Sex Roles

DOI 10.1007/s11199-009-9598-4

ORIGINAL ARTICLE

Heterosexual, Lesbian, and Gay Preadoptive Parents’

Preferences About Child Gender
Abbie E. Goldberg

# Springer Science + Business Media, LLC 2009

Abstract Little research has explored the child gender that it has largely examined the preferences of biological
preferences of preadoptive parents. This study utilized a parents and nonparents; little research has examined the
mixed-methods approach to explore child gender prefer- preferences of preadoptive parents, whose ideas about the
ences (and individuals’ reasons for such preferences) in a gender of their future child may be shaped by their unique
geographically diverse, US sample of 93 heterosexual, 61 route to parenthood. Further, research focuses largely on
lesbian, and 48 gay male preadoptive couples. Heterosexual heterosexual men and women; the child gender preferences
men were the least likely to demonstrate a gender of lesbians and gay men, who are becoming parents at
preference and gay men were the most likely. Individuals increasing rates (Gates and Ost 2004) have rarely been
in heterosexual relationships were more likely to prefer studied. It is expected that the unique contexts of adoption
girls than individuals in same-gender relationships. In and sexual orientation may have distinct implications for
explaining their preferences, sexual minorities often em- men’s and women’s gender preferences. Specifically, it is
phasized gender socialization considerations (e.g., their expected that (a) preadoptive parents’ gender preferences
perceived inability to socialize a child of the opposite may differ in unique ways from those of biological parents
gender) and concerns about heterosexism (e.g., some gay as reported in the literature; (b) the gender preferences of
men preferred girls because they felt a boy would encounter lesbian/gay preadoptive parents might differ in key ways
more harassment). from those of heterosexual preadoptive parents; and (c) the
reasons cited by lesbian/gay preadoptive parents for their
Keywords Adoption . Child gender . Gay . Lesbian . gender preferences might differ in meaningful ways from
Preferences those mentioned by heterosexual preadoptive parents.
Next, the literature on biological parents’ and non-
parents’ gender preferences is discussed, followed by the
Introduction limited research on the preferences of adoptive parents and
lesbian/gay parents.
The purpose of the current study is to examine the child
gender preferences of preadoptive parents, using quantita- The Child Gender Preferences of Biological Parents
tive and qualitative data from a geographically diverse, US and Nonparents
sample of 93 heterosexual, 61 lesbian, and 48 gay male
couples who were awaiting adoptive placement of their first A moderately large body of research has examined parents’
child. The research on child gender preferences is limited in and nonparents’ child gender preferences. This research has
often been motivated by interest in how such preferences
may influence reproductive decision-making and fertility
A. E. Goldberg (*) patterns (e.g., parents with a strong son preference may be
Department of Psychology, Clark University,
motivated to have more children if they bear only
950 Main St.,
Worcester, MA 01610, USA daughters; Arnold 1997). Some of the research in this area
e-mail: agoldberg@[Link] has been motivated by concerns that antenatal technology
 Sex Roles

such as amniocentesis is increasingly being used for sex strengths, and interests (Williamson 1976) which may
determination or sex selection purposes, which could, inform parents’ gender preferences. Indeed, parents partic-
theoretically, result in an unbalanced sex ratio favoring ipate in a range of activities with their children, some of
boys (Oomman and Ganatra 2002). The research on child which are traditionally gendered (e.g., sports; discussing
gender preferences suggests that both men and women feelings; Pollard and Morgan 2002). It follows, then, that if
often prefer boys (Hortacsu et al. 2001) with fathers demon- a parent values some of these activities in particular, s/he
strating a stronger preference (Coombs and Fernandez 1978; might have a greater preference for a male or female child
Steinbacher and Gilroy 1990). Sons are continually preferred (although, as Pollard and Morgan point out, increasing
as first children (McDougall et al. 1999; Williamson 1976) structural opportunities for women and the weakening of
and once parents have obtained one or more sons, they are gender stratification in society should help to reduce the
less likely to have subsequent children (Hank and Kohler relevance of a child’s gender to the activities of that child
2003). and its parents).
Many studies suggest that in Western societies, a son Other factors, in addition to enjoyment of gender-typed
preference tends to persist, although it is not as pronounced activities, may also motivate gender preferences. Boys are
as in developing countries (McDougall et al. 1999). Further, often valued because they will carry on the family name,
many prospective parents do not demonstrate any gender particularly in traditional societies (Arnold and Kuo 1984;
preferences (Steinbacher and Gilroy 1985; Hank and Callan and Kee 1981). Sons are also valued for their ability
Kohler 2003; Swetkis et al. 2002; Walker and Conner to assist with male tasks around the house, and for more
1993) and a few studies have found evidence for a girl psychologically oriented reasons such as their capacity to
preference, particularly among women pregnant with their provide companionship (Callan and Kee 1981) particularly
first child (Marleau and Saucier 2002). Illustrating these to fathers (Arnold and Kuo 1984). Girls are sometimes
two points, Steinbacher and Gilroy (1985) studied Amer- preferred because they are presumed to embody character-
ican women during their first pregnancy and found that 82 istics such as neatness, cuteness, and helpfulness (Arnold
(58%) expressed no gender preferences. Of the remaining and Kuo 1984; Williamson 1976); girls are also expected to
58, 33 (24%) desired girls and 25 (18%) desired boys. The provide companionship (Callan and Kee 1981). As Hank
authors argue that their findings may be evidence of a and Kohler (2003) argue, since in industrialized societies
gradual move from boy preference to no preference and a children no longer provide economic utility but have become
gradual weakening of societal bias against females. Walker a source of time and monetary investment, they are
and Conner (1993) studied the gender preferences of 243 increasingly valued for social-psychological reasons such
pregnant women and found that 81% had a gender as expansion of the self, affiliation, and accomplishment.
preference; of these, 52% wanted a girl. Hank and Kohler Men’s and women’s motivations for preferring boys (and
(2003) studied German residents and found that childless girls) may differ. Williamson (1976) suggests that men’s
persons tended to have no gender preferences (35%) or to preference for boys (e.g., Steinbacher and Gilroy 1990)
favor a balanced gender mix (47%). An equal proportion may arise from their greater concern with the continuity of
(9%) preferred girls and boys. When respondent gender the family name or their feeling that having sons is a
was taken into account, however, this revealed that women signifier of masculinity. Consistent with this, Hammer and
were more likely to have a preference, and to favor girls. McFerran (1988) found that men who desired boys seemed
Some studies suggest that “one or more of each” is to be motivated by a desire to vicariously gratify their need
increasingly the ideal in the USA and other developed for achievement through their male child. Men may also
countries, such that many parents wish for an even prefer sons because they believe sons will provide more
distribution of boys and girls (Arnold 1997; Hank and opportunities for joint activities; indeed, fathers tend to
Kohler 2003). This, Okun (1996) argues, “may be a feature spend more time with their sons than their daughters (Raley
of a more modern society in which couples reproduce and Bianchi 2006). Women who prefer boys sometimes
primarily for purposes of deriving satisfaction from explain that they want sons to please their husbands
children, rather than for traditional purposes such as (Williamson 1976). Because men tend to be more involved
investment or old-age support” (p. 470). with their sons than their daughters, some women may
Of interest is why gender preferences exist. In explaining prefer sons out of a desire to draw their husbands into the
gender preferences, scholars have often emphasized the parenting process and to ensure security in their marriage
societal “gender system”: the socially constructed set of (Swetkis et al. 2002).
expectations for the differential behaviors and capabilities Women who prefer daughters may be motivated by a
of men and women, which has shaped a societal discourse desire for companionship (Warren 1985) and a need for
that privileges men over women (Katzev et al. 1994). In mutual identification (Notman 2006). They may also see
turn, boys and girls are seen as having different traits, girls as easier to raise or as more rewarding companions,
Sex Roles

perhaps because of their gender similarity (Marleau and (Warren 1985). Likewise, it is possible that, like biological
Saucier 2002) and, thus, their perception of girls as easier to fathers-to-be, adoptive fathers-to-be tend to prefer boys out
relate to (Hammer and McFerran 1988). Men who prefer of a desire to fulfill both affiliative and status needs
daughters may be motivated by a yearning to gratify (Coombs and Fernandez 1978). However, it is also possible
“relationship needs”: that is, Hammer and McFerran that, in the absence of a biological connection between
(1988) found that men who wanted daughters often parent and child, sons are less likely to be regarded as
highlighted their desire to have a close relationship with vehicles of status by adoptive parents, particularly fathers.
their child, and felt that females tended to be more oriented That is, a son’s accomplishments do not carry the same
toward relationships and were more capable of showing “pride reinforcing qualities” as they would if they could be
affection than males. attributed to some heritable set of characteristics, and, thus,
the possibility of a male progenitor might be less important
The Child Gender Preferences of Heterosexual to fathers. In other words, the inability to continue on the
Adoptive Parents male genealogical lineage may diminish adoptive fathers’
investment in having a son, and men may therefore
Unlike biological parents, adoptive parents can, theoreti- demonstrate few gender preferences; or, they may tend to
cally, choose their child’s gender. Interestingly, though, prefer girls, if they perceive their wives as strongly desiring
little research has explicitly examined adoptive parents’ of daughters.
preferences about child gender. One exception is the Many heterosexual couples who seek to adopt, however,
National Survey of Family Growth (2002), which surveyed may not espouse any preferences for their child’s gender, in
women only and found that among women who were light of their unique route to parenthood. Heterosexual
seeking to adopt, 36.5% had no preference about child couples that seek to adopt typically do so because they have
gender, 34.6% preferred a girl, and 28.9% preferred a boy been unsuccessful in conceiving (Daniluk 2001), and prior
(Jones 2008). Notably, census data show that girls are to pursuing adoption have often spent months or years
overrepresented among adopted children in the USA trying to conceive, often with the assistance of fertility
(Kreider 2003). This is likely in part due to the fact that treatments. Upon arriving at adoption after years of longing
girls are more likely than boys to be relinquished for for a child, heterosexual couples may feel that “beggars
adoption, both domestically and abroad (Bachrach et al. can’t be choosers” or feel that “any child will do.” In this
1992). However, the fact that women who have a gender way, failed conception efforts may create a context in which
preference tend to prefer girls may also be a factor. Adam couples come to downplay the significance of their child’s
Pertman, the executive director of the Evan B. Donaldson gender. Specifically, their urgent desire to parent may have
Adoption Institute, has stated that “the extent to which quashed any fantasies regarding their child’s gender.
women are the driving force in most adoptions is probably Alternatively, they may simply feel less strongly about
a factor….It is usually true that the women are filling out gender than about other characteristics of their child, such
the paperwork, going to the conferences, the support as health, age, and race. Indeed, research on heterosexual
groups” (Gravois 2004). This notion is consistent with parents suggests that they typically prefer to adopt
research indicating that in couples who experience infertil- inracially (Brodzinsky and Pinderhughes 2002) and prefer
ity, women are typically the “driving force” in seeking out infants and toddlers over older children (Kreider 2003). In
and making decisions about fertility treatments (Greil et al. the context of other valued characteristics, then, adopters
1988). In turn, wives may also be the more assertive may view gender as relatively insignificant.
partners in the adoption process, and may have greater
influence over the child that couples adopt. The Child Gender Preferences of Lesbians and Gay Men
Thus, these data suggest that (a) most women who are
seeking to adopt do not have a preference for their child’s Unlike individuals in heterosexual relationships, persons in
gender, but, when they do, they are somewhat more likely same-gender relationships confront the reality that only one
to prefer girls, and (b) girls tend to be adopted into families gender is represented in the parental unit, which may have
at higher rates than boys, due to a variety of factors. Little is implications for their gender preferences. Several studies
known about the gender preferences of heterosexual men have examined the gender preferences of lesbians who
who are seeking to adopt. Additionally, the reasons for became parents via insemination. Gartrell et al. (1996)
adoptive parents’ gender preferences (or lack of prefer- studied 154 lesbians during the insemination or pregnancy
ences) are not well understood. It is quite possible that stage and found that 38% had no preference about their
adoptive mothers-to-be may be particularly desiring of child’s gender; 55% of those with a preference, and 88% of
daughters for similar reasons as biological mothers-to-be: those with a strong preference, wanted a girl. Herrmann-
i.e., for companionship and to fulfill affiliative needs Green and Gehring (2007) studied 105 lesbian mothers in
 Sex Roles

Germany and found that among those who planned to have suggesting that they may be more open to adopting children
a second child, most had no gender preference (62%); most with special needs. Sexual minorities’ apparent flexibility
of those with a preference desired a girl (32% of the with respect to their future child’s characteristics may
sample). Although no data are available on lesbian/gay reflect their status as “family outlaws” (Calhoun 1997).
adoptive parents’ preferences, analysis of US census data That is, their existence outside of traditional definitions of
indicates that among female same-gender couples with “the family” may lead them to experience fewer pressures
adopted children, 56% of children are female, and among and expectations regarding how they build their families
male same-gender couples with adopted children, 54% of (and, in turn, to embrace the possibility of parenting a wide
children are male (G. Gates, personal communication, range of children); whereas heterosexual couples, in
October 29, 2008). contrast, may tend to be more influenced by traditional
These data suggest that (a) many lesbians who become definitions of “family” as heterosexual, biologically related,
parents via insemination do not have a gender preference, and physically similar (Parry 2005). Alternatively, sexual
and (b) when they do have a preference, they prefer girls. minorities’ family outlaw status may sensitize them to the
There is also some evidence that lesbian adopters may tend reality that they are seen as less optimal parents by adoption
to prefer girls and gay male adopters may tend to prefer professionals, agencies, and birth mothers (Matthews and
boys, although the extent to which actual adoption rates Cramer 2006) which may lead them to express openness to
mirror parents’ preferences is unknown. The tendency for a broad range of children.
lesbians to prefer girls may reflect same-gender couples’ In sum, heterosexual, lesbian, and gay male couples
awareness that, as a parental unit, they lack representation arrive at adoptive parenthood with distinct sets of experi-
of both genders (Goldberg and Allen 2007). Lesbians (and ences and expectations that may differ from one another
gay men) inevitably confront the societal belief that the and from those of expectant biological parents, which may
presence of a same-gender parent in the household is have implications for their gender preferences. For exam-
essential to children’s gender identity development (Juni et ple, heterosexual couples’ infertility experiences and lack of
al. 1985) and they are likely aware of societal concerns that biological connection to their child clearly distinguish them
the different-gender children of sexual minorities will lack from expectant biological parents, and may have implica-
same-gender role models with whom to identify. Lesbians tions for how they consider their future child’s gender.
and gay men may internalize such concerns about gender Lesbians and gay couples are uniquely shaped by their
socialization and, in turn, espouse a preference for a same- marginalized status as sexual minorities and their status as a
gender child. Alternatively, lesbians might be more likely to two-woman/two-man parental unit, both of which may
prefer girls (and gay men to prefer boys) simply as a affect their gender preferences. Finally, all adopters are
function of their own gender socialization: like heterosexual shaped by their own gender identities and socialization,
women, they may imagine that a same-gender child might which may have implications for their gender preferences.
be easier to raise based on their shared experiences as
females. It is also possible that the strong stigma against
Research Questions
male homosexuality in particular (Hicks 2006) may cause
some gay men to ponder whether it is easier for a girl to be
In light of the existing literature, the following questions
raised by two men than a boy. Indeed, adoption workers
and hypotheses were posed:
sometimes question the motives of gay male adopters,
reflecting pervasive stereotypes of gay men as child abusers 1. What are the gender preferences of preadoptive
and pedophiles (Hicks 2006). parents? Does either respondent’s sexual orientation
Presumably, not all sexual minorities will espouse (gay/heterosexual) or gender (male/female) impact
gender preferences; and, in fact, some research suggests preferences? Hypothesis: I expect that a large propor-
that prospective lesbian/gay adopters may be particularly tion of persons in both heterosexual and same-gender
non-restrictive with respect to their future child’s gender. couples will report no gender preference. Among those
Specifically, there is evidence that lesbian/gay adopters are with a preference, I expect that women will tend to
open to a broader range of children than heterosexual prefer girls, and men will tend to prefer boys. In other
adopters. For example, census data suggest that same- words, I expect that similar proportions of persons in
gender couples are significantly more likely to have heterosexual and same-gender couples will report no
adopted internationally than heterosexual couples (Gates gender preferences (i.e., there will be no significant
et al. 2007), suggesting that they may be more open to main effects or interactions between sexual orientation
adopting transracially. Census data also suggest that same- and gender). Among those with a preference, I expect a
gender couples are more likely to be fostering a child with a main effect of gender, such that women are more likely
disability than heterosexual couples (Gates et al. 2007), to prefer girls.
Sex Roles

2. To what extent do partners agree in their gender Participation entailed an individual telephone interview
preferences? Does sexual orientation or gender impact and completion of a questionnaire packet within 1–2 weeks
agreement? Hypothesis: I expect that partners in same- of the interview, to be completed before couples were
gender couples will be more likely to agree in their placed with a child. Couples were mailed two packets,
preferences than heterosexual couples. That is, because consent forms, and postage-paid envelopes, and partners
of their shared gender and shared status as stigmatized were asked to return their consent form with their packet.
minorities, partners in same-gender couples may have They then completed individual interviews.
many of the same concerns and beliefs about parenting
a same- or opposite-gender child and will show higher Description of the Sample
agreement. Thus, I expect a significant main effect of
sexual orientation on agreement. Heterosexual, lesbian, and gay couples did not differ in
3. What are individuals’ explanations for their preferen- geographical region. Thirty-four percent of the sample
ces? Does sexual orientation or gender impact their resided on the West Coast, 44% lived on the East Coast,
explanations? Hypothesis: This question will be ex- 8% lived in the Midwest, and 14% lived in the South.
plored qualitatively; thus, it is not appropriate to Heterosexual couples’ average relationship duration was
specify formal predictions regarding themes and pat- longer than lesbians’, but not gay men’s, F(2,199)=3.81, p<
terns. However, in light of existing theory and research, .05. Heterosexual, lesbian, and gay male couples had been
I cautiously posit that sexual minorities will explicitly in their current relationships for an average of 9.40 years
and uniquely consider their own sexual orientation and (SD=4.86), 7.55 years (SD=3.63) and 8.16 years (SD=
marginalized status in explaining their preferences. 3.27) respectively. There were no differences in age among
heterosexual women, men, lesbians, and gay men; the
average age of participants was 38.07 years old (SD=6.10).
The majority of participants were Caucasian: namely, 92%
Method of heterosexual women, 91% of heterosexual men, 88% of
lesbians, and 86% of gay men were Caucasian. These data
A total of 93 heterosexual couples (93 women, 93 men), 61 are comparable to national estimates, which indicate that
lesbian couples (122 women), and 48 gay male couples (96 approximately 76% of married heterosexual couples and
men) were included in the current study. All couples were 73% of same-gender couples with adopted children are
adopting for the first time, and no couples had prior Caucasian (Gates et al. 2007).
biological children. Heterosexual women, heterosexual men, lesbians, and
gay men had similar educational levels. Eight percent of the
Participant Recruitment sample had a high school education or less, 12% had an
associate’s degree/some college, 36% graduated college,
Inclusion criteria were: (a) couples must be adopting their 33% had a master’s degree, and 11% had a PhD/JD/MD.
first child; and (b) both partners must be becoming parents Significant differences in personal income emerged, F
for the first time. Adoption agencies throughout the USA (3,400)=8.95, p<.001. Post-hoc tests revealed that gay
were asked to provide study information to clients who had men earned a higher annual salary than lesbians (M=
not yet adopted. Census data were utilized to identify states $86,693, SD=$75,468 vs. M=$60,201, SD=$58,088) and
with a high percentage of lesbians and gay men (Gates and heterosexual women (M=$46,736, SD=$39,736); hetero-
Ost 2004) and effort was made to contact agencies in those sexual men also earned a higher salary than heterosexual
states. Over 30 agencies provided information to their women (M=$75,404, SD=$51,616). This sample is more
clients, typically in the form of a brochure that invited them financially affluent compared to national estimates, which
to participate in a study of the transition to adoptive indicate that the average household incomes for same-
parenthood; clients were asked to contact the researcher for gender couples and heterosexual married couples with
study details. Both heterosexual and same-gender couples adopted children are $102,474 and $81,900, respectively
were targeted through agencies to facilitate similarity on (Gates et al. 2007).
geographical location and income. Because some same- Heterosexual couples were more likely to have tried to
gender couples may not be “out” to agencies about their conceive than lesbian couples, χ2(1,154)=15.80, p<.001.
sexual orientation, gay/lesbian organizations such as the A total of 77 of 93 heterosexual couples (83%) had tried,
Human Rights Campaign, a national gay political organiza- compared to 28 of 61 lesbian couples (46%) in which at
tion, also assisted in disseminating study information. For least one partner had tried to conceive. Among gay men, 14
example, the HRC posted study information on their Family- couples (29%) had considered but not pursued surrogacy,
Net listserv, which is sent to 15,000 people per month. typically because of cost; and two couples (4%) had
 Sex Roles

pursued surrogacy, but stopped after multiple failed more conceptual in nature than initial coding (Charmaz
attempts. There were no differences among the three types 2006), and the categories that emerge are those that best
of couples in the length of time they had been waiting for a synthesize the data. Codes were examined across gender
child placement: At the time of the study, couples had been and sexual orientation, to identify similarities and contrasts
waiting for an average of 7.59 months (SD=9.17). A total in the nature and meaning of coding categories. At this
of 48 heterosexual (52%), 30 lesbian (49%), and 32 gay stage, I enlisted a research assistant to independently code a
couples (67%) were pursuing private domestic open random selection of transcripts (one fourth of the narratives
adoptions; 11 heterosexual (12%), 19 lesbian (31%), and generated by heterosexual, lesbian, and gay male partic-
11 gay couples (23%) were pursuing public adoptions; and ipants), in an effort to verify the soundness of the emerging
34 heterosexual (36%), 12 lesbian (20%), and five gay scheme (Patton 2002). This process of check coding (Miles
couples (10%) were pursuing international adoptions. and Huberman 1994) is useful in helping to clarify
categories and definitions and to provide a reliability check.
Open-Ended Questions Initial intercoder agreement ranged from 70–85% across
coding categories (reliability = #agreements/# agreements +
Participants were interviewed by the principal investigator disagreements). Disagreements were discussed and led to
and graduate student research assistants during the several refinements in the scheme and clarification of the
years 2005–2008. Interviews were later transcribed, and coding definitions. I applied the revised scheme to all
pseudonyms were assigned to protect confidentiality. The narratives, and the secondary coder independently coded a
data are derived from the following open-ended questions: random selection of transcripts (again, one-fourth of the
narratives). Intercoder agreement of our final scheme
1. Do you have any preference at all regarding the sex of
ranged from 90–95%, providing evidence of the utility of
the child that you hope to adopt? (Probe: Do you have
the scheme for describing the data. The final coding scheme
a preference for a boy or a girl?)
is represented in Table 3. Counts of all categories are
2. (If yes) What sex do you prefer?
provided for the purpose of communicating the most
a. Why? (Probe: Why are you drawn toward a boy/
frequently endorsed themes. This approach (i.e., quantify-
girl?)
ing qualitative data) can enhance the validity of the data, if
3. (If no) Why don’t you have a preference, do you think?
it is carefully carried out and is clearly linked to the
respondents’ own way of ordering the world (Hesse-Biber
Data Analysis 1995).

First, the relationships of sexual orientation and gender to

gender preferences were examined using logistic regres-
sion. Second, participants’ reasons for their gender prefer- Results
ence (or lack thereof) were examined through qualitative
analysis. This mixed-methods approach was appropriate Question 1: Preferences Regarding Child Gender
given my goal of generating a comprehensive portrait of an
understudied phenomenon (Bryman 2006). I utilized a To test my prediction that neither sexual orientation nor
modified grounded theory approach (Strauss and Corbin gender would affect gender preferences, I conducted a
1998) in analyzing the open-ended data. After reading logistic regression with sexual orientation, gender, and the
transcripts of each person’s data multiple times (i.e., 404 interaction of sexual orientation by gender as the indepen-
participant narratives) I initiated the coding process with dent variables, and gender preference (yes, no) as the
open coding, which involves examining each line of dependent variable. The logistic regression was significant,
narrative and defining events or actions within it. This led χ2(3,404)=11.60, p<.01, with sexual orientation (B=.95,
me to generate initial categories, which I reviewed and then SE=.30, Wald=9.90, p<.01), gender (B=.72, SE=.30,
often subsumed under more abstract categories. For Wald=5.63, p<.05), and the interaction (B=−.93, SE=.41,
example, persons who highlighted their interests in Wald=5.17, p<.05) all significantly contributing to whether
“cooking,” “crafts,” and “shopping” as reasons for wanting there was a gender preference (Table 1). Examination of the
girls were assigned the more general code of “enjoyment of significant interaction effect was done by a series of simple
female-typed activities.” chi-square analyses, as per Meyers et al. (2006). These
Next, I pursued focused coding, which uses initial simple comparisons showed that gay men (55.2%) were
themes that frequently reappear in order to sort the data significantly more likely than heterosexual men (32.3%) to
(e.g., concern about lack of opposite-gender role models as express a preference, χ2(1,189)= 10.10, p < .001; and,
a reason for wanting a same-gender child). This coding is among heterosexuals, females (49.5%) were significantly
Sex Roles

Table 1 Summary of logistic regression analyses. categorized as not in agreement if one partner had a
Variable B SE Wald statistic preference and one did not; or, if they both had a preference
but for different genders. To test my prediction that same-
Predicting Gender Preferences (Yes/No; n=404) gender couples would show higher agreement than hetero-
Sexual orientation .95 .30 9.90** sexuals, I conducted a logistic regression with sexual
Gender .72 .30 5.63* orientation, gender, and the interaction as the independent
Sexual orientation × gender −.93 .41 5.17* variables, and in agreement (yes, no) as the dependent
Predicting Actual Gender Preference (Boy/Girl; n=190) variable. Contrary to expectation, the logistic regression
Sexual orientation −1.04 .49 4.57* was not significant: agreement did not differ as a function
Gender .08 .52 .03 of sexual orientation or gender. A total of 63% of
Sexual orientation × gender .61 .64 .90 heterosexual couples agreed in their preferences, 56% of
lesbian partners agreed, and 51% of gay male partners
For sexual orientation, 1=gay/lesbian, 2=heterosexual; for gender, 1= agreed.
female, 2=male. For whether or not they had a gender preference, 1=
yes, 2=no; for the actual gender preference, 1=boy, 2=girl
Next, a logistic regression was performed only on
*p<.05. **p<.01
couples who were not in agreement. Disagreeing couples
were placed into two groups: those in which one partner
had a preference and one did not; and those with divergent
more likely than males (32.3%) to express a preference, preferences (i.e., one wanted a girl and one wanted a boy).
χ2(1,186) = 5.67, p < .05 (Table 2). Thus, contrary to Sexual orientation, gender, and their interaction were the
expectation, gay men were the most likely to demonstrate independent variables, and divergent/not divergent was the
a preference, and heterosexual men were the least likely. outcome. The logistic regression was not significant.
To investigate my hypothesis that respondent gender
would affect gender preferences, I conducted a logistic Question 3: Explanations for Preferences Regarding
regression among persons who had a preference regarding Child Gender
the gender of their child (n=190) using sexual orientation,
gender, and their interaction as the independent variables, Participants provided a range of reasons to explain their
and actual gender preference (boy, girl) as the dependent preferences for girls or boys, and their lack of preference
variable. The logistic regression was significant, χ2(3,190)= (Table 3). Of note is that some individuals reported more
8.77, p<.05, with only sexual orientation (B=−1.04, SE= than one reason.
.49, Wald=4.57, p<.05) significantly contributing to gender
preference; neither gender nor the interaction term contrib- Preferences for Girls
uted to gender preferences (Table 1). Contrary to prediction,
persons in heterosexual relationships (29%) were less likely Participants who preferred to adopt girls cited a number of
to prefer boys than persons in same-gender relationships reasons for this preference (Table 3). These reasons tended
(46%; see Table 2 for breakdowns of preferences by sexual to fall into several broad categories. Some categories were
orientation and gender). unique to same-gender couples; most, though, were cited
by all types of couples.
Question 2: Agreement within Couples Many participants expressed that they simply had an
inexplicable desire to adopt a girl. Of the 39 reasons cited
Of interest is how much agreement exists between partners by the 33 heterosexual women who preferred girls, 17 (43%
regarding their preferences, and whether there are differ- of the 39 reasons) concerned an inexplicable desire for
ences in agreement as a function of sexual orientation and girls. Similarly, five of the 21 reasons cited by the 21
gender. To examine this, couples were categorized as in heterosexual men who preferred girls (24%), seven of the
agreement if they both did not have a preference; or, if they 33 reasons cited by the 24 gay men who desired girls
both had a preference for the same gender. Couples were (21%), and nine of the 47 reasons cited by the 38 lesbians

Table 2 Percentage of parents who expressed gender preferences, by sexual orientation and gender.

Gay male Heterosexual male Lesbian female Heterosexual female

% with preference for gender of the child (n=404) 55.2a 32.3a,b 50.0 49.5b
% who preferred girls (n=190) 45.3 70.0 62.3 71.1

Percentages sharing subscripts are significantly different from each other, p<.05
 Sex Roles

Table 3 Reasons for preferring girls, boys, and no gender preferences.

Type of couple

Heterosexual women (n, %) Heterosexual men (n, %) Lesbians (n, %) Gay men (n, %)

Reasons for preferring girls

Inexplicable desire 17 (43%) 5 (24%) 9 (20%) 7 (21%)
Gender identity considerations 4 (10%) 1 (5%) 3 (6%) 2 (6%)
Perceived characteristics of girls 3 (8%) 5 (24%) 3 (6%) 2 (6%)
Perceived characteristics of relationship 2 (5%) 4 (19%) 1 (2%) 1 (3%)
Gender socialization considerations 3 (8%) 2 (10%) 22 (47%) 6 (18%)
Societal heterosexism considerations 2 (4%) 14 (43%)
Miscellaneous 10 (26%) 4 (18%) 7 (15%) 1 (3%)
Total # reasons 39 21 47 33
Total # participants who preferred girls 33 21 38 24
Reasons for preferring boys
Inexplicable desire 3 (24%) 2 (14%) 6 (21%) 8 (18%)
Gender identity considerations 4 (29%) 12 (41%) 2 (4.5%)
Perceived characteristics of boys 2 (17%) 2 (14%) 1 (3%) 2 (4.5%)
Perceived characteristics of relationship 2 (17%) 4 (9%)
Gender socialization considerations 2 (17%) 1 (7%) 4 (14%) 24 (55%)
Societal heterosexism considerations 2 (7%)
Internalization of patriarchal norms 1 (8%) 5 (36%) 2 (4.5%)
Miscellaneous 2 (17%) 4 (14%) 2 (4.5%)
Total # reasons 12a 14 29 44
Total # participants who preferred boys 13 9 23 29
Reasons for no gender preferences
Child gender not that important 32 (68%) 48 (74%) 40 (60%) 24 (43%)
Health of child more important 4 (9%) 1 (2%) 4 (7%) 6 (10%)
Leaving it up to fate 1 (2%) 4 (6%) 3 (6%)
Not stereotypically male 3 (4%) 1 (2%)
Positives/negatives of both boys and girls 3 (6%) 2 (3%) 11 (16%) 6 (11%)
Initial preference; now don’t care 3 (6%) 5 (8%) 5 (7%) 8 (15%)
Eventually want one of each 4 (9%) 2 (3%) 7 (10%) 7 (13%)
Total # reasons 47 65 67 55
Total # participants with no gender prefs 47 63 61 43
a
There was missing data for one heterosexual woman

who preferred girls (20%) emphasized an inexplicable Gender identity considerations were cited by several
desire. These individuals could not cite any specific reasons members of all three types of couples in explaining their
except to say that they had a strong “leaning” or preference for girls. These individuals emphasized that their
“inclination” towards girls; a daughter was something that interests and the activities that they enjoyed were feminine
they intuitively wanted but “couldn’t explain.” Explained (or at least not traditionally masculine), and, thus, they felt
Roy, a gay man, “I cannot tell you [why]. It is just some that they might experience greater enjoyment parenting a
very visceral desire I have and I cannot articulate it beyond female child. Indeed, both lesbians (n = 3; 6%) and
that.” Said Ben, a heterosexual man, “I don’t know why, I heterosexual women (n=4; 10%) emphasized their own
just kind of always wanted a daughter.” Notably, hetero- feminine interests and “girly-ness” in explaining their
sexual women were particularly likely to say that they had preference for daughters. For example, Alicia, a heterosex-
always “imagined” raising a daughter (five of 17 women): ual woman, said, “I would like a girl only because when I
“When I imagine my child, I just usually imagine a girl.” sort of imagined the things that we would do—I like to do
Sex Roles

crafts, and cook, and that kind of stuff I could do with a own—and their partners’—socialization as females may
girl.” Two gay men (6%) and one heterosexual man (5%) represent as an asset, and a limitation, with regard to raising
specifically noted their lack of interest in sports in explain- children. Between both partners, they possess extensive
ing why they might enjoy parenting a boy less than a girl. knowledge of and sensitivity to the experiences and
Stated Damian, a gay man: “I think our fear is, if we challenges of being female; by extension, they lack
had a boy and he ended up being a jock, what would firsthand knowledge of what it is like to be a man in
we do? (laughs). How would we teach him how to play society. Further, unlike heterosexual women, they cannot
football and all? Our interests are just more on the count on a male coparent to fill this gap. A total of 22
bookish side.” lesbians (47%) emphasized their greater knowledge, com-
fort, preparation, and confidence in raising girls as
Perceived characteristics of girls were also cited as a compared to boys, based upon their own personal experi-
reason for preferring girls. Five heterosexual men (24%), ences and socialization as females. Four of these women
three heterosexual women (8%), three lesbians (6%), and also noted that they had grown up around mostly women
two gay men (6%) espoused particular beliefs or attribu- (e.g., with all sisters) and thus felt that they knew “next to
tions about girls, all of which were positive. Several nothing” about boys. Stated Suzanne: “I prefer a girl. I can
participants desired daughters because of their perception relate to a girl more. I've been there, I am a girl, and I know
that girls were “easier” to rear than boys (two heterosexual what girls go through.”
women, one heterosexual man, one lesbian, one gay man). In some cases, women noted that not only would it be
Additionally, several participants shared their belief that easier for them to raise a girl, but that, additionally, it might
girls were more “interesting” and “complex” than boys in be easier for a girl to be raised by two mothers than a boy,
an effort to explain their preference for a daughter (one who would lack a same-gender male parent with whom to
heterosexual woman, one heterosexual man, one lesbian). A identify. Emphasized Laurie:
few participants perceived girls as “safer” and “less
Any kid’s gonna have obstacles having two lesbian
physically challenging than boys, who tend to break things”
mothers. Having a girl, it might be easier for them, in
(one heterosexual man, one lesbian). Several other positive
terms of having two parents to identify with. In some
attributions about girls were named, including the percep-
ways it’s easier to identify with someone of your own
tion that “girls are less likely to have ADHD than boys”
gender than the other. A boy wouldn’t have that.
(one gay man), “girls are cute” (one heterosexual man), and
“there are more opportunities in the world for girls than Furthermore, some women felt downright unprepared to
boys” (one heterosexual man). deal with the challenges of raising an opposite-gender child,
particularly during adolescence:
Perceived characteristics of the parent–daughter relationship
I’m a girl, and I have three nieces, and I guess I would
were also cited as a reason for preferring girls. That is,
know more how to deal with things from a girl’s
participants expressed beliefs about the nature of the
perspective, if that makes sense. How comfortable is a
relationship that girls enjoy with their fathers/mothers
little boy going to be coming to ask Mom about, um,
which led them to experience greater excitement about the
his first wet dream or something like that, you know?
possibility of raising a daughter. Four heterosexual men
How do we deal with that, other than calling up Uncle
(19%) and one gay man (3%) expressed that they liked the
Eric and saying, “Talk to him!”
idea of “daddy’s little girl”—that is, they were reassured by
the belief that “all girls love their daddies.” Two hetero- Some gay men (n=6; 18%) also emphasized socializa-
sexual women (5%) described positive relationships with tion/preparation factors in explaining their preference for a
their own mothers, noting that they wished to replicate daughter. Namely, several gay men cited good relationships
those relationships with a daughter: “I knew I wanted to with females (sisters, friends) who could serve as role
have a daughter who would have the same relationship with models as a reason why they felt they could be particularly
me that I have with my mom.” One lesbian (2%) noted that good parents to girls (three men; 9%), and several men
“girls hold families together, generation through genera- emphasized their significant experience caring for girls (e.g.,
tion” and felt that a daughter would remain more committed babysitting younger sisters) (three men; 9%). These factors—
to her parents than a son. the availability of female role models, and their own personal
experience caring for young girls—enabled them to feel well-
Gender socialization considerations were particularly com- equipped to meet the socialization needs of potential
mon among lesbians. Lesbian couples who seek to adopt daughters, and to express a high level of comfort with the
are inevitably aware of their status as a two-woman parental idea of raising girls. Explained Shawn: “I feel there are
unit, and thus may be sensitive to the ways in which their enough female role models in our lives, women that will be in
 Sex Roles

our lives, to help out with talking about periods and all the rest Two lesbian women (4%) also cited societal heterosex-
of that stuff.” Likewise, Kevin noted, “I spent most of my ism in explaining why they preferred girls. Specifically,
babysitting time with girls. I’m a little more comfortable both women described their observation that boys with
babysitting girls so I kind of want a girl over a boy.” lesbian mothers seemed to encounter less acceptance and
Three heterosexual women (8%) and two heterosexual more teasing (e.g., regarding their masculinity) than girls
men (10%) also emphasized prior experiences babysitting with lesbian mothers, thereby leading them to favor the idea
or caring for female relatives in explaining why they felt of adopting a girl.
well-suited to parent girls. Thus, unlike lesbians, hetero- Several other miscellaneous reasons were cited by
sexual women did not tend to emphasize their own female participants in explaining their desire for girls. Six lesbians
socialization as a reason for preferring girls. Heterosexual (13%) and two heterosexual women (5%) emphasized the
women (who were married to men) were simply not as preponderance of boys in their extended families, and
concerned about or aware of the ways in which their own acknowledged a desire to “balance things out” by adopting
gender socialization would both benefit and limit them as a daughter. Four heterosexual women (10%) and two
parents. Likewise, heterosexual men obviously did not invoke heterosexual men (9%) noted that their partners’ desire for
the presence of female role models in explaining their a girl had led them to prefer a girl, for their partner’s sake.
preference for girls; their wives necessarily functioned as For three heterosexual women (8%) and one heterosexual
“role models” by virtue of their presence as female coparents. man (4.5%), the wish for a daughter was fueled by a sense
of duty to rescue “abandoned” girls from China. Similarly,
Consideration of societal heterosexism was mentioned one lesbian (2%) described her perception that male adopt-
mainly by gay men in explaining their preference for a ees were more “in demand,” which led her to prefer girls,
girl. That is, some men considered whether society might whom she regarded as “less desired.” One heterosexual
be less accepting of a boy with two fathers than a girl with woman (3%) revealed that a prior failed adoption had been
two fathers. Indeed, 14 men (43%) preferred girls because with a boy; the sting of this experience led her to want
of their belief that a girl with two dads would be less prone something “different.” One heterosexual man (4.5%) stated
to teasing than a boy with two dads. These men understood that he “liked the idea of an older sister” in explaining his
that “all kids may get teased a little for having gay parents, desire for a girl. In contrast, one gay man (3%) preferred to
but for girls it seems like it would be a lot less.” Said adopt a daughter first because he did not want a boy to
Matthew, “I really want a girl. If we get a boy it’ll be ok, “buy into the masculine construction” of needing to protect
but honestly I think a girl would be easier to raise. A boy his younger sister.
having two fathers—when he gets to his early teens he is Thus, to summarize, the most common reason for
going to get his butt busted. It is going to be harder for him preferring a girl among heterosexual women was their
than for a girl.” “inexplicable desire” for a daughter, whereas heterosexual
Some of these men also noted that they felt that a men most frequently invoked their inexplicable desire for a
daughter would be more accepting of and less threatened by girl, perceived characteristics of girls, and perceived charac-
her two gay dads than a son. Observed Orson: teristics of the father-daughter relationship. Lesbians, on the
other hand, tended to emphasize gender socialization consid-
I feel that that women, girls, are less critical of gay erations in explaining their desire for a girl, and gay men most
men. I think that in today’s society, it’s become frequently cited concerns about societal heterosexism.
almost cool to be gay to a girl. I think that a male
would be more threatened by it. I just think it would Preferences for Boys
be a better situation for a girl—you know, if you have
a boy, and they’re playing sports and things like Participants who preferred to adopt boys named a range of
that…gay dads might not be okay [to them]. reasons for their preferences (Table 3). Like those who
preferred girls, many participants who espoused a prefer-
Likewise, Greg put it like this:
ence for boys explained this preference in terms of an
I sort of feel like the whole idea of sexual fluidity, or inexplicable desire to raise a son. Of the 12 reasons cited by
comfort around gender identity issues, is more pro- the 13 heterosexual women who preferred boys, three
nounced in women than in men. And I think that’s (24%) focused on an intuitive, inexplicable desire. Similar-
probably not gender-based but society-based. So I feel ly, six of the 29 reasons cited by the 23 lesbians who
like that’s where my feeling that women would be more preferred girls (21%), eight of the 44 reasons named by the
okay with gay dads comes from. It’s not necessarily that 29 gay men who preferred boys (18%), and two of the 14
men are innately more discriminatory, but there’s more reasons cited by the nine heterosexual men who preferred
support for women to be more open-minded. boys (14%) centered around an inexplicable desire. These
Sex Roles

men and women stated that they were “intuitively” drawn “useful” than a daughter (e.g., he imagined that a son might
towards boys or “leaned towards boys for no particular be more capable of cutting the lawn).
reason.” Barry, a gay man stated, “I guess my intuitive
sense is just wanting a boy but I am not sure why.” Characteristics of the parent–son relationship were also
Likewise, Moira, a heterosexual woman, mused, “If I could cited by several participants. For example, four gay men
pick, I’d choose a boy. Why is that? I don’t know.” (9%) imagined that they would enjoy a particularly special
As with participants who preferred girls, gender identity bond with a son. These men all noted that there was
considerations were cited by some participants who something special about the “father–son bond” that they
preferred to adopt boys. Lesbians in particular were likely wished to experience; two of them also noted that they
to emphasize that their own interests tended to be more wished to have the kind of relationship with a son that their
masculine and “tomboyish,” not “frilly” or “fru-fru” in fathers did not have with them. Said Peter, “I feel more of a
explaining their boy preference. Indeed, 12 lesbians (41%) bond, more of a natural kinship towards the whole father–
emphasized their personal lack of interest in “girly things,” son atmosphere. And that’s probably, if you want to get all
and, in turn, their interest in stereotypically masculine activi- psychological, a reflection of a failed father–son relation-
ties (e.g., sports, outdoor activities, “roughhousing”). Stated ship I have with my own father.” Two heterosexual women
Marta: “I was a real tomboy, so I just think there’s a way in (17%) also imagined something special about the bond that
which I can imagine very vividly the fun of having a boy they would have with a son. For one woman, her strained
child. And, I’m female, but, like, I was always into sports and relationship with her own mother led her to be “terrified” of
playing outside compared to like the girly girl type of things.” raising a daughter. She believed that the mother–son
Similarly, Edie explained: dynamic would be more harmonious and less contentious.
I think [I prefer a boy] because I was a tomboy growing Another woman similarly explained, “I prefer boys because
up. It’s really hard for me to relate to—all my nieces are boys love their mothers, always. When they’re teenagers
very fru-fru and everything has to be pink and glittery they still love their moms, whereas girls really don’t.”
and I have no concept of that (laughs). They don’t like
to get dirty. It is just not who I was or who I am. In Gender socialization considerations were cited by many
some ways I’d rather have the boy that comes home participants, particularly gay men. Namely, 24 gay men
with the frog in his pocket than the girl that says, “Oh (55%), two heterosexual women (17%), four lesbians
my god, I have dirt on my shoes!” and freaks out. (14%), and one heterosexual man (7%) emphasized that
they felt more confident about their ability to raise and
Four heterosexual men (29%) and two gay men (4.5%)
socialize boys than girls. These individuals perceived
noted that their own interests tended to be more masculine
themselves as possessing the skills and experiences, as
(e.g., they enjoyed fishing, sports, and “outdoor stuff”) and
well as the relevant socialization (in the case of male
therefore looked forward to raising sons. For example,
participants) to raise male children. Both heterosexual
Robert, a gay man, stated, “I’m not, like, macho, but I like
women and lesbians emphasized their experiences growing
all the boy toys. I get more excited thinking about baseball
up with, and often caring for, brothers and male relatives;
and Tonka trucks than I do about dolls. [So I want a boy]
thus, males were their “reference point” and they felt
from a selfish perspective.” These men, then, felt that,
comfortable with the idea of raising sons. Gay men and
given their own masculine interests, they would have more
heterosexual men, in contrast, referred to their own personal
fun parenting boys than girls.
experiences and socialization, as boys, and now as men, in
explaining their greater comfort with raising sons. Howev-
Perceived characteristics of boys were also cited by several
er, unlike heterosexual men, gay men do not have partners
participants in explaining their preference for sons. Specif-
whom they can count on to ensure that a daughter receives
ically, two heterosexual women (17%), two heterosexual
proper female socialization. Thus, in addition to consider-
men (14%), two gay men (4.5%), and one lesbian (3%)
ing their own male socialization, they also considered their
described gendered attributions in explaining why they
status as a two-dad, no-mom family, which ultimately led
preferred boys. A few participants described their percep-
them to conclude that they might be better prepared and
tion that boys seemed less “complex and complicated” than
more well equipped to raise a son than a daughter. Stated
girls (one heterosexual woman, one lesbian, one gay man).
Eugene:
One heterosexual man and one heterosexual woman felt
that “girls are more difficult to relate to around puberty” We both get kind of squeamish around girly things,
which led them to prefer boys. One gay man voiced his uncomfortable when my niece starts talking about
perception that boys were simply more “fun” than girls. girly things….We’ve never had that same comfort
Finally, one heterosexual man felt that a son might be more with girly stuff. So we thought we’d be better parents
 Sex Roles

for a boy. And we thought our interests and heterosexual man, explained, “My oldest brother has no kids,
experiences are more geared towards, would probably and, well, he’s never gonna have kids. My middle brother has
be more helpful for a boy. And also, we are going to two girls. So there, there’s a little bit of pressure to pass on the
be a two dad family. There aren’t gonna be any family name. He’ll be the little heir, you know?” An
women to help out with the girly stuff. So that was the additional two heterosexual men noted that they preferred
deciding factor….It’s just what we thought we’d be their first-born children, specifically, to be sons, although
comfortable with. they could not exactly articulate why (e.g., they noted that
“it’s a typical guy thing” and “I don’t know; it just seems
Thus, as highlighted in Eugene’s narrative, many men really important”). Finally, when asked why he preferred a
who preferred boys over girls not only noted their own boy, one gay man simply acknowledged that “we live in a
greater comfort in raising boys (e.g., because of their own sexist culture, therefore I guess we tend to prefer boys.”
socialization as boys, their familiarity with male body parts, A number of other miscellaneous reasons were also
and so on), but also emphasized their consideration of the cited. Namely, two heterosexual women (17%), four
child’s perspective. That is, they wondered whether a son lesbians (14%), and two gay men (4.5%) named other
might benefit from having two male parents that understood reasons for preferring boys. Three lesbians described their
and could relate to their experiences, and, by extension, perception that girls were more “in demand” as potential
whether a girl might suffer in the absence of a female parent adoptees, leading them to develop a strong inclination
with whom they could identify, particularly during adoles- towards boys. One gay man and one heterosexual woman
cence. Indeed, some men noted a deficiency in sisters and noted that they preferred to adopt a boy because of their
female friends who could serve as role models to a partner’s strong desire for a son. One gay man preferred a
daughter, which served to magnify their concerns about boy because he enjoyed a close, special relationship with
raising a girl. In their eyes, not only did they lack personal his niece and felt a relationship with a girl might “compete”
experience as females, and not only would their child grow with this relationship. Finally, one heterosexual woman
up without a mother, but they had limited access to women preferred a son because a prior failed adoption had been
outside the family. Explained Jim: with a boy, and one lesbian woman preferred a son because
of the large number of girls in her extended family.
Because we’re both males, I think it would be sort of
Thus, to summarize, the most common reason for
hard. We don’t have many female friends in our lives;
preferring boys among heterosexual women was an
we have our sisters but they live out of state. I think of
“inexplicable desire” for a son (although there were no
girls as needing their mothers when they reach a certain
notable patterns in women’s responses), whereas hetero-
age so they can understand about their female—you
sexual men’s desire for a son most frequently reflected
know, when they reach womanhood, they need their
patriarchal norms and gender identity considerations.
mothers to explain it all to them. She wouldn’t have a
Lesbians most frequently invoked gender identity consid-
mother and I don’t know how I could actually explain it
erations (i.e., their own atypical gender identities), and gay
all to a girl.
men most often highlighted gender socialization consider-
Thus, Jim and others felt insecure about their ability to ations in explaining their preference for a boy.
meet the socialization needs of a daughter, especially
during adolescence. In contrast, they viewed their knowl- No Gender Preference
edge of the “mechanics” of both male psychology and
anatomy as excellent preparation for raising a son. Many participants did not endorse any preferences regard-
ing child gender: they were equally interested in the
Consideration of societal heterosexism was emphasized by prospect of parenting boys and girls. These participants
lesbians only (n=2; 7%). One woman wanted a boy did not always have specific explanations for their lack of
because she did not want to be accused of sexually abusing preference (Table 3). When probed as to what drove their
a girl; the other wanted a son because she did not want to lack of preference, many simply said that they just didn’t
be accused of trying to “make” her daughter a lesbian. care: they were “open to either” and “it just doesn’t matter.”
Some participants—namely, five heterosexual men Of the 65 reasons named by the 63 heterosexual men that
(36%), one heterosexual woman (8%), and two gay men did not have a preference, 48 (74%) converged around this
(4.5%)—explained their reasons for preferring a boy in theme. Likewise, 32 of the 47 reasons named by the 47
ways that reflected the internalization of patriarchal norms heterosexual women who did not have a preference (68%),
and ideals. Specifically, three heterosexual men, one 40 of the 67 reasons provided by the 61 lesbians who did
heterosexual woman, and one gay man desired sons in not have a preference (60%), and 24 of the 55 reasons
order to “continue on the family name.” As Timothy, a named by the 43 gay men who did not have a preference
Sex Roles

(43%) also fell into this category. For example, Tony, a and disadvantages were named by participants. For exam-
heterosexual man, exclaimed, “No, not at all. I have no ple, one gay man acknowledged that having a boy would be
preference. I would just be happy with whatever was given something that he would understand more “intimately and
to us. I don’t care.” Sherry, a lesbian, said, “I just don’t immediately,” but noted that his father and his sisters had a
care. All I want is a child. Ten fingers, ten toes. I’d be fine “great relationship,” leading him to feel that he would be
with a boy, I’d be fine with a girl. I’m open.” For many, equally interested in parenting boys and girls. Another gay
child gender was simply unimportant to them in light of man noted that “with girls, you have to worry about them
their long-awaited goal of being parents. dating; with boys you have to worry about them getting
Several individuals explicitly noted that the health of into trouble.” Two lesbians mused that boys were suppos-
their future child was most important to them; in this edly more difficult in childhood, but easier as teenagers,
context, they explained, child gender made little difference. when girls became “scary.” Two heterosexual women felt
Namely, six gay men (10%), four heterosexual women that while girls seemed “easier,” boys seemed “more fun,”
(9%), four lesbians (7%), and one heterosexual man (2%) leading them to feel excited about both.
stated that child gender was insignificant in comparison to Eight gay men (15%), five heterosexual men (8%), five
child health. Stated Lucy, a heterosexual woman: “Oh, I lesbians (7%), and three heterosexual women (6%) noted
don’t care. I really couldn’t care. As long as it’s healthy, I that they had had initial gender preferences which had
really couldn’t care.” gradually evaporated over time, leaving them feeling as
Several participants noted that they felt uncomfortable though “either would be fine.” Thus, they experienced
declaring a gender preference when “in nature you don’t get initial gender preferences that eventually disappeared, as
to make those decisions.” They noted that if they were time went on and their desire for a child—any child—grew.
having a biological child they would be unable to choose, For example, several participants (both male and female)
and thus preferred to leave it to fate to decide their future noted that early on, they had believed that raising a girl
child’s gender. Four heterosexual men (6%), three gay men would be easier and more fun; however, as they went
(6%), and one heterosexual woman (2%) endorsed this as through the adoption process, they became more comfort-
explanation for why they had no preference. Said Luke, a able with the possibility of raising a child of either gender,
heterosexual man: concluding that “it just doesn’t matter; a child is a child,” as
one gay man stated. Similarly, several men (gay and
You can’t control that in nature, so, I’m trying to look
straight) noted that early on they had felt more “comfort-
at this as much as possible as, you know, whatever the
able” with the prospect of raising a boy “because we
natural course of things will be. It’s not like we can
understand boys better,” but had ultimately decided that the
breed specifically for a boy or girl so I don’t care if
gender of their child did not make a difference. Said Mike,
we get a boy or girl. My take on it is, you don’t get to
a gay man:
choose when you do it naturally, so we might as well
take the same attitude with adoption. At first we both thought of a boy just because we
understand boys better. But now we don’t care. I’m not
Three heterosexual men (4%) and one gay man (2%)
sure why that changed. I guess, as we go through this
attributed their openness regarding child gender to the fact
process, our comfort level just grows; we get more
that, in their estimation, they were not stereotypically male
comfortable with whatever. Whatever the race of the
(e.g., in their interests, personality, etc.), and therefore they
child, the gender of the child, it doesn’t really matter.
felt equally interested in and open to parenting boys and
girls. Said Lars, a heterosexual man, “I’m not a big sports Similar sentiments were shared by several women,
fan, so I don’t have dreams of my kid playing all these lesbian and heterosexual, who voiced their experience that
sports. A girl would be just as much fun as a boy. I’m not early on, they had preferred girls, but that as time went on,
like some guys, who have it in their minds that they want a they came to feel that “the gender thing just doesn’t
boy so they can play football and blah blah blah.” matter…we’d be excited about either one.”
Some participants grounded their lack of gender prefer- Finally, seven gay men (13%), seven lesbians (10%),
ences in their perception that there were positives and four heterosexual women (9%), and two heterosexual men
negatives associated with parenting boys and girls. Eleven (3%) noted that they eventually wanted at least one of each;
lesbians (16%), six gay men (11%), three heterosexual thus, it did not matter to them whether a boy or girl came
women (6%), and two heterosexual men (3%) noted that first. Said Erin, a heterosexual woman: “We definitely
boys and girls both “come with advantages and disadvan- would like to have or adopt several children, boys and girls,
tages” and “have their own challenges and benefits.” In so for the first we don’t care.”
turn, their cost–benefit analysis left them with no strong Thus, to summarize, the majority of participants—
preferences regarding child gender. A variety of advantages heterosexual, lesbian, and gay—who did not espouse any
 Sex Roles

gender preference explained that they “just didn’t care” about who had a gender preference tended to prefer girls. Further,
their future child’s gender; that is, gender was insignificant in heterosexual men also tended to prefer daughters. Perhaps
the context of their larger goal of becoming parents. women are the “agents” of adoption, and therefore
influence their husbands’ preferences via their own, such
that men come to assume their wives’ preferences. Indeed,
Discussion the absence of a biological connection between father and
child may free heterosexual men in particular from valuing
This study represents the first investigation known to date and desiring sons as status symbols. Unable to continue on
that explores the child gender preferences of both hetero- the male genealogical lineage biologically, they may simply
sexual and sexual minority preadoptive parents. The data not have strong preferences regarding their child’s gender,
suggest that both the adoption context and the sexual and may therefore tend to espouse no preference, or, a
orientation context may have implications for how men and preference for a girl—if that is what they perceive their
women think about the gender of their future children. wives as preferring (indeed, of note is that, contrary to
Of note is that many participants did not voice an expectation, heterosexual partners were the most likely to
explicit preference for the gender of their adoptive child. In agree in their preferences, a fact that may reflect some
studies of biological parents and nonparents, the percentage men’s tendency to assume their wives’ preferences).
of individuals with “no preference” has ranged from less Individuals in same-gender relationships were significantly
than 20% (Walker and Conner 1993) to 35% (Gartrell et al. less likely than heterosexuals to prefer girls, although it is
1996; Hank and Kohler 2003; Jones 2008) to 60% notable that the majority of lesbians with a preference
(Herrmann-Green and Gehring 2007; Steinbacher and (62%) did desire daughters, which is consistent with
Gilroy 1985). It is important to note that studies vary research on lesbians who became parents via insemination
considerably in how they query participants about gender (Gartrell et al. 1996; Herrmann-Green and Gehring 2007).
preferences, and thus, it is difficult to draw comparisons The fact that 55% of gay men desired sons is remarkably
among studies. That being said, it is notable that the consistent with census data showing that 54% of same-
percentages of gay men, lesbians and heterosexual women gender adoptive male couples are the parents of boys, and
with no gender preferences fall within this upper range suggests the importance of sexual orientation and the
(45%, 50%, and 50%) and the percentage of heterosexual couple context (i.e., the presence of two men versus a
men with no preference exceeds this upper limit (68%). man and a woman) in dictating gender preferences.
Elements of the adoption context may help to explain Of interest, of course, is how participants explained their
why so many of the men and women in the sample gender preferences. Consistent with prior research, hetero-
espoused a lack of gender preferences. For individuals that sexual men who desired girls often cited their desire for a
experienced infertility—mainly heterosexual couples—the particular type of relationship with a daughter (Hammer and
salience of child gender may pale in comparison to the McFerran 1988) and also described various attributes of
long-awaited goal of having a child. Similarly, both daughters (e.g., cute, nondestructive; Arnold and Kuo 1984;
heterosexual and sexual minority participants may feel that Williamson 1976). Also consistent with prior research, men
gender is relatively unimportant in comparison to other child who desired boys often emphasized their own orientation
characteristics, such as health, race, and age (Brodzinsky and toward masculine activities and interests (Pollard and
Pinderhughes 2002). Lesbians and gay men in particular Morgan 2002; Williamson 1976). Heterosexual men’s most
may feel that as “less-than-ideal” adoptive parents, they are common reason for preferring boys, however, was the
in no position to set a gender criterion for their future child desire to carry on the family name (Arnold and Kuo 1984)
(Matthews and Cramer 2006); although notably, this and their desire for their “first born” to be male, suggesting
explanation was not explicitly mentioned by lesbian and that, for a minority of men, sons continue to be valued as
gay participants. Additionally, it is notable that some gay signifiers of masculinity (Katzev et al. 1994; Williamson
men (and a few lesbians) observed that they had possessed 1976). Thus, although most heterosexual men had no
initial gender preferences which gradually diminished over gender preferences or desired girls, those men who desired
time. Perhaps these individuals—who, for the most part, boys were sometimes motivated by very traditional con-
did not experience fertility problems—experienced during cerns: namely, the extent to which their son would be a
the adoption process what heterosexual couples experi- symbol of status (Hammer and McFerran 1988).
enced during infertility: a gradual loosening of their Heterosexual women who preferred girls typically
imagined, ideal child, as the wait for a child stretched on explained this in terms of an inexplicable desire—that is,
and their yearning for a child grew. they weren’t sure why, but they had always longed for or
Consistent with nationally representative data on women imagined a girl. It is very likely that what these women
considering adoption (Jones 2008), heterosexual women were not able to fully articulate was their desire for
Sex Roles

companionship (Warren 1985) and mutual identification girl (which is consistent with data suggesting that sons of
(Notman 2006) based on their gender similarity. Indeed, lesbians may be more likely to be teased about their
only a few women cited gender identity considerations (i.e., sexuality than daughters of lesbians: Tasker and Golombok
I like to cook so I hope for a girl so we can cook together) 1997). These data suggest that sexual minorities, particu-
and gender socialization considerations (i.e., I am a girl so I larly gay men, are highly aware of the ways in which
think I’d be a better parent to a girl) in explaining their homosexuality continues to represent a threat to masculinity
preference for a girl. Perhaps those women who cited an (Kane 2006), such that boys are expected to experience
“inexplicable desire” or “inclination” were also motivated greater attacks on their sexuality than girls as a function of
by these types of considerations but simply could not having gay parents.
articulate them, in part because of the difficulty explaining Interestingly, almost half of the lesbians who preferred to
what for many women seems to be a completely intuitive adopt boys explained this desire in terms of their own
process (i.e., their yearning for a girl). No consistent opposite-gender interests and orientations. Indeed, some
patterns emerged in women’s reasons for preferring boys, research suggests that lesbians and gay men are more likely
which may in part be a function of the small number of to demonstrate androgynous or cross-gender interests
heterosexual women who preferred boys (13 women). (Carlson and Steuer 1985); in turn, women’s self-
In contrast to heterosexual women, lesbians who perceived nonconformity with the norms associated with
preferred girls frequently invoked gender socialization their gender led them to wonder whether they might be
considerations; likewise, in contrast to heterosexual men, better parents to (and/or have greater enjoyment raising) an
gay men who preferred boys also tended to highlight opposite-gender child. Thus, as prior research has shown, a
concerns about gender socialization. This distinction clearly desire to participate in particular (gender-typed) activities
reflects lesbians’ and gay men’s awareness of their unique may underlie some individuals’ gender preferences
status as a parental unit in which only one gender is (Pollard and Morgan 2002; Williamson 1976), although,
represented. Both lesbians and gay men worried about in this case, it was lesbians’ preference for opposite-
procuring opposite-gender role models for their opposite- gender activities that led them to prefer an opposite-
gender children, and expressed concern about how to gender child.
handle delicate subjects such as pubertal changes with an
opposite-gender child. Such worries about gender sociali- Limitations, Conclusions, and Future Directions
zation appear to be both child-centered (e.g., they wish to
“do right” by their child) and parent-centered (e.g., they worry The current study is limited in several ways. First, the
about their capacity to construct their children’s gender sample is well-educated, affluent, and mostly White. While
identity according to societal norms; Kane 2006). That is, these characteristics are typical of many adopters, they do
they are aware that they will be held accountable by both not describe the entire spectrum of preadoptive parents
their future child and society for their (in)ability to provide (Gates et al. 2007). Less educated adopters and racial
proper gender socialization (Kane 2006). By extension, minority adopters may have very different expectations and
many lesbians and gay men felt particularly confident about ideals regarding their child’s gender. Second, this study did
their ability to parent a child of the same gender, given their not include a sample of non-adoptive heterosexual, lesbian,
own and their partners’ “insider knowledge” regarding body or gay parents-to-be, which significantly limits our ability
parts, gender development, and gender socialization. This to draw conclusions regarding the extent to which the
confidence, in turn, led them to believe that they might be adoption context per se is responsible for the observed
particularly good parents to a same-gender child. patterns of gender preference. Third, the preadoptive
Gay men who preferred girls were often uniquely parents in this sample were all in committed relationships;
motivated by concerns about societal heterosexism and thus, findings regarding their gender preferences may not
scrutiny. Almost half of the men who desired daughters generalize to single prospective adoptive [Link],
expressed the belief that a boy raised by two fathers would the current study does not include post-adoptive data.
encounter more stigma and resistance in society than a girl Future research should investigate preadoptive parents’
raised by two fathers. Further, some of these men felt that a gender preferences longitudinally, in order to determine
daughter would be more accepting of her fathers’ homo- how impactful these preferences are in determining the
sexuality, given females’ greater freedom to embrace a actual gender of the child that couples adopt. Finally, this
diversity of gender and sexual expressions, and thus their study did not assess the magnitude of individuals’ gender
tendency to be less homophobic than males (D’Angelo et preferences. It is likely that for some, these preferences are
al. 1998). This sentiment was shared by several lesbians mild, whereas for others they are intense. The magnitude of
who preferred girls, who voiced their feeling that a boy such preferences is important, as they might have implications
with two mothers might have a more difficult time than a for couples’ decision-making during the adoption process
 Sex Roles

(e.g., whether a couple turns down a potential placement Arnold, F., & Kuo, E. (1984). The value of daughters and sons: A
comparative study of gender preferences of parents. Journal of
because it is the “wrong” gender). Future research should
Comparative Family Studies, 15, 299–318.
(a) assess the intensity of adoptive parents’ gender Bachrach, C. A., Stolley, K. S., & London, K. A. (1992).
preferences and (b) evaluate the consequences of such Relinquishment of premarital births: Evidence from the national
preferences for parents’ decision-making. survey data. Family Planning Perspectives, 24, 27–32, 48.
Brodzinsky, D., & Pinderhughes, E. (2002). Parenting and child
Despite these limitations, this study provides evidence
development in adoptive families. In M. Bornstein (Ed.),
that both the adoption context and the sexual orientation Handbook of parenting (pp. 279–311). Mahwah, NJ: Erlbaum.
context may have implications for how individuals think Bryman, A. (2006). Integrating quantitative and qualitative research:
about and imagine their child’s gender, and, in turn, makes How is it done? Qualitative Research, 6, 97–113.
Calhoun, C. (1997). Family outlaws. Philosophical Studies, 85, 181–193.
several unique contributions. First, many heterosexual and Callan, V. J., & Kee, P. K. (1981). Sons or daughters? Cross-cultural
same-gender couples do not appear to have a preference comparisons of the sex preferences of Australian, Greek, Italian,
regarding their future child’s gender, which may in part Malay, Chinese, and Indian parents in Australia and Malaysia.
reflect their position as individuals for whom parenthood is Population and Environment, 4, 98–108.
Carlson, H., & Steuer, J. (1985). Age, sex-role categorization, and
long-awaited, and for whom other characteristics, such as
psychological health in American homosexual and heterosexual
child race, may emerge as more important. Research is men and women. Journal of Social Psychology, 125, 203–211.
needed that articulates how preadoptive parents weigh the Charmaz, K. (2006). Constructing grounded theory: A practical guide
importance of child gender against other selectable child through qualitative analysis. London: Sage.
Coombs, L. C., & Fernandez, D. (1978). Husband–wife agreement
characteristics. Further, heterosexual men in particular
about reproductive goals. Demography, 15, 57–73.
showed a lack of gender preference, and, when they did D’Angelo, R. J., McGuire, J. M., Abbott, D. W., & Sheridan, S.
have a preference, tended to prefer girls. This suggests that (1998). Homophobia and perceptions of people with AIDS.
men’s preferences regarding their child’s gender may not be Journal of Applied Social Psychology, 28, 157–170.
Daniluk, J. C. (2001). “If we had to do it over again…”: Couples’
as strong when patrimony is not an issue (Gravois 2004), reflections on their experiences of infertility treatments. The
which could have implications for parent–child relation- Family Journal, 9, 122–133.
ships (perhaps adoptive fathers will be less likely to favor Gartrell, N., Hamilton, J., Banks, A., Mosbacher, D., Reed, N.,
their sons than biological fathers: Raley and Bianchi 2006). Sparks, C. H., et al. (1996). The national lesbian family study: 1.
Interviews with prospective mothers. American Journal of
Another important finding that emerges from these data
Orthopsychiatry, 66, 272–281.
is that lesbian and gay prospective adopters must negotiate Gates, G., & Ost, J. (2004). The lesbian and gay atlas. Washington,
a unique set of gender- and sexuality-related concerns when DC: Urban Institute Press.
they envision raising a child of the same- or opposite- Gates, G., Badgett, M. V. L., Macomber, J. E., & Chambers, K.
(2007). Adoption and foster care by gay and lesbian parents in
gender. They must consider societal heterosexism, the
the United States. Washington, DC: The Urban Institute.
availability of male/female role models (Goldberg and Goldberg, A. E., & Allen, K. R. (2007). Lesbian mothers’ ideas and
Allen 2007), and the implications of their same-gender intentions about male involvement across the transition to
parental status for gender socialization. Prior research parenthood. Journal of Marriage and Family, 69, 352–365.
Gravois, J. (January 16, 2004). Bringing up babes: Why do adoptive
suggests that once sexual minorities become parents, they
parents prefer girls? Slate. Retrieved October 21, 2008 from
continue to harbor concerns about socializing their child- [Link]
ren’s gender identities according to societal norms (Kane Greil, A., Leitko, T., & Porter, K. (1988). Infertility: His and hers.
2006) but may address these concerns by enlisting other Gender & Society, 2, 172–199.
Hammer, M., & McFerran, J. (1988). Preference for sex of child: A
agents of gender socialization (e.g., opposite-gender
research update. Individual Psychology, 44, 481–491.
friends) (Goldberg and Allen 2007). Of interest is how Hank, K., & Kohler, H. P. (2003). Sex preferences for children
lesbians and gay men who strongly desire same-gender revisited: New evidence from Germany. Population, 58, 131–143.
children cope when they are placed with an opposite-gender Herrmann-Green, L. K., & Gehring, T. M. (2007). The German
lesbian family study: Planning for parenthood via donor
child (e.g., to what extent does this motivate efforts to line
insemination. Journal of GLBT Family Studies, 3, 351–395.
up male/female friends?) To examine this and other Hesse-Biber, S. (1995). Unleashing Frankenstein’s monster: The use
questions, studies should follow sexual minorities from of computers in qualitative research. Studies in Qualitative
pre-adoption into parenthood to assess how attitudes about Methodology, 5, 25–41.
Hicks, S. (2006). Maternal men—Perverts and deviants? Making
gender preferences change over time, and the implications
sense of gay men as foster carers and adopters. Journal of GLBT
of these preferences for parent and family adjustment. Family Studies, 2, 93–114.
Hortacsu, N., Bastug, S. S., & Muhammetberdiev, O. B. (2001).
Desire for children in Turkmenistan and Azerbaijan: Son
preference and perceived instrumentality for value satisfaction.
References
Journal of Cross-Cultural Psychology, 32, 309–321.
Jones, J. (2008). Adoption experiences of women and men and demand
Arnold, F. (1997). Gender preferences for children. Demographic and for children to adopt by women 18–44 years of age in the United
Health Surveys Comparative Studies, no. 23. States, 2002. Vital and Health Statistics, Series 23, Number 27.
Sex Roles

Juni, S., Rahamim, E. L., & Brannon, R. (1985). Sex role Parry, D. C. (2005). Women’s experiences with infertility: The fluidity
development as a function of parent models and oedipal fixation. of conceptualizations of ‘family.’. Qualitative Sociology, 28,
Journal of Genetic Psychology, 146, 89–99. 275–291.
Kane, E. (2006). “No way my boys are going to be like that!”: Patton, M. (2002). Qualitative evaluation and research methods.
Parents’ responses to children’s gender nonconformity. Gender & Newbury Park, CA: Sage.
Society, 20, 149–176. Pollard, M. S., & Morgan, S. P. (2002). Emerging parental gender
Katzev, A. R., Warner, R. L., & Acock, A. C. (1994). Girls or boys? indifference? Sex composition of children and the third birth.
Relationship of child gender to marital instability. Journal of American Sociological Review, 67, 600–613.
Marriage and the Family, 56, 89–100. Raley, S., & Bianchi, S. (2006). Sons, daughters, and family
Kreider, R. (2003). Adopted children and stepchildren, 2000. Census processes: Does gender of children matter? Annual Review of
2000, special reports. Washington, DC: US Bureau of Census. Sociology, 32, 401–421.
Marleau, J. D., & Saucier, J. F. (2002). Preference for a first-born boy Steinbacher, R., & Gilroy, F. (1985). Preference for sex of child
in western societies. Journal of Biosocial Science, 34, 13–27. among primiparous women. The Journal of Psychology, 119,
Matthews, J., & Cramer, E. (2006). Envisaging the adoption process 541–547.
to strengthen gay and lesbian headed families: Recommendations Steinbacher, R., & Gilroy, F. (1990). Sex selection technology: A
for adoption professionals. Child Welfare, 85, 317–340. prediction of its use and effect. The Journal of Psychology, 124,
McDougall, J., DeWit, D. J., & Ebanks, G. E. (1999). Parental 283–288.
preferences for sex of children in Canada. Sex Roles, 41, 615– Strauss, A., & Corbin, J. (1998). Basics of qualitative research.
626. Thousand Oaks, CA: Sage.
Meyers, L. W., Gamst, G., & Guarino, A. J. (2006). Applied Swetkis, D., Gilroy, F., & Steinbacher, R. (2002). Firstborn preference
multivariate research: Design and interpretation. Thousand and attitudes toward using sex selection technology. The Journal
Oaks, CA: Sage. of Genetic Psychology, 163, 228–238.
Miles, M. B., & Huberman, A. M. (1994). Qualitative data analysis: Tasker, F., & Golombok, S. (1997). Growing up in a lesbian family:
An expanded sourcebook. Thousand Oaks, CA: Sage. Effects on child development. New York: Guilford.
Notman, M. K. (2006). Mothers and daughters as adults. Psychoan- Walker, M. K., & Conner, G. K. (1993). Fetal sex preference of
alytic Inquiry, 26, 137–153. second-trimester gravidas. Journal of Nurse Midwifery, 38, 110–
Okun, B. S. (1996). Sex preferences, family planning and fertility. An 113.
Israeli subpopulation in. transition. Journal of Marriage and the Warren, M. (1985). Gendercide: The implications of sex selection.
Family, 58, 469–475. Totowa, NJ: Rowman & Allanheld.
Oomman, N., & Ganatra, B. R. (2002). Sex selection: The systematic Williamson, N. (1976). Sex preferences, sex control, and the status of
elimination of girls. Reproductive Health Matters, 10, 184–187. women. Signs, 1, 847–862.