less sensitive to aberrant data than the mean), bisexual women and men report similar numbers of partners, who are more often of the other sex, both for women and for men. These results challenge the notion that bisexuality should be understood specifically in relation to homosexuality, as a transitional period on the way to homosexuality or as denied homosexuality. Bisexuality can also be understood in relation to heterosexuality and its norms. From a sexual viewpoint, bisexual identification may be understood as a sexuality that is expanded to include same-sex partners but that often remains centred on different-sex partners. Because women’s and men’s sexual socializations continue to be distinct, this broadening is even more marked among bisexual women. They had an earlier sexual debut than other women and reported a larger number of sexual partners than heterosexual women. Bisexual identification among women is thus associated with sexual trajectories that deviate more generally from the norms for women’s sexuality.

Bisexual women are younger; bisexual men are less likely to have a partner

11 While those who describe themselves as homosexual (particularly men) tend to be younger and more educated, the situation of bisexual individuals is different. Nearly half of bisexual women are under age 30, and their mean level of education is higher than that of heterosexual women but lower than that of lesbians. These differences have also been found in other surveys [7]. This may be a sign that identification as bisexual is better tolerated at younger ages. Socially understood as a transitional identification, this may become de facto the case for women’s bisexuality, as they are obliged to give up an identification that is decreasingly tolerated as they become older. It may also be that women in recent cohorts have become increasingly likely to identify as bisexual. This difference would then reflect an increase in women’s sexual autonomy.

12 The situation of bisexual men is distinguished from that of both homosexuals and heterosexuals. More than half are over 50 years old; they have completed less formal education than heterosexual men and much less than homosexual men. We can hypothesize that in certain male cohorts and social groups, it has been easier for men to identify as bisexual than as homosexual. The gay lifestyles that developed during the 1960s and 1970s in France [9] are doubtless not always either accessible or desirable to all men who have sex with men.

13 Like their sexuality, the conjugality of bisexual people who are in a relationship is mainly oriented towards the other sex. The partner of bisexuals who are in a relationship is typically a person of the other sex. This is particularly the case of bisexual women, half of whom have a male partner. The proportion of bisexual women who do not have a partner (42%) is relatively high compared to heterosexual women (27%) but slightly lower than lesbians (44%). The proportion for bisexual men (70%) is very high, much higher than for both heterosexual (25%) and homosexual men (45%). This applies to bisexual men at all ages.

14 These results might reflect greater difficulties in finding a partner or a lesser interest in having one. Being single can be a choice; a portion of bisexual men may perceive conjugality as an obstacle to the fulfilment of their desires. These results may also reflect a conjugal vulnerability specific to bisexual men, as seen in the results of other surveys [10]. Identification as bisexual may have a deterrent effect, with these men socially considered as less trustworthy partners by both women and men.

15 The characteristics of people who identify as bisexual do not really suggest indifference about the gender of sexual or conjugal partners, but rather the broadening of a sexuality that remains centred on members of the other sex to include same-sex individuals as well. The high proportion of bisexual people who do not have a partner suggests that this broadening carries a cost. The differences between women’s and men’s bisexualities also show that identification as bisexual depends on gendered differences in the spaces of sexual possibility, as well as generational differences. Self-description as bisexual is not only the expression of an individual desire; it is an identification whose meanings vary with differences in sexual socialization, between groups and over time.