Content uploaded by Andrew L. Whitehead
Author content
All content in this area was uploaded by Andrew L. Whitehead on Feb 01, 2018
Content may be subject to copyright.
http://jfi.sagepub.com/
Journal of Family Issues
http://jfi.sagepub.com/content/early/2014/05/23/0192513X14536564
The online version of this article can be found at:
DOI: 10.1177/0192513X14536564
published online 25 May 2014Journal of Family Issues
Andrew L. Whitehead and Samuel L. Perry
Relative Effects of Religious Tradition, Practices, and Beliefs
Religion and Support for Adoption by Same-Sex Couples: The
Published by:
http://www.sagepublications.com
can be found at:Journal of Family IssuesAdditional services and information for
http://jfi.sagepub.com/cgi/alertsEmail Alerts:
http://jfi.sagepub.com/subscriptionsSubscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
What is This?
- May 25, 2014OnlineFirst Version of Record >>
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
Journal of Family Issues
1 –25
© The Author(s) 2014
Reprints and permissions:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/0192513X14536564
jfi.sagepub.com
Article
Religion and Support
for Adoption by
Same-Sex Couples:
The Relative Effects
of Religious Tradition,
Practices, and Beliefs
Andrew L. Whitehead1 and Samuel L. Perry2
Abstract
This article examines the relative impact of religious factors on Americans’
attitudes toward adoption by same-sex couples. Drawing upon national
survey data, we fit logistic regression models and compute standardized
logistic regression coefficients to estimate the relative net effects of religious
tradition, practices, and beliefs about the Bible on support for same-sex
adoption. Findings reveal that religious factors are among the strongest
predictors of opposition to same-sex adoption, but that religious tradition
has no significant effect on support for same-sex adoption once frequency of
religious practice and beliefs about the Bible are held constant. Americans who
more frequently engage in practices such as religious service attendance and
sacred text reading are less supportive of same-sex adoption, and compared
to biblical literalists, those who believe the Bible requires interpretation,
contains human error, or is a book of history/legends are all more likely to
support same-sex adoption. Findings suggest that religious affiliations matter
less for predicting same-sex adoption attitudes than how Americans practice
and hold their faith.
1Clemson University, Clemson, SC, USA
2University of Chicago, Chicago, IL, USA
Corresponding Author:
Andrew L. Whitehead, Department of Sociology and Anthropology, Clemson University, 132
Brackett Hall, Clemson, SC 29634, USA.
Email: andrewhitehead@gmail.com
536564JFIXXX10.1177/0192513X14536564Journal of Family IssuesWhitehead and Perry
research-article2014
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
2 Journal of Family Issues
Keywords
same-sex adoption, same-sex families, gay and lesbian adoption,
homosexuality, religion, religious tradition, beliefs, biblical literalism
Public opinion toward gay and lesbian romantic and family relationships
serves as an important barometer of future voting patterns and policy deci-
sions regarding the legal recognition of same-sex unions (Haider-Markel &
Joslyn, 2005, 2008; Olson, Cadge, & Harrison, 2006). Landmark legislation
or judicial decisions concerning family formation, particularly at the federal
level, are most often precipitated by shifting opinions among the American
public (Frank & Mceneaney, 1999; Haider-Markel & Joslyn, 2005; Rosenfeld,
2007). Consequently, investigations into what social and ideological factors
tend to predict support or opposition toward same-sex relationships—and
just as important, which factors matter more than others—help adumbrate the
axes on which future public debates regarding same-sex families will turn.
Over the last three decades, a voluminous body of research examining the
social correlates of attitudes toward same-sex romantic relationships finds
that religious factors such as conservative Protestantism (vis-á-vis other reli-
gious affiliations or no affiliation) and frequent church attendance are among
the most consistent predictors of opposition to same-sex sexuality, marriage,
and civil unions (Adamczyk & Pitt, 2009; Burdette, Ellison, & Hill, 2005;
Haider-Markel & Joslyn, 2008; Lubbers, Jaspers, & Ultee, 2009; Olson et al.,
2006; Schulte & Battle, 2004; Whitehead, 2010, 2014; Whitehead & Baker,
2012). Far less attention, by comparison, is given to the ways in which reli-
gious factors predict Americans’ attitudes toward same-sex adoption. This is
a significant oversight since, to a large degree, cultural, political, and legal
debates surrounding same-sex marriage are increasingly centered on the
(often adopted) children of such unions (Briggs, 2012; Clarke, 2001;
Rosenfeld, 2007), with religious groups at the forefront of the debate (Briggs,
2012; Wilson, 2004). Indeed, feminist historian Laura Briggs (2012) con-
tends, “[Gay adoption] has become [the defining issue in the gay and lesbian
freedom movement]—to a significant extent the question of gay marriage is
centrally about the raising of children, for its supporters among LGBT folk,
for its Christian Right opponents, and even for the judges writing decisions
about it” (p. 241).
Among the few studies that include religion measures in their analysis of
support for gay and lesbian adoption, little effort is made to understand exactly
how religious factors such as religious tradition, practices, and beliefs shape
public opinion toward such adoptions. There is also little attempt to under-
stand which religious factors matter most and the substantive significance of
those factors relative to other relevant correlates. Rather, within the vast
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
Whitehead and Perry 3
majority of studies that include religious predictors of support for same-sex
adoption, the measures are only included as control variables in multivariate
models, and thus, never receive sufficient analytical attention. Moreover, due
to data limitations, the religion measures that are used are typically sparse and
overaggregated, yielding limited information about the relationship between
religious life and Americans’ views toward same-sex adoptive families. As a
result, previous research on this topic potentially omits important nuance and
insight with respect to how various dimensions of religious life, in all its com-
plexity, shapes attitudes toward same-sex adoptions and which factors are of
central importance.
The current study fills this gap in the literature on attitudes toward same-
sex adoption in two important ways. First, we include a more comprehensive
battery of religion measures, including religious affiliations, various religious
practices, and beliefs about the Bible, in order to examine the net effects of
religious factors on support for same-sex adoption, while controlling for rel-
evant sociodemographic and ideological factors. And second, we compute
standardized logistic regression coefficients in order to examine which pre-
dictors of support for same-sex adoption matter more than others. This
research thus contributes to the literature on public opinion toward same-sex
families by providing greater insight into the substantive significance of reli-
gious factors for explaining Americans’ attitudes toward same-sex adoptive
families.
Religion and Attitudes Toward Same-Sex
Romantic and Family Relationships
A growing amount of research focuses on the social and ideological corre-
lates of support for same-sex romantic and family relationships. Drawing
from a variety of data sources, researchers consistently find that persons who
are more supportive of same-sex intimate relationships tend to be younger,
female, non-Southern, urban, politically liberal, more educated, more exposed
to diversity, less religious, and hold to the belief that homosexuality is innate,
not a choice (Adamczyk & Pitt, 2009; Burdette et al., 2005; Haider-Markel &
Joslyn, 2008; Olson et al., 2006; Perry, 2013a, 2013b; Petersen &
Donnenwerth, 1998; Whitehead, 2010, 2014).
Among these predictors of attitudes toward same-sex intimacy, religious fac-
tors—most often religious affiliation and church attendance, and less frequently,
views about the Bible—tend to be among the strongest predictors of attitudes
toward gay civil rights both in the United States and abroad. Conservative
Protestants tend to be the least supportive of same-sex romantic relationships
(Andersen & Fetner, 2008; Lubbers et al., 2009; Olson et al., 2006; Sherkat,
Powell-Williams, Maddox, & Mattias de Vries, 2011; Whitehead, 2013), and
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
4 Journal of Family Issues
others have found that Muslims also tend to express opposition to such relation-
ships (Adamczyk & Pitt, 2009). By contrast, mainline Protestants, Jews, and the
religiously unaffiliated tend to be more supportive of gay and lesbian relation-
ships (Besen & Zicklin, 2007; Perry, 2013a, 2013b). Beyond religious affilia-
tion, religious belief is associated with attitudes toward homosexuality as well.
Individuals who hold traditional religious beliefs (e.g., “The Bible should be
read literally, word for word”) are consistently less supportive of homosexuality
or gay civil rights (Burdette et al., 2005; Froese, Bader, & Smith, 2008; Hill,
Moulton, & Burdette, 2004; Sherkat, Mattias de Vries, & Creek, 2010; Sherkat
et al., 2011; Whitehead, 2010, 2014).
The effect of church attendance—generally used as a proxy for “religios-
ity,” religious devotion, or religious practice—also receives consistent sup-
port. A number of studies report a significant net effect of church attendance
on opposition to same-sex couples and families (Andersen & Fetner, 2008;
Lubbers et al., 2009; Olson et al., 2006; Sherkat et al., 2010; Whitehead,
2010, 2014). Accounting for this effect, it is likely through regular interaction
at worship services with other like-minded individuals that the plausibility
structures that undergird attitudes toward various issues, like homosexuality
or adoption by same-sex couples, are constructed and maintained (Berger,
1967; Petersen & Donnenwerth, 1998). Participation in religious institutions
can even influence gays and lesbians to support same-sex marriage at lower
levels (Baiocco, Argalia, & Laghi, 2014). One weakness of how the majority
of these studies operationalize religious behavior is their reliance on religious
service attendance only. For many individuals it may be that religious service
attendance is much less important while other types of religious behavior like
reading sacred scriptures or devotional prayer play a primary role (Ellison,
Wolfinger, & Ramos-Wada, 2013; Lubbers et al., 2009).
Some studies find that the effects of religious belief, behavior, and affiliation
on attitudes toward homosexuality moderate one another (Burdette et al., 2005;
Ellison & Musick, 1993; Schulte & Battle, 2004). These findings encourage the
inclusion of multiple measures of religion into each analysis to identify which
aspects of religion are more closely associated with attitudes toward “morality
issues” as broadly construed in public rhetoric, like homosexuality or same-sex
adoption (Ellison et al., 2013; Lubbers et al., 2009). Whitehead (2010) makes
the case that when investigating attitudes toward moral issues “it is not enough
to account for just one aspect of religiosity” (p. 74) as a measure of the effect of
religion. In the current analysis multiple measures of religion (belief, behavior,
and affiliation), in effect, triangulate the effect of religion as a whole on respon-
dents’ attitudes toward adoption by same-sex couples.
We situate the findings of prior research on religion and attitudes toward
homosexuality and our analysis of religion’s effect on attitudes toward same-sex
adoption, in the theoretical framework of structuration theory drawing on the
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
Whitehead and Perry 5
concept of “schemas.” Introduced by Giddens (1984) and elaborated by Sewell
(2005), structuration theory maintains that structures denote the tendency of pat-
terns of relations to be reproduced even when actors engaging in them are
unaware of the patterns or even when they do not desire their reproduction.
Structures shape actors’ practices but it is also these practices that comprise and
reproduce structures (Sewell, 2005). Structures, therefore, do not simply con-
strain behavior but also enable it; there is a duality of structure. Structures are
composed of both schemas (or “rules” according to Giddens) and resources.
Resources are “anything that can serve as a source of power in social interac-
tions” (Sewell, 2005, p. 132). Schemas pattern practices and are “generalizable
procedures applied in the enactment/reproduction of social life” (Sewell, 2005,
p. 131). They serve as templates or rules for acting in the social world. A key
quality of schemas is that they are transposable (Sewell, 2005). This transpos-
ability allows for their application to new situations and contexts.
The concept of schemas and their transposable nature helps explicate the
relationship between religious belief and attitudes toward same-sex adoption,
and by extension religious affiliation and same-sex adoption attitudes. Biblical
literalists and Evangelical Protestants are much more likely to oppose same-
sex adoption due to the schemas directing their actions toward homosexuality
generally. These schemas are predicated on their interpretation of specific bib-
lical passages that in their interpretation explicitly oppose any type of same-
sex intimacy.1 The transposable nature of schemas allows biblical literalists
and Evangelical Protestants to apply this particular schema to adoption by
same-sex couples. Biblical literalists and Evangelicals also valorize the idea of
the “traditional family,” another schema, which is assumed to be made up of a
man and wife with their biological children (Denton, 2004; Gallagher, 2003;
Gallagher & Smith, 1999; Hoffmann & Bartkowski, 2008). Same-sex couples
would obviously violate this ideal, leading Evangelicals and biblical literalists
to oppose their adoption of children. The concept of schemas has proven use-
ful in prior studies examining religion’s influence on various social attitudes
and outcomes (Denton, 2004; Edgell, Gerteis, & Hartmann, 2006; Hoffmann
& Bartkowski, 2008; Miller & Hoffmann, 1999; Sherkat, 1998).
Religion and Attitudes Toward Same-Sex Adoption
Despite the now enormous amount of research that focuses on the social cor-
relates of support for gay and lesbian romantic relationships, relatively little
research highlights support for same-sex adoption explicitly. This is likely for
at least two reasons. First, the majority of legal controversy surrounding
same-sex relationships is ostensibly about whether or not society should
legally recognize the romantic unions of gay and lesbian couples. In this
sense, same-sex adoption is viewed as a subsidiary to this broader issue. And
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
6 Journal of Family Issues
second, researchers neglect to recognize the ways in which same-sex adop-
tion is a different type of relationship than same-sex romantic unions, both in
terms of the legal status awarded to each2 and the inherent power differen-
tials.3 Consequently, studies of public opinion regarding gay and lesbian rela-
tionships often combine measures of support for same-sex adoption and
marriage together (e.g., Schulte & Battle, 2004; van den Akker, van der
Ploeg, & Scheeper, 2013). As a result, there are relatively few studies that
give overt attention to the various social and ideological factors that shape
Americans’ views toward same-sex adoption.
Among the few exceptions, similar to research regarding support for same-
sex romantic relationships, the majority of these studies utilize church atten-
dance and religious tradition as measures of religious life. However, the
influence of religious factors is not the focus of these studies, and researchers
never specify hypotheses regarding the effects of religion measures. MacLeod,
Crawford, and Zechmeister (1999) studied a group of heterosexual college stu-
dents and found that more frequent church attendance was weakly associated
with students believing that the adopted child of a gay male parent would ben-
efit from custody reassignment and would suffer from sexual orientation and
gender confusion if she or he remained with the gay parent. These findings
washed out in multivariate models, however, likely due to the sample size (N =
77). Using a larger sample of 364 college students, Lambert, Ventura, Hall, and
Cluse-Tolar (2006) found that students who attended church more often were
less likely to agree that gay/lesbian parents should be allowed to adopt. Church
attendance, however, was not associated with students’ attitudes toward gay/
lesbian foster parents, and students’ self-reported importance of religion was
not associated with their views regarding gay/lesbian adoption or foster parent-
ing. Drawing upon a random sample of 413 registered voters in Florida, Ryan,
Bedard, and Gertz (2004) found that Floridians who identified as Christian held
a significantly lower opinion of gay and lesbian adoptive parents than persons
who did not espouse Christian faith, net of other factors. Using the same data,
the authors (Ryan, Bedard, & Gertz, 2007) later reported that being a Christian
was highly predictive of a Floridian voter opposing a boy or girl being legally
adopted by a lesbian or gay parent. More recently, Averett, Strong-Blakeney,
Nalavany, and Ryan (2011) studied the attitudes of 776 adoptive parents toward
gay and lesbian adoption, finding membership in a Christian denomination was
associated with negative attitudes toward gay and lesbian adopters among mar-
ried adoptive fathers and mothers, but not adoptive singles, and greater religi-
osity (church attendance and prayer frequency) was associated with lower
levels of support for gay and lesbian adoption among all adoptive parents.
Although these studies provide some empirical foundation for generating
hypotheses regarding the relationship between religion and support for same-
sex adoption, their findings are limited in that they are based on either
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
Whitehead and Perry 7
convenience samples or subpopulations (college students, Florida voters,
adoptive parents); their samples are rather small; and their measures of reli-
gious tradition (Christian vs. other) or religious practice are somewhat nar-
row. Drawing upon data from a national probability sample, Besen and
Zicklin (2007) reported that persons who are “born again Christians” (broadly,
Evangelical) or Roman Catholic are less likely to support gay adoption. They
also found an interaction between gender and religiosity with more religious
males being even less likely to support gay adoption than those who are less
religious or female. Also using national-level data, Perry (2013a) included a
number of religion measures to predict support for same-sex sexuality, mar-
riage, and adoption. He found that, net of other factors, Catholics, mainline
Protestants, and persons of “other” religious faiths were all more likely than
Evangelical Protestants to support same-sex adoption. Black Protestants,
however, were no more likely than Evangelicals to support such adoptions.
He also found that persons who more frequently engaged in religious prac-
tices and strict biblical literalists were both relatively less likely to support
same-sex adoption. Perry’s measure of biblical literalism, however, was lim-
ited to a dichotomous measure (literalist vs. other) and thus was likely over-
aggregated. Moreover, his analyses did not include the strongest predictor of
support for same-sex relationships, attribution (homosexuality—innate or by
choice), which may have biased his findings.
Hypotheses
Building upon findings from previous research regarding the link between
religious factors, homosexuality, and support for same-sex adoption specifi-
cally, we formulate the following hypotheses:
Hypothesis 1: Evangelical Protestants will be less supportive of same-sex
adoption than all other religious or non-religious groups.
Hypothesis 2: Persons who more frequently engage in religious activities
(including religious service attendance, prayer, and sacred text reading)
will be less supportive of same-sex adoption.
Hypothesis 3: Persons who are strict biblical literalists will be less sup-
portive of same-sex adoption than persons who hold different beliefs
about the Bible.
An additional strength of our study is that we are able to compare the sub-
stantive net effects of each religion measure on support for same-sex adop-
tion. Based on previous research suggesting the primary importance of
religious practices and beliefs over and against religious affiliation (Burdette
et al., 2005; Ellison & Musick, 1993; Schulte & Battle, 2004), we predict
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
8 Journal of Family Issues
Hypothesis 4: The net effects of religious practices and theological con-
servatism will have a stronger influence on support for same-sex adoption
than religious affiliation.
Data
Data for these analyses are drawn from the third wave (2010) of the Baylor
Religion Survey (BRS). The 2010 BRS is a random, national sample of 1,714
U.S. citizens administered by the Gallup Organization. The survey utilized a
mixed-mode sampling design consisting of two phases. Both phases resulted
in a total of 3,500 individuals screened and 2,556 possible respondents. A
total of 1,714 questionnaires were returned resulting in a response rate of
49% (1,714/3,500) among all individuals screened, and a response rate of
67% (1,714/2,556) for those who agreed to receive a mailed survey. Previous
BRS data compare favorably to other national surveys (Bader, Mencken, &
Froese, 2007) and comparisons between the 2010 BRS and the 2010 General
Social Survey are available on request. The 2010 BRS is useful for the exist-
ing research question because it contains a breadth of religion measures as
well as questions pertaining to attitudes toward adoption by same-sex cou-
ples and various other sociodemographic controls. The combination of these
variables is not found in any other large, national random survey.
Dependent Variable
The dependent variable for this study focuses on respondents’ views toward
adoption by same-sex couples. The question utilized asks, “How do you feel
about the morality of the following? Adoption of children by homosexual
couples.”4 Possible responses include “Always wrong,” “Almost always
wrong,” “Only wrong sometimes,” and “Not wrong at all.” In order to predict
which respondents report unequivocal support for adoption by same-sex cou-
ples, this variable was recoded such that 1 = Not wrong at all, with all other
responses recoded as 0. Close to 40% of Americans declare clear support for
adoption by same-sex couples (see Table 1).
Independent Variables of Interest
In order to provide a more complete picture of religion’s relationship with
attitudes toward same-sex adoption, we utilize a collection of religion mea-
sures that offer greater breadth and depth than those found in past research.
Previous analyses measured religious affiliation using a “Christian” versus
“Other” dichotomy (Averett et al., 2011; Ryan, Bedard, & Gertz, 2004, 2007).
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
Whitehead and Perry 9
However, Steensland et al. (2000) demonstrate that there are seven distinct
religious traditions operating in the United States and individuals’ affiliation
with these traditions predicts their views on a number of outcomes. Their
RELTRAD typology categorizes all religious groups into seven distinct cat-
egories: Evangelical Protestant, Mainline Protestant, Black Protestant,5
Catholic, Jewish, other, and no affiliation.6 This analysis follows the
RELTRAD coding strategy and places all respondents into each of these cat-
egories. Because prior research consistently finds Evangelical Protestants the
Table 1. Descriptive and Bivariate Statistics (Multiple Imputation Data).
Variable Description Mean or % SD
Correlation with
same-sex adoption
Support same-sex adoption 1 = Not wrong at all 38.3% — —
Religious service attendance 1 = Never to 9 = Several
times a week
4.09 2.97 –0.37***
Frequency of reading sacred
scriptures
1 = Never to 9 = Several
times a week
4.35 3.08 –0.36***
Prayer frequency 1 = Never to 6 = Several
times a day
4.02 1.82 –0.31***
Biblical literalista1 = Biblical literalist 20.9% — –0.30***
Biblical interpretation 1 = Interpret Bible 34.5% — –0.15***
Bible contains errors 1 = Bible contains errors 12.5% — 0.13***
Bible ancient book 1 = Bible full of legends 23.8% — 0.32***
Bible undecided 1 = Don’t know 8.2% — 0.05*
Evangelical Protestanta1 = Evangelical Protestant 31.0% — –0.10***
Mainline Protestant 1 = Mainline Protestant 24.9% — –0.04
Black Protestant 1 = Black Protestant 2.4% — –0.01
Catholic 1 = Catholic 24.3% — –0.04
Jewish 1 = Jewish 1.6% — 0.04
Other 1 = Other 5.4% — 0.00
No religion 1 = No religion 10.3% — 0.25***
Age Age in years 55.9 16.22 –0.19***
Female 1 = Female 54.2% — 0.07**
Non-White 1 = Non-White 5.3% — 0.03
Married 1 = Married 62.8% — –0.09***
Education 1 = 8th grade or less, 7 =
postgraduate work/degree
4.62 1.62 0.17***
Income 1 = $10,000 or less to 7 =
$150,000 or more
4.26 1.61 0.11***
South 1 = South 35.2% — –0.10***
Urban 1 = Urban 17.0% —
Politically conservative 1 = Extremely liberal to 7 =
Extremely conservative
4.43 1.66 –0.59***
Homosexuality innate 1 = Homosexuality innate 57.6% — 0.44***
Source. Baylor Religion Survey (2010).
aContrast category.
*p < .05. **p < .01. ***p < .001.
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
10 Journal of Family Issues
most opposed to homosexuality and gay rights, we exclude them as the con-
trast category.
In order to measure religious behavior this analysis includes three differ-
ent measures: frequency of prayer/meditation, frequency of religious service
attendance, and frequency of reading sacred scriptures. The items for reading
sacred scriptures and religious service attendance range from 1 = Never to
9 = Several times a week. The item concerning prayer ranges from 1 = Never
to 6 = Several times a day. A benefit of including these three measures of
religious behavior is that they account for both public (religious service atten-
dance) and private (prayer/meditation, reading sacred scriptures) activities.
In addition to religious affiliation and religious behavior, these analyses
also incorporate respondents’ religious beliefs. One widely used measure of
religious belief is the degree to which individuals believe the Bible should be
read literally. A question in the 2010 BRS asked, “Which one statement comes
closest to your personal beliefs about the Bible?” Possible responses were
“The Bible means exactly what it says. It should be taken literally, word-for-
word, on all subjects”; “The Bible is perfectly true, but it should not be taken
literally, word-for-word. We must interpret its meaning”; “The Bible contains
some human error”; “The Bible is an ancient book of history and legends”;
and “I don’t know.” We recoded this measure to create a series of dichotomous
variables. Because biblical literalists tend to be most opposed to homosexual-
ity and gay rights, we exclude them as the contrast category in the full model.
We perform additional analyses (reported below) where we rotate each of the
other views of the Bible into the reference category position.
Control Variables
Building upon past studies analyzing attitudes toward adoption by same-sex
couples, this study includes a host of sociodemographic and ideological con-
trols. The socio-demographic measures include age (in years), gender (1 =
Female), race (1 = Non-White), marital status (1 = Married), region (1 =
South), size of city (1 = Urban), educational attainment (1 = 8th grade or less
to 7 = postgraduate work/degree), and income (1 = $10,000 or less to 7 =
$150,000 or more). Ideological controls include political ideology (1 =
Extremely liberal to 7 = Extremely conservative), and whether the respondent
believes homosexuality is innate (1 = Innate).
Method
Table 1 displays descriptive statistics for all of the measures utilized in this
analysis. It also includes the bivariate associations between the dependent
variable and each of the independent variables. Table 2 displays the various
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
11
Table 2. Logistic Regression Analysis of Support for Same-Sex Adoption (MI Data).
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Variables βOR βOR βOR βOR βOR βOR
Age –0.17*** 0.98 –0.15*** 0.98 –0.11** 0.99 –0.13*** 0.99 –0.14** 0.98 –0.18*** 0.98
Female 0.11*** 1.47 0.13*** 1.58 0.18*** 1.93 0.18*** 1.92 0.12** 1.53 0.10* 1.42
Non-White 0.01 — 0.02 — 0.04 — 0.05 — –0.01 — −0.00 —
Married –0.19*** 0.49 –0.18*** 0.50 –0.16*** 0.55 –0.16*** 0.56 –0.13** 0.62 –0.14** 0.60
Education 0.14*** 1.17 0.15*** 1.18 0.20*** 1.25 0.15*** 1.19 0.12** 1.14 0.10* 1.12
Income 0.15*** 1.18 0.13** 1.15 0.09* 1.11 0.07 — 0.13* 1.16 0.12* 1.15
South –0.12*** 0.64 –0.09** 0.71 –0.06 — −0.05 — –0.07 — –0.08* 0.73
Urban 0.08** 1.48 0.07* 1.43 0.07* 1.40 0.08* 1.44 0.06 — 0.06 —
Religious affiliationa
Mainline Protestant 0.02 — 0.01 — –0.02 — –0.01 — −0.01 —
Black Protestant –0.01 — 0.00 — 0.00 — 0.01 — 0.03 —
Catholic 0.02 — –0.01 — –0.02 — –0.02 — –0.02 —
Jewish 0.04 — 0.02 — 0.02 — 0.00 — –0.00 —
Other 0.02 — 0.02 — –0.01 — –0.01 — 0.00 —
No religion 0.27*** 4.92 0.09* 1.75 0.06 — 0.04 — 0.05 —
Religious behavior
Attendance –0.22*** 0.87 –0.15** 0.91 –0.14** 0.92 –0.11* 0.93
Reading sacred scriptures –0.26*** 0.86 –0.19*** 0.90 –0.16** 0.91 –0.12* 1.03
Prayer –0.05 — −0.01 — 0.04 — 0.03 —
(continued)
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
12
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Variables βOR βOR βOR βOR βOR βOR
Bible beliefsb
Biblical interpretation 0.21*** 2.26 0.20*** 2.15 0.14* 1.71
Bible contains errors 0.32*** 5.67 0.25*** 4.01 0.20*** 3.03
Bible ancient book 0.39*** 5.21 0.27*** 3.18 0.20** 2.38
Bible undecided 0.18*** 3.34 0.13** 2.34 0.09* 1.86
Ideological beliefs
Politically conservative –0.62*** 0.51 –0.58*** 0.53
Homosexuality innate 0.39*** 4.18
Intercept –0.569 –0.937** 0.070 –1.008* 1.919*** 1.280*
N1,714 1,714 1,714 1,714 1,714 1,714
PRE 0.073 0.107 0.184 0.217 0.331 0.373
Note. MI = multiple imputation; OR = odds ratio; β = standardized coefficient; PRE = proportional reduction in error.
aEvangelical Protestant is contrast category.
bBiblical Literalist is contrast category.
*p < .05. **p < .01. ***p < .001 (two-tailed significance tests).
Table 2. (continued)
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
Whitehead and Perry 13
multivariate models. Due to the dichotomous coding of the dependent vari-
able, logistic regression is utilized. The first model contains only the socio-
demographic factors. Model 2 includes the RELTRAD typology. Model 3
adds the religious practice measures while Model 4 includes the biblical lit-
eralism series of dichotomous variables. Models 5 and 6 include political
conservatism and belief that homosexuality is innate, respectively.7 All of the
variables included in the final models except religious tradition and region
had missing information. Multiple imputation was used to correct for missing
data (Rubin, 1987).8 To allow for the interpretation of substantive signifi-
cance alongside statistical significance, standardized coefficients for the
logistic regression models were estimated
[(/)
]
*
Bbss
yx yx
xy
= (Pampel,
2000).9 Table 3 displays the results of rotating the excluded category for the
biblical literalism series of dichotomous variables. Doing so allows for an
examination of how each distinct response differs from every other response.
Table 3 does not display the standardized coefficients for the other religion
and control variables in order to focus specifically on the relationships
between the discrete categories on the biblical literalism measure. There are
no substantive or statistical differences for the control and other religion vari-
ables between Model 6 displayed in Table 2 and Models 5 through 8 dis-
played in Table 3.10
Results
In Table 1 we find that 38.3% of Americans fully support adoption by same-
sex couples. We also find strong correlations between the various religion
Table 3. Odds Ratios Comparing Differences of Support for Adoption by Same-
Sex Couples between Biblical Literalism Responses (MI Data).
Model 4
(identical to Table 2) Model 5 Model 6 Model 7 Model 8
Biblical Literalist — 0.58* 0.34*** 0.43** NS
Interpret Bible 2.26*** — 0.59** NS NS
Bible contains errors 5.67*** 1.78** — NS NS
Bible ancient book 5.21*** NS NS — NS
Bible undecided 3.34*** NS NS NS —
Note. MI = multiple imputation; NS = not significant. Results show Biblical Literalism series of
dummy variables only; full model was included in each logistic regression. When rotating the
excluded category for this series of dummy variables there were no substantive or statistical
changes in the other variables included in the full model. Therefore, we only display the
Biblical Literalism response categories. Results are available on request.
*p < .05. **p < .01. ***p < .001.
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
14 Journal of Family Issues
measures and support for adoption by same-sex couples. Individuals who
practice their religion regularly, read the Bible literally, believe the Bible is
perfectly true but must be interpreted, or are Evangelical Protestants are
much more likely to oppose adoption by same-sex couples. Those who
believe the Bible contains human error, believe the Bible is an ancient book
full of history and legend, do not know what they believe about the Bible, or
are unaffiliated with a religious tradition are more supportive of adoption by
same-sex couples.
Table 2 presents the multivariate analyses. Model 1 contains the sociode-
mographic controls. Consistent with previous research, older adults, married
adults, and those from the South are all less supportive of adoption by same-
sex couples. Being female, living in an urban area, and increasing levels of
education and income are all predictive of greater support for adoption by
same-sex couples.
Model 2 includes the RELTRAD typology of religious affiliation. Net of
all the effects of the relevant control variables, only the religiously unaffili-
ated are significantly different from Evangelical Protestants. The unaffiliated
are much more likely to support adoption by same-sex couples compared to
Evangelicals, with odds 2.26 times greater. The net effects of the control vari-
ables remain largely unchanged from Model 1 to Model 2. Model 3 incorpo-
rates three measures of religious practice. Individuals who practice their
religion more frequently by reading sacred scriptures or attending religious
services are much more likely to oppose same-sex adoption. Frequency of
prayer, however, is not significantly associated with attitudes toward same-
sex adoption. When religious behavior is included in the model the difference
between the unaffiliated and Evangelicals is largely attenuated, but not com-
pletely. Controlling for public and private religious behavior does account for
the statistical significance of region of the country in prior models. Identical
to Models 1 and 2, age, gender, marital status, education level, income, and
size of place—all maintain their previous associations.
Model 4 includes the biblical literalism measures. As predicted, individu-
als who read the Bible literally are much more likely to oppose adoption by
same-sex couples compared with all the other groups. Also as expected, even
when controlling for views of the Bible, attendance at religious services and
frequency of Bible reading are still negatively and significantly related to
views toward same-sex adoption. With religious beliefs and practices
accounted for in the model, there are now no significant differences between
Evangelicals and every other religious tradition on support for same-sex
adoption. Model 5 introduces a measure of political conservatism which is
significantly and negatively associated with support for adoption by same-
sex couples.
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
Whitehead and Perry 15
Model 6 includes a measure of whether respondents’ believe homosexual-
ity is innate (not a choice) and represents the full model. Believing homo-
sexuality is innate is a robust and significant predictor of support for same-sex
adoption. Regarding our variables of interest, in the full model, the odds of
supporting same-sex adoption for those who interpret the Bible are 1.71 times
greater than biblical literalists. Similarly, the odds of supporting same-sex
adoption for those who believe the Bible contains errors or that it is a book
full of history and legends are 3 times and 2.38 times greater, respectively.
The odds of supporting adoption by same-sex couples for those who do not
know what they believe about the Bible are 1.86 times greater than biblical
literalists. Frequent attendance at religious services and frequent reading of
sacred scriptures are each negatively and significantly associated with sup-
port of same-sex adoption. Frequency of prayer, however, is not. There con-
tinue to be no significant differences between conservative Protestants and
every other religious tradition. Age, gender, marital status, education, income,
and political ideology all maintain their previous associations.
Substantively, political conservatism (β = −0.58) and believing homosex-
uality is innate (β = 0.39) are the two strongest predictors in the full model.
The standardized effect of the difference between biblical literalists and those
who believe the Bible contains errors is the third strongest predictor (β =
0.20), along with the differences between biblical literalists and those who
believe the Bible is an ancient book (β = 0.20). Age is the fourth strongest
predictor (β = −0.18) followed by the difference between biblical literalists
and those who interpret the Bible (β = 0.14) and marital status (β = −0.14).
Frequency of attending religious services (β = −0.12) and reading sacred
scriptures (β = −0.11) are the next strongest predictors. These findings
strongly suggest that, when predicting attitudes toward same-sex adoption,
religious factors are quite important. Yet, as predicted in our fourth hypothe-
sis, knowing where an individual worships (e.g., an Evangelical, Catholic, or
Mainline church, or no church) is not as important as knowing how often they
practice both publically and privately, or the types of beliefs they hold.
Table 3 provides a more comprehensive examination of how the different
views of the Bible relate to support for adoption by same-sex couples. The
first column, Model 4, shows biblical literalists as the contrast category
which is identical to Model 4 from Table 2 and whose results were discussed
above. Individuals who believe the Bible is perfectly true but it requires inter-
pretation are the contrast category in Model 5. The odds of supporting adop-
tion for biblical literalists decrease 72% compared with interpreters.11 The
odds for those who believe the Bible contains errors are 78% greater com-
pared with interpreters. There are no significant differences between inter-
preters and those who believe the Bible is an ancient book, or those who do
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
16 Journal of Family Issues
not know what they believe about the Bible. In Model 6, those who believe
the Bible contains errors are the contrast category. The odds of biblical liter-
alists and interpreters supporting adoption by same-sex couples are 2.9 and
1.7 times less than those who believe the Bible contains errors. There are no
significant differences between those who view the Bible as an ancient book,
are undecided about their view of the Bible, and those who believe the Bible
contains errors. In Model 7 we find that the only significant difference for
those who believe the Bible is an ancient book is with biblical literalists,
whose odds are 2.3 times smaller. There are no significant differences for
those who believe the Bible is an ancient book and interpreters, those who
believe that the Bible contains errors, and the undecided. Finally, Model 8
focuses on those who are unsure of their view of the Bible. Regarding support
for adoption by same-sex couples, there are no significant differences
between this group and every other view of the Bible.
Discussion and Conclusions
The results from this analysis demonstrate religion’s strong and multifaceted
association with attitudes toward adoption by same-sex couples. While previ-
ous studies generally relied on overaggregated measures of religious affilia-
tion or used a single measure of religious behavior in worship service
attendance, only by using a more complete array of religion measures does
the true nature of the relationship between religion and same-sex adoption
attitudes stand out. First, individuals who practice their religion more fre-
quently by reading sacred scriptures or attending worship services are much
more likely to oppose same-sex adoption. However, one private religious
behavior, prayer, is not significantly associated with attitudes toward adop-
tion by same-sex couples. While it is through both public and private reli-
gious practice that individuals’ plausibility structures are constructed and
preserved, and their attitudes toward adoption by same-sex couples formed
and sustained (Berger, 1967; Petersen & Donnenwerth, 1998), not all reli-
gious behavior portends opposition toward adoption by same-sex couples.
Accounting for this difference, religious service attendance and sacred text
reading both expose adherents to traditional religious teachings on family
relationships, and consequently, traditionalist schemas opposing homosexu-
ality. In contrast, prayer is often done in private, and does not imply exposure
to antihomosexuality schemas. The nonsignificant effect of prayer frequency
on support for same-sex adoption suggests that it is not religious devotion per
se that inclines persons to oppose same-sex family forms, but inculcation
with religiocultural schemas that denounce such relationships. Biblical liter-
alists, compared to those who believe the Bible should be interpreted, that it
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
Whitehead and Perry 17
contains errors, that it is an ancient book of legends, or who are unsure of
what they believe about the Bible, are also much more likely to oppose adop-
tion by same-sex couples. Biblical literalists appear to be much more likely
to apply passages concerning homosexuality to the present day (Burdette
et al., 2005; Froese et al., 2008; Hill et al., 2004; Sherkat et al., 2010; Sherkat
et al., 2011; Whitehead, 2010). Their schemas concerning homosexuality and
the traditional family transpose onto the issue of adoption by same-sex
couples.
This analysis also makes clear that once religious practice and religious
belief are accounted for, the differences between Evangelical Protestants and
the unaffiliated disappear. When considering attitudes toward adoption by
same-sex couples, religious behavior and religious belief are most important
while the differences between religious affiliations fade to the background.
Our models suggest that the differences observed between Evangelical
Protestants and the unaffiliated regarding support for same-sex families are
due to their (obvious) underlying differences in religious behavior and Bible
beliefs. The overall lack of effects of religious tradition could also be due to
nominal affiliation patterns by those who claim to affiliate but are relatively
inactive in the tradition. While religious traditions exhibit no significant dif-
ferences in the attitudes individuals hold toward same-sex adoption in multi-
variate models, those influences are possibly channeled through the schemas,
religious beliefs, and behaviors the religious traditions tend to encourage.
Future analyses could utilize structural equation modeling to test if religious
tradition has an indirect effect on attitudes toward same-sex adoption.
An even finer-grained analysis of Bible beliefs reveals more interesting
findings. First, there is a definitive separation between biblical literalists and
all other possible views of the Bible concerning views toward same-sex
adoption. However, differences also exist between the other views of the
Bible. Those who interpret the Bible are more likely to support same-sex
adoption compared to biblical literalists, even though they still hold the Bible
in very high regard, but are less likely to support adoption by same-sex cou-
ples compared to those who believe the Bible contains errors. Interpreters
occupy a middle ground on this issue between biblical literalists and less
traditional views of the Bible. These groups appear to approach this issue
through different sets of schemas. This is an important finding when consid-
ering that more than a third of Americans (34.5%) support a view of the Bible
that encourages interpretation, by far the largest group (see Table 1).
Substantively, we find that in addition to Bible beliefs and behavior being the
most important religious predictors of attitudes toward same-sex adoption,
they are also among the most important predictors in the entire model. Future
research examining attitudes toward adoption by same-sex couples must
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
18 Journal of Family Issues
account for the multifaceted nature of religion. Only accounting for religious
affiliation, or religious service attendance, may overlook the complex influ-
ence of religion as a whole.
A number of additional findings warrant discussion. First, political views
are consistently the strongest predictor of attitudes toward same-sex adop-
tion. This finding is in line with prior research that highlights the importance
of political ideology on attitudes toward homosexuality generally, and same-
sex unions specifically (Hill et al., 2004; Sherkat et al., 2011). The beliefs
individuals have about the cause of homosexuality are also strongly associ-
ated with views of same-sex adoption. Various studies point out that believ-
ing homosexuality is a choice or an innate characteristic informs individuals’
views toward the rights that should be afforded gays and lesbians (Haider-
Markel & Joslyn, 2008; Whitehead, 2010). However, attribution beliefs in
past research on attitudes toward homosexuality and same-sex unions are
consistently the strongest variables in the model (Haider-Markel & Joslyn,
2008; Whitehead, 2010, 2014). When considering attitudes toward same-sex
adoptions, however, attribution beliefs are second to political views. This
suggests that an individual’s political views influence her attitudes toward
same-sex adoption more so than whether they believe sexuality is an innate
characteristic or not.
The consistent negative association between marital status and support of
adoption by same-sex couples stands out as well. Married individuals are
much less likely than the unmarried to endorse the building of families by
same-sex couples. Individuals with stakes in conformity and a vested interest
in maintaining the status quo, which past theorists assume married individu-
als have, are less likely to favor any changes that might provide an alternative
to their current situation (Burdette et al., 2005; Hill et al., 2004; McVeigh &
Diaz, 2009; Powell, Bolzendahl, Geist, & Steelman, 2010). Marriage could
also operate as a type of plausibility structure where more traditional beliefs
toward family life become solidified (Berger, 1967). The effects of age, gen-
der, education, income, and region follow expectations drawn from past lit-
erature on attitudes toward same-sex sexuality, legal unions, and adoption
(Andersen & Fetner, 2008; Averett et al., 2011; Ryan et al., 2007; Sherkat
et al., 2010; Sherkat et al., 2011; Whitehead, 2010).
Several limitations of the analysis must be mentioned. First, due to the
cross-sectional nature of the 2010 BRS, causal direction cannot be determined.
It may be that individuals’ views toward same-sex adoption influence their
interpretation of the Bible, or how often they attend religious services, or pray.
Nevertheless, past research supports the theoretical progression we identify in
this study that for most individuals, religious beliefs and practices occupy a
central position in their lives that serves to influence their views on a number
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
Whitehead and Perry 19
of issues. Second, the dependent variable does not distinguish between atti-
tudes toward “gay couples” and “lesbian couples.” Despite this weakness,
public rhetoric surrounding gays and lesbians tends to focus on homosexuality
as a whole. Social movements both supporting and opposing gay rights tend to
either seek equality for both gays and lesbians or oppose them equally (Herek
2002). In this way, the dependent variable does measure a portion of this larger
set of attitudes held by many (see footnote 4). Finally, future research could
use structural equation models to tease out the relationships among the reli-
gious and political factors, which the stepwise multivariate models in this
analysis are largely unable to do (see Sherkat et al. 2010, Sherkat et al. 2011).
Despite these limitations, this study advances the literature on attitudes
toward adoption by same-sex couples in two important ways. First, it pro-
vides a comprehensive look at the multifaceted effects of religious belief,
behavior, and affiliation. How individuals view the Bible and how often they
practice their religion both publicly and privately can tell researchers a great
deal about how they will view same-sex adoption. Where a person worships
matters much less. Second, this analysis shows that the effects of religious
belief and behavior are among the strongest influences on a person’s beliefs,
net of all other possible effects. While political orientation and an individu-
al’s view of the cause of homosexuality exhibit the strongest associations
with attitudes toward same-sex adoption, taken together, the combined effects
of religion on an individual’s views of adoption by same-sex couples is of
central importance. As the legal recognition of various family structures are
debated, religion will continue to influence and inform the attitudes of the
public and the policy makers. Understanding the multifaceted effect of reli-
gion on American adults is an essential element in comprehending their views
toward adoption by same-sex couples.
Beyond the research implications of our study, the results presented here
clarify for gay and lesbian couples seeking to adopt the sources of religious
opposition. Our findings suggest that religious men and women tend to
oppose the adoption of children by same-sex couples not because of religious
affiliations or even personal religious devotion, but more likely because of
greater inculcation with religiocultural schemas (found in particular congre-
gations and sacred scriptures) that oppose same-sex relationships. Thus, for
the gay community, attenuating this opposition would not necessarily require
the denouncement of religious adherence per se, but rather an engagement
with faith communities, providing countervailing narratives and encouraging
such communities to embrace their deeper teachings about tolerance, equal-
ity, and love. Research on the importance of interpersonal contact with gays
and lesbians for increasing support for gay rights suggests that engagement
with faith communities may have positive results (Rosenfeld, 2007).
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
20 Journal of Family Issues
Acknowledgments
The authors would like to thank Constance Shehan and the three anonymous review-
ers for their constructive comments, and Kelly Whitehead and Jill Perry for their
ongoing support.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research,
authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publi-
cation of this article.
Notes
1. For example: Leviticus 18:22; Leviticus 20:13; Romans 1:26-27; 1 Corinthians
6:9-11.
2. Same-sex romantic unions have been the focus of more explicit legal sanction
than same-sex adoption. At the time of this writing, full marriage between same-
sex partners is legal in 13 states and the District of Columbia, with civil unions
being granted in another five states. A full 35 states prohibit full same-sex mar-
riage by statute or in their constitutions, with 17 of these banning both same-sex
marriage and civil unions. By contrast, only two states (Mississippi and Utah)
have laws that explicitly prohibit same-sex couples from adopting. Twenty-one
states and the District of Columbia allow for same-sex couples to jointly peti-
tion to adopt statewide while 18 states and the District of Columbia allow for
second-parent adoption. A handful of other states, however, create obstacles for
same-sex parents interested in adopting that hinge on the fact that they cannot be
legally married in these states (Human Rights Campaign, 2013).
3. For example, same-sex romantic relationships occur between social and legal
peers who, under normal circumstances, both voluntarily agree to enter into a
relationship. Same-sex adoption, by contrast, involves a different power dynamic
between two adults and a minor child who (depending on his or her age at adop-
tion) had little to no choice on her or his adoptive parents. Relatedly, because of
the power dynamic, the direction of influence in same-sex adoption is more uni-
lateral (parents → child(ren)) compared with influence in a same-sex romantic
relationship (partner ↔ partner). Lastly, while there is typically no developmen-
tal or financial need for an adult to be in a romantic relationship, minor children
are vulnerable and in need of care and support for healthy development, and thus,
require being placed in the care of legal guardians of some form.
4. A weakness of this measure is that it asks for respondent’s attitudes toward
“homosexual couples” instead of either “gay couples” or “lesbian couples.”
Prior research shows that when asking about attitudes toward “homosexuality”
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
Whitehead and Perry 21
the stronger negative reaction that gay men receive overwhelms the less nega-
tive reaction that lesbians generally receive, leading to more negative reactions
to homosexuality in general (Herek, 2002; Kite & Whitley, 1996). While spe-
cifically asking about “gay couples” or “lesbian couples” is ideal, Herek (2002)
points out that public rhetoric commonly utilizes “homosexuality” with no
distinction between gay men and lesbians. Likewise, most groups opposed to
homosexuality do not distinguish between gay men and lesbians. The lack of
distinction between gay men and lesbians is true for the gay movement as well;
they desire equality for both. Herek (2002, p. 42) points out that in the pub-
lic discourse gay men and lesbians share a “common characteristic that makes
them members of a distinct quasi-ethnic group with its own culture and political
concerns.” This means that while the question used cannot distinguish between
attitudes toward gay couples and lesbian couples, it does measure an aspect of
the broader attitudes maintained by a majority of the public.
5. The RELTRAD typology recommended by Steensland et al. (2000) categorizes
Black Protestants using historically black denominations. However, doing so
misses a great deal of variation within Black Protestantism, as Sherkat et al.
(2010) show. While we follow a wealth of past research that uses the RELTRAD
typology, readers should interpret these results with caution.
6. For details about religious tradition coding in the Baylor Religion Survey, see
Dougherty, Johnson, and Polson (2007).
7. Due to the number of religion variables in the full model we performed collin-
earity diagnostics. Despite the correlation between the religion measures, mul-
ticollinearity does not appear to be adversely affecting the models. Also, the
maximum likelihood estimation procedure used in the logistic regression models
in these analyses is less susceptible to issues arising from multicollinearity com-
pared to standard ordinary least squares techniques.
8. The multiple imputation (MI) procedure generates five imputations using mul-
tiple Markov Chains based on all variables included in each model, resulting in
an overall N of 8,570 (1,714 × 5). All results use the MI data set. The correla-
tions reported in Table 1 and all of the results reported in Tables 2 and 3 are from
the MIANALYZE procedure in SAS. This procedure combines all of the results
from each of the five imputations resulting in overall estimates, standard errors,
and significance levels. The standardized coefficients and odds ratios for each
model were calculated using these overall estimates. The proportional reduction
in error (PRE) reported in Table 2 for each model are the average of the PRE for
each individual iteration.
9. This analysis follows Pampel’s (2000) assumption that the standard deviation of
logit(y) = 1.8138.
10. This analysis also tested for a host of moderating effects and interactions between
the religion measures themselves, and between the religion measures and the
socio-demographic control variables. However, no significant interactions were
present. Further detail on which interactions were tested is available from the
authors.
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
22 Journal of Family Issues
11. In order to calculate the percent change in odds for the measures with negative
odds ratios, 1 is divided by each ratio. This corrects for negative odds ratios
being bounded between 0 and 1. Thus, biblical literalists = 1/0.58 = 1.72.
References
Adamczyk, A., & Pitt, C. (2009). Shaping attitudes about homosexuality: The role of
religion and cultural context. Social Science Research, 38, 338-351.
Andersen, R., & Fetner, T. (2008). Cohort differences in tolerance of homosexuality:
Attitudinal change in Canada and the United States, 1981-2000. Public Opinion
Quarterly, 72, 311-330.
Averett, P., Strong-Blakeney, A., Nalavany, B., & Ryan, S. (2011). Adoptive parents’
attitudes towards gay and lesbian adoption. Journal of GLBT Family Studies, 7,
30-48.
Bader, C., Mencken, F., & Froese, P. (2007). American piety 2005: Content, methods
and selected results from the Baylor Religion Survey. Journal for the Scientific
Study of Religion, 46, 447-463.
Baiocco, R., Argalia, M., & Laghi, F. (2014). The desire to marry and attitudes toward
same-sex family legalization in a sample of Italian lesbians and gay men. Journal
of Family Issues, 35, 181-200. doi:10.1177/01925X12464872
Berger, P. (1967). The sacred canopy: Elements of a sociological theory of religion.
New York, NY: Anchor Books.
Besen, Y., & Zicklin, G. (2007). Young men, religion, and attitudes towards homo-
sexuality. Journal of Men, Masculinities, and Spirituality, 1, 250-266.
Briggs, L. (2012). Somebody’s children: The politics of transnational and transracial
adoption. Durham, NC: Duke University Press.
Burdette, A., Ellison, C., & Hill, T. (2005). Conservative Protestantism and tolerance
toward homosexuals: An examination of potential mechanisms. Sociological
Inquiry, 75, 177-196.
Clarke, V. (2001). What about the children? Arguments against lesbian and gay par-
enting. Women’s Studies International Forum, 24, 555-570.
Denton, M. (2004). Gender and marital decision making: Negotiating religious ideol-
ogy and practice. Social Forces, 82, 1151-1180.
Dougherty, K., Johnson, B., & Polson, E. (2007). Recovering the lost: Remeasuring
U.S. religious affiliation. Journal for the Scientific Study of Religion, 46,
483-499.
Edgell, P., Gerteis, J., & Hartmann, D. (2006). Atheists as “other”: Moral boundaries
and cultural membership in American society. American Sociological Review,
71, 211-234.
Ellison, C., & Musick, M. (1993). Southern intolerance: A fundamentalist effect?
Social Forces, 72, 379-398.
Ellison, C., Wolfinger, N., & Ramos-Wada, A. (2013). Attitudes toward marriage,
divorce, cohabitation, and casual sex among working-age Latinos: Does religion
matter? Journal of Family Issues 34, 295-322.
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
Whitehead and Perry 23
Frank, D., & Mceneaney, E. (1999). The individualization of society and the liberal-
ization of state policies on same-sex sexual relations, 1984-1995. Social Forces,
77, 911-943.
Froese, P., Bader, C., & Smith, B. (2008). Political tolerance and god’s wrath in the
United States. Sociology of Religion, 69, 29-44.
Gallagher, S. (2003). Evangelical identity and gendered family life. New Brunswick,
NJ: Rutgers University Press.
Gallagher, S., & Smith, C. (1999). Symbolic traditionalism and pragmatic egalitari-
anism: Contemporary evangelicals, family, and gender. Gender & Society, 13,
211-233.
Giddens, A. (1984). The constitution of society: Outline of the theory of structuration.
Cambridge, England: Polity Press.
Haider-Markel, D., & Joslyn, M. (2005). Attributions and the regulation of marriage:
Considering the parallels between race and homosexuality. PS: Political Science
& Politics, 38, 233-240.
Haider-Markel, D., & Joslyn, M. (2008). Beliefs about the origins of homosexuality
and support for gay rights: An empirical test of attribution theory. Public Opinion
Quarterly, 72, 291-310.
Herek, G. (2002). Gender gaps in public opinion about lesbians and gay men. Public
Opinion Quarterly, 66, 40-66.
Hill, T., Moulton, B., & Burdette, A. (2004). Conservative Protestantism and atti-
tudes toward homosexuality: Does political orientation mediate this relationship?
Sociological Focus, 37, 59-70.
Hoffmann, J., & Bartkowski, J. (2008). Gender, religious tradition and biblical literal-
ism. Social Forces, 86, 1245-1272.
Human Rights Campaign. (2013). Second parent adoption. Retrieved from http://
www.hrc.org/resources/entry/second-parent-adoption
Kite, M., & Whitley, B. (1996). Sex differences in attitudes toward homosexual
persons, behaviors, and civil rights: A meta-analysis. Personality and Social
Psychology Bulletin, 22, 336-353.
Lambert, E. G., Ventura, L. A., Hall, D. E., & Cluse-Tolar, T. (2006). College stu-
dents’ views on gay and lesbian issues. Journal of Homosexuality, 50, 1-30.
Lubbers, M., Jaspers, E., & Ultee, W. (2009). Primary and secondary socialization
impacts on support for same-sex marriage after legalization in the Netherlands.
Journal of Family Issues, 30, 1714-1745.
McLeod, A. C., Crawford, I., & Zechmeister, J. (1999). Heterosexual undergraduates’
attitudes toward gay fathers and their children. Journal of Psychology & Human
Sexuality, 11, 43-62.
McVeigh, R., & Diaz, M. D. (2009). Voting to ban same-sex marriage: Interests, val-
ues, and communities. American Sociological Review, 74, 891-915.
Miller, A., & Hoffmann, J. P. (1999). The growing divisiveness: Culture wars or a war
of words? Social Forces, 78, 721-745.
Olson, L. R., Cadge, C., & Harrison, J. T. (2006). Religion and public opinion about
same-sex marriage. Social Science Quarterly, 87, 340-360.
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
24 Journal of Family Issues
Pampel, F. C. (2000). Logistic regression: A primer. Thousand Oaks, CA: Sage.
Perry, S. L. (2013a). Multiracial church attendance and support for same-sex romantic
and family relationships. Sociological Inquiry, 83, 259-285.
Perry, S. L. (2013b). Racial diversity, religion, and morality: Examining the moral
views of multiracial church attendees. Review of Religious Research, 55,
355-376.
Petersen, L., & Donnenwerth, G. (1998). Religion and declining support for tradi-
tional beliefs about gender roles and homosexual rights. Sociology of Religion,
59, 353-371.
Powell, B., Bolzendahl, C., Geist, C., & Steelman, L. C. (2010). Counted out: Same-
sex relations and American’s definitions of family. New York, NY: Russell Sage
Foundation.
Rosenfeld, M. (2007). The age of independence: Interracial unions, same sex unions,
and the changing American family. Cambridge, MA: Harvard University Press.
Rubin, D. B. (1987). Multiple imputation for nonresponse in surveys. New York, NY:
Wiley.
Ryan, S., Bedard, L., & Gertz, M. (2004). Florida’s gay adoption ban: What do
Floridians think? University of Florida Journal of Law and Public Policy,
15,261-283.
Ryan, S., Bedard, L., & Gertz, M. (2007). The influence of gender on the placement of
children with gay or lesbian adoptive parents. Journal of GLBT Family Studies,
3, 15-34.
Schulte, L., & Battle, J. (2004). The relative importance of ethnicity and religion in
predicting attitudes towards gays and lesbians. Journal of Homosexuality, 47,
127-142.
Sewell, W.H. (2005). Logics of History: Social theory and social transformations.
Chicago, IL: University of Chicago Press.
Sherkat, D. (1998). Counterculture or continuity? Competing influences on baby
boomers’ religious orientations and participation. Social Forces, 76, 1087-1115.
Sherkat, D., Mattias de Vries, K., & Creek, S. (2010). Race, religion, and opposition
to same-sex marriage. Social Science Quarterly, 91, 80-98.
Sherkat, D., Powell-Williams, M., Maddox, G., & Mattias de Vries, K. (2011).
Religion, politics, and support for same-sex marriage in the United States, 1988-
2008. Social Science Research, 40, 167-180.
Steensland, B., Park, J., Regnerus, M., Robinson, L., Wilcox, W. B., & Woodberry,
R. (2000). The measure of American religion: Toward improving the state of the
art. Social Forces, 79, 291-318.
van den Akker, H., van der Ploeg, R., & Scheeper, P. (2013). Disapproval of homo-
sexuality: Comparative research on individual and national determinants of dis-
approval of homosexuality in 20 European countries. International Journal of
Public Opinion Research, 25, 64-86.
Whitehead, A. L. (2010). Sacred rites and civil rights: Religion’s effect on attitudes
toward same-sex unions and the perceived cause of homosexuality. Social
Science Quarterly, 91, 63-79.
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
Whitehead and Perry 25
Whitehead, A. L. (2013). Religious organizations and homosexuality: The acceptance
of gays and lesbians in American congregations. Review of Religious Research,
55, 297-317.
Whitehead, A.L. (2014). Politics, religion, attribution theory, and attitudes toward
same-sex unions. Social Science Quarterly, DOI: 10.1111/ssqu.12085.
Whitehead, A. L., & Baker, J. (2012). Homosexuality, religion, and science: Moral
authority and the persistence of negative attitudes. Sociological Inquiry, 82,
487-509.
Wilson, R. (2004). A matter of conviction: Moral clashes over same-sex adoption.
BYU Journal of Public Law, 22, 475-497.
at CLEMSON UNIV on July 16, 2014jfi.sagepub.comDownloaded from
