G Model

SOCSCI-1351; No. of Pages 12 ARTICLE IN PRESS

The Social Science Journal xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

The Social Science Journal

journal homepage: www.elsevier.com/locate/soscij

Global liberalization on homosexuality: Explaining the

African gap
Nicholas O. Alozie ∗ , Kathy Thomas, Patience Akpan-Obong
Faculty of Social Science, College of Integrative Sciences and Arts, Arizona State University-Poly Campus,
7271 E. Sonoran Arroyo Mall – MC 2780, Mesa, AZ 85212-6415, USA

a r t i c l e i n f o a b s t r a c t

Article history: We are trapped in two divergent worlds when it comes to global views on homosexuality.
Received 28 February 2016 There is the liberal world epitomized by Spain and other nations, where homosexuality
Received in revised form 26 August 2016 is increasingly accepted; gays and lesbians are claiming their human rights; and laws
Accepted 2 September 2016
are changing to codify that transformation. The second is the extremely anti-gay world
Available online xxx
symbolized by Africa, the Middle East and parts of Asia, where attitudes are favorable to
criminalization. This research explains the “African Gap” in attitudes toward homosexuality
Keywords:
in a comparative analysis of six African nations and Argentina and Canada, South and North
Africa
America’s most liberal nations on gay rights. Using Pew’s 2015 Spring Global Attitudes Sur-
Argentina
Canada vey data, we find that the major variables have essentially similar effects on opinion in any
Homosexuality context. Africa’s distinction is explained by its comparatively higher levels of factors such
Gay rights as religion, morality dogma, and low socioeconomic status that generally retard support
Support homosexuality for homosexuality, at the same time of lower levels of factors such as education, urban-
LGBTI Africa ization, and personal liberty that increase gay support. Africa’s extreme anti-gay outlook
Anti-gay is mutable. Two social forces will facilitate this softening: expansion of liberalizing agents
such as education and urbanization, and repositioning away from “traditionalism” toward
modernism emphasizing individualism, civil rights, and personal liberties.
© 2016 Western Social Science Association. Published by Elsevier Inc. All rights reserved.

1. Introduction
Following its 39-nation, 2013 Spring Global Attitudes
Survey (SGAS), Pew reported that there is a global divide
on homosexuality (Pew, 2013a). While Pew’s account was
not surprising given anti-gay disputation often carried in
the global media (e.g., BBC, 2011; Michaelson, 2014) and
decades of public opinion tracking intimating a gulf (Smith,
Son, & Kim, 2014; Yang, 1997), it crystallized both the facts
and extent of the global split by incorporating nations not
previously surveyed. As data from the survey mapped in
Fig. 1 evidence, while levels of liberalization vary within
∗ Corresponding author.
regions (see also Štulhofer & Rimac, 2009), that global
gulf is largely tracking two worlds. One is the most lib-
eral epitomized by Spain, Germany, Australia, Philippines,
Canada, Argentina, and Russia, where homosexuality is rel-
atively highly accepted; gays and lesbians are emerging
from the “closet” and claiming their human rights, par-
taking in public life; and laws are evolving to codify those
shifting realities. One of these budding laws was the U.S.
Supreme Court’s historic ruling on June 26, 2015 legal-
izing same-sex marriage nationwide (Obergefell et al. v.
Hodges, 2015). While not analogous to policymaking by
Congress directly accountable to the people, and in fact
producing a backlash of internal demand protests where
several states have recently passed laws that restrict gay
rights, it is federal law. Earlier, Canada, 2005; Spain, 2005;

http://dx.doi.org/10.1016/j.soscij.2016.09.001
0362-3319/© 2016 Western Social Science Association. Published by Elsevier Inc. All rights reserved.

Please cite this article in press as: Alozie, N. O., et al. Global liberalization on homosexuality: Explaining the African gap.
The Social Science Journal (2016), http://dx.doi.org/10.1016/j.soscij.2016.09.001
G Model
SOCSCI-1351; No. of Pages 12 ARTICLE IN PRESS
2 N.O. Alozie et al. / The Social Science Journal xxx (2016) xxx–xxx
Fig. 1. Should homosexuality be accepted by society?
Source: Calculated from Pew’s 2013 SGAS data.
Belgium, 2003; and the Netherlands, 2001, had all legal-
ized same-sex marriage (Hogg, 2006). In all, “nearly two
dozen countries currently have national laws allowing gays
and lesbians to marry, mostly in Europe and the Americas”
(Pew, 2015).
The second is the extremely anti-gay world epito-
mized more vividly by Sub-Saharan Africa, Middle East,
and some Asian nations. Of course, there are also what can
be considered middle-ground nations such as Italy, Greece,
Poland, Israel, Japan, South Korea, Venezuela, and Bolivia.
In much of Sub-Saharan Africa, opposition to homosexual-
ity is to all basic civil rights (including the right to life) and
public debates take the decidedly violent and politicized
bent of executing and imprisoning homosexuals (Cheney,
2012; Epia, 2014; M’Baye, 2013; Smith, 2014; Tamale,
2014). Nigeria, Uganda, Kenya, Gambia, Liberia, Burundi,
South Sudan, Tanzania, Mauritania, and Zimbabwe have
all enacted statutes criminalizing homosexuality. Not only
are gay rights more limited in Africa than in any other
region (Smith, 2014), but homosexuality is actually pro-
scribed in at least 38 African nations (Jolly, 2008), although
“state homophobia” in Africa is diverse (Ireland, 2013, p.
59).1 Here is how Epia (2014) justifies Nigeria’s sensa-
tional anti-gay law: “the massive public approval of the
bill by Nigerians is highly disproportionate to the minor-
ity who disapprove. No doubt, this is a demonstration of
democracy where the majority has its way and the minor-
ity had its say.” Whenever rebuke of their anti-gay posture
starts gaining traction, most Africans seek refuge under this
1
These countries include: Algeria, Angola, Botswana, Burundi,
Cameroon, Comoros, Egypt, Eritrea, Ethiopia, Gambia, Ghana, Guinea,
Kenya, Liberia, Libya, Malawi, Mauritania, Mauritius, Morocco, Namibia,
Nigeria, Senegal, Sierra Leone, Seychelles, Somalia, South Sudan, Sudan,
Swaziland, Tanzania, Tunisia, Togo, Uganda, Zambia, and Zimbabwe.
Please cite this article in press as: Alozie, N. O., et al. Global liberalization on homosexuality: Explaining the African gap.
The Social Science Journal (2016), http://dx.doi.org/10.1016/j.soscij.2016.09.001
majoritarian dogma, as if conscience is subject to majority
rule (Abah, 2016).
Homosexuality may well be highly contested across
the world, but Africa is particularly deeply uncomfortable
with it (e.g., Dunton & Palmberg, 1996, p. 5; Epia, 2014;
Pew, 2013a; Smith et al., 2014; Ward, 2002). The per-
vasive legalization of homophobia in the region and the
rising tide of anti-gay sentiment have become so violent
that the International Lesbian, Gay, Bisexual, Trans and
Intersex Association (ILGA)2 and the broader civil rights
community have raised alarm (BBC, 2013; Kutsch, 2013;
NYT, 2014), even if these laws are not enforced system-
atically across countries (Pflanz, 2014). Homosexuality is
one of those “hush–hush” taboos that Africans would rather
avoid, although it is not a new phenomenon in the region
(Epprecht, 2008, 2013; McNamara, 2014; Murray & Roscoe,
2001). While the American or Western European might
avoid the subject of homosexuality over the negative con-
sequences and censure of being “politically incorrect” given
the legal and social legitimacy it has attained (e.g., Button,
Wald, & Rienzo, 1999; Kane, 2007; Pan, Meng, & Zhou,
2010; Tushnet, 1992; Wald, Button, & Rienzo, 1996), the
African would steer clear mostly as a social, cultural, and
religious taboo (Epia, 2014).3 Even in South Africa, sup-
posedly at the frontier of the battle for gay rights, mass
approval of the lifestyle is minimal (Dunton & Palmberg,
1996; Human Rights Watch, 2011).
2
Founded in 1978 and headquartered in Brussels, Belgium, ILGA is a
worldwide federation of 1100 member organizations from 110 countries
campaigning for lesbian, gay, bisexual, trans, and intersex rights (see,
ilga.org).
3
This hazard of potential sanctions is not limited to individuals. It has
been extended to what some Western countries have labeled as “anti-gay
countries” including, Nigeria and Uganda (BBC News, 2011; see also Epia,
2014).
G Model
SOCSCI-1351; No. of Pages 12 ARTICLE IN PRESS
N.O. Alozie et al. / The Social Science Journal xxx (2016) xxx–xxx
Africa may be the “most homophobic continent on
earth” (Smith, 2014), but it is not the only region with
extreme anti-gay posture: Iran, Saudi Arabia, and Yemen in
the Middle East have the death penalty for homosexuality
(BBC, 2014). Africa attracts attention because of recent state
anti-gay activism that has manifested in harsh anti-gay
laws and enabled environments where violence is routinely
perpetrated against gays with impunity (Human Rights
Watch, 2011). With the long list of embedded interests
vying for supremacy (defiant LGBTI groups with Western
governments and international civil rights organizations
protecting and urging them forward on one side, and
local politicians and media, conservative U.S. evangelicals,
and ordinary fanatics poised to impose street-level justice,
standing on the opposite side and ensuring they do not
succeed), the battle lines are drawn. Africa has become the
“symbolic ground-zero”, the last stand for the battle on gay
rights that none of the gladiators is willing, or can afford,
to lose (Smith, 2014).
This research focuses on a question of profound schol-
arly interest, one that has been a subject of considerable
debate and speculation: what accounts for the extreme
prevalence of anti-gay attitudes in Africa compared to the
more liberal regions? The study addresses this question
for North and South America. We develop hypotheses on
the correlates of micro-level position on homosexuality
and then undertake comparative analyses of those factors
in Africa, Argentina, and Canada, the latter two countries
being the most liberal on homosexuality in South and North
America respectfully. Homosexuality is not well-studied in
Africa, especially as it pertains to establishing the causal
factors associated with position (Ireland, 2013, p. 57). This
may be due, perhaps, to the dearth of support in the region
(Pew, 2013a). It could also emanate from the inherent risk
in engaging such a forbidden subject (Amory, 1997). What-
ever the explanation, this is a significant void at a time of
budding interest on the interaction of public opinion and
development of laws on the subject in the region (Epia,
2014) and elsewhere (Witt & McCorkle, 1997). The SGAS
data which we analyze is the most extensive inter-country
micro-level probability data collected in Africa so far con-
cerning society’s acceptance of homosexuality. Even then,
our study still suffers from data limitations as paucity of
support precludes maximization of design features tar-
geting position within countries. Moreover, Pew’s survey
covers only the six Sub-Saharan countries we examine,
inviting caution with generalization of our findings. How-
ever, we are still able to pool the data estimating causal
effects.
2. The literature: mass attitudes toward
homosexuality
Although significant strides are being made to extend
the research (e.g., Adamczyk & Pitt, 2009; Smith et al.,
2014), much of what we know about factors that shape atti-
tudes toward homosexuality derives from the avalanche
of research that has amassed on Western industrial-
ized countries. As Adamczyk and Pitt (2009, p. 338)
have observed, “very little research has been done to
explain variation in attitudes toward homosexuality in
Please cite this article in press as: Alozie, N. O., et al. Global liberalization on homosexuality: Explaining the African gap.
The Social Science Journal (2016), http://dx.doi.org/10.1016/j.soscij.2016.09.001
3
non-Western nations”. Two research streams can be dis-
cerned from this extant, largely Western literature. The first
strand is the literature that traces mass attitudes toward
homosexuality. From the tracking, we surmise that both
opinions and attitudes on all aspects of the homosexuality
question have continued to soften over time, as more and
more people report willingness to extend rights (Andersen
& Fetner, 2008; Avery et al., 2007; Brewer, 2003; Lax &
Phillips, 2009; Linneman, 2003; Loftus, 2001; Pew, 2013a;
Scott, 1998; Smith et al., 2014; Yang, 1997). Notably, the
same tracking is discovering that most of the developing
world, especially the countries of Africa and parts of Asia
have not been too enthusiastic about jumping on the liber-
alization bandwagon (Pew, 2013a; Smith et al., 2014), not
even with threats from donor countries to tie foreign aid to
liberalization (BBC, 2011; Gunkel, 2013; Smith, 2014).
The second research genre interrogates factors asso-
ciated with mass attitudes toward homosexuality. While
robust, the narrow focus of much of this research, almost
exclusively on advanced countries, as Adamczyk and
Pitt (2009) noted, limits its utility to understanding the
dynamics of mass attitudes toward homosexuality in non-
Western contexts. Adamczyk and Pitt make the additional
crucial observation in particular, that in examining the
relevance of religion in the formation of attitudes on
homosexuality, the lack of sizeable Muslim populations
has angled much of this Western literature toward overly
emphasizing the supposed overwhelming effect of con-
servative Christian principles and values, to the virtual
exclusion of the weight of other religions and faiths. In real-
ity, as Adamczyk and Pitt demonstrate, there are strong
reasons to expect Muslims to have more conservative
social values than some Christian denominations gener-
ally (Yuchtman-Yaar & Alkalay, 2007) and about sexual
morality in particular (Finke & Adamczyk, 2008). Thus, the
breadth of this literature, while serving as an important
roadmap, is only tangential until more concrete research
emerges on non-Western, especially developing, contexts.
Altogether, two broad themes can be deciphered to guide
our comparative investigation: (1) religious conservatism;
and (2) social liberalization attached to such agents as
education, ideology, socioeconomic status, urbanism, and
disposition to civil rights and individual freedoms.
2.1. Religious conservatism
Religion has a long tradition as a determinant of both
secular human behavior and group solidarity (Driskell,
Embry, & Lyon, 2008, p. 295; Durkheim, 1912; Weber,
1958). Perhaps more than any singular factor, religion has
been linked to opposition to homosexuality (Adamczyk
& Pitt, 2009; Burdette, Ellison, & Hill, 2005; Gaines &
Garand, 2010; Merino, 2013; Olson, Cadge, & Harrison,
2006; Sherkat, Powell-Williams, Maddox, & de Veries,
2011; Štulhofer & Rimac, 2009; Wilcox & Jelen, 1990; Witt
& McCorkle, 1997). In the USA, “conservative Protestants
are typically seen as having the least accepting attitudes”
(see literature covered in Adamczyk and Pitt, 2009, p.
339). Although conservative Judeo-Christian denomina-
tions have long been the cornerstone of opposition in
the West (both organized and unorganized), the resilient
G Model
SOCSCI-1351; No. of Pages 12 ARTICLE IN PRESS
4 N.O. Alozie et al. / The Social Science Journal xxx (2016) xxx–xxx
negative effect of religion is not the exclusive domain
of Christendom (Adamczyk & Pitt, 2009; Pew, 2013a,
and citations). The fact is that no major world religion
endorses homosexuality or codifies it as a measure of
“orthodox” human behavior. Instead, as Adamczyk and Pitt
(2009, p. 339) have noted citing Yip (2005), most reli-
gions “categorize behaviors associated with homosexuality
as ‘unnatural’, ‘ungodly’, ‘impure’ and subject to ‘eter-
nal damnation’. Ultimately, this notion of an anathema
becomes the foundation of conservative faith opposi-
tion. While particular faiths may differ in the intensity
of their opposition (Reimer & Park, 2001), many inter-
pret their opposition as a routine practice of their religion.
Accordingly, religion should impact attitude formation
on homosexuality in Africa given the centrality of reli-
gious beliefs in people’s lives on the continent (Pew,
2010).
2.2. Social liberalization
The second major locus of explanation is what can
be grouped into the common sticker of “social liberaliza-
tion and progressivism”. Socially liberal individuals tend
to be more tolerant than conservatives (Pew, 2014) as
well as think more progressively (Freeman, 2011). While
the factors associated with social liberalization such as
education and urbanism are often presented as disparate
variables in research, perhaps because their liberalizing
effects derive from different theoretical foundations, they
actually speak to an underlying factor: disposition to
accepting the new and unfamiliar. Education is said to
advance liberalism through provision of information, bet-
ter information, and enlightenment (Weil, 1985), what in
the decision-making literature is called “bounded ratio-
nality” (Williamson, 1981). Urbanism is said to expose
the individual to heterogeneity, higher and more com-
plex levels of division of labor, and more contact with and
reliance on strangers (Claude, 1995, 1975; Wirth, 1938),
thus advancing liberalism. Age has been linked to liberal-
ism in three separate ways (Schwadel, 2011). The first is
an “age effect” that posits that people get more conserva-
tive as they age (Goerres, 2008; New American Foundation,
2008). The second is the “cohort effect” that holds that peo-
ple raised in different generations exhibit different liberal
dispositions (Inglehart, 1977; Pampel, 2011; Schatz, 2002).
The third is the “periodic effect”, which describes how eras
of socialization such as the Great Depression instills differ-
ent behavioral traits (Andersen & Fetner, 2008; Braungart &
Braungart, 1986; Treas, 2002). Also, higher socioeconomic
status is likely to increase social density and provide access
to more group memberships (Himmelstein & McRae, 1988;
Jetten, Haslam, & Barlow, 2013).
Overall, literature suggests a profile of agents of liber-
alization enhancing support for homosexuality as follows:
more education, younger age, female gender, and urban-
ism (Adamczyk & Pitt, 2009; Pew, 2013a; Smith et al.,
2014, p. 13). The gender effect comes from the differing
socialization of the sexes. It could also emanate from the
disparate effects of the relevant factors on men and women
(Manganaro & Alozie, 2015). The later possibility justifies
disaggregation of the analyses by gender to probe potential
Please cite this article in press as: Alozie, N. O., et al. Global liberalization on homosexuality: Explaining the African gap.
The Social Science Journal (2016), http://dx.doi.org/10.1016/j.soscij.2016.09.001
gender interactions. Gender interaction seems likely given
the disparities in education, employment, and marital sta-
tus for Africa.
2.3. Explaining the African gap
Public opinion data reveal that the nations “softening”
on homosexuality, even in Europe and the Americas, have
not always been that receptive. As of 1789, the farthest
date for which data are available, homosexuality was illegal
in 126 countries (BBC, 2014). Evidently, the softening has
occurred over time (Andersen & Fetner, 2008; Yang, 1997),
nudged by a complex mixture of changing social condi-
tions (Avery et al., 2007; Brewer, 2003). Nations do not
begin with the acceptance or inheritance of homosexuality
as part of popular culture. Within Africa, the explanation
of the origins of homophobia and handling of homosexu-
ality varies markedly (Ireland, 2013; M’Baye, 2013), with
nations having contradictory laws (BBC, 2014), and oth-
ers not legislating on it all (Ireland, 2013). Therefore, the
claim by some Africans that homosexuality is antithetical
to their cultures and indigenous traditions and that that
alone explains the extreme hostility manifest in the region
(Epia, 2014) is not persuasive. If that were the singular
explanation, no major nation would tolerate it.
Some, contrary to evidence that homosexuality is not
un-African (Epprecht, 2008; M’Baye, 2013; Murray &
Roscoe, 2001; Smith, 2014), have even floated the narra-
tive that phantom colonialists “imported” homosexuality
into Africa as part of some conspiracy to “poison”, even
destroy, the African way of life (Evaristo, 2014; Smith,
2014). Yet, as Smith (2014) notes, buttressed by data pub-
lished by the BBC (2014) showing dates of criminalization
of homosexuality, “in many cases, the very laws being
imposed so zealously were introduced by the European
empires that carved up and plundered Africa . . . (and were)
enforced legally by colonial administrators and ideolog-
ically by Christian missionaries” (see also M’Baye, 2013;
Mama, 2007). Smith further notes that “Africans have not
only internalized homophobia that came as part of colo-
nial oppression and now proclaim it as their own authentic
African tradition,” but Western evangelicals are actually
being blamed for “whipping up homophobia”, the same
“imported evil” many Africans now proclaim as part of
indigenous political platforms that must be eradicated.
There must be a more plausible explanation
than Africa’s ethnocentric, dismissive, anti-culture/
indigenousness rendering. Part of that explanation is polit-
ical expediency, as homosexuals “have been constructed
as scapegoats” (M’Baye, 2013, p. 109). Gay rights have
been woven into Africa’s contemporary narrative and
social construction of neo-colonialism (Gunkel, 2013) and
exploited by desperate politicians in search of a populist
distraction from poorly performing and failing govern-
ments (Coly, 2013; Ireland, 2013; Tamale, 2013). A more
compelling explanation is found in classical sociology.
Durkheim’s (1893) treatise on the development of societies
is central. Most African societies are still bastions of what
Durkheim describes as “mechanical solidarity.” That is,
they are largely traditional enclaves held together by tribal,
kinship, cultural, and religious dictates that have little
G Model
SOCSCI-1351; No. of Pages 12 ARTICLE IN PRESS
N.O. Alozie et al. / The Social Science Journal xxx (2016) xxx–xxx
room for complex division of labor and social complexity
found in Western societies. These societies are rigid, con-
servative, and are not predisposed to change. According
to Durkheim, all societies tend toward heterogeneity and
“organic solidarity,” characterized by emphasis on division
of labor, human rights, human dignity, social justice,
secularity, and personal freedom. These societies are more
progressive and amenable to uncertainty and risk-taking
in embracing the unfamiliar. Thus, as African societies
move toward organic solidarity, where gay rights are seen
in light of human rights, secularity, and social justice (as
opposed to culture and religion), opinions will soften.
A contemporary rendition of Durkheim’s treatise as
it applies to personal “emancipation” and liberalism
on homosexuality was offered by Adamczyk and Pitt
(2009), envisaging something of a double whammy for
Africa. Drawing from Inglehart, Norris, and Welzel (2002),
Inglehart and Baker (2000), Inglehart (1977) and Adamczyk
and Pitt (2009, pp. 339–340) distinguish between two kinds
of societal contexts. The first are societies undergoing a
process of industrialization and modernization where,
attitudes and values shift away from concerns about
physical and economic security to world-views that
are increasingly rational, tolerant, and trusting. . .. With
this new orientation, people can more easily afford to
tolerate new ideas and non-normative behaviors like
homosexuality. [The second context are nations] reg-
ularly faced with political and economic uncertainty
and insecurity, [where] people are more likely to sup-
port values and norms that emphasize the familiar. As
a result people in nations that are characterized by a
strong survivalist orientation may be less tolerant of
nontraditional ideas and lifestyles.
Most African states are still beset by economic, social,
and political crises that preclude industrialization and
modernization, fostering environments of uncertainty and
insecurity. Predictably, African peoples are relegated to
“survival” as opposed to “self-expressive” orientation,
although movement to the latter may not necessar-
ily obviate religion’s negative effect (Adamczyk & Pitt,
2009). Comparing African nations to Argentina and Canada,
this model suggests that citizens of the latter nations
have reached higher levels of self-expressive orientation,
while Africans lack industrialization/modernization and
face political and economic uncertainly. It predicts less lib-
eralism by Africans than either Argentines or Canadians. It
further predicts more liberalism for Canadians than Argen-
tines.
3. Data and methodology
Our comparative inquiry is based on the SGAS data gen-
erated from survey conducted between March and April,
2013 (Pew, 2013b). The surveys relied on national proba-
bility samples of adults 18 years or older. Six Sub-Saharan
African countries (Ghana, Kenya, Nigeria, Senegal, South
Africa, and Uganda) were included. These six countries
do not represent all of Sub-Saharan Africa, but they are
some of the most significant countries of the region. In
all, 5043 respondents were interviewed in Africa, 819 in
Please cite this article in press as: Alozie, N. O., et al. Global liberalization on homosexuality: Explaining the African gap.
The Social Science Journal (2016), http://dx.doi.org/10.1016/j.soscij.2016.09.001
5
Argentina, and 701 in Canada.4 Pooling the survey data
for these African countries is appropriate since the same
multi-stage cluster sampling and stratification techniques
were employed throughout to generate the country sam-
ples. Moreover, while Pew used the random digit dialing
technique in some countries, face-to-face interviews were
utilized in these six countries. Our research design com-
pares factors associated with support for homosexuality
in these six African countries (8.7%) to the two most lib-
eral countries in North (Canada, 79.5%) and South America
(Argentina, 73.0%), with a goal of explaining the African
gap. The questions related to the variables of interest were
asked in both countries.
3.1. Dependent variable
3.1.1. Measuring support for homosexuality
As Yang (1997) has noted (see also Smith et al., 2014,
p. 2; Takács & Szalma, 2013), the concept of support for
homosexuality is porous. To ask whether homosexuality is
moral is not analogous to asking whether it should be legal-
ized. Yang’s (1997) tracking of public responses to opinion
polls over time clearly suggests that the public does not
think about them unidimensionally. The item we use for
measuring support was chosen for two reasons. The first is
its simplicity in lacking the normative ambiguity endemic
in many questions that measure position on homosexual-
ity. The second is its relevance to securing civil rights to the
gay community which, after all, is at the core of gays’ quest
for “complete and total acceptance”. Question #27 in Pew’s
survey asked: “And which one of these comes closer to your
opinion, number 1 or number 2? Number 1: Homosexuality
should be accepted by society. Number 2: Homosexuality
should not be accepted by society.” We coded option 1 as
“1” for support, and 2 as “0” for lack of support.
3.2. Independent variables
3.2.1. Religion
Since none of the major world religions endorses
homosexuality, we did not sort respondents into reli-
gious groupings. Indeed, Christians and Muslims have
acted together in Nigeria and Kenya to oppose homosex-
uality (Epia, 2014). Like Adamczyk and Pitt (2009), we
relied on question #178 that taps the importance of reli-
gion: “How important is religion in your life?” We coded
very and somewhat important as “1”, while not too and
not at all important were coded as “0”.5 Education. Edu-
cation is listed incrementally according to country. We
coded it progressively from 1 to 10, 10 being the high-
est level of attainment. Ideology. Question #84 presents a
list of seven items (using contraceptives; getting a divorce;
4
The country samples are as follows: Ghana 799, men 408, women
391; Kenya 798, men 458, 340; Nigeria 1031, men 517, women 514; Sene-
gal 800, men 358, women 442; South Africa 815, men 408, women 407;
Uganda 800, men 388, women 412; Argentina, 819, men 403, women,
416; Canada, 701, men 349, women, 352.
5
We considered a measure of how often one prayed as a surrogate
for religiosity, but our diagnoses indicated it was largely dormant in our
models. Consequently, we dropped the variable.
G Model
SOCSCI-1351; No. of Pages 12 ARTICLE IN PRESS
6 N.O. Alozie et al. / The Social Science Journal xxx (2016) xxx–xxx
having an abortion; drinking alcohol; married people hav-
ing an affair; gambling; sex between unmarried adults) and
asked the respondent to say whether the item is morally
acceptable, morally unacceptable, or whether the item is
not a moral issue. We coded morally unacceptable as “1”,
indicating conservatism and the other categories as “0” rep-
resenting liberalism. Addition of scores across the seven
items produces a variable with a count ranging from 0
to 7, where respondents scoring 7 will represent extreme
conservatism and those scoring 0 and 1 will represent lib-
eralism.
3.2.2. Socioeconomic status
We measure socioeconomic status using employment
and poverty status. Employment status is coded as “1”
for the unemployed, and “0” for employed. Although the
income variable was included in the survey, we had two
concerns regarding its ability to capture level of consump-
tion. The first is that it was denominated according to each
country’s local currency. We could standardize by trans-
lating to some international value using a currency such
as the dollar, but we might not capture real consumption
positions with the translation. Moreover, personal income
data are highly unreliable in many developing countries.
Accordingly, we do not use the income variable. We use a
measure of poverty status as follows: Question 182a asked:
Have there been times during the last year when you did
not have enough money to buy food for your family? We
preferred the cleanliness of this question as a measure of
affordability and proceeded with the “poverty” variable
instead of income. The variable is coded “1” for yes, and
“0” for no.
3.2.3. Urbanism
Urbanism is coded “1” for urban resident; and “0” oth-
erwise. Orientation toward Personal Freedoms (Instrumental
Variable). Question 71 asked: “How important is it to
you that people have access to the internet without gov-
ernment censorship?” Responses of very and somewhat
important are coded as “1”, while not too and not impor-
tant at all are coded as “0”. The other variables added to the
analysis are age, gender, and marital status. Age is the actual
age of the respondent. Some literature suggests that the
age effect may be non-linear, in the sense that older adults
raised in more conservative times may be more reluctant to
support homosexuality. This links to the broader theoreti-
cal question of whether the effect of age is a matter of the
life-course or of cohort replacement (Smith et al., 2014).
Accordingly, we added the quadratic term, age.2 Gender is
coded as “1” for women; and “0” for men. Marital status is
coded as “1” if married/cohabiting; and “0” otherwise.
3.3. Analytical methods
We conduct two levels of analyses. The first addresses
the level of support for homosexuality, using descriptive
statistics to establish the African gap. Then, we apply mul-
tivariate analyses to identify correlates of position across
the regions. Since the dependent variable is binary, we use
binary logistic regression to estimate the models. Two lev-
els of multivariate analyses are conducted. The first is a
Please cite this article in press as: Alozie, N. O., et al. Global liberalization on homosexuality: Explaining the African gap.
The Social Science Journal (2016), http://dx.doi.org/10.1016/j.soscij.2016.09.001
macro-level specification for the entire country/regional
samples. The second are disaggregated models exploring
gender interactions. The models estimated are of the gen-
eral form: Ŷ = a + bxn + e, where Ŷ is the probability that
the respondent supports homosexuality, and x is the vec-
tor of independent variables explaining that probability.
Our robust sample sizes generated moderate inter-item
correlations that allowed for efficient estimation of the
models.
4. Empirical findings
The profiles of the 5043 respondents across the six
Sub-Saharan African nations, 819 in Argentina, and 701
in Canada are displayed in Table 1. The profile captures
what can be expected of a stratified random sample in Sub-
Saharan Africa: Older adults of middle age (mean age = 35)
by the region’s life expectancy standard of 56 at birth.
African men are far more likely to be employed, while
women are both poorer and slightly more likely to be mar-
ried. Moreover, there is evidence and clarification of the
gender gap in education. African women are less likely to
have formal education—where they do at all, men hold an
advantage in higher education. A very high proportion of
Africans reported that religion is important in their lives.
The mean scores on the morality quotient for Africa are
moderate (4.5 out of 7, or 64%), with no noticeable gen-
der gap. Finally, African men are far more enthusiastic
about liberalism than the women. The average Argentine
(approximately 43 years old) and Canadian (49 years old)
in their country samples are much older than the African in
the African sample, tended to be more educated, employed,
and married. The African is poorer, more religious, and less
liberal. There are gender gaps on religion between both the
Argentine and Canadian samples with women being more
religious.
4.1. Descriptive findings: support for homosexuality
(establishing the African gap)
Table 2 and Fig. 2 indicate that support for homosex-
uality in Africa is rather sparse. Only 423 (8.7%) of the
final sample of 4906 of African respondents used in the
analysis say homosexuality should be accepted by society.
Among the 2472 men, 223 (9.1%) say homosexuality should
be accepted, while the comparable figure for women is
200 (8.3%) of 2434. Across the six countries, only in South
Africa with a 32% approval level did there emerge what
can be considered a meaningful variation in position. The
approval levels in the other five countries are as fol-
lows: Kenya, 8.0%; Uganda, 4.0%; Ghana, 3.0%; Senegal,
3.0%; and Nigeria, 1.0%.6 These statistics are instructive.
First, they reinforce the notion that homosexuality has
very little popular support in Africa, and explain the
6
These distributions made it difficult to implement our original
research design which called for both country fixed-effects and country-
specific analyses at the level of multivariate specifications. The limitation
allows for the kind of exploratory analysis reported here only by pooling
the available data. With 423 respondents approving, sufficient deviation
emerges for analysis.
G Model
SOCSCI-1351; No. of Pages 12 ARTICLE IN PRESS
N.O. Alozie et al. / The Social Science Journal xxx (2016) xxx–xxx 7

Table 1
Descriptive statistics for variables (weighted means).

Variable Africa Argentina Canada

All Men Women All Men Women All Men Women

Gender – – 49.7% – – 50.8% – – 50.2%

Age 35.4 36.1 34.7 42.9 41.9 44.0 49.2 47.6 50.8
(14.6) (15.2) (14.0) (17.5) (17.8) (17.1) (17.4) (17.6) (17.1)
Unemployed 59.7% 53.2% 66.2% 46.1% 33.5% 58.2% 37.2% 30.4% 44.1%
Married 50.3% 56.7% 57.3% 52.5% 50.4% 54.6% 60.1% 61.9% 58.2%
Educationa 2.7 2.8 2.5 4.9 4.9 4.9 5.58 5.68 5.48
(0.89) (0.86) (0.90) (1.8) (1.8) (1.8) (2.05) (2.1) (1.9)
None 11.8% 8.9% 14.6% – – – – – –
Primary 23.5% 20.3% 26.8% – – – – – –
Secondary 47.1% 49.7% 44.5% – – – – – –
University 17.6% 21.1% 14.1% – – – – – –
Poor 58.2% 56.4% 59.9% 21.1% 19.1% 23.1% 8.7% 6.0% 11.4%
Urban 41.9% 40.8% 43.0% 96.5% 96.5% 96.6% 81.0% 83.4% 78.7%
Religion important 97.0% 96.1% 97.9% 64.5% 56.3% 72.4% 54.1% 46.9% 61.1%
Internet liberalization 65.4% 70.1% 60.7% 81.2% 82.1% 80.3% –b –b –b
Morality Quotient (7) 4.5 4.4 4.5 4.9 4.9 5.0 3.6 3.6 3.6
(2.2) (2.2) (2.2) (2.1) (2.1) (2.1) (2.2) (2.2) (2.2)
N 5043 2537 2506 819 403 416 701 349 352

Note: Figures in parentheses are standard deviations.

a
Education for Africa came in different measurements. We categorized these into four major groups, and coded 1–4, with 4 being the highest level of
attainment.
b
Question not asked in Canada.

Table 2
Should homosexuality be accepted by society?

Response Africa Argentina Canada

All Men Women All Men Women All Men Women

Yes 8.7% 9.1% 8.3% 73.0% 71.2% 74.0% 79.5% 77.9% 80.9%
N = 423 N = 223 N = 200 N = 598 N = 287 N = 308 N = 557 N = 272 N = 285

No 91.3% 90.9% 91.7% 27.0% 28.8% 26.0% 20.5% 22.1% 19.1%

N = 4483 N = 2249 N = 2234 N = 221 N = 116 N = 108 N = 144 N = 77 N = 69
Total (N) 4906 2472 2434 819 403 416 701 349 352

Fig. 2. Should homosexuality be accepted by society?

Source: Calculated from Pew’s 2013 SGAS data.

proliferation of anti-gay legislations in the region. Sec- from the complete absence (or is it silencing?) of voices
ondly, rejection of homosexuality in Africa is not gendered. of dissent to opposition to homosexuality. The persis-
Of course, the question has never been about mass sup- tent puzzle has been on how to explain position on the
port because the paucity of support has been evident issue.

Please cite this article in press as: Alozie, N. O., et al. Global liberalization on homosexuality: Explaining the African gap.
The Social Science Journal (2016), http://dx.doi.org/10.1016/j.soscij.2016.09.001
G Model
SOCSCI-1351; No. of Pages 12 ARTICLE IN PRESS
8 N.O. Alozie et al. / The Social Science Journal xxx (2016) xxx–xxx

We now turn to support for homosexuality in Argentina
and Canada, levels of which establish evidence of the
African gap and support Adamczyk and Pitt’s (2009)
self-expressive formulation. Overall, 73.0% (N = 598) of
Argentines say society should accept homosexuality. This
establishes a response gap of 64.3% with Africa. The per-
centage loadings for Argentine men (71.2%) and women
(74.0%) also establish firm gender gaps with African men
(gap = 62.1%) and women (gap = 65.7%). As would be pre-
dicted by Adamczyk and Pitt’s (2009) self-expressive
orientation formulation, the macro-support figure for
Canada (79.5%) is higher than Argentina’s 73.0%, but con-
siderably higher than Africa’s 8.7%. Also, the figures for
Canadian men (77.9%) and women (80.9%) display gaps
with Argentine men (gap = 6.7%) and women (gap = 6.9%),
respectively, but, again, the size of those gaps pale in
comparison to African men (gap = 68.8%) and women
(gap = 72.6%). These figures not only speak to what Pew
(2013a) calls “The Global Divide on Homosexuality”, but they
support the notion of an African gap.
4.2. Multivariate findings: predicting support for
homosexuality (explaining the African gap)
The results of the multivariate analyses are presented
in Tables 3 and 4. The results displayed in Table 3 are
for entire samples. The gender interaction models are
presented in Table 4. The Model Chi-Squares and per-
centage of cases correctly classified are robust, indicating
that the independent variables are successful in predict-
ing position on homosexuality across all models. Most
of the results presented in Table 3 are consistent with
the theorized effects of the independent variables and
hypotheses. Respondents who reported that religion is
important in their lives are less likely to support homosexu-
ality. This result is robust across all three regions, for Africa
(MLE = −1.324, p < .001), Argentina (MLE = −0.788, p < .001),
and Canada (MLE = −0.806, p < .001). The literature sug-
gests that religion is perhaps the singular most consistent
negative source of global disapproval of homosexuality.
The results reported in Table 3 are consistent with that

Table 3
Binary logit analysis of predictors of support for homosexuality (full samples).

Variable Africa Argentina Canada

MLE Odds ratio MLE Odds ratio MLE Odds ratio

Religion
Religion −1.324*** 0.27 −0.788*** 0.46 −0.806*** 0.45
(6.43) (3.96) (3.72)
Enlightenment
Education 0.137* 1.13 0.172*** 1.19 0.078 1.08
(2.01) (3.25) (1.50)
Ideology
Morality −0.209*** 0.81 −0.062 0.94 −0.126** 0.88
(9.09) (1.41) (2.68)
Socioeconomic status
Unemployed 0.086 – 0.354+ 0.72 −0.249 0.78
(0.74) (1.84) (0.50)
Poverty status −0.303** 1.09 −0.050 0.095 −0.240 0.78
(2.83) (0.24) (0.69)
Exposure to outsiders
Urbanism 0.548*** 1.73 0.184 1.20 0.028 0.91
(4.98) (0.42) (0.12)
Personal freedoms
Internet 0.422*** 1.53 0.588** 1.80 –a –a
(3.27) (2.89)
Demographics
Age 0.037+ 1.04 0.020 1.00 −0.029 0.97
(1.85) (0.72) (0.88)
Age2 0.000 0.000 1.02 0.000 1.0
(1.24) (0.00) (0.00)
Gender 0.022 – 0.182 1.20 0.403* 0.68
(0.20) (1.03) (2.01)
Married −0.007 – 0.068 1.07 −0.385+ 0.68
(0.06) (0.39) (1.81)

Intercept −1.899*** 0.508 2.874***

(3.67) (0.63) (3.20)
Nagelkerke R2 10.0 14.4 11.6
% Predicted 91.4 75.0 79.0
Model 2 223.73*** 85.28*** 53.79***
N 4906 819 701

Figures in parentheses are t-statistics.

a
Question not asked in Canada.
+
p < .10
*
p < .05
**
p < .01*** p < .001 (two-tailed).

Please cite this article in press as: Alozie, N. O., et al. Global liberalization on homosexuality: Explaining the African gap.
The Social Science Journal (2016), http://dx.doi.org/10.1016/j.soscij.2016.09.001
G Model
SOCSCI-1351; No. of Pages 12 ARTICLE IN PRESS
N.O. Alozie et al. / The Social Science Journal xxx (2016) xxx–xxx 9

Table 4
Binary logit analysis of predictors of support for homosexuality (by gender).

Variable Africa Argentina Canada

Men Women Men Women Men Women

Religion
Religion −1.276*** −1.487*** −0.724** −0.839** −0.548+ −1.130**
(4.89) (4.37) (2.79) (2.59) (1.95) (3.05)
Enlightenment
Education 0.109 0.170+ 0.209** 0.131+ 0.137* −0.018
(1.11) (1.75) (2.68) (1.70) (2.02) (0.22)
Ideology
Morality −0.216*** −0.198*** −0.067 −0.064 −0.130* −0150*
(7.20) (6.00) (1.09) (1.00) (2.06) (2.08)
Socioeconomic Status
Unemployed 0.324* −0.119 0.166 0.591* −0.037 −0.505
(1.98) (0.71) (0.56) (2.11) (0.11) (1.48)
Poverty status −0.357* −0.275+ −0.086 0.045 −0.600 −0.141
(2.38) (1.76) (0.29) (0.15) (1.14) (0.29)
Exposure to outsiders
Urbanism 0.610*** 0.507*** −0781 1.187+ −0.220 0.191
(3.96) (3.19) (1.09) (1.95) (0.60) (0.55)
Personal freedoms
Internet 0.452* 0.408* 0.860** 0.290 –a –a
(2.40) (2.28) (2.93) (0.99)
Demographics
Age 0.006 0.060+ 0.015 0.024 −0.077 0.009
(0.21) (1.94) (0.39) (0.59) (1.64) (0.19)
Age2 0.000 −0.001 0.000 −0.001 0.001 0.000
(0.12) (0.11) (0.00) (0.00) (0.00) (0.00)
Married 0.378* −0.239 0.040 −0.107 −0.253 −0.619+
(2.00) (1.48) (0.15) (0.42) (0.86) (1.88)

Intercept −1.481* −2.072** 0.878 0.582 3.637** 3.364*

(2.11) (2.66) (0.74) (0.49) (2.95) (2.41)
Nagelkerke R2 10.7 10.03 15.8 16.2 10.3 17.0
% Predicted 91.0 91.8 73.4 76.0 77.4 82.1
Model 2 123.32*** 110.66*** 46.79*** 48.67*** 24.25** 39.44***
N 2472 2434 403 416 349 352

Figures in parentheses are t-statistics.

a
Question not asked in Canada.
+
p < .10.
*
p < .05.
**
p < .01.
***
p < .001 (two-tailed).

trajectory and, in combination with the data reported for
religion in Table 1, present what appears to be one under-
lying explanation for the African gap. Overall, it is clear that
the negative effect of religion holds across the board. At the
same time, the data in Table 1 indicate that while the per-
centage of Africans reporting that religion is important in
their lives is as high as 97%, and 96.1% and 97.9%, respec-
tively for African men and women, only 64.5% of Argentines
(56.3% for men and 72.4% for women) and 54.1% of Canadi-
ans (46.91% for men and 61.1% for women) make that claim.
Thus, the higher levels of religion in Africa is a significant
distinguishing factor. This means religion’s discount effect
weighs more on Africa.
Education improves support in Africa and Argentina, but
not in Canada. Looking at the education data presented
in Table 1, it is apparent why education in the surveyed
African countries will not do as much as Argentina’s in
liberalization as Argentina’s educational levels are practi-
cally double what they are in Africa. Thus, education is a
positive factor, but its low supply in Africa diminishes its
potency. Orthodox morality is another source of the African
gap, which is not surprising given the strong effect of reli-
gion in the region, the latter being a derivative. Its negative
effect can be seen in Canada (MLE = −0.126, p < .01), but not
in Argentina (MLE = −0.062, n.s.). While morality levels in
Argentina and Africa appear similar (Table 1), its lack of
significance in the multivariate model means that it does
not discount in Argentina. Taken together, the results for
Africa, Argentina, and Canada suggest that morality is more
of an impediment in Africa than it is in either Argentina or
Canada.
Poverty (socioeconomic status) is another notice-
able explanation for the African gap. While its negative
coefficients are not statistically significant in Argentina
and Canada, it is salient in Africa (MLE = −0.303, p < .01).
This result, alone, is sufficient to conclude that poverty’s
negative effect weighs more on Africa than either on
Argentina or Canada. However, that conclusion would not
be sufficient without again calling attention to the poverty
statistics shown in Table 1 that reveal something of a dou-
ble jeopardy on the effect of the factor in Africa. While
58.2% of Africans (56.4% for men and 59.9% for women) are

Please cite this article in press as: Alozie, N. O., et al. Global liberalization on homosexuality: Explaining the African gap.
The Social Science Journal (2016), http://dx.doi.org/10.1016/j.soscij.2016.09.001
G Model
SOCSCI-1351; No. of Pages 12 ARTICLE IN PRESS
10 N.O. Alozie et al. / The Social Science Journal xxx (2016) xxx–xxx
classified as poor, only 21.1% of Argentines (19.1% for men
and 23.1% for women) and 8.7% of Canadians (6.0% for men
and 11.4% for women) are so classified. Thus, the discount
effect of poverty is greatest in Africa also because far more
Africans are in poverty than either Argentines or Canadians.
Urbanization produces a strong positive effect for Africa,
but not for Argentina or Canada. However, perusal of the
urbanization data in Table 1, 41.9% for Africa, 96.5% for
Argentina, and 81.0% for Canada reveals why the positive
urban effect will not be sufficient to close the African gap
vis-à-vis the comparative countries. Simply put, the urban-
ization level in Africa is comparatively low. Finally, the
personal freedom question was not asked in Canada, but
it produced strong positive effects in Africa (MLE = 0.422,
p < .001) and Argentina (MLE = 0.588, p < .01). Still, the lib-
eralism data in Table 1 suggest that Africa will not benefit
as much from this factor, not with its 60.7–70.1% range and
Argentina’s 80.3–82.1%.
The results show both gender and marital effects in
Canada that are not operative in Africa or Argentina, where
Canadian women are more supportive of homosexuality
(MLE = 0.403, p < .05), and their married or cohabiting coun-
terparts are less supportive (MLE = −0.385, p < .10). This
Canadian result is in accord with the research which sug-
gests that women are more supportive of homosexuality,
while married couples are less supportive (Pew, 2013a;
Smith et al., 2014). Finally, except for a weak African effect,
age is not a factor. The results suggest that the major vari-
ables have similar baseline effects across the countries, but
that contextual differences that produce something of a
double-edged sword largely explain the African gap. Essen-
tially, factors such as religion, morality, and poverty that
retard support exist at higher levels, at the same time that
education, urbanism, and personal freedom that increase
liberalization occur at lower levels.
Table 4 displays the results of the gender interaction
specifications. There are nuances in the effects of some
of the variables, although the overall results largely repli-
cate the general patterns noted in Table 3. For religion,
the results show no gendered effects, as men and women
reporting that religion is important in their lives are as
likely to reject homosexuality. Education is a particu-
larly relevant factor for Argentine and Canadian men, but
produces weak effects for African and Argentine women.
The results for morality are replicated with the nuance
that they are not gendered. Finally, the results for poverty
status, urbanism, and liberalism are replicated without
gendered effects, except for the weak effect of urbanism
noted for Argentine women. A nuance observable in Table 4
is the positive effect of unemployment for African men and
Argentine women, which produces a wash when consid-
ered in light of the percentages in Table 1.
5. Summary and implications
As Adamczyk and Pitt (2009, p. 338) observe, buttressed
by both Epia (2014) in Nigeria and Witt and McCorkle
(1997) for the USA, “since public opinion is an impor-
tant factor in developing laws and policies, understanding
the source of these attitudes can shed light on why some
nations have such liberal policies, and others continue to
Please cite this article in press as: Alozie, N. O., et al. Global liberalization on homosexuality: Explaining the African gap.
The Social Science Journal (2016), http://dx.doi.org/10.1016/j.soscij.2016.09.001
punish sexual behaviors like homosexuality”. Our study
does not tackle the matter of the incidence or appropriate-
ness of homosexuality, neither does it burden itself with
the many unsettled issues surrounding the subject matter,
such as origin and differential political responses within
jurisdictions. Rather, we address the compelling question
of what accounts for the extreme prevalence of anti-gay
attitudes in Africa compared to other nations: what we
call the African Gap. Our strategy was to identify factors
that shape position and by comparing those factors across
regions, at least account for the African gap across these
divides. We use the word “Africa” throughout this report
to designate the six African nations studied. While there is
no reason to suppose that our results, as inclusive as they
appear, are outliers, basic cannons still urge caution with
generalization.
We summarize the results of our study as follows: Reli-
gion contributes to explaining the African gap because
of its higher incidence in Africa vis-à-vis Argentina and
Canada. Orthodox morality, partly a derivative of religion,
also contributes to explaining the African gap because of its
higher occurrence in Africa vis-à-vis Argentina and Canada.
Poverty is more prevalent in Africa explaining its lopsided
negative effect. Education, urbanism, and personal freedom
are relevant, but their effects are not maximized because of
their comparatively low levels. Ultimately, our data explain
the African gap with Argentina and Canada as a double-
edged dynamic, where factors that increase support are in
shorter supply, while those that decrease it are in larger
supply.
Admittedly, the insights we provide, no matter how
incisive, is not the whole story as they cannot explain away
the entire African gap. There is an important dimension
about context that this kind of research cannot tap, and the
main implication of the research makes it apparent. The
obvious consequence of our findings is that, like elsewhere,
Africa’s conservative outlook is not immutable. Positions
most likely will soften over time. Two contextual factors
will champion this softening. One is expansion of education
and urbanism. Another will be disposition toward personal
liberties. All are linked to Adamczyk and Pitt’s (2009) self-
expressive formulation emanating from industrialization,
modernization and political and economic security. While,
as they caution, such changes may not necessarily obviate
the strong negative religious effect, its potential to liberal-
ize the African mind generally and moderate positions on
homosexuality in particular, cannot be underestimated.
Acknowledgements
This paper uses data from the 2013 Spring Pew Global
Attitudes Survey. The views expressed in the paper are
those of the authors only and have no connection to Pew,
its agents, or institutions. Earlier versions were presented
at the 58th Annual Meeting of Western Social Science
Association, April 13–16, 2016, Reno, Nevada; the 19th
Annual Meeting of the American Association of Behav-
ioral and Social Sciences, February 1–2, 2016, Las Vegas,
Nevada; the 96th Annual Meeting of the Southwestern
Social Science Association, March 23–26, 2016, Las Vegas,
Nevada; and the 57th Annual Meeting of Western Social
G Model
SOCSCI-1351; No. of Pages 12 ARTICLE IN PRESS
N.O. Alozie et al. / The Social Science Journal xxx (2016) xxx–xxx
Science Association, April 8–11, 2015, Portland, Oregon.
We thank the anonymous reviewers and conference par-
ticipants whose observations form part of this final draft.
References
Abah, S. (2016). Conscience, not subject to majority rule. The Guardian,
(June 3)
Adamczyk, A., & Pitt, C. (2009). Shaping attitudes about homosexuality:
The role of religion and cultural context. Social Science Research, 38,
338–351.
Amory, D. P. (1997). ’Homosexuality’ in Africa: Issues and debates. Issues:
A Journal of Opinion, 25(1), 5–10.
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.
Avery, A., Chase, J., Johansson, L., Litvak, S., Montero, D., & Wydra, M.
(2007). America’s changing attitudes toward homosexuality, civil
unions, and same-gender marriage: 1977–2004. Social Work, 52,
71–79.
BBC News. (2014). Where is it legal to be gay? London: BBC News. February
10.
BBC News. (2013). Amnesty International condemns “homophobia” in Africa.
London: BBC News. June 25.
BBC News. (2011). Cameron threat to dock some aid to anti-gay nations.
London: BBC News. October 30.
Braungart, R. G., & Braungart, M. M. (1986). Life-course and generational
politics. Annual Review of Sociology, 12, 205–231.
Brewer, P. R. (2003). The shifting foundations of public opinion about gay
rights. Journal of Politics, 65, 1208–1220.
Burdette, A. M., Ellison, C. G., & Hill, T. D. (2005). Conservative protes-
tantism and tolerance toward homosexuals: An examination of
potential mechanisms. Sociological Inquiry, 75, 177–196.
Button, J. W., Wald, K. D., & Rienzo, B. A. (1999). The election of openly
gay public officials in American communities. Urban Affairs Review,
35, 188–209.
Cheney, K. (2012). Locating neo colonialism, “tradition” and human rights
in Uganda’s “gay death penalty”. African Studies Review, 55(2), 77–95.
Claude, F. S. (1995). The subcultural theory of urbanism: A twentieth-year
assessment. American Journal of Sociology, 101(3), 543–577.
Claude, F. S. (1975). Toward a subcultural theory of urbanism. American
Journal of Sociology, 80(6), 1319–1341.
Coly, A. A. (2013). Introduction. African Studies Review, 56(2), 21–30.
Driskell, R., Embry, E., & Lyon, L. (2008). Faith and politics: The influence
of religious beliefs on political participation. Social Science Quarterly,
89(2), 294–314.
Dunton, C., & Palmberg, M. (1996). Human rights and homosexuality in
Southern Africa. Nordic African Institute.
Durkheim, E. (1912). The elementary forms of religious life, Fields, K.E. trans.
New York: Free Press.
Durkheim, E. (1893). The division of labour in society. New York: The Free
Press.
Epia, O. (2014). Public diplomacy and the homosexuality debate. Pre-
mium Times. February 5. http://www.premiumtimesng.com/opinion/
154650-public-diplomacy-homosexuality-debate-oke-epia.html
Epprecht, M. (2013). The making of “African sexuality”: Early sources cur-
rent debates. In S. N. Nyeck, & M. Epprecht (Eds.), Sexual diversity in
Africa (pp. 54–66). Mc-Gill Queens University Press.
Epprecht, M. (2008). Heterosexual Africa? The history of an idea from the age
of exploration to the age of aids. Columbus: Ohio University Press.
Evaristo, B. (2014). The idea that African homosexuality was a colonial import
is a myth. The Guardian. Saturday, March 8.
Finke, R., & Adamczyk, A. (2008). Cross-national moral beliefs: The influ-
ence of national religious context. Sociological Quarterly, 49, 615–650.
Freeman, S. (2011). Capitalism in the classical and high liberal traditions.
Social Philosophy and Policy, 28, 19–55.
Gaines, S. N., & Garand, J. C. (2010). Morality, equality, or locality: Ana-
lyzing the determinants of support for same-sex marriage. Political
Research Quarterly, 63, 553–567.
Goerres, A. (2008). The grey vote: Determinants of older voters’ party
choice in Britain and West Germany. Electoral Studies, 27(2), 285–304.
Gunkel, H. (2013). Some reflections on postcolonial homophobia, local
interventions, and LGBTI Solidarity Online: The politics of global peti-
tions. African Studies Review, 56(2), 67–81.
Hogg, P. W. (2006). Canada: The constitution and same-sex marriage.
International Journal of Constitutional Law, 4(4), 712–721.
Himmelstein, J. L., & McRae, J. A., Jr. (1988). Social issues and socioeco-
nomic status. Public Opinion Quarterly, 52, 492–512.
Please cite this article in press as: Alozie, N. O., et al. Global liberalization on homosexuality: Explaining the African gap.
The Social Science Journal (2016), http://dx.doi.org/10.1016/j.soscij.2016.09.001
11
Human Rights Watch. (2011). South Africa: LGBT rights in name only?
http://www.hrw.org/news/2011/12/05/south-africa-lgbt-rights-
name-only
Inglehart, R. (1977). Long term trends in mass support for European uni-
fication. Government and Opposition, 12(2), 150–177.
Inglehart, R., Norris, P., & Welzel, C. (2002). Gender equality and democ-
racy. Comparative Sociology, 1, 321–345.
Inglehart, R., & Baker, W. E. (2000). Modernization, cultural change, and
the persistence of traditional values. American Sociological Review, 65,
19–51.
Ireland, P. R. (2013). A macro-level analysis of the scope, causes, and
consequences of homophobia in Africa. African Studies Review, 56(2),
31–45.
Jetten, J. S., Haslam, S. A., & Barlow, F. K. (2013). Bringing back the system.
Social Psychological and Personality Science, 4(1), 6–13.
Jolly, J. (2008). Africa’s lesbians demand change. London: BBC News.
February 27.
Kane, M. D. (2007). Timing matters: Shifts in the causal determi-
nants of sodomy law discrimination, 1961–1998. Social Problems, 54,
211–239.
Kutsch, T. (2013). Homosexuality still a crime in 38 African countries.
Aljazeera America,. November 14.
Lax, J. R., & Phillips, J. (2009). Gay rights in the states. American Political
Science Review, 103, 367–386.
Linneman, T. J. (2003). Weathering change: Gays and lesbians, Christian
conservatives and everyday hostilities. New York: New York University
Press.
Loftus, J. (2001). America’s liberalization in attitudes toward homosexu-
ality, 1973–1998. American Sociological Review, 66, 762–782.
Mama, A. (2007). Is it ethical to study Africa? African Studies Review, 50(1),
1–26.
Manganaro, L., & Alozie, N. (2015). The political gender gap in Afghanistan.
Journal of Women, Politics & Policy, 36, 285–310.
M’Baye, B. (2013). The origins of Senegalese homophobia: Discourses on
homosexuals and transgender people in colonial and postcolonial
Senegal. African Studies Review, 56(2), 109–128.
McNamara, T. (2014). Not the Malawi of our parents: Attitudes toward
homosexuality and perceived westernization in Northern Malawi.
African Studies, 73(1), 84–106.
Michaelson, J. (2014). Why Africa’s turning anti-gay. The Daily
Beast—World,. March 31.
Merino, S. M. (2013). Contact with gays and lesbians and same-sex mar-
riage support: The moderating role of social context. Social Science
Research, 42(4), 1156–1166.
Murray, S. O., & Roscoe, W. (2001). Boy-wives and female-husbands: Studies
in African-American homosexualities. New York: Palgrave Macmillan.
New American Foundation. (2008). The millennial pendulum: A new gener-
ation of voters and the prospects for a political realignment. Washington,
DC: New American Foundation.
New York Times. (2014). Editorial: Demonizing gays in Africa. February 11.
Obergefell et al. v. Hodges. (2015). Slip opinion. Washington, DC: Supreme
Court of the United States. http://www.cnn.com/2015/06/26/politics/
scotus-opinion-document-obergefell-hodges/index.html
Olson, L. R., Cadge, W., & Harrison, J. T. (2006). Religion and public opinion
about same-sex marriage. Social Science Quarterly, 87, 340–360.
Pampel, F. (2011). Cohort changes in socio-demographic determinants of
gender egalitarianism. Social Forces, 89, 961–982.
Pan, P., Meng, J., & Zhou, S. (2010). Morality or equality: Ideological framing
in news coverage of gay marriage legitimization. Social Science Journal,
47(3), 630–645.
Pew Research Center. (2015). Gay marriage around the world..
http://www.pewforum.org/2015/06/26/gay-marriage-around-
the-world-2013/
Pew Research Center. (2014). Teaching the children: Sharp ideological dif-
ferences some common ground. Washington, DC: Pew.
Pew Research Center. (2013a). The global divide on homosexuality. Wash-
ington, DC: Pew.
Pew Research Center. (2013b). 2013 spring global attitudes survey. Wash-
ington, DC: Pew.
Pew Research Center. (2010). Tolerance and tension: Islam and Christianity
in Sub-Saharan Africa. Pew.
Pflanz, M. (2014). Anti-gay laws are political boon across Africa. The Chris-
tian Science Monitor. March 7.
Reimer, S., & Park, J. (2001). Tolerant (In)civility? A longitudinal analysis
of white conservative protestants’ willingness to grant civil liberties.
Journal of the Scientific Study of Religion, 4, 735–745.
Schatz, S. (2002). Age cohort voting effects in the breakdown of single-
party rule. Journal of Aging Studies, 16, 199–219.
Schwadel, P. (2011). Age, period, cohort effects on religious activities and
beliefs. Social Science Research, 40, 181–192.
G Model
SOCSCI-1351; No. of Pages 12 ARTICLE IN PRESS
12 N.O. Alozie et al. / The Social Science Journal xxx (2016) xxx–xxx
Scott, J. (1998). Changing attitudes to sexual morality: A cross-national
comparison. Sociology, 32, 815–845.
Sherkat, D. E., Powell-Williams, M., Maddox, G., & de Veries, K. M. (2011).
Religion, politics, and support for same-sex marriage in the United
States, 1988–2008. Social Science Research, 40(1), 167–180.
Smith, D. (2014). Why Africa is the most homophobic continent. The
Guardian,. Saturday, Feb.22.
Smith, T. W., Son, J., & Kim, J. (2014). Public attitudes towards homosexuality
and gay rights across time and countries. NORC at the University of
Chicago. November 7.
Štulhofer, A., & Rimac, I. (2009). Determinants of homonegativity in
Europe. Journal of Sex Research, 46, 24–32.
Takács, J., & Szalma, I. (2013). How to measure homophobia in an inter-
national comparison? Druzboslovne Razprave, 73, 11–42.
Tamale, S. (2014). Homosexuality is not un-African. Aljazeera America. April
26.
Tamale, S. (2013). Confronting the politics of nonconforming sexualities
in Africa. African Studies Review, 56, 31–45.
Treas, J. (2002). How cohorts, education, and ideology shaped a new sexual
revolution on American attitudes toward nonMarital sex, 1972–1998.
Sociological Perspectives, 45, 267–283.
Tushnet, M. (1992). Political correctness, the law, and the legal academy.
Yale Journal of Law and the Humanities, 4(1). Article 6.
Please cite this article in press as: Alozie, N. O., et al. Global liberalization on homosexuality: Explaining the African gap.
The Social Science Journal (2016), http://dx.doi.org/10.1016/j.soscij.2016.09.001
Wald, K. D., Button, J. W., & Rienzo, B. A. (1996). The politics of gay rights
in American communities: Explaining antidiscrimination ordinances
and policies. American Journal of Political Science, 40, 1152–1178.
Ward, K. (2002). Same-sex relations in Africa and the debate on homosex-
uality in East African Anglicanism. Anglican Theological Review, 84(1).
Weber, M. (1958). The protestant ethic and the spirit of capitalism. New
York: Charles Scribner’s Sons [1904–1905].
Weil, F. D. (1985). The variable effects of education on liberal attitudes.
American Sociological Review, 50, 458–474.
Wilcox, C., & Jelen, T. (1990). Evangelicals and political tolerance. American
Politics Quarterly, 18, 25–46.
Williamson, O. E. (1981). The economics of organization. American Journal
of Sociology, 87(3), 548–577.
Wirth, L. (1938). Urbanism as a way of life. American Journal of Sociology,
44, 1–24.
Witt, S. L., & McCorkle, S. (Eds.). (1997). Anti-gay rights: Assessing voter
initiatives. Santa Barbara, CA: Praeger.
Yang, A. S. (1997). Trends: Attitudes toward homosexuality. Public Opinion
Quarterly, 61(3), 477–507.
Yip, A. (2005). Queering religious texts. Sociology, 39(1), 47–65.
Yuchtman-Yaar, E., & Alkalay, Y. (2007). Religious zones, economic devel-
opment and modern value orientations: Individual versus contextual
effects. Social Science Research, 36, 789–807.