You are on page 1of 15

The Journal of Sex Research

ISSN: 0022-4499 (Print) 1559-8519 (Online) Journal homepage: http://www.tandfonline.com/loi/hjsr20

Demographic Comparison of American Individuals


in Polyamorous and Monogamous Relationships

Rhonda N. Balzarini, Christoffer Dharma, Taylor Kohut, Bjarne M. Holmes,


Lorne Campbell, Justin J. Lehmiller & Jennifer J. Harman

To cite this article: Rhonda N. Balzarini, Christoffer Dharma, Taylor Kohut, Bjarne M. Holmes,
Lorne Campbell, Justin J. Lehmiller & Jennifer J. Harman (2018): Demographic Comparison
of American Individuals in Polyamorous and Monogamous Relationships, The Journal of Sex
Research, DOI: 10.1080/00224499.2018.1474333

To link to this article: https://doi.org/10.1080/00224499.2018.1474333

Published online: 18 Jun 2018.

Submit your article to this journal

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at


http://www.tandfonline.com/action/journalInformation?journalCode=hjsr20
THE JOURNAL OF SEX RESEARCH, 00(00), 1–14, 2018
Copyright © The Society for the Scientific Study of Sexuality
ISSN: 0022-4499 print/1559-8519 online
DOI: https://doi.org/10.1080/00224499.2018.1474333

Demographic Comparison of American Individuals in


Polyamorous and Monogamous Relationships
Rhonda N. Balzarini, Christoffer Dharma, and Taylor Kohut
Department of Psychology, University of Western Ontario

Bjarne M. Holmes
Department of Psychology, Champlain College
Lorne Campbell
Department of Psychology, University of Western Ontario
Justin J. Lehmiller
The Kinsey Institute, Indiana University, Bloomington, IN

Jennifer J. Harman
Department of Psychology, Colorado State University

Research on polyamorous relationships has increased substantially over the past decade. This
work has documented how polyamory is practiced and why individuals might pursue such
arrangements. However, there is a lack of a systematic investigation of who is in polyamorous
relationships and how they might differ from individuals in monogamous relationships. The
present study is one of the first to address this by comparing the demographic backgrounds of
individuals in polyamorous (N = 2,428) and monogamous (N = 539) relationships in the United
States. Compared to participants in monogamous relationships, those in polyamorous relation-
ships were more likely to report minority sexual identities. Despite similar age distributions,
individuals in polyamorous relationships were more likely to report being in a civil union, being
divorced, and earning less than $40,000 per year compared to individuals in monogamous
relationships. People in polyamorous relationships were also more likely to select “other”
options for most demographic characteristics, suggesting that they tend to choose less tradi-
tional response options in general. The current research highlights several demographic differ-
ences that need to be considered and potentially controlled for in future comparisons of
polyamorous and monogamous relationships.

In a monogamous relationship, it is generally not acceptable to Growing public interest regarding polyamory (Plummer,
seek out sexual interactions or emotional intimacy with any 1995) is reflected not just in increased coverage in the popular
person other than one’s partner (see Jonason & Balzarini, press (e.g., magazines and television programs) (see Barker &
2016). In contrast, polyamory refers to the practice or accep- Langdridge, 2010) but also in significantly increased numbers
tance of having multiple emotionally close relationships that of online searches for polyamory-related material between
may or may not be sexual in nature, with the consent of 2006 and 2015 (Moors, 2016).
everyone involved (Barker & Langdridge, 2010). While Research on polyamory has addressed how this relation-
monogamous relationships remain the norm in North ship style has the potential to prevent psychological suffoca-
America, polyamory is on the rise. Formally recognized as a tion of a single partner by distributing the needs that people
form of relationship in the 1990s (World Heritage expect from their relationships across many partners (Conley
Encyclopedia, 2016), polyamory is receiving an increasing & Moors, 2014; Finkel, Hui, Carswell, & Larson, 2014).
amount of attention in popular culture and society at large. Research has also examined the robust and pervasive stigma
surrounding this practice (e.g., Conley, Moors, Matsick, &
Ziegler, 2013; Moors, Matsick, Ziegler, Rubin, & Conley,
Correspondence should be addressed to Rhonda N. Balzarini,
University of Western Ontario, Social Science Centre, Department of
2013) and assessed how relationships and experiences of
Psychology Main Office, Room 7418, London, Ontario, N6A 5C2 stigma differ among primary and secondary partners in poly-
Canada. E-mail: rbalzari@uwo.ca amorous relationships (Balzarini et al., 2017). Although these
MONOGAMOUS AND POLYAMOROUS RELATIONSHIPS

areas of inquiry are important, a significant antecedent ques- in varying demographic groups (gender, race/ethnicity, sexual
tion has been overlooked: Are there demographic differences orientation, and age) were over- or underrepresented in CNM
between those who are in polyamorous relationships and and monogamous relationships. No significant differences
those who are in monogamous relationships? were found between gender, race, sexual orientation, and age
Considerable research attention has been given to con- among individuals in monogamous and CNM relationships
sensual nonmonogamy (CNM) more broadly, a term which (Rubin et al., 2014). This research suggested that the demo-
encompasses polyamory and several distinctly different graphic backgrounds of individuals in CNM relationships
practices, such as swinging and open relationships. Recent (including polyamory, as well as swinging and open relation-
research efforts have examined the potential benefits and ships) were similar to those in monogamous relationships.
costs associated with CNM relationships and considered Loving More (LM)—a polyamory support and advocacy
how they may be similar to or different from monogamous organization—administered a survey to more than 4,000
relationships (Conley, Ziegler, Moors, Matsick, & Valentine, CNM participants and compared the results with the
2012; Moors, Matsick, & Schechinger, 2017; Ziegler, 2010–2014 General Social Survey (GSS; a national repre-
Matsick, Moors, Rubin, & Conley, 2014). Research also sentative sample of U.S. residents). The study was designed
has assessed the demographic differences among those to explore potential demographic differences among indivi-
who are currently in CNM and monogamous relationships duals who self-identified as practicing or as open to practi-
(Cox, 2016; Rubin, Moors, Matsick, Ziegler, & Conley, cing some form of CNM relationship relative to the general
2014) and among those who have practiced CNM and U.S. population (Cox, 2016). CNM individuals indicated
monogamous relationships (Haurpert et al., 2017). To date, that they had slightly more adults living in their household
however, most of what we know about differences between (CNM 78.9% of persons in household were adults; GSS
individuals in these relationships is based on studies exam- 77.5% of the total household were adults) but were less
ining CNM relationships more broadly, rather than polya- likely to have children (CNM 19.3%; GSS 23.4%) (D.
morous relationships specifically. Cox, personal communication, October 2017). CNM indi-
Our research is important for several reasons. First, it pro- viduals were more likely to have had sexual partners of both
vides much-needed descriptive information about polyamorous sexes (46.3% females, 18.8% males) and were better edu-
relationships in particular (e.g., information about political cated (62.4% completed a bachelor’s degree or higher) than
affiliation, education, religious affiliation), which allows us to the general population (28.3%) (Cox, 2017; Cox,
test several popular assumptions about persons in these relation- Fleckenstein, & Bergstrand, 2013). The LM survey offers
ships, such as the idea that polyamorists are more likely to be helpful insights into the potential demographic differences
White, bisexual, and politically liberal than the rest of the of individuals who identify as CNM and monogamous, but
population. Second, this research can help us to better under- it was limited in scope and did not include critical demo-
stand why some people are more likely than others to practice graphic questions (e.g., sexual orientation identity, ethnicity,
polyamory. Third, this study offers potential insights into why employment information, income, marital status, religious
some people may be more successful or satisfied with poly- affiliation, political affiliation) or distinguish among types of
amorous relationships than monogamous relationships. Fourth, CNM (e.g., swinging, open, or polyamory).
this work can help us identify potential covariates to consider in Haupert et al., (2017) assessed the prevalence of CNM
future research, which may be important when comparing indi- (including but not limited to polyamory) using two separate
viduals in polyamorous relationships to those in monogamous Census-based samples of single adults in the United States
relationships (e.g., if polyamorous and monogamous indivi- (Study 1: N = 3,905; Study 2: N = 4,813). This research
duals differ in their sexual orientation, controlling for the stigma found that approximately 21% to 22% of individuals from
experienced as a result of one’s sexual orientation would be the United States have engaged in a CNM relationship at
important when assessing stigma toward the relationship orien- some point in their lives. The proportion of people who
tation specifically). Finally, if we identify differences or inequal- reported having engaged in a CNM relationship did not
ities experienced by those in polyamorous or monogamous differ by age, education level, income, religion, geographic
relationships, these findings could have implications for therapy region, political affiliation, or race, but it did vary by gender
or treatment of persons in these relationships. Specifically, and sexual orientation. Men (compared to women) and
knowledge of the stigmas or inequalities faced by people in people who identified as gay, lesbian, or bisexual (compared
polyamorous relationships and an understanding of the inter- to those who identified as heterosexual) were more likely to
sectionality between polyamory and other identities could report previous participation in a CNM relationship
potentially aid clinicians when working with these clients. (Haupert et al., 2017).
Although these studies provide some insight into partici-
pation or identification with CNM, they offer little insight
Demographic Comparisons Among CNM and into those who have been involved in polyamorous relation-
Monogamous Samples ships specifically. It is possible that results would diverge
among various CNM relationship subtypes, given other
Using data from two large online samples, Rubin and differences documented among these groups. Research sug-
colleagues (2014) examined the extent to which individuals gests that CNM individuals favor their own specific

2
BALZARINI, DHARMA, KOHUT, HOLMES, CAMPBELL, LEHMILLER, AND HARMAN

relationship orientation and stigmatize other relationship States. Demographic differences in terms of age, sexual
orientations (see Balzarini, Shumlich, Kohut, & Campbell, orientation, gender identity, education, religious affiliation,
2018). Swingers and polyamorous individuals, for example, political affiliation, parental demographics, ethnicity, and
are often critical of each other (Frank & DeLamater, 2010). household-related variables (e.g., number of children and
Polyamorists critique swingers’ supposed focus on recrea- divorces) were assessed. We did not specify hypotheses for
tional sex and the stereotypically gendered nature of swing- each of these variables; instead, we sought to explore
ing (Barker & Langdridge, 2010; Frank & DeLamater, whether differences or similarities emerge across these
2010). Swingers criticize the purported “conservative” atti- groups with the goal of informing future research in this
tudes that polyamorists have about sex and polyamorists’ area.
view that love can occur outside of a couple (Barker &
Langdridge, 2010; Frank & DeLamater, 2010). In a similar
vein, Ritchie (2010) reviewed news reports on polyamory
and found that interviewees who were quoted in these METHOD
stories presented polyamory as more meaningful than
swinging because polyamory was based on love rather Participants
than casual sex. Our study used data from two large online convenience
samples examining individuals in polyamorous (N = 3,530)
and monogamous relationships (N = 1,358). One sample
Demographic Comparisons Among Polyamorous
came from a study in which we told participants we were
Samples
interested in investigating the perceptions of partners among
Noël (2006) conducted a detailed content analysis of 12 individuals in polyamorous relationships; the other came
polyamory relationship guides (i.e., self-help books written from a study in which we told participants we were inter-
for the polyamory community) and critiqued these guides in ested in the perceptions of partners among individuals in
terms of their consideration of diversity. She found that the monogamous relationships. Participants for both studies
authors of these books presumed that their polyamorous were recruited online from various Internet forums; dating
audience consisted of homogenously White, educated, mid- sites; lesbian, gay, bisexual, transgender, and queer
dle-class, able-bodied citizens of the United States (Noël, (LGBTQ) groups; and Facebook group pages. Many of
2006; Wheeler, 2011). Other qualitative research by Sheff these Web sites and groups were specifically geared toward
and colleagues corroborates some of these assumptions by either a polyamorous or monogamous audience, as were the
indicating that individuals who participate in polyamorous advertisements for recruitment. Recruitment materials spe-
relationships are demographically homogenous with respect cified that eligibility required that participants be in a poly-
to race, profession, and education (Sheff, 2005; Sheff & amorous or monogamous relationship. Participants in
Hammers, 2011). polyamorous relationships were asked to indicate whether
The conclusions drawn by critics of these relationship each partner they listed was aware of and had consented to
guides are based on interpreted material or assumptions rather the participant having other partner(s) (asked separately for
than empirical research. One cannot rule out whether the lack each partner). Only participants whose partners were aware
of diversity found had to do with general base rates in these of and consented to their other partners were included in the
factors (e.g., 77% percent of Americans are White; Census current study in order to provide a clear delineation between
Bureau, 2016), sampling strategies, or actual differences in polyamory and infidelity. Because some response options
distributions of demographic variables among those who iden- applied only to participants in the United States (e.g., poli-
tify as polyamorous (e.g., if monogamists had been studied, tical affiliation, income in U.S. dollars), we restricted our
would they also appear homogenous on these traits, or would analyses to those living in the United States, which resulted
they differ from polyamorists?). Further, the number of parti- in a sample of n = 2,428 individuals in a polyamorous
cipants in these studies was small, with Sheff and Hammers relationship, and n = 539 individuals in a monogamous
(2011) reporting results for 81 polyamorous participants and relationship. The number of excluded participants from out-
Sheff (2005) reporting results for 40 polyamorous participants. side the Unted States is presented in the appendix.
Qualitative research is a proper approach to some, but not all,
research questions. When trying to acquire estimates of demo-
graphics for a population, a larger sample should be obtained Measures
and quantitative data are ideal.
Most of the demographic items were modified versions
of those used successfully in the GSS (Smith, Marsden,
THE CURRENT STUDY Hout, & Kim, 2015) and LM survey (Cox et al., 2013); a
The current research aimed to fill a gap in the literature few were developed by the researchers where indicated.
by critically examining demographic differences between
individuals who are currently in polyamorous and monoga- Age. Participants were asked to indicate their age in
mous relationships from two samples collected in the United years.

3
MONOGAMOUS AND POLYAMOROUS RELATIONSHIPS

Gender identity. Participants reported their current (e) Computers/information technology, (f) Marketing, (g)
gender identity. The four response options were (a) Male, Retail/sales, (h) Service, health care, (i) Homemaker or
(b) Female, (c) Transgender, and (d) Other. stay-at-home parent, (j) Unemployed, or (k) “Other.”

Sexual orientation. Participants identified their Marital status. Participants indicated their marital
current sexual orientation in terms of (a) Heterosexual, (b) status using one or more of the five response options: (a)
Lesbian/gay, (c) Bisexual, (d) Pansexual, and (e) Other. Currently married, (b) Widowed, (c) Divorced, (d)
Separated, or (e) Never been married. Participants who
Education. Participants reported their highest level of had been divorced were asked to indicate how many
education with six response choices: (a) Some high school, (b) divorces they had had (completed or in progress).
High school diploma/GED, (c) Vocational training certificate,
(d) Associate’s degree, (e) Bachelor’s degree, and (f) Master’s Household members and children. Participants
or doctoral degree. They also were asked to report the highest reported the number of people living in their household
education level from both parents using identical response for three days a week or more (including the participant).
choices. When parents' education differed, respondents Because polyamorous relationships can consist of numerous
selected the highest level of education that applied for each people who may not spend their entire week in one
parent. Because there were very few participants who did not dwelling, we defined household members as those who
finish high school (n = 12 polyamorous and 3 monogamous live together for three days a week or more. Participants
participants), we categorized education into three groups: also were asked to indicate how many children they had
High school or less, Associate’s degree/vocational training, from 0 to 5+, and how many children lived with them full
and Bachelor’s degree or higher. time.

Religious affiliation. Religious affiliation was


Procedure
assessed using eight response choices: (a) Agnostic, (b)
Atheist, (c) Buddhist, (d) Christian, (e) Hindu, (f) Jewish, Prospective participants were provided a link that redirected
(g) Muslim, or (h) Other. Participants were also asked to them to a secure survey hosted on Qualtrics. The first page of
report the religious affiliation of both parents using identical the survey requested informed consent, which was received
response choices, and to select all affiliations that applied. from each participant digitally. Participants were also asked to
confirm their eligibility by indicating that they were either
Political affiliation. Political affiliation was assessed polyamorous or monogamous and currently in a committed
using six response choices: (a) Democrat, (b) Republican, relationship or dating. Those who met the eligibility require-
(c) Independent, (d) Libertarian and Green Party, ments were asked to indicate the number of romantic or dating
(e) Unaffiliated, or (f) “Other.” They were also asked to partners with whom they were currently involved. Only mono-
report the political affiliation of their parents using identical gamous participants who reported one current partner and
response choices. polyamorous participants who reported at least two concurrent
partners were retained for analyses. Participants in polyamorous
Ethnicity. Participants were asked to identify their and monogamous relationships answered identical question-
ethnicity from eight response choices and to select all that naires that included demographic questions about themselves
applied: (a) Asian/Asian American, (b) African/African and their partners, and additional questions about their relation-
American, (c) Hispanic, (d) Native American/Native ship experiences with their partners (e.g., jealousy, communica-
Alaskan, (e) Pacific Islander, (f) White (non-Hispanic), (g) tion, satisfaction). Participation in the study was voluntary; no
biracial/multiracial, and/or (h) “Other.” Because some compensation was provided. Completion of the survey could be
ethnicity categories had a very small number of done at participants’ convenience (median time = 59 minutes).
participants, we created three mutually exclusive Only the demographic assessments were used in the present
categories: White only, Multiethnic, and Non-White. study. This research was conducted in accordance with the
Multiethnic included all participants who identified with ethical guidelines of the American Psychological Association.
more than one of the available ethnic categories (including The materials and procedure for data collection were approved
multiethnic). by our institutions’ ethics research boards.

Income. Participants were asked to report their annual


Analytic Strategy
income in U.S. dollars and were provided eight choices,
ranging incrementally from Less than $20,000 to $150,000 Because the polyamorous sample was much larger than
and higher. the monogamous sample, we conducted a power analysis
before we registered and ran our analyses with unequal
Profession/work. Current profession was assessed allocation (N2/N1 = 4) to resemble the ratio of eligible
using 11 options: (a) Student, (b) Education (nonstudent), participants in the data set. A minimum of 1,335 and 339
(c) Military, government/civil service, (d) Banking/finance, individuals in each group, respectively, was required to

4
BALZARINI, DHARMA, KOHUT, HOLMES, CAMPBELL, LEHMILLER, AND HARMAN

obtain a power of .95, assuming a small effect size (d = 0.2) polyamorous relationships (M = 35.01) (t = 0.45;
(power estimated using G-Power 3.1; Faul, Erdfelder, p = .665; Cohen’s d = .02).
Buchner, & Lang, 2009). This power analysis ensured that
the current study had sufficient participants to make demo-
Gender Identity
graphic comparisons across various categories.
To answer our primary research question concerning the There was a significant difference in the distribution of
demographic differences between participants in polyamor- gender identity between participants in monogamous and
ous and monogamous relationships, Pearson χ2 tests for polyamorous relationships (χ2 = 28.4, p < .001; Cramer’s
cross-tabulations (or Fisher’s exact test where appropriate) V = .10). Although most participants in both samples identi-
were used for all categorical comparisons. When Fisher’s fied as female, more participants in monogamous relationships
exact test was required for a table with a large number of indicated that their gender was female (polyamorous, 57.8%;
categories, p values were simulated using Monte Carlo monogamous, 63.3%; p = .020). Furthermore, compared
methods (Patefield, 1981). When chi-square tests detected to participants in monogamous relationships, a higher
an overall significant difference, post hoc analyses com- proportion of participants in polyamorous relationships
paring participants in polyamorous and monogamous rela- selected the “other” (polyamorous, 5.3%; monogamous,
tionships were conducted for each category option. 0.74%; p < .001) or transgender identity options (polyamor-
Cramer’s Vs are presented to illustrate the strength of the ous, 2.1%; monogamous, 0.56%; p < .001). Similar propor-
difference when chi-square tests were applied; values ran- tions of participants in polyamorous and monogamous
ging from 0.1 to 0.2 are considered weak associations, 0.2 relationships identified their gender as male (polyamorous,
to 0.3 are considered moderate, 0.3 to 0.4 are considered 34.8%; monogamous, 35.4%; p = .802). For both groups,
strong; and above 0.5 are considered redundant (Cohen, the most common open-ended responses for the gender ques-
1988). A chi-square test was used to compare the distribu- tion included “genderqueer” (n = 51), “genderfluid/non-bin-
tion of sexual orientation identities within different gender ary” (n = 13), “FTM/MTF/transwoman/transman” (n = 7), and
identities, separated between polyamorous and monoga- “agender” (n = 4).
mous participants. An independent t test also was used to
compare the mean ages between the two groups. Cohen’s
Sexual Orientation
d is presented to illustrate the magnitude of the
differences, with 0.2 considered a small effect size, 0.5 a There was a significant difference in the distribution of
medium effect size, and 0.8 a large effect size. The sig- sexual orientation responses selected by participants in mono-
nificance level for all tests was kept at the standard .05 gamous and polyamorous relationships (χ2 = 281.96,
level rather than adjusting for multiple comparisons. p < .001; Cramer’s V = .31). The most commonly reported
Because of the exploratory nature of these analyses, we sexual orientation among participants in monogamous rela-
were willing to accept the inflated experimentwise Type 1 tionships was heterosexual, which was much lower among
error (Rothman, 1990). All analyses were conducted using participants in polyamorous relationships (monogamous,
R 3.3.1 (R Core Team, 2016). The rationale and the 74.0%; polyamorous, 36.4%; p < .001). By contrast, partici-
analytic plan for the demographic comparisons were pre- pants in polyamorous relationships were more likely to iden-
registered to the Open Science Framework (OSF) after the tify as bisexual (polyamorous, 32.5%; monogamous, 13.5%;
data were collected, but prior to conducting the analyses, p < .001), pansexual (polyamorous, 18.0%; monogamous,
and can be viewed at https://osf.io/76p7p/. 3.5%; p < .001), or as another orientation not listed (poly-
amorous, 9.2%; monogamous, 3.2%; p < .001). However,
participants in polyamorous and monogamous relationships
did not differ in their rate of selecting “gay/lesbian” (poly-
RESULTS
amorous, 3.9%; monogamous, 5.8%; p = .059). For both
groups, common “other” responses included “hetero-flexible”
A total of 539 monogamous and 2,428 polyamorous
(n = 51), “queer” (n = 114), “bicurious”/“curious” (n = 16),
participants completed the survey and were eligible for the
and “asexual” (n = 7).
current analyses (see Table 1 for their demographic distribu-
tion and for total sample characteristics). Most participants
self-identified as White females between 25 and 44 years Sexual Orientation and Gender Identity
old (Mage = 34.97) with a bachelor’s degree or higher, a
The distribution of participants’ gender identity and sex-
Democratic- or Independent-leaning political affiliation, and
ual orientation stratified by their relationship orientation
reported an annual income of less than $60,000.
(polyamorous versus monogamous) is presented in
Table 2. There is an overall significant difference in the
distribution of sexual orientation responses by gender iden-
Age
tities within individuals in polyamorous relationships
Overall, there was no significant difference in the mean (χ2 = 671.81, p < .001; Cramer’s V = .30) and within
age of participants in monogamous (M = 34.76) and individuals in monogamous relationships (χ2 = 86.61,

5
MONOGAMOUS AND POLYAMOROUS RELATIONSHIPS

Table 1. Demographic Distribution of Participants

Polyamorous Monogamous Total


Demographics (n = 2,428) (n = 539) (n = 2,967) p Valuea χ2/t Statistics Cramer’s V/ Cohen’s db

Age (mean) 35.01 34.76 34.97 .665 0.45 .02


Gender identity (n = 2,426) (n = 539) < .001 28.42 .10
Male 844 (34.8) 191 (35.4) 1,035 (34.9) .803
Female 1,403 (57.8) 341 (63.3) 1,744 (58.8) .020
Transgender 51 (2.1) 3 (0.56) 54 (1.8) .012
Other 128 (5.3) 4 (0.7) 132 (4.5) < .001
Sexual orientation (n = 2,422) (n = 539) < .001 281.96 .31
Heterosexual 882 (36.4) 399 (74.0) 1,281 (43.3) < .001
Lesbian/gay 95 (3.9) 31 (5.8) 126 (4.3) .059
Bisexual 788 (32.5) 73 (13.5) 861 (2.9) < .001
Pansexual 435 (18.0) 19 (3.5) 454 (15.3) < .001
Other 222 (9.2) 17 (3.2) 239 (8.1) < .001
Highest education completed (n = 2,426) (n = 539) .043 6.28 .05
High school or less 427 (17.6) 88 (16.4) 515 (17.4) .530
Associate’s degree/vocational training 459 (18.9) 80 (14.9) 539 (18.2) .030
Bachelor’s degree or higher 1,540 (63.5) 370 (68.8) 1,910 (64.4) .022
Parental education (n = 2,422) (n = 538) .177 3.47 .03
High school or less 500 (20.6) 127 (23.6) 627 (21.2) ―
Associate’s degree/vocational training 380 (15.7) 91 (16.9) 471 (15.9) ―
Bachelor’s degree or higher 1,542 (63.7) 320 (59.5) 1,862 (62.9) ―
Religious affiliation (n = 2,405) (n = 539) < .001 182.36 .25
Agnostic 553 (23.0) 157 (29.4) 710 (24.2) .002
Atheist 684 (28.4) 118 (22.1) 802 (27.9) .003
Buddhist 81 (3.4) 14 (2.6) 95 (3.2) .420
Christian 260 (10.8) 157 (29.4) 417 (14.2) < .001
Hindu 6 (0.25) 1 (0.19) 7 (0.24) 1.000
Jewish 78 (3.2) 17 (3.2) 95 (3.2) 1.000
Muslim 0 (0.00) 2 (0.37) 2 (0.06) .033
Other 743 (30.9) 68 (12.7) 811 (27.6) < .001
Parental religious affiliationc (n = 2,423) (n = 538) ― ― ―
Agnostic 499 (20.6) 93 (17.3) 592 (20.0) .094
Atheist 292 (12.0) 61 (11.3) 353 (11.9) .699
Buddhist 59 (2.43) 9 (1.67) 68 (2.30) .341
Christian 1,714 (70.6) 390 (72.4) 2,104 (71.1) .432
Hindu 11 (0.45) 2 (0.37) 13 (0.44) 1.000
Jewish 214 (8.8) 35 (9.7) 249 (8.4) .097
Muslim 16 (0.7) 2 (0.4) 18 (0.61) .637
Other 278 (11.5) 48 (8.9) 326 (11.0) .103
Political affiliation (n = 2,421) (n = 539) < .001 36.07 .11
Democrat 1,045 (43.2) 262 (48.6) 1,307 (44.2) .024
Republican 73 (3.0) 30 (5.6) 103 (3.5) .006
Libertarian 172 (7.1) 23 (4.3) 195 (6.6) .016
Green Party 134 (5.5) 8 (1.5) 142 (4.8) < .001
Independent 879 (36.3) 201 (37.3) 1,080 (36.5) .692
Other 118 (4.9) 15 (2.8) 133 (4.5) .038
Parental political affiliationc (n = 2,389) (n = 534) ― ― ―
Democrat 1,384 (57.0) 288 (53.4) 1,672 (57.2) .143
Republican 992 (40.9) 218 (40.5) 1,210 (41.4) .899
Green Party 42 (1.7) 2 (0.37) 44 (1.5) < .001
Libertarian 93 (3.8) 8 (1.3) 101 (3.5) .002
Independent 387 (15.9) 111 (20.6) 498 (17.0) < .001
Other 119 (4.9) 12 (2.2) 131 (4.5) < .001
Ethnicity (categorized) (n = 2,416) (n = 539) .005 10.43 .06
Non-White, single race 143 (5.9) 40 (7.4) 183 (6.2) .222
White only 2,010 (83.2) 463 (86.1) 2,473 (83.7) .118
Multiethnic 263 (10.9) 35 (6.5) 298 (10.1) .003
Ethnicityc (n = 2,416) (n = 539) ― ― ―
Asian 57 (2.4) 19 (3.5) 76 (2.6) .131
African 64 (2.7) 13 (2.4) 77 (2.6) .881
Hispanic 106 (4.4) 24 (4.5) 130 (4.4) .908
Native 109 (4.5) 8 (1.5) 117 (4.0) .001

(Continued )

6
BALZARINI, DHARMA, KOHUT, HOLMES, CAMPBELL, LEHMILLER, AND HARMAN

Table1. (Continued)

Polyamorous Monogamous Total


Demographics (n = 2,428) (n = 539) (n = 2,967) p Valuea χ2/t Statistics Cramer’s V/ Cohen’s db

Pacific Islander 13 (0.5) 0 (0.00) 13 (0.44) .143


White 2,205 (91.3) 483 (89.6) 2,688 (91.0) .371
Multiethnic 125 (5.2) 17 (3.2) 142 (4.8) .057
Other 50 (2.1) 3 (0.55) 53 (1.8) .018
Income (US$) (n = 2,408) (n = 539) .003 21.21 .09
< $20,000 697 (29.0) 124 (23.1) 821 (27.9) .007
$20,000–$40,000 626 (26.0) 120 (22.4) 746 (25.3) .089
$40,000–$60,000 389 (16.2) 106 (19.8) 495 (16.8) .048
$60,000–$80,000 247 (10.3) 71 (13.3) 318 (10.8) .046
$80,000–$100,000 155 (6.4) 29 (5.4) 184 (6.3) .430
$100,00–$120,000 107 (4.4) 27 (5.0) 134 (4.6) .567
$120,000–$150,000 56 (2.3) 18 (3.36) 74 (2.5) .170
> $150,000 131 (5.4) 41 (7.65) 172 (5.8) .053
Profession (n = 2,426) (n = 539) < .001 73.13 .16
Student 370 (15.3) 96 (17.8) 466 (15.7) .150
Education (nonstudent) 178 (7.3) 82 (15.2) 260 (8.8) < .001
Military 18 (0.74) 0 (0.00) 18 (0.61) .058
Government 74 (3.0) 20 (3.7) 94 (3.2) .416
Banking 58 (2.4) 18 (3.3) 76 (2.6) .227
Information technology 432 (17.8) 56 (10.4) 488 (16.5) < .001
Marketing 46 (1.9) 11 (2.1) 57 (1.9) .862
Retail 131 (5.4) 22 (4.1) 153 (5.2) .237
Service 126 (5.2) 39 (7.2) 165 (5.6) .077
Health care 188 (7.8) 55 (10.2) 243 (8.2) .068
Homemaker 112 (4.6) 25 (4.6) 137 (4.6) 1.00
Unemployed 104 (4.3) 10 (1.9) 114 (3.8) .006
Other 589 (24.3) 105 (19.5) 694 (23.4) .018
Marital statusc (n = 2,419) (n = 537) ― ― ― ―
Married 992 (40.9) 236 (43.8) 1,228 (41.5) .230
Civil union 244 (10.0) 29 (5.5) 273 (9.2) .009
Separated 111 (4.6) 3 (0.56) 114 (3.9) < .001
Divorced 420 (17.3) 40 (7.5) 460 (1. 6) < .001
Widowed 33 (1.4) 6 (1.1) 39 (1.3) .810
Never married 839 (34.6) 233 (43.2) 1,072 (36.7) < .001
Children in the household (n = 2,413) (n = 537) .805 2.31 .03
0 1,481 (61.6) 321 (60.3) 1,802 (61.4) ―
1 337 (14.0) 75 (14.1) 412 (14.0) ―
2 333 (13.9) 76 (14.3) 409 (14.0) ―
3 149 (6.2) 33 (6.2) 182 (6.2) ―
4 60 (2.5) 19 (3.6) 79 (2.7) ―
5+ 44 (1.8) 8 (1.5) 52 (1.8) ―
Members of the household (n = 2,404) (n = 532) < .001 32.82 .11
1 317 (13.1) 50 (9.3) 367 (12.4) .014
2 802 (33.2) 231 (43.0) 1,033 (35.0) < .001
3 552 (22.9) 119 (22.2) 671 (22.8) .776
4 390 (16.2) 91 (17.0) 481 (16.3) .652
5 183 (7.6) 24 (4.5) 207 (7.0) .009
6 72 (3.0) 15 (2.8) 87 (3.0) .889
7 49 (2.0) 3 (0.6) 52 (1.8) .017
8+ 48 (2.0) 4 (0.7) 52 (1.8) .005
Number of divorces (n = 2,428) (n = 539) < .001 38.11 .11
0 1,712 (70.5) 448 (83.1) 2,160 (72.8) < .001
1 541 (22.3) 76 (14.1) 617 (20.8) < .001
2 132 (5.4) 13 (2.41) 145 (4.9) .110
3 27 (1.1) 2 (0.4) 29 (0.98) .135
4 11 (0.5) 0 11 (0.37) ―
5 2 (0.08) 0 2 (0.07) ―
8+ 3 (0.01) 0 3 (0.10) ―
a
Significant p values are bolded (< .05).
b
For the pairwise within-group comparisons, due to small expected number of cells, we utilized Fisher’s exact test; χ2 and Cramer’s V cannot be presented.
c
Will not add up to 100% because participants may select multiple options.

7
MONOGAMOUS AND POLYAMOROUS RELATIONSHIPS

Table 2. Sexual Orientation and Gender Identity of Participants difference between participants in polyamorous and mono-
gamous relationships whose highest education was high
Gender Identity
N (%)a school or less (polyamorous, 17.6%; monogamous, 16.4%;
p = .529).
Participants Male Female Transgender Other Total

Polyamorous Parental Education


Heterosexual 545 (64.6) 329 (23.5) 1 (2.0) 6 (4.7) 882
Gay/lesbian 28 (3.3) 57 (4.1) 3 (5.9) 7 (5.5) 95 There were no statistically significant differences
Bisexual 152 (18.0) 611 (43.6) 8 (15.7) 17 (13.3) 788 between participants in polyamorous and monogamous rela-
Pansexual 76 (9.0) 282 (20.1) 18 (35.3) 59 (46.1) 435
tionships in their reported parental education (χ2 = 3.47,
Other 43 (5.1) 119 (8.5) 21 (41.2) 39 (30.5) 222
Total 844 1403 51 128 p = .177; Cramer’s V = .03). Just like the participants
χ2 = 671.81; p < .0001; Cramer’s V = 0.30 themselves, the highest level of education attained by either
Monogamous parent was typically a bachelor’s degree or higher (polya-
Heterosexual 152 (79.6) 245 (71.9) 1 (33.3) 1 (25.0) 399 morous, 63.7%; monogamous, 59.5%).
Gay/lesbian 17 (8.9) 14 (4.1) 0 0 31
Bisexual 17 (8.9) 55 (16.1) 1 (33.3) 0 73
Pansexual 3 (1.6) 12 (3.5) 1 (33.3) 3 (75.0) 19 Religious Affiliation
Other 2 (1.1) 15 (4.4) 0 0 17
Total 191 341 3 4 There was a difference in the distribution of religious
χ2 = 86.61; p < .0001; Cramer’s V = 0.23 affiliations between the participants in monogamous and
Note. Table shows the distribution of sexual orientation identities (column) polyamorous relationships (χ2 = 182.36, p < .001;
and gender identities (row) in the study. Cramer’s V = .25). The largest difference was for
a
Percentages are calculated as subtotals of the columns. Christianity, such that participants in monogamous relation-
ships were more likely to identify as Christian (monoga-
mous, 29.4%; polyamorous, 10.8%; p < .001). By contrast,
p < .001; Cramer’s V = .23). Among female-identified participants in polyamorous relationships were more likely
participants, bisexual was the most commonly reported sex- to choose the option for “other” religion than were partici-
ual orientation identity for those who were in polyamorous pants in monogamous relationships (polyamorous, 30.9%;
relationships (43.6%), whereas heterosexual was the most monogamous, 12.7%; p < .001). For both groups, common
commonly reported sexual orientation for women who were open-ended responses included “secular humanist” (n = 3),
in monogamous relationships (71.9%). Among male parti- “Wiccan” (n = 34), “spiritual” (n = 70), “pagan” (n = 230),
cipants, heterosexual was the most commonly endorsed and “none” (n = 47). The proportions of agnostic (poly-
sexual orientation category for both relationship types (poly- amorous, 23.0%; monogamous, 29.4%; p = .002) and athe-
amorous, 64.6%; monogamous, 79.6%). Pansexual was an ist affiliations (polyamorous, 28.4%; monogamous, 22.1%;
orientation that was frequently mentioned by polyamorous p = .003) differed significantly between the two groups,
females (20.1%), but this orientation was mentioned less such that participants in polyamorous relationships were
often by polyamorous males (9.0%, p < .001). Gay/lesbian less likely to identify as agnostic and more likely to identify
was the least mentioned sexual orientation category among as atheist compared to participants in monogamous relation-
polyamorous males and females (males, 3.3%; females, ships. Finally, there were no participants in polyamorous
4.1%; p = .337). Across both groups, those who were relationships in this sample who identified as Muslim, and
transgender or identified with other genders often reported very few, yet significantly more participants in monogamous
pansexual or “other” sexual orientations. Due to the sparse relationships, who identified as Muslim (polyamorous,
number of participants in some of these categories (e.g., 0.00%; monogamous, 0.37%; p = .033).
monogamous gay/lesbian transgender participants), further
statistical inferences were not performed. (Data for an enu-
meration of all possible pairwise comparisons are available Parental Religious Affiliation
upon request or on the OSF: https://osf.io/z8cme/.) Participants most commonly reported that their parents
were Christian (polyamorous, 70.6%; monogamous, 72.4%;
p = .432), followed by agnostic (polyamorous, 20.6%;
Highest Level of Education
monogamous, 17.3%; p = .094) and atheist (polyamorous,
Participants in polyamorous and monogamous relation- 12.0%; monogamous, 11.3%; p = .699). There were no
ships differed in their highest level of education achieved statistically significant differences between the two groups
(χ2 = 6.27, p = .043; Cramer’s V = .05). Most participants regarding parental religion.
reported having at least a bachelor’s degree or higher,
although participants in monogamous relationships were
Political Affiliation
more likely to have a bachelor’s degree or higher than
participants in polyamorous relationships (polyamorous, Participants in polyamorous and monogamous relation-
63.5%; monogamous, 68.8%; p = .022). There was no ships differed in their political affiliation (χ2 = 36.07,

8
BALZARINI, DHARMA, KOHUT, HOLMES, CAMPBELL, LEHMILLER, AND HARMAN

p < .001; Cramer’s V = .11). While Democrat was the most (polyamorous, 5.9%; monogamous, 7.4%; p = .222), and
commonly endorsed party by both groups, participants in the most common non-White ethnicity was Hispanic (poly-
monogamous relationships selected this option more fre- amorous, 4.4%; monogamous, 4.5%; p = .907), followed by
quently than did participants in polyamorous relationships Asian (polyamorous, 2.4%; monogamous, 3.5%; p = .131),
(monogamous, 48.6%; polyamorous, 43.2%; p = .024). and African American (polyamorous, 2.7%; monogamous,
There were a few Republicans in the current samples for 2.4%; p = .881). “Other” responses include “mixed” (n = 8),
both groups, although participants in polyamorous relation- “Jewish” (n = 9), and “Middle Eastern”/“Arab”/“Persian”
ships were less likely to select Republican than were partici- (n = 5). The “other” option was more often selected by
pants in monogamous relationships (polyamorous, 3.0%; participants in polyamorous relationships than by partici-
monogamous, 5.6%; p = .006). There were not many partici- pants in monogamous relationships (polyamorous, 2.1%;
pants who selected other parties, but polyamorous participants monogamous, 0.55%; p = .018).
were more likely to identify with smaller parties, such as the
Green Party (polyamorous, 5.5%; monogamous, 1.5%;
Income
p < .001) and Libertarian (polyamorous, 7.1%; monogamous,
4.3%; p = .016). Independent was an affiliation that was There was a difference in the distribution of income
equally endorsed by both groups (polyamorous, 36.3%; between the two groups (χ2 = 21.21, p = .003; Cramer’s
monogamous, 37.3%; p = .692), yet slightly more participants V = .09) such that participants in polyamorous relationships
in polyamorous relationships selected “other” political affilia- tended to report a lower income bracket than did those who
tions than did participants in monogamous relationships were in monogamous relationships. Participants in polyamor-
(polyamorous, 4.9%; monogamous, 2.8%; p = .038). For ous relationships were more likely to make less than $20,000
both groups, “other” affiliations included “anarchist” per year compared to participants in monogamous relation-
(n = 21), “progressive” (n = 10), “none” (n = 11), and ships (polyamorous, 29.0%; monogamous, 23.1%; p = .007).
“Democratic socialist”/“socialist liberal” (n = 20). Conversely, participants in monogamous relationships were
more likely to report making more than $100,000 per year
compared to participants in polyamorous relationships (mono-
Parental Political Affiliation
gamous, 16.1%; polyamorous, 12.1%; p = .020).
Participants could select more than one option for their
parents’ political affiliation. Just like the participants them-
Profession/Work
selves, the most common parental political affiliation
selected was Democrat (polyamorous, 57.0%; monoga- There was a difference in the distribution of professions
mous, 53.4%; p = .143). Unlike the participants, though, between participants in polyamorous and monogamous rela-
many parents of participants in both monogamous and poly- tionships (χ2 = 73.13, p < .001; Cramer’s V = .16). Participants
amorous relationships were Republican. However, the pro- in polyamorous relationships were more likely to work in
portion of Republican parents did not differ significantly information technology (polyamorous, 17.8%; monogamous,
between groups (polyamorous, 40.9%; monogamous, 10.4%; p < .001), whereas participants in monogamous rela-
40.5%; p = .899). With regard to third parties, there were tionships were more likely to work in education (monoga-
significant differences in the proportions of Libertarian par- mous, 15.2%; polyamorous, 7.3%; p < .001). Participants in
ents (polyamorous, 3.8%; monogamous, 1.3%; p = .005), polyamorous relationships were more likely to be unemployed
Independent parents (polyamorous, 15.9%; monogamous, (polyamorous, 4.3%; monogamous, 1.9%; p = .006).
20.6%; p < .001), and those who selected “other” (polya- Participants in polyamorous relationships also were more
morous, 4.9%; monogamous, 2.2%; p = .008), indicating likely to select “other” for their profession (polyamorous,
that the parents of participants in polyamorous relationships 24.3%; monogamous, 19.5%; p = .018); these responses com-
were more likely to be Libertarian (although this percent is monly included “writer,” “transportation,” “wholesale,” “self-
low), more likely to identify with another party, and less employed,” “retired,” and “legal.”
likely to identify as Independent.
Marital Status
Ethnicity
Participants could select all marital statuses that applied.
There was a significant difference in the distribution of Many of the participants in this study reported that they
ethnicity between the two groups (χ2 = 10.43, p = .005; were married (polyamorous, 40.9%; monogamous, 43.8%;
Cramer’s V = .06). Overall, most participants were White p = .230). Although the proportion of married individuals
(polyamorous, 83.2%; monogamous, 86.1%; p = .118), did not significantly differ across groups, participants in
although participants in polyamorous relationships were polyamorous relationships were more likely to be in a
more likely to identify as multiethnic than were participants civil union than were participants in monogamous relation-
in monogamous relationships (polyamorous, 10.9%; mono- ships (polyamorous, 10.0%; monogamous, 5.5%; p < .001).
gamous, 6.5%; p = .003). There was a low number of single Participants in a polyamorous relationship also were more
ethnicity, non-White participants in both groups likely to report separation (polyamorous, 4.6%;

9
MONOGAMOUS AND POLYAMOROUS RELATIONSHIPS

monogamous, 0.56%; p < .001) and divorce (polyamorous, Diversity within the polyamorous sample is evident in
17.3%; monogamous, 7.5%; p < .001), and less likely to participants’ reports of their gender identity. For example,
report never being married (polyamorous, 34.6%; monoga- over 7% of participants in polyamorous relationships iden-
mous, 43.2%; p < .001). Differences were found between tified as transgender or “other” genders, whereas only 1% of
the two groups when comparing the number of divorces participants in monogamous relationships identified as such.
between participants in a polyamorous and monogamous Findings concerning sexual orientation were consistent with
relationship (χ2 = 38.11, p < .001). Participants in polya- past results among CNM participants (Haupert, Gesselman,
morous relationships were less likely to report never being Moors, Fisher, & Garcia, 2017; Manley, Diamond, & van
divorced than were participants in a monogamous relation- Anders, 2015) in that individuals in polyamorous relation-
ship (polyamorous, 70.5%; monogamous, 83.1%; p < .001), ships were more likely to identify as bisexual or pansexual,
and participants in polyamorous relationships were more or to indicate an “other” sexual orientation, whereas indivi-
likely to report at least one divorce (polyamorous, 22.3%; duals in monogamous relationships were more likely to
monogamous, 14.1%; p < .001). Very few participants identify as heterosexual. In fact, the difference in the pro-
reported two divorces (polyamorous, 5.4%; monogamous, portion of individuals in polyamorous and monogamous
2.4%; p < .001); of those participants who did report two relationships who identified as bisexual in our study was
divorces, participants in polyamorous relationships were close to 20%, whereas the difference for heterosexual iden-
more likely to have been divorced twice. tification was approximately 40%. One possibility is that
bisexual individuals are more predisposed to enter and
remain in polyamorous relationships. Alternatively, it is
Household Members and Children
possible that being in a polyamorous/CNM relationship
The distribution of number of children did not signifi- leads individuals to be more likely to acknowledge bisexual
cantly differ between participants in polyamorous and attractions or to experience sexual fluidity.
monogamous relationships (χ2 = 2.31, p = .804; Cramer’s These group differences in gender identity and sexual
V = .03). Most participants did not have children living in orientation complement one another in that people who
the household (polyamorous, 61.6%; monogamous, 60.3%), identify as trans or gender nonbinary tend to be more likely
although there was a significant difference in the distribu- to endorse nontraditional sexual orientation categories (e.g.,
tion of number of household members between participants pansexual), which do not have assumptions of a binary
in polyamorous and monogamous relationships (χ2 = 32.82, gender (Kuper, Nussbaum, & Mustanski, 2012). It is also
p < .001; Cramer’s V = .11). Participants in polyamorous noteworthy that the least endorsed sexual orientation among
relationships were less likely to have two household mem- individuals in polyamorous relationships was gay/lesbian
bers than were participants in monogamous relationships rather than heterosexual. In fact, similar proportions of
(polyamorous, 33.2%; monogamous, 43.0%; p < .001), participants in polyamorous and monogamous relationships
and they were more likely to have five or more household endorsed a gay/lesbian identity. This finding contrasts with
members than those who were in monogamous relationships the results of the LM study, where heterosexual was the
(polyamorous, 7.6%; monogamous, 4.4%; p = .017). least commonly endorsed orientation category among CNM
participants (Cox et al., 2013). While it is unknown why this
is the case, one may argue that our study’s recruitment
DISCUSSION strategies, which focused on polyamorous groups, may
have underrecruited gay/lesbian polyamorists, whose poly-
Our study explored demographic differences between amorous identities may be less salient and/or who are pos-
individuals in polyamorous and monogamous relationships sibly more active in LGBTQ-oriented groups. Some
recruited from two online samples of people living in the literature has shown that a high proportion of gay men
United States. Overall, differences emerged for gender, sex- practice CNM, most commonly in the form of open relation-
ual orientation, level of education, religious affiliation, poli- ships and swinging (Barker & Langdridge, 2010). An alter-
tical affiliation, parental political affiliation, ethnicity, native possibility is that gay men’s increased propensity to
income, employment, and marital status. No differences practice CNM does not extend to the practice of polyamory.
were observed for age, parental education level, and paren- We also examined differences in the political and reli-
tal religious affiliation. This study extends previous descrip- gious affiliations of our respondents, as well as those of
tive findings by focusing on polyamory exclusively (rather their parents. Of the political and religious affiliation com-
than CNM overall) and by assessing differences in areas that parisons, the association between religious affiliation and
have not been considered in previous research (e.g., relationship orientation was the strongest (Cramer’s
employment type, income, marital status, religious affilia- V = .25). Individuals in polyamorous relationships (10.8%)
tion, political affiliation, and parental background). The were much less likely to identify as Christian compared to
within-group differences among our participants call into individuals in monogamous relationships (29.4%). This
question the view that polyamorous participants represent finding is not surprising in light of strong Christian prohibi-
a homogenous group (see Noël, 2006; Sheff, 2005; Sheff & tions against nonheterosexuality and the high rate of non-
Hammers, 2011; Wheeler, 2011). heterosexuality we observed among persons in polyamorous

10
BALZARINI, DHARMA, KOHUT, HOLMES, CAMPBELL, LEHMILLER, AND HARMAN

relationships. Given this difference, one might suspect that participants in polyamorous relationships were more
individuals in polyamorous relationships would be more likely to be multiethnic compared to participants in
inclined to identify as atheist or agnostic. Although this monogamous relationships. It is possible that compar-
was true for atheism, participants in polyamorous relation- able results have not been reported because of how
ships were less likely to identify as agnostic compared to questions about race/ethnicity were asked in past studies.
individuals in monogamous relationships. Individuals in In the current study, the option of “multiracial” was a
polyamorous relationships were much more likely to have stand-alone choice. This approach has been taken in one
selected “other” religions. previous study involving polyamory (Rubin et al., 2014),
Although the association was not as strong as it was with although many studies on gender and sexual minorities
religion, we also found that polyamorous participants were in the United States have not provided the option of
less likely to be Republican and more likely to be affiliated multiracial (e.g., Conron, Scott, Stowell, & Landers,
with the Green Party than those who were monogamous. 2012; Kuper et al., 2012).
This finding contrasts with previous findings for swingers, Our findings also suggest that individuals in polya-
who are typically moderates or conservatives and tend to morous relationships had more turnover in their rela-
vote Republican (Jenks, 1998). This difference may stem tionships compared to individuals in monogamous
from the fact that polyamorists (relative to swingers) seem relationships. Individuals in polyamorous relationships
disproportionately likely to identify as sexual and gender not only were more likely to have been divorced or
minorities. Given that sexual and gender minorities have separated but also reported a higher number of divorces
historically been excluded by the Republican party than did individuals in monogamous relationships.
(McGee, 2016), this could explain why they are less likely Because individuals in polyamorous relationships
than swingers to affiliate with that party. reported having more relationships than individuals in
Another novel contribution of our study involved the monogamous relationships, we might expect that there
assessment of differences among the parents of respondents. would be more separations and divorces among poly-
While no differences in parental religious affiliation and amorous participants. This is not to say that polyamor-
education level between monogamous and polyamorous ous relationships are less strong or stable in general;
individuals’ parents emerged, polyamorous respondents’ rather, it may be a reflection of the fact that with more
parents were more likely to report their affiliation as Green relationship experiences subsequently come more break-
Party and Libertarian, and less likely to report their affilia- ups. Caution in interpreting this finding is required due
tion as Independent. Results need to be interpreted with to the intricacies in the structure of certain polyamorous
caution due to the small number of participants who relationships. For instance, one may legally “divorce”
reported parents affiliated with these parties. one’s current partner in order to legally marry a new
Across analyses involving gender identity, sexual orien- partner, while maintaining commitment to both partners
tation, religious affiliation, political affiliation, and ethnicity, (Musumeci, 2017). The prevalence of such practice is
we noted a tendency for participants in polyamorous rela- unknown and its contribution to the high rates of
tionships to choose the “other” option more frequently than divorce within the polyamorous community cannot be
did participants in monogamous relationships. To our estimated in this study.
knowledge, there have been no other studies that have Our data also suggest that individuals in polyamor-
reported similar findings across demographic variables ous relationships have been in more civil unions than
among people in CNM relationships. We speculate that have participants in monogamous relationships. This
this finding may reflect polyamorists’ preferences to reject tendency toward civil unions may be driven by the
or deviate from traditional group labels, similar to how trans lower desirability of marriage, especially in current
people tend to reject traditional sexual orientation labels Western societies which prohibit formal polyamorous
(Kuper et al., 2012) and how both sexual and gender mino- marriages (Berkowitz, 2007; Duvall-Antonacopoulos,
rities continue to challenge the traditional definition of MacRae, & Paetsch, 2005) and enforce “socially
families as exclusively involving monogamous heterosexual imposed monogamy” (Eastwick, 2013), or it may be
couples who have children together (Giddens, 1992). another sign of their attraction to nontradition labels.
Rejection of labels in our data is corroborated by the However, this explanation on its own seems insufficient,
open-ended responses for the ethnicity question, where given that 41% of individuals in polyamorous relation-
some participants explicitly expressed their rejection of ships either were married currently or had been in the
these traditional labels with such comments as “Race is a past. Thus, civil unions may be a way to signify com-
social construction.” mitment to partners other than marriage, after a mar-
It is also important to note that we have replicated riage has occurred, or once a participant became
recent findings from Rubin and colleagues (2014), which polyamorous. Historically, sexual minorities have been
challenged the prevailing belief that CNM (Jenks, 2014; on the forefront of family changes (Giddens, 1992), and
Jenks, 1998) and polyamorous groups (Noël, 2006; the differences found in numbers of divorces, estimates
Sheff, 2005; Sheff & Hammers, 2011; Wheeler, 2011) of civil unions, and number of adults in a household
are homogenously White. We also found that may suggest a restructuring of the family.

11
MONOGAMOUS AND POLYAMOROUS RELATIONSHIPS

LIMITATIONS AND FUTURE DIRECTIONS (2014) support this hypothesis: Individuals who participated
in CNM and who were recruited from online communities
Because of the exploratory nature of these analyses, we did
tended to be more homogenous than participants recruited by
not adjust for multiple comparisons. As a result, the analyses
other methods. It is also important to note that these two
were more sensitive to potential differences (i.e., to maintain
studies recruited CNM participants broadly instead of poly-
low experiment-wise Type 2 error). As such, the interpretation
amorous participants specifically. Taken together, it appears
of the differences we found should be constrained by the
possible that individuals in polyamorous and monogamous
knowledge of a higher than typical experiment-wise Type 1
relationships are more alike in terms of age, education, and
error rate in this study. Because there are no more than 10
ethnicity than previously suggested.
pairwise comparisons per family of comparisons, we argue
To our knowledge, our findings on income and employ-
that only low p values (.05/10 = .005, thus p < .005) should be
ment inequalities within the polyamorous community have
considered reliable (see Table 1). Those with higher p values
not been identified in prior research. More research should
should be considered tentative until replicated in an indepen-
be conducted to investigate whether such inequality exists
dent confirmatory study with well-controlled Type 1 error.
due to the marginalized status of one’s relationship or some-
Future studies will benefit from using validated and more
thing else, such as lifestyle choices, or the fact that indivi-
inclusive questionnaires to measure certain constructs, such
duals in polyamorous relationships are more likely to be
as gender identity (Bauer, Braimoh, Scheim, & Dharma,
sexual and gender minorities. With the rise in the practice of
2017; Broussard, Warner, & Pope, 2017), as well as asking
CNM (Haupert et al., 2017; Moore, 2016), it is important to
other relevant questions about religiosity and religious prac-
establish whether such inequalities are reliable, and, if so,
tices (Freiheit, Sonstegard, Schmitt, & Vye, 2006; Koenig &
the impact on those experiencing them.
Büssing, 2010). For example, it might be useful to assess
Polyamory seems to be gaining some societal popularity
the importance of religion in one’s life and one’s frequency
and interest as a potential relationship alternative (Barker &
of attending religious services. Also, because we restricted
Langdridge, 2010; Moors, 2016). With this in mind, social
our sample to individuals living in the United States, there
scientists are also increasingly turning their attention to the
are important cultural constructs to take into account if
study of CNM and polyamory. While research has begun to
comparisons are to be made among individuals across cul-
document how polyamory is practiced and why individuals
tures in future studies. For example, the institutional and
pursue such relationships, little research has examined who is
social sanctioning of consensually nonmonogamous rela-
in polyamorous relationships and how they might compare to
tionships would be important to consider when determining
or differ from individuals in more traditional monogamous
the kinds of legal (e.g., marriage) or public commitments
relationships. Despite limitations, this is one of the first studies
(e.g., commitment ceremonies) that polyamorous indivi-
to examine demographic differences between individuals in
duals are able to make with their partners.
polyamorous and monogamous relationships. Our findings
Another limitation of the current study is generalizability.
suggest there are notable similarities between individuals in
Our comparisons may not apply to populations that were not
polyamorous and monogamous relationships but also some
included in our final sample, such as those who are not living in
key demographic differences that need to be considered and
the United States and those who are not connected online.
potentially controlled. We trust this insight and our findings
Among those who are in the United States, our findings may
will be useful when designing future work.
not apply to individuals who are in monogamous relationships
in rural areas with more religious and Republican leanings,
given that political conservatives seem to be underrepresented
in the current survey. The proportion of heterosexuals among REFERENCES
participants in monogamous relationships was also much lower
Balzarini, R. N., Campbell, L., Kohut, T., Holmes, B. M., Lehmiller, J. J.,
than the general population, which suggests that individuals in Harman, J. J., & Atkins, N. (2017). Perceptions of primary and
monogamous relationships in the current sample are more likely secondary relationships in polyamory. PLoS ONE, 12. doi:10.1371/
to be sexual minorities compared to the general population. journal.pone.0177841
Although generalizability might be limited, we can be more Balzarini, R. N., Shumlich, E., Kohut, T., & Campbell, L. (2018). Dimming
assured that differences found in the current study were not the halo around monogamy: Re-assessing stigma surrounding consen-
sually non-monogamous romantic relationships as a function of per-
due to sampling strategy as both samples were recruited sonal relationship. Frontiers in Psychology. Advance online
similarly. This strategy may partially explain why some of publication. doi: 10.3389/fpsyg.2018.00894
our results differ from past studies. For example, the LM study Barker, M., & Langdridge, D. (2010). Whatever happened to non-mono-
used GSS data for its monogamous comparison group, which gamies? Critical reflections on recent research and theory. Sexualities,
involved a weighted, probability-based sample of U.S. house- 13, 748–772. doi:10.1177/1363460710384645
Bauer, G. R., Braimoh, J., Scheim, A. I., & Dharma, C. (2017).
holds and assumed all participants in relationships were mono- Transgender-inclusive measures of sex/gender for population surveys:
gamous. Although their sample was more representative of the Mixed-methods evaluation and recommendations. PLoS ONE, 12,
general population, the differences observed in the LM study e0178043. doi:10.1371/journal.pone.0178043
may reflect the different strategies in recruiting monogamous Berkowitz, J. D. (2007). Beneath the veil of mormonism: Uncovering the
and CNM participants. Results from Rubin and colleagues truth about polygamy in the United States and Canada. The University
of Miami Inter-American Law Review, 38, 615–640.

12
BALZARINI, DHARMA, KOHUT, HOLMES, CAMPBELL, LEHMILLER, AND HARMAN

Broussard, K. A., Warner, R. H., & Pope, A. R. (2017). Too many boxes, or Jonason, P. K., & Balzarini, R. N. (2016). Unweaving the rainbow of
not enough? Preferences for how we ask about gender in cisgender, human sexuality: A review of one-night stands, serious romantic
LGB, and gender-diverse samples. Sex Roles, 78, 1–19. relationships, and the relationship space in between. In K. Aumer
Census Bureau (2016). Quick facts. Retrieved from https://www.census. (Ed.), The psychology of love and hate in intimate relationships (pp.
gov/quickfacts/fact/table/US/PST045216. 13–28). New York City, NY: Springer International Publishing.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences Koenig, H. G., & Büssing, A. (2010). The Duke University Religion Index
(2nd ed.). Hillsdale, NJ: Erlbaum. (DUREL): A five-item measure for use in epidemological studies.
Conley, T. D., & Moors, A. C. (2014). More oxygen please!: How polyamorous Religions, 1, 78–85. doi:10.3390/rel1010078
relationship strategies might oxygenate marriage. Psychological Inquiry, Kuper, L. E., Nussbaum, R., & Mustanski, B. (2012). Exploring the
25, 56–63. doi:10.1080/1047840X.2014.876908 diversity of gender and sexual orientation identities in an online
Conley, T. D., Moors, A. C., Matsick, J. L., & Ziegler, A. (2013). The fewer sample of transgender individuals. Journal of Sex Research, 49,
the merrier?: Assessing stigma surrounding consensually non-mono- 244–254. doi:10.1080/00224499.2011.596954
gamous romantic relationships. Analyses of Social Issues and Public Manley, M. H., Diamond, L. M., & van Anders, S. M. (2015). Polyamory,
Policy, 13, 1–30. doi:10.1111/asap.2013.13.issue-1 monogamy, and sexual fluidity: A longitudinal study of identity and
Conley, T. D., Ziegler, A., Moors, A. C., Matsick, J., & Valentine, B. sexual trajectories. Psychology of Sexual Orientation and Gender
(2012). A critical examination of popular assumptions about the ben- Diversity, 2, 168–180. doi:10.1037/sgd0000098
efits and outcomes of monogamous relationships. Personality and McGee, R. W. (2016). Do Republicans have different views on homosexu-
Social Psychology Review, 2, 124–141. ality than Democrats? An empirical study of opinion in the United
Conron, K. J., Scott, G., Stowell, G. S., & Landers, S. J. (2012). States. SSSN. Retrieved from https://ssrn.com/abstract=2799871
Transgender health in Massachusetts: Results from a household prob- Moore, P. (2016). Young Americans are less wedded to monogamy than
ability sample of adults. American Journal of Public Health, 102, their elders. Retrieved from https://today.yougov.com/news/2016/10/
118–122. doi:10.2105/AJPH.2011.300315 03/young-americans-less-wedded-monogamy/.
Core Team, R. (2016). R: A language and environment for statistical Moors, A. C. (2016). Has the American public’s interest in information related
computing. R foundation for statistical computing, Vienna, Austria. to relationships beyond “the couple” increased over time? Journal of Sex
Retrieved from https://www.R-project.org/. Research, 54, 677–684. doi:10.1080/00224499.2016.1178208
Cox, D. W. (2016) Exploring the health of consensually non-monogamous Moors, A. C., Matsick, J. L., & Schechinger, H. A. (2017). Unique and
individuals: A mixed methods approach. (Unpublished doctoral dis- shared relationship benefits of consensually non-monogamous and
sertation). University of Oklahoma, Norman, OK. Retrieved from monogamous relationships. European Psychologist, 22, 55–71.
https://shareok.org/handle/11244/47030 doi:10.1027/1016-9040/a000278
Cox, D. W. (Personal communication, October 6, 2017). Moors, A. C., Matsick, J. L., Ziegler, A., Rubin, J. D., & Conley, T. D.
Cox, D. W., Fleckenstein, J., & Bergstrand, C. R. (2013). “What do polys (2013). Stigma toward individuals engaged in consensual nonmono-
want?: An overview of the 2012 loving more survey.” Loving More gamy: Robust and worthy of additional research. Analyses of Social
Magazine. Retrieved from http://www.lovemore.com/polyamorousam Issues and Public Policy, 13, 52–69. doi:10.1111/asap.2013.13.issue-1
ory-articles/2012-lovingmore-polyamory-survey/. 1–19. Musumeci, A. (2017, April 25). Couple wants to divorce each other to marry
Duvall-Antonacopoulos, K., MacRae, L., & Paetsch, J. J. (2005). An live-in girlfriend. New York Post. Retrieved from http://nypost.com/2017/
international review of polygamy: Legal and policy implications for 04/25/couple-wants-to-divorce-each-other-to-marry-live-in-girlfriend/
Canada. Polygamy in Canada: Legal and Social Implications for Noël, M. J. (2006). Progressive polyamory: Considering issues of diversity.
Women and Children. Sexualities, 9, 602–620. doi:10.1177/1363460706070003
Eastwick, P. W. (2013). Cultural influences on attraction. In J. A. Simpson Patefield, W. M. (1981). Algorithm AS 159: An efficient method of gen-
& L. Campbell (Eds.), Handbook of close relationships (pp. 161–182). erating random r × c tables with given row and column totals. Applied
New York, NY: Oxford University Press. Statistics, 30, 91–97. doi:10.2307/2346669
Faul, F., Erdfelder, E., Buchner, A., & Lang, A. G. (2009). Statistical power Plummer, K. (1995). Telling sexual stories: Power, change and social
analyses using G* Power 3.1: Tests for correlation and regression worlds. London: Routledge.
analyses. Behavior Research Methods, 41, 1149–1160. doi:10.3758/ Ritchie, A. (2010). Discursive constructions of polyamory in mono-norma-
BRM.41.4.1149 tive media culture. In M. Barker & D. Langdridge (Eds.),
Finkel, E. J., Hui, C. M., Carswell, K. L., & Larson, G. M. (2014). Understanding Non-Monogamies (pp. 46–54). New York, NY:
The suffocation of marriage: Climbing Mount Maslow without Routledge.
enough oxygen. Psychological Inquiry, 25, 1–41. doi:10.1080/ Rothman, K. J. (1990). No adjustments are needed for multiple compar-
1047840X.2014.863723 isons. Epidemiology, 1, 43–46. doi:10.1097/00001648-199001000-
Frank, K., & DeLamater, J. (2010). Deconstructing monogamy: Boundaries, 00010
identities, and fluidities across relationships. In M. Barker & D. Rubin, J. D., Moors, A. C., Matsick, J. L., Ziegler, A., & Conley, T. D.
Langdridge (Eds.), Understanding Non-Monogamies (pp. 9–22). New (2014). On the margins: Considering diversity among consensually
York, NY: Routledge. non-monogamous relationships. Journal Für Psychologie, 22, 1–23.
Freiheit, S. R., Sonstegard, K., Schmitt, A., & Vye, C. (2006). Religiosity Sheff, E. (2005). Polyamorous women, sexual subjectivity, and power.
and spirituality: A psychometric evaluation of the Santa Clara strength Journal of Contemporary Ethnography, 34, 251–283. doi:10.1177/
of religious faith questionnaire. Pastoral Psychology, 55, 27–33. 0891241604274263
doi:10.1007/s11089-006-0029-y Sheff, E., & Hammers, C. (2011). The privilege of perversities: Race, class
Giddens, A. (1992). The transformation of intimacy. Cambridge, England: and education among polyamorists and kinksters. Psychology &
Polity Press. Sexuality, 2, 198–223. doi:10.1080/19419899.2010.537674
Haupert, M., Gesselman, A., Moors, A., Fisher, H., & Garcia, J. (2017). Smith, T. W., Marsden, P. V., Hout, M., & Kim, J. (1972–2012). General
Prevalence of experiences with consensual non-monogamous relation- Social Surveys, 1972–2012 [Machine-readable data file]. Storrs, CT:
ships: Findings from two nationally representative samples of single Roper Center for Public Opinion Research, University of Connecticut
Americans. Journal of Sex & Marital Therapy, 43, 424–440. [distributor].
Jenks, R. J. (1998). Swinging: A review of literature. Archives of Sexual Wheeler, S. C. (2011). Poly-tically incorrect: Women negotiating identity,
Behavior, 27, 507–521. doi:10.1023/A:1018708730945 status, and power in polyamorous relationships. (Unpublished mas-
Jenks, R. J. (2014). An on-line survey comparing swingers and polyamor- ter’s thesis). San Diego State University, San Diego, CA. Retrieved
ists. Electronic Journal of Human Sexuality 17, Retrieved from http:// from http://sdsu-dspace.calstate.edu/bitstream/handle/10211.10/1413/
www.ejhs.org/volume17/swing.html Wheeler_Sarah.pdf?sequence=1

13
MONOGAMOUS AND POLYAMOROUS RELATIONSHIPS

World Heritage Encyclopedia. (2016). Terminology within polyamory. Polyamorous (N = 3,517 Participants Who Indicated
Retrieved from http://community.worldheritage.org/articles/ Their Place of Residence)
Terminology_within_polyamory
Ziegler, A., Matsick, J. L., Moors, A. C., Rubin, J. D., & Conley, T. D.
(2014). Does monogamy harm women? Deconstructing monogamy ● United States (n = 2,428; 69%)
with a feminist lens. Journal Für Psychologie, 22, 1–18. ● Canada (n = 335; 9.5%)
● United Kingdom (n = 177; 5%)
● Europe, excluding United Kingdom (n = 356; 10.1%)
APPENDIX ● Australia and New Zealand (n = 119; 3.4%)
● South America (n = 52, 1.5%)
GEOGRAPHIC ORIGIN OF INCLUDED AND ● Mexico and Central America (n = 26; 0.7%)
EXCLUDED PARTICIPANTS ● Asia (n = 12; 0.3%)
● Africa (n = 7; 0.2%)
In the current study, polyamorous and monogamous
● US Territories (n = 3; 0.1%)
participants were recruited from all over the world.
● Middle East (n = 2; 0.1%)
However, for analytic purposes, we included only
those who were presently living in the United States
to ease interpretation for certain questions (e.g., we only Monogamous (N = 1,110 Participants Who Indicated
inquired about American political affiliations and Their Place of Residence)
income in U.S. dollars). Those who currently lived in
other countries but were U.S. citizens were included ● United States (n = 539; 48.6%)
(e.g., those who wrote “currently in Bolivia but primar- ● Canada (n = 486; 43.8%)
ily live in the U.S.”). In other publications based on this ● United Kingdom (n = 16; 1.4%)
data set that were not concerned with demographic ● Europe, excluding United Kingdom (n = 41; 3.7%)
differences, all participants were retained for analyses ● Australia and New Zealand (n = 15; 1.4%)
regardless of geographic location (Balzarini, et. al., ● South America (n = 4, 0.4%)
2017). The breakdown of participants based on country ● Asia (n = 6; 0.5%)
or world region is as follows: ● Africa (n = 3; 0.3%)

14

You might also like