Birth Order and Sexual Orientation in a National Probability Sample.

We do not know what causes the development of sexual orientation, the erotic inclination/attraction for the opposite sex (heterosexuality), the same sex (homosexuality), or both sexes (bisexuality). However, there has been a substantial amount of empirical work and theory on this question in recent years (e.g., Bell, Weinberg, & Hammersmith, 1981; Bem, 1996; Blanchard, 1997; Ellis & Ames, 1987; Hamer & Copeland, 1994; LeVay, 1996; Singh, Vidaurri, Zambarano, & Dabbs, 1999). Scientific interest in this question occurs because sexual orientation is a fundamental aspect of human sexual variability, and thus researching this question can be an important pursuit for scientists interested in personality and individual differences. In addition, sexual orientation is related to some aspects of gender roles (Bailey & Zucker, 1995; Bell et al., 1981; Green, 1987; Zuger, 1988), and thus, as noted by Bailey, Dunne, and Martin (2000), research on the origins of sexual orientation may reveal the causes of sex differences. Furthermore, research on the origins of sexual orientation can lead to a better understanding of the mechanisms of sexual attraction in general, including a greater understanding of pathological sexual variation, conditions that cause major distress in society. For example, some of our recent work on sexual orientation development has directly led to new research ideas on the origins of pedophilia and other sexual disorders (see Bogaert, Bezeau, Kuban, & Blanchard, 1997; Lalumiere, Harris, Quinsey, & Rice, 1998).

BIRTH ORDER AND SEXUAL ORIENTATION IN MEN

One factor that seems to play a role in sexual orientation development in men is birth order, or the position one has within a family, be it first born, last born, or somewhere in-between. A number of studies have suggested that homosexual men have, on average, a later birth order than population norms or comparable groups of heterosexual men (for a review, see Blanchard, 1997). One of the more impressive elements of the birth order effect is its generalizability. Not only has the effect been demonstrated in samples of men from different eras (see Blanchard & Bogaert, 1996a, 1996b), from different countries (see Blanchard & Bogaert, 1996b; Blanchard, Zucker, Cohen-Kettenis, Gooren, & Bailey, 1996), and with different ethnicities (Bogaert, 1998), but it has also been found to occur in clinical (e.g., Bogaert et al., 1997) and nonclinical (e.g., Blanchard & Bogaert, 1996b) samples. Recently, we have demonstrated that the birth order difference between homosexual and heterosexual men is primarily the result of homosexual men being born later among their brothers, relative to heterosexual men (Blanchard & Bogaert, 1996a, 1996b, 1997; Bogaert et al., 1997).

Although the birth order effect seems to be one of the more robust findings in modern sexology, the effect has never been demonstrated in a representative sample of adults. To date, the samples have consisted of clinical groups of one kind or another (e.g., homosexual versus heterosexual pedophiles; see, for example, Bogaert et al., 1997) or nonclinical volunteer/convenience samples gathered by, among other techniques, soliciting participation from homophile and similar (predominantly) heterosexual organizations (see, for example, Blanchard & Bogaert, 1996b). These sampling strategies raise doubts about whether the birth order effect can generalize to a representative sample of adults, because volunteers for sex research studies differ in important ways from the general population. For example, volunteers for sex research are more sexually liberal and experienced (e.g., Morokoff, 1986; Strassberg & Lowe, 1995; Wiederman, 1993) and differ in personality (e.g., higher in sensation seeking) than more general nonvolunteer samples (Bogaert, 1996). The importance of sampling issues in the context of sexual orientation development has recently been demonstrated by Bailey et al. (2000), who showed that the concordance rates of existing published twin studies on sexual orientation are likely to be inflated due to an ascertainment/volunteer bias. Moreover, Bailey et al. (2000), although finding support for familial factors, raised doubts about well-publicized findings on the importance of genetic factors for sexual orientation development. Thus, demonstrating the birth order effect in a large representative sample of adults is an important advance in research on sexual orientation development.

BIRTH ORDER AND SEXUAL ORIENTATION IN WOMEN

Birth order and sexual orientation in women has been less studied, and the results from early and/or clinical studies are inconsistent (see Blanchard, 1997; Bogaert, 1997). To help clarify this relation, Bogaert (1997) recently studied data from a very large sample of women who were interviewed by investigators at the Kinsey Institute from 1938 to 1963. The women were dichotomously classified as lesbian (n = 257) or heterosexual (n = 5,008). No significant effects for birth order, birth order among brothers, or birth order among sisters were observed. These null results are noteworthy given the large sample size and the resulting power to detect very small effects (see Blanchard, Zucker, Sielegman, Dickey, & Klassen, 1998, for a subsequent null result in a smaller sample gathered during the years 1968-1970). These results, if they prove reliable, are also important for understanding sexual orientation development in at least two ways. First, they add to the literature on sexual orientation development in women. Second, they advance theoretical accounts of the birth order effect in men. In particular, they suggest a gender-specific (i.e., male only) developmental mechanism for the birth order effect in men.

Although existing research suggests no relation between birth order and sexual orientation in women, there are a number of reasons to continue to study this issue. First, there is less research on women's (vs. men's) sexual orientation development, including the possible importance of birth order. Second, research on sexual orientation and birth order in women, like that in men, needs to investigate these questions in a more representative sample. Existing research on women have used clinical samples of one kind or another or typical volunteer/convenience samples. For example, the Kinsey data (see Bogaert, 1997), although being the largest sample of its kind in the world and having a unique historical significance, has been criticised for having rather haphazard sampling methods (e.g., Laumann, Gagnon, Michael, & Michaels, 1994, pp. 44-45). Third, one explanation for why the birth order effect is found only in samples of men concerns the age of the women's samples. To my knowledge, no published study has investigated birth order and sexual orientation in a nonclinical sample of women collected after 1975. Yet the sexual attitude and behavior literature suggests that sex differences have narrowed, some substantially, from the 1960s to the 1980s for a number of sexual variables (Oliver & Hyde, 1993). Analyzing recent samples of women is, therefore, worthwhile because the birth order mechanism--whatever it may be--may be more likely to operate across sex now than prior to 1975.

THEORIES/MECHANISMS UNDERLYING (FRATERNAL) BIRTH ORDER

Various theories, both biological and psychosocial in nature, have been proposed to account for the birth order effect in men (see Blanchard, 1997), but little testing of these theories has occurred and no theory to date has emerged as the only plausible explanation. One recently proposed psychosocial theory (see Jones & Blanchard, 1998) builds on the idea that sexual activity among brothers during childhood is a predisposing factor in homosexual development (e.g., Slater, 1958). To explain the older brother effect, however, this explanation has the added assumption that the effect of same-sex activity is greater on the younger brother (rather than the older brother) involved, leading to a tendency for gay men to be born later in their sibships. Examples of such sibling sexual activity (or at least an older brother having a sexual influence) include mutual sex play among brothers, sexual fantasy among brothers, an older brother initiating or encouraging a younger brother to engage in same-sex activity, (eroticized) sexual instruction by an older brother toward a younger brother, or even sexual abuse by an older brother against a younger brother. Note that this explanation is a special case of the behavioral contagion model, which has been used to explain sibling effects for antisocial behavior (D. R. Jones & Jones 1992; M. B. Jones & Jones, 1994, 1995), although, in principle, it applies equally well to sibling effects for more neutral and prosocial behaviors, such as sports involvement or volunteering.

Although indirect, there is some evidence in favor of the contagion model of the birth order effect. First, as mentioned, the model has been supported in the context of sibling effects for other behaviours, particularly for delinquency (D. R. Jones & Jones 1992; M. B. Jones & Jones, 1994, 1995). Second, there is evidence of a correlation between same-sex behaviour in childhood and later homosexuality (e.g., Van Wyk & Geist, 1984; but see Wellings, Field, Johnson, & Wadsworth, 1994, pp. 204-206), although the direction of the effect is open to interpretation (e.g., Gebhard, Gagnon, Pomeroy, & Christenson, 1965, pp. 329, 457). Such a correlation might mean that sexual interactions among boys (e.g., two friends or brothers) cause them (particularly the younger of the boys involved) to develop a homosexual orientation, but it may be interpreted merely as evidence that sexual orientation is determined very early in development (e.g., in utero) and that a "prehomosexual" boy is simply more interested than a "preheterosexual" boy in sexual interaction with other males. Despite only modest evidence currently in its favor, the contagion model remains the most popular of the psychosocial theories of the birth order effect (Purcell, Blanchard, & Zucker, in press), perhaps because it seems inherently plausible to many people.

THE PRESENT STUDY

In the present study the relation between birth order and sexual orientation, along with a possible mechanism underlying the birth order effect--sexual activity among siblings--is investigated in men and women using a large (and recent) national probability sample of the U. S. population. The sample is the National Health and Social Life Survey (Laumann et al., 1994), which contains 3,432 cases. This study has been praised for its sampling methods and is considered the best sexuality survey of the general U.S. population (e.g., see Hyde & DeLamater, 2000, p. 59). It also contains relevant information on birth order, sexual orientation, and sibling same-sex activity/influence, in both men and women, and thus provides a unique opportunity to examine these research questions.

METHOD

Sample

The Laumann et al. (1994) study used a probability sample of households in the U.S. This sampling procedure ensured that less than 3% of Americans were excluded (e.g., prisoners, and students in college dormitories). All other adults between the ages of 18 and 59 were eligible. The researchers obtained a 79% cooperation rate, which is very impressive given that general surveys (e.g., political opinions) typically have response rates of only 75%. Each participant was interviewed, which lasted about 90 minutes, and then was given a small supplementary questionnaire for certain sensitive topics (e.g., masturbation). Participants were paid $35 for their time. The final sample was comprised of 1,921 women and 1,511 men.

Birth Order and Other Relevant Demographics 

For birth order the Laumann et al. (1994) data are less than ideal, with only three categories (first born, last born, and middle child) and no information on the participant's position relative to his or her brothers or sisters (e.g., no information on number of older brothers). As such, the present study provides a conservative test of whether birth order and sexual orientation is related in both men and women, given the recent evidence that the birth order effect is likely to be an incremental older brother effect in men (e.g., Blanchard & Bogaert, 1996b) and, thus, may result from a similar gender-based birth order effect in women (if such an effect exists). However, these data do contain information on additional relevant family demographic variables (i.e., number of brothers and number of sisters), and these variables are included in the analyses. Laumann et al. (1994) included both biological and nonbiological siblings (e.g., stepsiblings, adopted siblings, etc.) in their counts of number of brothers, number of sisters, and birth order, and there is no way to distinguish between biological and nonbiological siblings in these data. In addition, for number of brothers and sisters, Laumann et al. (1994) collapsed 6 through 10 siblings into 6, and 11 or more siblings were collapsed into 11. For example, if a participant happened to have 8 brothers and 12 sisters, his or her scores on these variables would be 6 and 11, respectively. Thus, possible values for these sibling variables were 0, 1, 2, 3, 4, 5, 6, and 11. For those with no brothers and no sisters (n = 214 or 6.2% of the sample), the birth order question was not asked and they were given a missing value for this measure.

Finally, other demographics assessed included age in years, education (1 = 8th grade or less to 8 = Advanced degree), and ethnicity/race (1 = White, 2 = Black, 3 = Native American, 4 = Asian/Pacific Islander, 6 = Hispanic).

Sexual Orientation

Most researchers use a psychological definition of sexual orientation and, as such, define sexual orientation as one's degree of sexual attraction to men or women (e.g., Bailey et al., 2000). Sexual attraction in the Laumann et al. (1994) interview protocol was assessed using a 5-point scale: "In general, are you sexually attracted to only men, mostly men, both men and women, mostly women, or only women?" where 1 = Only men, 2 = Mostly men, 3 = Both men and women, 4 = Mostly women, and 5 = Only women. This is the main measure of sexual orientation used in the present study. Overt sexual behavior is correlated with one's attraction, but for a variety of social and other factors, sexual orientation (i.e., one's "attraction" to men or women) and actual sexual behavior can have a less than perfect correspondence. However, given that a number of sexual orientation studies do include a measure of behavior, we also included the following behavioral measure: "Have your sex partners in the last 5 years been: 1 = Exclusively male, 2 = Both male and female, or 3 = Exclusively female. The attraction and behavioral measures were significantly correlated in both men, [r.sub.s] (1336) = .627, p [is less than] .001, and women, [r.sub.s] (1673) = .459, p [is less than] .001, using Spearman's Rank correlation.

Sibling Sexual Interaction/Influence

Three items from the Laumann et al. (1994) interview protocol are relevant for the assessment of sibling sexual activity/influence. One concerned whether a brother or sister taught the participant about sex: "When you were growing up, in which of the following ways did you learn about sexual matters?" Possible responses were from a brother (1 = yes or 0 = no) or from a sister (1 = yes or 0 = no). A related question was: "From which one of those [from the above list] did you learn the most?" Possible responses were from a brother (1 = yes and 0 = no) or from a sister (1 = yes or 0 = no). These variables should capture, at least to some degree, whether, for example, an older brother initiated or encouraged a younger brother to engage in same-sex activity or whether there may have been (eroticized) sexual instruction by an older brother toward a younger brother. It is of note, however, that these two questions in the Laumann et al. data do not indicate which sibling (i.e., older/younger) is the "teacher" or role model with regard to sex instruction/influence, although it is likely to be the older (and not the younger) sibling who acts as the "teacher" or a role model in such interactions because older siblings (relative to younger) are, on average, more experienced, cognitively advanced, and dominant. Finally, a question asked whether someone had sexually touched the participant prior to puberty, and what was this person's relationship to participant: "Before the age of puberty (or prior to 12 or 13), did anyone touch you sexually (1 = yes or 0 = no)?" and "What was his/her relationship to you?" A possible response was an older brother (1 = yes or 0 = no), but few if any (older) sisters were listed as having touched the respondent, so older sister was not a coded response in the data set. This question should capture whether, for example, mutual sex play occurred among brothers or whether sexual abuse occurred by an older brother against a younger brother.

RESULTS

The mean age for the men was 36.0 years (SD = 10.8) and their average educational level was 4.3 (SD = 1.7), where 4 = vocational/trade school. Seventy-eight percent were White, 14% were Black, and the remainder (approx. 8%) were another race/ethnicity. The mean age for women was 36.7 (SD = 11.0), and their average educational level was 4.2 (SD = 1.6), where 4 = vocational/trade school. Seventy-four percent were White, 18% were Black, and the remainder (approx. 8%) were another race/ethnicity.

Thirty-nine men reported sexual attraction to only men, 11 reported sexual attraction to mostly men, 9 reported sexual attraction to men and women equally, 43 reported sexual attraction to mostly women, and 1,402 reported sexual attraction to only women. For sexual behavior for men, 33 reported having only male partners in the last 5 years, 24 reported having both male and female partners in the last 5 years, and 1,280 reported having only female sexual partners in last 5 years. Small but significant correlations occurred between age and sexxual attraction, [r.sub.s] (1,503) = .06, p [is less than] .05, and education and sexual attraction, [r.sub.s] (1,497) = -.06, p [is less than] .05 using Spearman's Rank correlation, but here was no relationship between race (recoded as 1 = White and 2 = non-White) and sexual attraction, [X.sup.2] (4, N = 1504) = 4.13, ns. There were also small but significant correlations between age and sexual behavior, [r.sub.s] (1,338) = .07, p [is less than] .05, and education and sexual behavior [r.sub.s] (1,331) = .08, p [is less than] .01, using Spearman's Rank correlation, although again there was no significant relationship between race and sexual behavior, [X.sup.2] (2, N = 1,338) = 3.52, ns.

For women, 10 reported sexual attraction to only women, 11 reported sexual attraction to mostly women, 15 reported sexual attraction to both men and women equally, 51 reported sexual attraction to mostly men, and 1,814 reported sexual attraction to only men. For sexual behavior, 27 reported having only female sexual partners in the last 5 years, 13 reported both male and female sexual partners in the last 5 years, and 1,649 reported having only male partners in the last 5 years. There was a small but significant correlation between sexual attraction and education, [r.sub.s] (1885) = .13, p [is less than] .001, but there was no correlation between sexual attraction and age, [r.sub.s] (1898) = -.03, ns, using Spearman's Rank correlation. There also was no relationship between race and sexual attraction, [X.sup.2] (4, N = 1899) = 5.18, ns. There was no relationship between sexual behavior and education, [r.sub.s] (1675) = .04, ns, and sexual behavior and age, [r.sub.s] (1687) = -.03, ns, using Spearman's Rank correlation; there was also no significant relationship between between race and sexual behavior, [X.sup.2] (2, N = 1687) = 0.45, ns.

The men had an average (mean) of 1.9 brothers (SD = 1.7) and 1.7 sisters (SD = 1.5), with 439 (29.1%) being first born, 592 (39.2%) being middle children, 389 (25.7%) being last born, and 91 (6.0%) having no siblings and given a missing value on birth order. The women had an average (mean) of 1.9 brothers (SD = 1.7) and 1.8 sisters (SD = 1.7), with 506 (26.3%) being first born, 785 (40.9%) being middle children, 507 (26.4%) being last born, and 123 (6.4%) having no siblings and given a missing value on birth order.

For the sibling sexual influence variables, 311 men (20.7%) indicated that they had learned about sex from a brother (i.e., responded with a yes), 64 men (4.3%) indicated that they had learned most about sex from a brother, 6 men (0.4%) indicated that an older brother had touched them sexually before they were 12 or 13 years of age, 87 men (5.8%) indicated that they had learned sex from a sister, and 6 men (0.4%) indicated that they had learned most about sex from a sister. One hundred and seventeen women (6.1%) indicated that they had learned about sex from a brother, 13 women (0.7%) indicated they learned most about sex from a brother, 24 women (1.3%) reported that an older brother had touched them sexually before they were 12 or 13 years of age, 337 women (17.6%) indicated that learned about sex from a sister, and 82 women (4.3%) indicated that they had learned most about sex from a sister.

Birth Order and Sexual Orientation

To assess whether a relationship exists between birth order and sexual orientation in men, a linear regression was undertaken, with sexual attraction as the criterion, birth order and the three brother interaction/influence variables as simultaneously entered predictors, and family size (total number of brothers and sisters) as a weighted variable in a weighted least squares model. This weighted least squares model ensures that large families have more of an impact on the analysis. This is an important consideration because birth order effects are seen more clearly in large families. Such weighted least squares analyses have been used before for research on birth order and sexual orientation (e.g., Blanchard, Zucker, Bradley, & Hume, 1995), in particular for nonlinearly transformed and/or truncated birth order data, such as in the present study, where family size is obscured. Finally, age and education were controlled for because they were, as mentioned, related to sexual attraction.

As shown in Table 1, birth order significantly (p = .001) predicted sexual attraction, where later borns were significantly more likely to have same-sex attraction. Using the sexual behavior measure and controlling for age and education, which were also, as mentioned, related to sexual behavior, produced similar results, where later borns were significantly (p = .01) more likely to have same-sex behavior. Note that in both analyses, there were no sibling (i.e., brother) sexual activity/influence effects.(1)

[TABULAR DATA 1 NOT REPRODUCIBLE IN ASCII]

To assess whether a relationship exists between birth order and sexual orientation in women, similar weighted least squares linear regression analyses were undertaken, with sexual attraction and sexual behavior as the criteria; birth order, sibling interaction/influence variables, and the other relevant demographic variable (i.e., education for sexual attraction) as simultaneously entered predictors; and family size (total number of brothers and sisters) as the weighted variable. Instead of 3 sibling interaction/influence variables, 5 sibling interaction/influence variables were entered--the 3 brother variables (i.e., learned sex from brother, learned the most sex from brother, older brother touched me) and 2 sister variables (i.e., learned about sex from sister, learned the most about sex from sister). Both the brother and sister influence/interaction variables were included because, unlike in men, no clear fraternal or sororal birth order effect has yet been demonstrated in women.

As shown in Table 2, no significant birth order effect occurred for women. It is of note that the sister and brother sexuality/influence variables were also eliminated from a second regression analysis, in case these variables may have mediated and thus eliminated a possible birth order effect once they were controlled for. There was no significant birth order effect once these variables were removed from the analyses (both ps [is greater than] .3). Interestingly, however, there was a relationship between sexual orientation and whether a brother taught the female participant about sex, such that higher same-sex attraction and behavior was related to more brother influence.

[TABULAR DATA 2 NOT REPRODUCIBLE IN ASCII]

DISCUSSION

Using measures of both attraction and behavior to define sexual orientation, the present study demonstrates that homosexual men are later born relative to heterosexual men. These findings support previous studies on birth order and sexual orientation (for a review, see Blanchard, 1997), although the present study is an advance in this research program because it demonstrates the effect in a national probability sample. Sampling problems have raised questions about the reliability of some findings on the etiology of sexual orientation (e.g., Bailey et al., 2000). Evidently, sampling problems do not apply to research on sexual orientation and birth order in men. This study also reaffirms the utility of certain specialized and volunteer samples for studying certain sexual issues (cf., Saunders, Fisher, Hewitt, & Clayton, 1985). It is clear, for example, that the original Kinsey data evince the same pattern of results for birth order and sexual orientation as do the Laumann et al. (1994) data--a birth order effect only in men (Blanchard & Bogaert, 1996a; Bogaert, 1997)--despite the fact that Kinsey's sampling methods were haphazard and his sample is 30 to 60 years older than the Laumann et al. study. Thus, although it is clear that Kinsey and his colleagues did over-sample homosexual men and women and that these data should not be used for estimating the incidence of same-sex attraction/behavior in the population (e.g., Laumann et al., 1994, pp. 287-290), the heterosexual and homosexual men and women in these data may not have any particular bias that would exclude them from judicious comparisons/analyses.

This study is also important because it examines birth order and sexual orientation in women in a modern nonclinical sample; this study also has the advantage of using a national probability sample to test for such effects. However, despite a comparable level of power between samples of men and women, there was no birth order effect in women. With the present data, added to that of previous samples, one of which was very large and had the power to detect very small effects (Bogaert, 1997), a reasonable conclusion is that there is no relationship between birth order and sexual orientation in women. Thus, despite the fact that sex differences in sexuality have narrowed, some substantially, from the 1960s to the 1980s (Oliver & Hyde, 1993), the birth order mechanism--whatever it may be--continues to operate only in men.

Interestingly, there was a relationship between sexual orientation and whether a brother taught the female participant about sex, such that higher same-sex attraction was related to more brother influence. One interpretation of this relationship is that bisexual and lesbian women may seek out sexual information from (and establish close relationships with) brothers because they share with their (heterosexual) brothers sexual desires for women and perhaps have more stereotypically masculine interests (e.g., Bailey & Zucker, 1995), relative to heterosexual women.

The present study also advances theoretical accounts of the birth order effect in two ways. First, it demonstrates that the birth order effect in men is likely to be independent of same-sex activity/influence among siblings, as suggested by the "contagion" model of sibling effects (Jones & Blanchard, 1998). The birth order effect was still clearly significant controlling for all three variables and comprising an index of older brother sexual activity/influence. The contagion model had been considered the most popular of the psychosocial theories to account for the birth order effect in men (e.g., Purcell et al., in press), but the present data provide no support for this idea (c.f., Dawood, Pillard, Horvath, Revelle, & Bailey, 2000). In light of these nonsupportive results, another birth order study is noteworthy: Blanchard et al. (1995) examined a sample of probably prehomosexual boys who, like adult gay samples, had a relatively late birth order; however, given that these boys' mean age was only 8.46 years, it is unlikely that a significant number of them had already been conditioned to homosexuality by sexual activity/influence with or by an older brother or anyone else.

Second, this study advances theoretical accounts of the birth order effect because a number of theories now become less plausible in light of the null birth order findings in women. This criterion may not eliminate theories, but it raises serious questions about why a similar mechanism does not operate in women as it does in men. For example, in addition to the contagion model, related models of same-sex fantasy/activity are relevant in this regard. One such theory comes from Sulloway (1996, pp. 433-434, 488), who hypothesized that the birth order effect reflects later bores' greater "openness to experience," which predisposes them to experiment sexually, including with same-sex activity. Another recent across-gender theory of the birth order effect has been put forward by Bem (1996). In this view, feminine (or perhaps just nonmasculine) boys feel they are different from other, more masculine boys; these feelings of differentness lead to eroticization of other males (i.e., homosexuality). The more older brothers a boy has, the more likely he is to develop a sense of being different from (and thus being attracted to or eroticizing) other males. Both Sulloway's and Bem's theory are, in principle, meant to explain sexual orientation in the same way for both males and females, and therefore they predict that homosexual women will have more older siblings (i.e., sisters) than heterosexual women. However, as indicated, all existing data, including this contemporary national probability sample, suggests there is no birth older (i.e., older sister) effect in women.

There is clearly something unique about male development that predisposes later born males to homosexuality. Why is this the case? Although psychosocial explanations, such as limited versions of Sulloway's (1996) and Bem's (1996) theories, are worthy of further study, the answer may lie in the prenatal sexual differentiation process. This process is, of course, very different for males and females, and there is evidence that males are more likely than females to experience complications and, ultimately, alterations in typical development (Gualtieri & Hicks, 1985). Interestingly, like homosexuality in men, some of these atypical development patterns show a birth order effect (i.e., they become more pronounced in later born offspring). One of the leading candidates for an explanation of this gender-specific effect in atypical development, along with its correlation to birth order, is a maternal immune response to succeeding male pregnancies (Gualtieri & Hicks, 1985). This explanation is partly based on the idea that a woman's immune system would appear to be the most capable of "remembering" the number of male (but not female) fetuses she has previously carried and of progressively altering its response to the next fetus according to the current tally of preceding males. Presumably, female fetuses are not likely to induce an immune response because the mothers themselves are female. If the immune hypothesis is correct in the case of male homosexuality (Blanchard & Bogaert, 1996b; Blanchard & Klassen, 1997; Ellis & Ames, 1987; MacCulloch & Waddington, 1981), then the connection between the mother's immune reaction and the child's future sexual orientation would likely be some effect of maternal antibodies on the sexual differentiation of the fetal brain. Early formulations of a maternal immune response tot male homosexuality concentrated on testosterone as the relevant fetal antigen, but more recent formulations focus on male-specific, Y-linked H-Y antigen as the relevant fetal antigen (Blanchard & Bogaert, 1996b; Blanchard & Klassen, 1997). Although no direct support for a maternal immune response explanation of male homosexuality exists (see Blanchard, 1997, for a review of indirect evidence), it should be noted that this explanation does have the advantage of positing an inherently gender-specific mechanism.

Limitations of the present investigation should be mentioned. First, the variables testing the contagion/sibling influence model were not ideal. For example, the relevance of the siblings-teaching-about-sex variables is open to some interpretation, because although a sibling who takes the role of the teacher with regard to sex is likely to be the older (and not the younger) sibling, there is no way to insure that all respondents who indicated yes on these variables were, in fact, referring to an older brother, as the contagion model of the birth order effect would predict. Second, the birth order variable in the Laumann et al. (1994) data was problematic, with only three categories (first born, last born, and middle child) and no information on the participant's position relative to his or her brothers or sisters (e.g., no information on number of older brothers or sisters). As such, as mentioned, the present study provides a conservative test of whether birth order and sexual orientation are related in men and women. Third, although the majority of the participants are likely to come from biologically-intact families, there was no way to determine which (and how many) participants were part of biological or nonbiological families (e.g., adopted, step-siblings). As such, speculations about a biological mechanism (e.g., a maternal immune response) are not directly testable with these data.

Interestingly, one important and relatively definitive way of advancing theoretical accounts (e.g., a maternal immune response) of the birth order effect would be to study gay and heterosexual men raised in primarily nonbiological families. Indeed, such research would directly assess the relative importance of biological versus psychosocial mechanisms. For example, if gay men raised in nonbiological families (e.g., adoptees) still show a later "birth order" (and more older brothers) relative to a well-matched sample of heterosexual men (e.g., also adoptees), then it is clear that family dynamics/rearing factors are responsible for the older brother/birth order effect. Such a result would indicate that the older brother/birth order effect is in fact a sibling position effect and not the result of a biological factor related to one's birth, conception, or gestation order (e.g., maternal immune response). Indeed, to the best of my knowledge it would be the first study to definitively link a specific environmental factor--in this case, sibling position--to sexual orientation development. If, however, no detectable older brother/birth order effect is found for gay men raised in nonbiological families, relative to a well-matched sample of heterosexual men, then a biological mechanism is favoured (e.g., one related to the order of conception, gestation, or birth), particularly if the sample is of sufficient size to detect small effects. A primary challenge for this type of study would, of course, be recruiting enough (gay and bisexual) participants from nonbiological families, although community-based samples of gay/bisexual men can yield reasonably high numbers of participants from nonbiological families (e.g., 20%; Hafer, Bogaert, & McMullen, in press).

Finally, it may be of some interest to discuss this research within the context of the current academic discourse on sexual orientation. At one end of the spectrum are social constructionists who argue that the concept of sexual orientation is a culture-bound and context-specific notion (De Cecco & Elia, 1993). At the other end are essentialists, who argue that sexual orientation is a relatively stable characteristic inherent to the individual. The essentialists are currently dominated by biological-oriented researchers who link sexual orientation to genetic, hormonal, and neuroanatomical factors. Biological essentialism is currently dominant, in part because experience-based models of sexual orientation (e.g., early conditioning) have not received much empirical support (e.g., Bell et al., 1981). The present birth order research (at least for men) seems most consistent with the essentialist perspective. As mentioned, the birth order effect has shown impressive generalizability, being replicated across samples of men from different eras (see Blanchard & Bogaert, 1996a, 1996b), from different countries (see Blanchard & Bogaert, 1996b; Blanchard et al., 1996), and from different ethnicities (Bogaert, 1998). It has even been replicated in men with unusual age-oriented sexual preference systems, such as homosexual versus heterosexual pedophiles (e.g., Bogaert et al., 1997), and in unusual gender identities, such as homosexual versus heterosexual male-to-female transsexuals (e.g., Blanchard & Sheridan, 1992). And now, with the present results, the birth order effect has been replicated with a modem national probability sample of the U. S. population. Thus, to the extent that it has been studied (which is extensive relative to other research questions in sexology), the birth order/sexual orientation relation does not seem to be culture- or context-specific. Whether the mechanism responsible is, however, ultimately biological or experienced-based in origin remains to be seen.

(1) Instead of weighting so that larger families have more of an impact on the analysis, this effect can be modeled by including a Family Size x Birth Order interaction term in the regression analysis. Using this method, very similar results occurred, where a later birth order was significantly associated with same-sex attraction, but only in larger families; these effects were also significant controlling for sibling sexual influence/activity variables, which were not related to the criteria.

REFERENCES

Bailey, J. M., Dunne, M. P., & Martin, N. G. (2000). Genetic and environmental influences on sexual orientation and its correlates in an Australian twin sample. Journal of Personality and Social Psychology, 78, 524-536.

Bailey, J. M., & Zucker, K. J. (1995). Childhood sex-typed behavior and sexual orientation: A conceptual analysis and quantitative review. Developmental Psychology, 31, 43-55.

Bell, A. P., Weinberg, M. S., & Hammersmith, S. K. (1981). Sexual preference: Its development in men and women. Bloomington, IN: Indiana University Press.

Bem, D. J. (1996). Exotic becomes erotic: A developmental theory of sexual orientation. Psychological Review, 103, 320-335.

Blanchard, R. (1997). Birth order and sibling sex ratio in homosexual versus heterosexual males and females. Annual Review of Sex Research, 8, 27-67.

Blanchard, R., & Bogaert, A. F. (1996a). Biodemographic comparisons of homosexual and heterosexual men in the Kinsey interview data. Archives of Sexual Behavior, 25, 551-579.

Blanchard, R., & Bogaert, A. F. (1996b). Homosexuality in men and number of older brothers. American Journal of Psychiatry, 153, 27-31.

Blanchard, R., & Bogaert, A. F. (1997). Additive effects of older brothers and homosexual brothers in the prediction of marriage and cohabitation. Behavior Genetics, 27, 45-54.

Blanchard, R., & Klassen, P. (1997). H-Y antigen and homosexuality in men. Journal of Theoretical Biology, 185, 373-378.

Blanchard, R., & Sheridan, P. M. (1992). Sibship size, sibling sex ratio, birth order, and parental age in homosexual and nonhomosexual gender dysphorics. Journal of Nervous and Mental Disease, 180, 40-47.

Blanchard, R., Zucker, K. J., Bradley, S. J., & Hume, C. S. (1995). Birth order and sibling sex ratio in homosexual male adolescents and probably prehomosexual feminine boys. Developmental Psychology, 31, 22-30.

Blanchard, R., Zucker, K. J., Cohen-Kettenis, P. T., Gooren, L. J. G., & Bailey, J. M. (1996). Birth order and sibling sex ratio in two samples of Dutch gender-dysphoric homosexual males. Archives of Sexual Behavior, 25, 489-508.

Blanchard, R., Zucker, K. J., Sielegman, M., Dickey, R., & Klassen, P. (1998). The relation of birth order to sexual orientation in men and women. Journal of Biosocial Science, 30, 511-519.

Bogaert, A. F. (1996). Volunteer bias in human sexuality research: Evidence for both sexuality and personality differences in males. Archives of Sexual Behavior, 25, 125-140.

Bogaert, A. F. (1997). Birth order and sexual orientation in women. Behavioral Neuroscience, 111, 1395-1397.

Bogaert, A. F. (1998). Birth order, sibling sex ratio, and sexual orientation in nonwhite men. Archives of Sexual Behavior, 27, 467-473.

Bogaert, A. F., Bezeau, S., Kuban, M., & Blanchard, R. (1997). Birth order, sexual orientation, and pedophilia. Journal of Abnormal Psychology, 106, 331-335.

Dawood, K., Pillard, R. C., Horvath, C., Revelle, W., & Bailey, J. M. (2000). Familial aspects of male homosexuality. Archives of Sexual Behavior, 29, 155-164.

De Cecco, J. P., & Elia, J. P. (Eds.). (1993). If you seduce a straight person, can you make them gay? Issues in biological essentialism versus social constructionism in gay and lesbian identities. New York: Harrington Park Press.

Ellis, L., & Ames, M. A. (1987). Neurohormonal functioning and sexual orientation: A theory of homosexuality-heterosexuality. Psychological Bulletin, 101, 233-258.

Gebhard, P. H., Gagnon, J. H., Pomeroy, W. B., & Christenson, C. V. (1965). Sex offenders: An analysis of types. New York: Harper & Row.

Gualtieri, T., & Hicks, R. E. (1985). An immunoreactive theory of selective male affliction. Behavioral and Brain Sciences, 8, 427-441.

Green, R. (1987). The "sissy boy" syndrome and the development of homosexuality. New Haven, CT: Yale University Press.

Hafer, C. L., Bogaert, A. F., & McMullen, S. (in press). Belief in a just world and high risk sexual behaviour in gay men. Journal of Applied Social Psychology.

Hamer, D. H., & Copeland, P. (1994). Science of desire: The search for the gay gene and the biology of behavior. New York: Simon & Shuster.

Hyde, J. S., & DeLamater, J. (2000). Understanding human sexuality (7th ed.). Boston: McGraw-Hill.

Jones, M. B., & Blanchard, R. (1998). Birth order and male homosexuality: An extension of Slater's index. Human Biology, 70, 775-787.

Jones, D. R., & Jones, M. B. (1992). Behavioral contagion in sibships. Journal of Psychiatric Research, 26, 149-164.

Jones, M. B., & Jones, D. R. (1994). Testing for behavioral contagion in case-control designs. Journal of Psychiatric Research, 28, 35-55.

Jones, M. B., & Jones, D. R. (1995). Preferred pathways of behavioral contagion. Journal of Psychiatric Research, 29, 193-209.

Lalumiere, M. L., Harris, G. T., Quinsey, V. L., & Rice, M. E. (1998). Number of older brothers among sexual offenders. Sexual Abuse: A Journal of Research and Treatment, 10, 5-15.

Laumann, E. O., Gagnon, J. H., Michael, R. T, & Michaels, S. (1994). The social organization of sexuality: Sexual practices in the United States. Chicago: University of Chicago Press.

LeVay, S. (1996). Queer science: The use and abuse of research into homosexuality. Cambridge, MA: MIT Press.

MacCulloch, M. J., & Waddington, J. L. (1981). Neuroendocrine mechanisms and the aetiology of male and female homosexuality. British Journal of Psychiatry, 139, 341-345.

Morokoff, P. J. (1986). Volunteer bias in the psychophysiological study of female sexuality. The Journal of Sex Research, 22, 35-51.

Oliver, M. B., & Hyde, J. S. (1993). Gender differences in sexuality: A metaanalysis. Psychological Bulletin, 114, 29-51.

Purcell, D. W., Blanchard, R., & Zucker, K. J. (in press). Birth order in a contemporary sample of gay men. Archives of Sexual Behavior.

Saunders, D. M., Fisher, W. A., Hewitt, E. C., & Clayton, J. P. (1985). A method for empirically assessing volunteer selection effects: Recruitment procedures and responses to erotica. Journal of Personality and Social Psychology, 49, 1703-1712.

Singh, D., Vidaurri, M., Zambarano, R. J., & Dabbs, J. M. (1999). Lesbian erotic role identification: Behavioral, morphological, and hormonal correlates. Journal of Personality and Social Psychology, 76, 1035-1049.

Slater, E. (1958). The sibs and children of homosexuals. In D. R. Smith & W. M. Davidson (Eds.), Symposium on nuclear sex (pp. 79-83). London: Heinemann Medical Books.

Strassberg, D. S., & Lowe, K. (1995). Volunteer bias in sexuality research. Archives of Sexual Behavior, 24, 369-382.

Sulloway, F. J. (1996). Born to rebel: Birth order, family dynamics and creative lives. New York: Pantheon Books.

Van Wyk, P. H., & Geist, C. S. (1984). Psychosocial development of heterosexual, bisexual, and homosexual behavior. Archives of Sexual Behavior, 13, 505-544.

Wellings, K., Field, J., Johnson, A., & Wadsworth, J. (1994). Sexual behavior in Britain: The National Survey of Sexual Attitudes and Lifestyles. London: Penguin Books.

Wiederman, M. W. (1993). Demographic and sexual characteristics of nonresponders to sexual experience items in a national survey. The Journal of Sex Research, 30, 27-35.

Zuger, B. (1988). Is early effeminate behavior in boys early homosexuality? Comprehensive Psychiatry, 29, 509-519.

Manuscript accepted August 28, 2000

This research was supported by Social Sciences and Humanities Research Council of Canada Grant 410-99-0521 to Anthony F. Bogaert

The author wishes to thank Ray Blanchard, John Cairney, Nancy DeCourville, and Carolyn Haler for their help at various stages of this research.

Address correspondence to Anthony F. Bogaert, Departments of Community Health Sciences and Psychology, Brock University, St. Catharines, Canada, L2S 3A1; e-mail: [e-mail].