Social Science 2016

听听听听听听听听听听听 Walking home from a soccer game on an empty street in Cape Town, Mvuleni Fana was ambushed by several men. They jumped her in the back of an alley, each man beating and raping her one at a time. The perpetrators knew about her sexual orientation, and before leaving her unconscious told her, 鈥淎fter everything we're going to do to you, you're going to be a real woman, and you're never going to act like this again鈥� (Strudwick 2014). A similar story was told by Lungile Dladla who was walking home one evening with her fellow gay companion when, out of the dark, an armed man approached them. He tied their hands and feet, telling them 鈥�Ja, today I want to show you that you鈥檙e girls,鈥� and proceeded to rape them both. Dladla鈥檚 case was not taken seriously at the police station because of her 鈥渂utch鈥� appearance, and more tragically she nearly died years later because she had contracted HIV from the encounter (Hunter-Galt 2012). In South Africa, 鈥渃orrective rape鈥� represents a daily threat for many lesbians. In fact, gay women are assaulted nearly twice as often as their heterosexual counterparts (Hunter-Galt 2012).听Even more disheartening, South Africa is seen as one of the most gay-friendly countries in Africa. It is only one of two countries in the region to have legalized gay marriage, and is the only one to include protections for gay citizens in its constitution. In other countries in the region homosexual acts are penalized by life imprisonment or even the death penalty.

听听听听听听听听听听听 Today, Sub-Saharan Africa is the most homophobic region in the world in terms of public opinion (Horowitz et. al. 2013; McCarthy 2014). At the same time as gay marriage鈥檚 legalization in 35 of 50 U.S. states, and the adoption of LGBT rights by the international community as a human rights issue, many African countries continually embrace ever-harsher penalties for homosexual acts. At this moment homosexuality is illegal in over 70% of African countries (鈥淢aking Love a Crime鈥� 2013). This begs the question: 鈥渨hat explains homophobia within Sub-Saharan Africa today?鈥� Current literature largely places the blame on colonial institutions, which originally installed most of the homophobic laws in the region (Asal 2013; Gupta 2008; Han and O鈥橫ahoney 2014; Ireland 2013; Itaborahy and Zhu 2013; 鈥淢aking Love a Crime鈥� 2013; Sanders 2009), but this explanation fails to illustrate the larger picture. The interpretation hinges on a Euro-centric view that the actions of underdeveloped countries revolve around Western powers, and completely disregards the fact that Third World countries control their own fate. In the contemporary era, a uniquely homophobic culture has developed within the region. African governments have largely decided to go the opposite direction of their former colonizers who have become much more liberal in terms of the rights granted to the LGBT community. Countries such as Burundi, Ethiopia, Gambia, Ghana, Kenya, Nigeria, and Zimbabwe have all instituted either new or revised laws within the twenty-first century, which carry prison terms for homosexual behavior (Itaborahy and Zhu 2013).

听听听听听听听听听听听 Clearly, a broader framework needs to be presented to explain homophobia within Sub-Saharan Africa. This paper argues that Ronald Inglehart鈥檚 post-materialist theory, which postulates that economic development leads to greater social tolerance of issues such as homosexuality, can incorporate both colonial and contemporary factors to explain variations in African homophobia. To accomplish this analysis, I isolate factors previously identified as central to the development of global homophobia and use linear regression models to see if the factors鈥� relationships to variations in African homophobia agree with Inglehart鈥檚 predictions. Previous studies have addressed economic development (Anderson and Fetner 2008; Han and O鈥橫ahoney 2014; Wike and Horowitz 2007), education (Bobo and Licari 1989; Herek 1991), religion (Asal 2013; Epprecht 2013; Han and O鈥橫ahoney 2014; Shoko 2010), and colonial rule (Asal 2013; Gupta 2008; Han and O鈥橫ahoney 2014; Ireland 2013; Itaborahy and Zhu 2013; 鈥淢aking Love a Crime鈥� 2013; Sanders 2009) as some of the most salient aspects of homophobia within countries worldwide. I employ two types of homophobia (institutional and personal) to investigate the relationship between the aforementioned variables and Sub-Saharan homophobia. Through the statistical analysis, I posit that Inglehart鈥檚 theory largely supports variations in institutional and personal homophobia within Sub-Saharan Africa. It appears that both types of homophobia are largely influenced through religious cultural norms, which are upheld in underdeveloped states, and secularized in economically more developed African nations. It is posited that African governments lose their incentive to act in discriminatory ways through this process of secularization, and populations become more tolerant of homosexuality when they are both economically secure and no longer derive their moral convictions from traditional cultural norms. Colonial influence appears to have made an impact on the development of both types of homophobia, while the influence of education largely remains unclear.

听听听听听听听听听听听 In the following analysis, I will begin by first defining the two types of homophobia in order to set the boundaries of the study. Inglehart鈥檚 post-materialistic theory is then explained and postulated as an explanation for both types of homophobia within the Sub-Saharan region. Economic development, education, religion, and colonization are framed within the boundaries of Inglehart鈥檚 theory as a way to analyze the current homophobic climate of the region. Subsequently, several hypotheses are presented, which are tested by separate regression models for each type of homophobia. An individual-level analysis of the Sub-Saharan countries in the fifth and sixth World Values Surveys follows to corroborate the results taken from the analysis on personal homophobia. Finally, a discussion and conclusion highlight how this study鈥檚 findings can help understand institutional and personal homophobia in Sub-Saharan Africa.

Defining Homophobia and the Theoretical Scope听听听听听听听听听听

听听听听听听听听听听听 As a term, homophobia has been utilized in a vast array of settings, which has contributed to a misleading meaning. While literally translating to an irrational fear of homosexuals, it has adopted a more encompassing framework that includes 鈥渁nger, hatred, bias, ignorance, jealousy, or other sources of antipathy toward queer persons鈥� (Thoreson 2014).[1] Authors have previously broken homophobia down into several levels. These generally include institutional, interpersonal, personal, and societal (Blumenfeld 1992, Raja and Stokes 1998). In the following analysis, I study two dimensions of homophobia, institutional and personal, in order to study related, but distinct, dimensions of the term simultaneously.

听听听听听听听听听听听 Institutional homophobia is the systemic exclusion and discrimination of homosexuals through the actions of governments or state organizations (Raja and Stokes 1998, 118). Institutional homophobia encompasses a broad perspective of society as it includes governmental actions and rhetoric, state-sponsored violence, and the behavior of organizations within a country, along with other structural aspects. On the other hand, personal homophobia is the intolerant beliefs held by individuals in society regarding homosexuals (Raja and Stokes 1998, 118). These beliefs include viewing homosexuality as immoral or having hateful feelings towards gay individuals. The two terms address both the system of discrimination as well as the intolerant views of communities within states that together lie beneath homophobic cultures. Viewed side by side, the terms more fully contribute to an understanding of homophobia as a whole. By separating homophobia into these two terms, this analysis mirrors other studies that have broken down other types of discrimination, such as sexism and racism, into both an individual-level and a broader society-wide level of analysis (Arreola et. al. 2014; Herek 1996; Fiske 2001; Raja and Stokes 1998).

听听听听听听听听听听听 Since the terms are fundamentally different, with diverging causal factors, they must be measured independently from one another. Having said this, institutional and personal homophobia are not completely independent of one another; laws can change public opinion, and public opinion has an effect on the language and severity of the laws. Therefore, it cannot be claimed that one type contributes more to the homophobic climate than the other; instead both must be looked at simultaneously in order to gain a holistic understanding of homophobia in the region.

The Colonial Legacy of Homophobia
听听听听听听听听听听听 Many scholars have argued that colonial rule has shaped the current landscape and polices of African countries especially in regards to issues concerning homosexuality (Asal 2013; Gupta 2008; Han and O鈥橫ahoney 2014; Ireland 2013; Itaborahy and Zhu 2013; 鈥淢aking Love a Crime鈥� 2013; Sanders 2009). Their case is bolstered by a British historical legacy of discriminatory policies and actions towards gays. This legacy started in the 19^th century when Britain imported the first anti-sodomy laws into their African landholdings from section 377 of the Indian Penal Code, which they had themselves written to bring 鈥榗ivilized鈥� behavior to the backwardness of their indigenous colonies (Gupta 2008). The laws made 鈥渃arnal intercourse against the order of nature with any man, woman or animal鈥� illegal for any person within the British domain and henceforth made rigid heterosexuality the sexual norm. Most African countries formerly colonized by the British still utilize this language in anti-homosexual laws today.
听听听听听听听听听听听 In contrast anti-sodomy laws were largely absent in places colonized by the French because homosexuality had been de-criminalized in the 18^th century with the Napoleonic Penal Codes, which legalized activities done in privacy (Sanders 2009). While the French enacted some types of anti-sodomy laws in a few colonies, such as in the modern areas of Benin, Cameroon, and Senegal, in order to assert social domination, most places remained free of the discriminatory measures taken by the British (Gupta 2008). Today, 38 out of 54 countries in Africa ban the practice of homosexuality, and the region has some of the harshest anti-gay laws in the world (鈥淢aking Love a Crime鈥� 2013). The vast majority of the countries banning homosexuality were formerly British landholdings, while most of the areas colonized by the French have no anti-gay laws in place (鈥淢aking Love a Crime鈥� 2013). Indeed, Patrick Ireland鈥檚 work on sodomy laws in the region found that a significant portion of variation in homophobic legislation was explained by being a former British landholding.

听听听听听听听听听听听 But while the colonial explanation for contemporary homophobia is compelling, it ignores the fact that modern discrimination is a complex phenomenon, which cannot be explained simply by a colonial past. Colonialism certainly appears to have set the stage for African homophobia, but fails to fully explain modern enactment of homophobic laws as well as modern forms of homophobia outside the scope of criminalization. For instance, several Sub-Saharan countries, including Burundi, Ethiopia, Gambia, Ghana, Kenya, Nigeria, and Zimbabwe, have all adopted either new or revised legislation with prison sentences for homosexual behavior within the last 15 years (Itaborahy and Zhu 2013). It is important to remember that African countries declared their independence from Europe decades ago, so there must be uniquely African and strictly contemporary reasons for homophobia to have become so entrenched in the region in the 21^st century. Therefore, it is paramount to investigate other causes of African homophobia in order to look beyond colonialism as the only explanation.

The Influence of Colonialism on Materialist Attitudes in Africa

听听听听听听听听听听听 This paper contends that the colonial legacy can be incorporated into the broader framework of Inglehart鈥檚 post-materialist theory in order to explain African homophobia. Instead of viewing colonialism as the cause of African homophobia, post-materialism sees it as setting the stage for the future homophobic culture of the region. Post-materialism suggests countries follow path-dependent avenues of development according to their cultural heritage (Inglehart 1999). They develop in this way because the cost of changing cultural norms is expensive, especially for underdeveloped countries, and so they remain relatively unchanged over time (Asal 2013). Through the developmental process, states鈥� interests shift in a predictable way from economic to social as they shift from materialist to post-materialist, but specific policies and cultural norms within societies reflect a common history.

听听听听听听听听听听听 Colonialism might also reflect on social attitudes in states in a path-dependent manner. Homophobic laws dating back to the 19^th century could help form both norms about the morality of homosexuality, and set the stage for later discriminatory institutions. African countries under the same European authority would therefore be expected to develop political institutions and social values that reflect similar feelings about homosexuality. Under this framework, areas with harsher anti-gay legal systems under colonial rule would be expected to retain that legal code in the future, and store the principles of those laws in their cultural beliefs. However, even with a shared culture, countries should be able to shift their social outlooks as they develop economically (Inglehart 1999). Previous research supports this notion, finding that countries formerly colonized by the British do in fact have more homophobic laws than countries not colonized by the British, but the research also suggests that they do not take any longer to de-criminalize homosexuality than other states (Han and O鈥橫ahoney 2014). This indicates that heritage may influence the political and social landscape of a country down the road, but does not determine it. Instead common heritage simply gives countries a similar developmental starting place and trajectory over time, which is influenced by modern conditions.

The Post-Materialist Explanation for African Homophobia

听听听听听听听听听听听 Ronald Inglehart鈥檚 post-materialist theory predicts countries will develop similarly over time according a country鈥檚 economic conditions (1987, 1990, 1997, 1999). Accordingly, populations within pre-industrialized countries exhibit materialist behavior, which reflects the needs of the emerging working class (Inglehart 1987). At an early stage in economic development, average citizens are more concerned about their own economic well-being than society-wide social injustices because of a day-to-day struggle to survive. As a result, newly industrializing societies tend to focus their resources on policies that will attain economic growth in order to improve society鈥檚 overall quality of life. Traditional values and norms characterize societies with dominantly materialist attitudes because they do not yet have the time or political/physical resources to revise their norms in socially salient ways.

听听听听听听听听听听听 Important to this theory is that populations will adopt traditional values and norms, which in many instances are religiously based, in order to provide a shared sense of purpose among their fledgling populations (Inglehart 1987). These norms then continue in a path-dependent manner, as they are fairly resistant to change, and continue to shape future cultural norms and behavior over time. As a result, materialist oriented populations almost uniformly conform to specific cultural norms and religious practices at the expense of individualism. This societal structure continues until development eventually yields widespread economic security, and new generations increasingly want to address neglected social inequity over economic growth (Inglehart 1999). Through this cycle, societal attitudes slowly transitions from collectively oriented, with mainly material interests, to individually minded, with primarily post-materialist and social interests. Unfortunately, minority and marginalized groups within materialist-oriented societies bear a significant burden in this stage of development. As a result of the entrenchment of traditional norms, citizens within areas of predominately materialist attitudes 鈥渇eel threatened by foreigners, by ethnic diversity and by cultural change. This leads to an intolerance of gays and other outgroups, an insistence on traditional gender roles, and an authoritarian political outlook鈥� (Inglehart 1999, 26).

听听听听听听听听听听听 While past studies have shown that post-materialism can explain variation in global homophobia (Adamczyk and Pitt 2009; Anderson and Fetner 2008; Henshaw 2014; Stulhofer and Rimac 2009), can it explain why some African countries are more homophobic than others? As one of the most undeveloped areas in the world, the countries of Africa almost uniformly fit Inglehart鈥檚 description of states where societal attitudes are predominantly materialist. Accordingly, the region clings to traditional values, and has extreme, even violent, reactions to anything it sees as culturally foreign, including homosexuality. Further, the most economically developed country in the region, South Africa, has the most progressive gay-rights policies in all of Africa. There is, therefore, reasonable justification to analyze homophobia within Sub-Saharan Africa through Inglehart鈥檚 theory. To perform this analysis, several concepts previously put forth as impactful to either African homophobia or homophobia in general (including economic development, education, religiosity, and colonialism) are explored further in the context of Inglehart鈥檚 theory.

Economic Development and Societal Interests

听听听听听听听听听听听 At the core of the post-materialistic theory is economics: as countries economically develop, their population鈥檚 interests should shift from material to social (Inglehart 1987, 1990, 1997, 1999). Interestingly, economic underdevelopment is widely thought to be a driving cause of global homophobia (Anderson and Fetner 2008; Han and O鈥橫ahoney 2014; Wike and Horowitz 2007). This belief comes from the observation that worldwide feelings towards homosexuals are generally more negative the poorer a country, suggesting that a country鈥檚 overall economic performance may be tied to a population鈥檚 opinion of the LGBT community. Previous research has found this to be true within European countries where the prosperous region of Western Europe has far greater tolerance towards homosexuals than less developed European countries, even when controlling for former Communist rule (Anderson and Fetner 2008).

听听听听听听听听听听听 But while economic development has been used to explain more tolerant views of homosexuality in the West, thus far, the same explanation has not been applied to clarify variations of homophobia within Sub-Saharan Africa. There is good reason to believe that both homophobic attitudes and state actions in the region can be explained through an economic lens. Since the region ranks within the world鈥檚 most underdeveloped, ubiquitous materialist attitudes would be expected within African states through Inglehart鈥檚 theory (1987). As predicted by post-materialism, African countries are both some of the poorest in the world in terms of GDP per capita (鈥淕DP per capita鈥� 2014), and the most homophobic overall in term of opinion surveys (Horowitz et. al. 2013; McCarthy 2014). If economic circumstances have not allowed the interests of most people to shift from material to social, individuals could be perpetuating existing homophobic cultural norms as a result of widespread materialist attitudes in the region. Underdevelopment and poverty may, therefore, be responsible for homophobic attitudes in Africa by allowing the continuation and formation of conservative cultures, which form the basis for individual鈥檚 views.

听听听听听听听听听听听 Sub-Saharan countries, as previously noted, also have some of the harshest anti-gay policies in the world (Itaborahy and Zhu 2013). African governments may be implementing more homophobic policies in underdeveloped conditions to align with both the homophobic interests of their populations and the cultural norms of the country. Several scholars have noted how the impoverished conditions of Africa have encouraged the enactment of discriminatory policies towards gays within thriving homophobic cultures (Biruk 2014; Shoko 2010; Thoreson 2014; Van Klinken 2013). In fact it could be that underdevelopment has created a vicious cycle of discrimination in Sub-Saharan Africa, whereby widespread materialist attitudes have spurred the establishment of homophobic state structures, which have then further reinforced homophobic societal views. Overall, if underdevelopment can explain the predominance of both homophobic views and state structures in the Sub-Saharan region, materialist attitudes could reasonably be placed at the heart of African homophobia.

Education as an Avenue for Tolerance

听听听听听听听听听听听 Inglehart suggests that a formal education can give people the ability to break down dominant cultural norms because it allows for 鈥渁utonomous decision-making鈥� (1999, 22). Education may be a way for individuals to break from homophobic cultural norms and standards, to which the rest of society adheres, especially in underdeveloped countries. In addition, as states develop, they may be able to encourage rising educational levels as more money is allowed to finance schools. Therefore, education may act in a way to reform both homophobic cultures and state institutions. There is support for this framework in other places in the world such as Ireland where greater spending on educational programs seems to have led to more tolerant attitudes towards homosexuals across the entire population (Denny, 2011).

听听听听听听听听听听听 In Western countries, education is viewed as an essential avenue to increase societal tolerance. In fact, previous studies of developed countries have found education to be one of the strongest predictors of social tolerance towards homosexuality (Anderson and Fetner 2008). Previous research has demonstrated that individuals with a higher level of education tend to be more tolerant toward other groups, even ones that they do not like (Bobo and Licari 1989; Whitehead 2010). It seems that individuals with a greater amount of 鈥渃ognitive sophistication鈥� tend to see the worth in granting rights of expression to people different from themselves even when they do not necessarily see eye to eye (Bobo and Licari 1989). If education could be an avenue to counter discriminatory cultural norms, it could potentially be an important factor to explain homophobia in less developed places.

听听听听听听听听听听听 Importantly for Africa, more educated Sub-Saharan countries could have greater overall tolerance for marginalized groups, including gays. Even if educational programs are not directly teaching about LGBT issues, possibly a greater 鈥渃ognitive sophistication鈥� would allow individuals to tolerate the views and lifestyles different than their own. With the extremely homophobic climate of Sub-Saharan Africa, education could possibly help to explain why some countries are less homophobic than others. While many scholars have suggested that African school systems actually worsen homophobic attitudes by labeling homosexuality as a sin (Johnson 2007; 鈥淢aking Love a Crime鈥� 2013), it is difficult to tell if views are perpetuated by the schools or by the broader culture. Potentially, a person in a homophobic culture may be able to form individual beliefs only after a certain number of years in school. If education has the same effect in underdeveloped countries as it does in developed ones, it would be expected to oppose the homophobic culture of African states. When a large enough segment of the population breaks from their country鈥檚 homophobic cultural ideals, governments should reform homophobic state structures to reflect more tolerant interests. Education could therefore be a force to challenge the homophobic culture of African states.

Religion in Africa

听听听听听听听听听听听 Religiosity has regularly been cited as a reason for the homophobic nature of countries (Asal 2013; Epprecht 2013; Shoko 2010). For instance, in a broad study of various international actors, Alexis Henshaw found that 鈥渆ven controlling for a country鈥檚 level of modernization and an individual respondent鈥檚 self-expression values, a respondent鈥檚 religious belief has a significant and negative relationship to their tolerance toward homosexuality.鈥� Inglehart鈥檚 theory would suggest that religiously based conservative norms provide the moral backbone for the majority of people in societies where materialist attitudes are standard, which tends to create stigmatization of certain types of self-expression including homosexuality. In addition, religious cultural norms can be impressed on governments who create policies to reflect those ideals, thereby codifying homophobia into law (Asal 2013). However, as states develop economically, their populations rely less heavily on the conservative cultural framework for support, which allows cultural norms to secularize. Without homophobic cultural norms, governments鈥� incentive for discriminatory policies disappears, and state institutions correspondingly reform over time. Unfortunately, in underdeveloped regions, such as Sub-Saharan Africa, where conservative cultural norms have not been reformed, religion may continue to perpetuate discriminatory state structures and individual views.

听听听听听听听听听听听 Religion may make a considerable impact on the level of homophobia within African states because of its importance in African culture. It has been noted that religion ties together communities and families, and plays a large role in children鈥檚鈥� upbringing and education in African cultures (Johnson 2007). Importantly, African populations are often taught that homosexuality is foreign, un-African, and sinful through a religious framework (鈥淢aking Love a Crime鈥� 2013). This is significant in a region where religion is a part of almost everyone鈥檚 lives. Accordingly, a recent survey found that African countries have some of the most religious populations in the world (Crabtree 2010). When asked if 鈥渞eligion [was] an important part of [their] daily [lives],鈥� 85% or more of every Sub-Saharan country surveyed responded 鈥測es鈥� (Crabtree 2010). Religion鈥檚 ubiquity in Sub-Saharan Africa may therefore influence the development of homophobic cultural norms that are then impressed on countries鈥� populations and state intuitions. 听

听听听听听听听听听听听 Importantly to this conversation, Christianity and Islam, whose most accepted doctrines around the world espouse anti-homosexual rhetoric, are the two major religions in most African states and serve as the dominant moral codes. Many African states have created systems with integral ties to either Christianity or Islam, which both have generally accepted scriptural interpretations that have made gays highly stigmatized and even hated by many Africans (Shoko 2010). For example, several African countries directly use the teachings of the Qur鈥檃n, in the form of sharia law, to punish convicted homosexual activity with death by stoning. A broad survey of Sub-Saharan Africa by the Pew Research Center found that 60% of African Christians 鈥渇avor making [the] Bible the official law of the land鈥� and 63% of African Muslims 鈥渇avor making sharia the official law of the land,鈥� demonstrating the widespread influence of the region鈥檚 religious culture (Cooperman et. al. 2010). Interestingly, previous research has found Protestants, especially Evangelical Protestants, and Muslims both espouse the most negative views of homosexuals in comparison to Catholics, Jews, other religious groups and non-affiliated individuals (Adamczyk and Pitt 2009; Ogland and Verona 2014; Olson, Cadge, and Harrison 2006; Whitehead 2010). With these two groups making up the majority of religious followers in the region (Cooperman et. al. 2010; Hackett et. al. 2012, 2011), the effects of the existing homophobic culture could be compounded in many African states.

听听听听听听听听听听听 Africa鈥檚 religious culture has also created a lack of separation between church and state in various countries, which has only exacerbated homophobia in those places. For example, in Uganda the integral relationship between religion, government, and general homophobia was painfully felt with the introduction of the Anti-Homosexuality Act to the country鈥檚 legislature in 2009 (Epprecht 2013, 144-145). The country鈥檚 government was convinced to introduce the law, known as the 鈥楰ill the Gays Bill鈥� because its maximum penalty for homosexuality was death, by missionaries Scott Lively and Don Schmierer who both toured across Uganda preaching about the gays鈥� 鈥渄ark agenda鈥� (Epprecht 2013, 144-145). The instant and overwhelming enthusiasm for the bill by the general population not only illustrated how deeply rooted religion was in the cultural fabric of the country, but also demonstrated how those norms have shaped people鈥檚 views of homosexuality. If a connection exists between Africa鈥檚 religious cultures and regional homophobia it is easy to see how both laws and attitudes could take on a discriminatory tone.

听

Hypotheses

听听听听听听听听听听听 With the above discussion in mind, I formulate several hypotheses about the role of economics, education, religion, and colonial rule on institutional and personal homophobia. My predictions follow the principles set out by Inglehart鈥檚 post-materialist theory; namely homophobia should proliferate in economically underdeveloped African states, which abide by religiously rooted traditional values. The homophobic systems of Africa should have been built upon the colonial past, which started many homophobic norms. In addition, education should be a way for African states to increase individual thought and decrease cultural homophobia. Through this framework institutional and personal homophobia should have the same underlying causes, as both came out of the same African historical legacy, and should be driven by the framework outlined by post-materialism.

听听听听听听听听听听听 The Sub-Saharan region is concurrently the poorest in the world (鈥淕DP per capita鈥� 2014), and also the harshest in terms of homophobic laws (Itaborahy and Zhu 2013). This is partly explained by the fact that structural changes are costly for societies as path-dependence would suggest (Asal 2013), and therefore changing old homophobic legal frameworks taken from previous decades may be difficult for poor countries. But economic underdevelopment also plays a critical role by both bolstering conservative cultural norms, which reinforce negative social attitudes and discriminatory institutions. Opinion surveys have consistently shown Sub-Saharan countries as the most disapproving of homosexuality of any area surveyed in the world (Horowitz et. al. 2013; McCarthy 2014). However it remains to be seen if there is a relationship between homophobic attitudes and economic development specifically within Sub-Saharan countries.

Hypothesis 1: Economic development is negatively correlated with institutional homophobia in Sub-Saharan Africa.
Hypothesis 2: Economic development is negatively correlated with the personal homophobia in Sub-Saharan Africa.

听听听听听听听听听听听 Inglehart suggests that education can help individuals distance their thinking from traditional cultural norms (1999), and therefore change dominant cultural values over time. This is supported by other studies, which have found that education increases individual鈥檚 tolerance towards homosexuals (Bobo and Licari 1989). Possibly an ability to view the world from different perspectives allows people with more education to see the value in giving freedoms, such as expression, to others even when they do not like them. Likely, greater levels of education allow citizens to learn about other groups in society through a more objective lens. In Sub-Saharan Africa, countries with the economic stability to finance educational institutions may have more tolerant attitudes and institutions within their states as a result of more educated populations.

Hypothesis 3: The overall educational level of a country is negatively correlated with institutional homophobia in Sub-Saharan Africa.
Hypothesis 4: The overall educational level of a country is negatively correlated with personal homophobia in Sub-Saharan Africa.

听听听听听听听听听听听 Most African countries have norms and legal institutions based on religious teachings. This would be expected of states where materialist attitudes dominate because they tend to have populations who follow conservative overarching cultural norms (Inglehart 1999). Islam and Christianity, the two major religions in Africa, both largely view homosexuality as a moral sin. If the cultural frameworks of states rely heavily on these teachings, populations should be more radicalized in their dislike of homosexuality, and African government institutions should concurrently take harder stances against homosexuality (Epprecht, 2013).

Hypothesis 5: The religiosity of a country is positively correlated with institutional homophobia in Sub-Saharan Africa.
Hypothesis 6: The religiosity of a country is positively correlated with personal homophobia in Sub-Saharan Africa.

听听听听听听听听听听听 The British have been accused of initiating the homophobic nature of their former landholdings with the discriminatory laws they instituted into those places (Asal 2013; Gupta 2008; Han and O鈥橫ahoney 2014; Ireland 2013; Itaborahy and Zhu 2013; 鈥淢aking Love a Crime鈥� 2013; Sanders 2009). Accordingly, as a result of path-dependence, the effects of colonialism should reflect on the current norms of former British colonies by having more discriminatory attitudes and institutions within those states.

Hypothesis 7: Being a former British colony is positively correlated to institutional homophobia in Sub-Saharan Africa.
Hypothesis 8: Being a former British colony is positively correlated to personal homophobia in Sub-Saharan Africa. 听

Measurements and Methods

The Need for a Better Indicator听听听听听听听听听

听听听听听听听听听听听 To date, a considerable portion of the research that has been done on homophobia in Africa has focused on the criminalization of homosexuality (Asal 2013; Itaborahy and Zhu 2013; 鈥淢aking Love a Crime鈥� 2013). Unfortunately, this fails to show the true variation in the degree of homophobia felt in Sub-Saharan African countries. Recently, Gallup compiled a worldwide survey, which asked: 鈥淚s the city or area where you live in a good place or not a good place to live for gay or lesbian people?鈥� (McCarthy 2013). Places such as Benin, Chad, and Gabon overwhelming said their country was a bad place, 91%, 94%, 95% respectively, for gay people to live (McCarthy 2013) even though those countries have no laws banning homosexuality. Angola and Botswana, which have up to three and seven year prison terms respectively, for homosexual acts, had two of the highest percentages of people agreeing that their countries were a good place for gays to live (McCarthy 2013).

听听听听听听听听听听听 Many of the discriminatory legal systems were instituted over a century ago and therefore may not accurately represent the true nature of homophobia in their countries. Some African laws were likely too expensive for the underdeveloped states to change and therefore remained through the years even if they did not reflect broad societal interests (Asal 2013). This is apparent in many Sub-Saharan countries, such as Sierra Leone and Angola, where harsh sentences for homosexual acts exist, but the states have never actually prosecuted individuals under the laws. Both Angola鈥檚 and Sierra Leone鈥檚 anti-gay laws have not been updated since the 1800s suggesting that they are less homophobic than they appear when only measuring homophobia by the severity of the existing legal system. Unfortunately, the narrow view of homophobia looked at through the lens of criminalization fails to account for the multifaceted nature of homophobia.

Operationalizing Homophobia

听听听听听听听听听听听 As a result of the inadequacies of using criminalization as an indicator of homophobia, this study divides homophobia into the categories of institutional and personal in order to more accurately operationalize the term. As stated previously, institutional homophobia is the discrimination or exclusion of homosexuals in society through structurally-sanctioned means. In this context, institutional homophobia is anti-gay discrimination perpetuated by African state institutions and organizations. Personal homophobia, on the other hand, is the social stigma of homosexuality held by the population of a state. In this case, intolerance emanates from socially held biases and views of collective groups of African individuals. By utilizing these two terms this analysis attempts to take a holistic view of homosexuality that was not previously attainable when defining homophobia by the severity of anti-gay laws. The two terms are investigated side by side in order to illustrate both of their individual and underlying causes. Both are utilized in this study because the causes of an institutionalized system and the social stigmas held by the general public are likely quite divergent. Conceptualizing homophobia at the micro and macro-level helps to conceptualize the term by giving two sides of homophobia simultaneously.听

听听听听听听听听听听听 Homophobia was operationalized using two composite scores. The institutional homophobia score was created using several questions that conferred the amount of structural discrimination towards homosexuals in a specific country. The composite score includes, on the one hand, the current status of laws regarding homosexuality, and, on the other, the amount of homophobic persecution allowed by the state. Personal homophobia was quantified using a composite score of various surveys from the last 10 years asking respondents how they felt about homosexuality. This allows for a variety of countries within the Sub-Saharan region to be analyzed, which gives a clearer picture of population鈥檚 feeling regarding homosexuality. Table A1 gives a description and statistics about the various variables utilized within the study (see appendix A). Table A2 shows the state-level data collected for the 49 Sub-Saharan countries (see appendix A).

Institutional Homophobia Dependent Variable听听听听听听听听听

听听听听听听听听听听听 The institutional homophobia dependent variable attempts to capture the degree to which societal institutions persecuted homosexuals. To quantify this measure several yes or no questions were asked to gauge the level of a country鈥檚 institutional homophobia. The questions were set up so that one point was awarded if the answer to the question was 鈥測es鈥� and zero if the answer was 鈥渘o.鈥� After points had been designated, they were added to give a final score. A greater score indicates a greater degree of structural discrimination towards gays, and questions were set up so that answering 鈥測es鈥� to any question conferred a higher homophobia score. One component of the score included the severity of laws regarding homosexuality (ranging from the death penalty to legalized gay marriage), [2] and the other sought to capture a greater breadth of institutional homophobia鈥檚 meaning through questions directed at structural discrimination or inequality within societies.

听听听听听听听听听听听 The questions making up the portion of the Institutional homophobia score demonstrating the severity of homophobic laws include: Does the country have the death penalty for homosexual behavior? Is there life imprisonment for homosexual behavior? Does the penalty for homosexual activity range from 10 to 100 years? Does the penalty for homosexual activity range from 1 month to 9 years? Is homosexuality illegal within the country in some form? Is gay marriage illegal within the country? Have anti-homosexuality laws been updated since 1950?[3] Is the government or other organizations legally allowed to discriminate against gays? Has anyone ever been arrested for homosexual activity? Has anyone been prosecuted for homosexual activity?[4]

听听听听听听听听听听听 The questions making up the second part of the Institutional Homophobia score include: Is the state involved in violence against LGBT individuals? Is there a lack of LGBT organizations within the country? Has the state discriminated against or interfered with LGBT organizations? Has the state prevented open demonstrations of support for LGBT issues within the country? [5] Is there a lack of support for homosexuals from individuals in a position of power within the country, including the current leader? [6]

听听听听听听听听听听听 The institutional homophobia variable ranges from 1 at the low end of the homophobic spectrum to 14 at the high end (Table 1). The data鈥檚 mean of just over around 8.6 shows that the countries are skewed towards being more institutionally homophobic.

Table 1: Institutional Homophobia Variable Statistics

Variable	Description	Mean	Std. Dev.	Max	Min
Institutional Homophobia Score	Composite score demonstrating the level of institutional discrimination towards homosexuals within a country.	8.571	3.048	14	1

听听听听听听听听听听听 Potential drawbacks of the variable include the limited availability of information for African countries and inaccuracies of the weight given to particular components of the score. In terms of Sub-Saharan Africa, there is always a question as to the reliability of information. Many instances of violence and persecution go unreported in countries with weak policing infrastructures and few credible news agencies. Especially in terms of the LGBT community, the degree of repression can be difficult to determine. This was the reasoning behind making the measure from multiple questions in order to possibly eliminate inaccuracies of the score given to any one question. Taking the vast majority of the data from either Itaborahy and Zhu鈥檚 report on worldwide anti-homosexuality laws or the U.S. State Department鈥檚 Country Reports on Human Rights Practices also makes the scoring information more credible and consistent. But even so, invariably some countries will have somewhat skewed scores based on missing or inaccurate information, but unfortunately this problem is difficult to avoid. The other difficulty to consider is that every question is given the same weight in the score. It is possible that some aspects should contribute more when quantifying institutionalized discrimination, but doing this only adds more subjectivity to the score. For this reason each question was weighted equally.

Personal Homophobia Dependent Variable

听听听听听听听听听听听 Personal homophobia sought to capture the level of socially held anti-homosexual stigma held by the populations of various African nations. It was quantified using eight opinion surveys that all asked respondents about their feelings regarding homosexuality. The surveys were taken from Gallup, the Pew Research Center, and the World Values Survey.[7] Each survey was given an individual score according to country by dividing the percent of total respondents showing dislike for homosexuality by the same number added to the total percent of respondents showing positive views of homosexuality. In order to get easier numbers to utilize the results were multiplied by 100.

Personal Homophobia Score = (a)/(a + b) * 100听 [8]

This effectively eliminated the percent of respondents who gave an answer along the lines of 鈥渄on鈥檛 know鈥� for which it is impossible to infer the true meaning. Each county鈥檚 score was averaged across the surveys, in order to get a single score for each country. The surveys ranged from the years 2003 to 2013 in order to include as many countries in the analysis as possible. It is both assumed that attitudes remained fairly constant during this time frame and that the questions asked were measuring the same thing, which is reasonable because a comparison of scores for identical countries showed a high degree of agreement across all surveys included. A high personal homophobia scores indicates a greater social rejection of homosexuality across a given population. The personal homophobia score ranges from around 59 to 96 (Table 2). The scores mean is around 90, which is considerably skewed towards the upper bound of the data. This reflects the overall negative attitude felt about homosexuality by the majority of Sub-Saharan populations.

听听听听听听听听听听听 The principle disadvantage of the personal homophobia variable comes from inaccuracies after averaging multiple surveys into one score. These surveys come from different years and ask varying questions. It is possible that while the questions asked were similar across surveys, they were not all investigating the same issue. In addition, the percentage of people having either a favorable or unfavorable view of homosexuality fluctuates from year to year, which could result in an unrealistic averaged score. Having said this, a high correlation between related country scores across surveys makes it far less likely that questions demonstrate vastly different issues or that there has been a dramatic change is opinions regarding homosexuality in the last several years. Therefore, these potential hurdles do not appear to be major problems when drawing meaningful conclusions from the variable.

Table 2: Personal Homophobia Statistics Summary

Variable	Description	Mean	Std. Dev.	Max	Min
Personal Homophobia Score	Composite survey score reflecting the percent of a country鈥檚 population who disproves of homosexuality.	88.82	7.857	95.592	58.971

听听听听听听听听听听听 It is worth mentioning that the two variables for homophobia are related by a Pearson correlation coefficient (r) of 0.625. This suggests that they may influence one another, although the relationship between the two is difficult to define. It is important to recognize that African societies with high levels of homophobic views tend to also have state structures that are strongly discriminatory towards gays. However, the analyses in this paper will not focus on the relationship between the two dependent variables, although one may exist.

Independent Variables

听听听听听听听听听听听 GDP per capita, PPP was used as proxy for economic development (see Table A1 for details). The measure was used instead of GDP per capita to account for variations in cost within different domestic markets. The gross domestic product effectively demonstrates the size of a country鈥檚 economy by measuring the total value of goods produced in the country within the year. By using GDP per capita, a fair approximation of a country鈥檚 relative economic performance can be demonstrated, which can be compared to other countries regardless of population size. One downside is that countries with a single lucrative industry, but with few tangible jobs, can give an inflated GDP per capita, which does not represent the country鈥檚 overreaching economic development. GDP per capita, PPP was taken from the World Bank from 2014.

听听听听听听听听听听听 To measure the overall level of education within a country, the mean years of schooling from the HDI score was utilized. This indicator was used because more years of school across a population on average would be expected to demonstrate a more educated population overall. A drawback of the variable is that it does not show the quality of the populace鈥檚 education. For the purposes of this study, it is assumed that populations with more years of schooling also have received a better education overall than other populations with less average years of education.

听听听听听听听听听听听 A measure of religiosity was difficult to determine by country. Not enough information was available to find the degree of religious adherence for most populations of Sub-Saharan countries. Studies normally rely on measures of individuals rate of church attendance or personal belief in God, but unfortunately this type of data is lacking for Africa. For this reason another measure of religiosity was used. Instead of religious commitment, the percent of a population who placed themselves in the 鈥渦naffiliated鈥� category in religious identification surveys was used as a proxy for the degree of secularity within a country. In the surveys, respondents were given the choice to identify with one of the major world religions, a folk religion, the 鈥渦naffiliated鈥� category, or an unlisted religion. For the purposes of this study, I am defining religiously unaffiliated respondents as non-religious. Therefore I am characterizing countries with more non-religious or unaffiliated individuals as more secular and less religious overall. By doing this, I am assuming that the percent of the population identifying as unaffiliated is inversely related to the relative religiosity of a country. Admittedly, this is not a perfect indicator, but on the whole it appears to be a sensible measure of a state鈥檚 secular nature. In the following analysis the religiosity of a country will be approximated using the inverse value of the religiously unaffiliated percentage of a population. The percentage of unaffiliated was taken from Hackett, Grim, and Cooperman鈥檚 report on the global religious landscape.

听听听听听听听听听听听 To measure the effect of colonial influence I use nominal dummy variables to account for the identity of colonizing countries. Since Britain and France were the two main colonizers, and had very different approaches in terms of the treatment of homosexuals in their landholdings, I compare them to all other former African colonizers as a baseline. Other colonizers include Portugal, Spain, Germany, Belgium, and Italy. The former colonizers of modern day Sub-Saharan countries were identified based on the European country controlling the majority of the territory directly before the country gained independence. I decided to assign colonizers in this fashion because the country last in control most likely set the anti-gay policies in the territory before independence. The identity of the former colonizers was taken form the U.N.鈥檚 report on colonial landholdings (鈥淭rust and Non-Self-Governing鈥� 1999).

听听听听听听听听听听听 Other controls were also utilized for the state-level of analysis, which have been included in similar analyses (Anderson and Fetner 2008; Han and O鈥橫ahoney 2014; Henshaw 2014; Stulhofer and Rimac 2009). One is regime type; it has been suggested that countries with democratic governments may have less repressive state structures towards homosexuals because of the voice given to minorities in democratic systems (Han and O鈥橫ahoney 2014). Polity scores were used to approximate regime type (Marshall 2014). [9]听 The polity score ranges from 10 to -10, where a score of 6 to 10 is defined as a democracy, 5 to -5 as an anocracy, and -6 to -10 as an autocracy. Another necessary control in any study of Africa is ethnic fractionalization because of the large number of ethnic groups present in Sub-Saharan Africa. It is important to show that the phenomenon being investigated is not explained by the region鈥檚 ethnic landscape. Alberto Alesina鈥檚 measure of ethnic fractionalization was used as a control because it encompassed the greatest number of African countries (Han and O鈥橫ahoney 2014). In addition, studies have found varying ages within populations to be an important predictor of homophobic attitudes (Anderson and Fetner 2008; Stulhofer and Rimac 2009). In developed countries, it has been observed that younger people tend to be more tolerant of homosexuality than older individuals, which is generally attributed to changes in social norms regarding homosexuality over different generations (Inglehart 1990). This may not be true of the younger generation in Africa since social norms have been extremely anti-gay (Albright 2003; McCarthy, 2014). It is also conceivable that younger populations, which are generally more politically active, may be seeking homophobic governmental policies and actions. This analysis employs the median age of a country鈥檚 population from the CIA鈥檚 World Factbook to control for the possibility that the age distribution within a country affects opinions about gays and the level of discriminatory actions by state institutions.

Country-Level Results

听听听听听听听听听听听 Several regression models are employed in order to predict variations in institutional and personal homophobia scores. Incremental building of the models is used to individually examine independent variables. Institutional and personal homophobia are both analyzed in almost identical respects, and the framework of incrementally building models is the same for both dependent variables. Models 1, 2, and 3 all isolate singular independent variables and employ bivariate regression models. Models 4, 5, and 6 gradually add more variables and controls in order to see if relationships between variables hold across different models.

听听听听听听听听听听听 Model 1 (addressing hypothesis 1 and 2) examines the effect of GDP per capita, PPP, model 2 looks at education (addressing hypothesis 3 and 4), model 3 (addressing hypothesis 5 and 6) illustrates the effect of religiosity, and model 6 (addressing hypothesis 7 and 8) investigates the influence of colonial rule. Model 6 uses the identity of the former colonizing countries before independence as dummy variables. This allows the nominal criteria of either being a British or French former colony to be used as an independent variable in the analysis. The baseline is countries formerly controlled by countries other than the French or British. Model 4 utilizes the first multivariate regression to compare all of the continuous independent variables. Model 5 adds controls to the models including the median age of the population, ethnic fractionalization, and country polity scores. 听

The Causes of Institutional Homophobia

听听听听听听听听听听听 Table 3, which illustrates the regression models created for institutionalized homophobia, shows tentative support for Inglehart鈥檚 theory (1999). The driving force of post-materialism, economic development (modeled using GDP per capita, PPP), illustrates a significant negative correlation through almost every model. The negative relationship supports hypothesis 1 and suggests, as expected, that wealthier Sub-Saharan countries tend to have less discriminatory state institutions. The relationship seems to support Inglehart鈥檚 prediction that economic development allows states to adopt more socially progressive institutions even within the generally homophobic states of Africa. The relationship holds through almost every model, demonstrating the importance of the variable. Only in Model 6 does GDP per capita, PPP fail to produce

significant results, which could be a result of the number of controls used for the relatively small sample size. The number of Sub-Saharan countries with similarly small GDP values also seems to have potentially inhibited a more robust relationship with the institutional homophobia score.

听听听听听听听听听听听 One of the most striking results of the analysis is religiosity鈥檚 strong and positive correlation through every model. The relationship suggests that more religious African societies will have correspondingly more discriminatory state institutions, which lends credence to hypothesis 5. In states where materialist attitudes are common, societal institutions should reflect the interests of dominant religious norms (Inglehart 1999), and this should be reflected in African societies where religious doctrine generally encourage anti-gay ideologies. By itself, religiosity predicts around a quarter of the variance (its beta coefficient is 0.440 compared to all other variables in the sixth model), which implies that African countries have developed discriminatory institutions partially as a result of religious influences. However, the predictive ability of the variable also indicates that religion does not tell the whole story. It is difficult to tell if there are any outliers in the analysis (Figure 1) because many of the countries have very high religiosity scores, which makes it difficult to determine if groups of countries are deviating from their predicted values. 听

Table 3: Regression Models for Institutional Homophobia (Standard Error in Parenthesis)

Factor	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
GDP Per Capita, PPP (000鈥檚 of US$)	-0.149* (0.061)	--------	--------	-0.185** (0.067)	-0.162* (0.063)	-0.112 (0.066)
Education (in Years)	--------	-0.153 (0.210)	--------	0.514* (0.067)	0.937*** (0.220)	0.600* (0.280)
Religiosity	--------	--------	0.324*** (0.081)	0.323*** (0.085)	0.333*** (0.078)	0.280** (0.079)
Polity Index	--------	--------	--------	--------	-0.109 (0.068)	-0.109 (0.065)
Median Age (in Years)	--------	--------	--------	--------	-0.503** (0.134)	-0.501*** (0.129)
Ethnic Fractionalization	--------	--------	--------	--------	-0.699 (1.501)	-0.831 (1.456)
UK	--------	--------	--------	--------	--------	2.092* (1.011)
France	--------	--------	--------	--------	--------	0.645 (0.861)
Other Colonizers	--------	--------	--------	--------	--------	Baseline
Constant	9.177*** (0.504)	9.130*** (1.064)	-20.803* (7.838)	-24.062** (8.610)	-16.432* (7.957)	-11.086 (8.051)
N	48	47	49	47	45	45
R-Squared	0.114	0.012	0.231	0.344	0.573	0.624

*P-Value < 0.05, **P-Value < 0.01, ***P-Value < 0.001听听听听听

Figure 1: Regression of Institutional Homophobia vs. Religiosity
听

听听听听听听听听听听听 Throughout most of the models, education appears to be a significant predictor of institutional homophobia. Only in model 2 does the variable not show a significant relationship. In the rest of the models, education has a fairly large (with a beta coefficient of 0.414 in the sixth model) and positive relationship, suggesting that African populations with more years of schooling, on average, live in places with more homophobic state structures (contradicting hypothesis 3). It is therefore possible that African school systems are actually reinforcing discriminatory cultures within the region. It seems strange that education鈥檚 significance varies so much throughout the models, which leads one to believe that the variable鈥檚 significance may be masked by other factors. In a separate analysis (not published), I find that the variable becomes especially significant with the addition of the median age variable. Possibly the relationship between institutional homophobia and a country鈥檚 average educational level differs depending on the country鈥檚 age distribution. It could be that in places with a younger population, the educational system can actually reinforce the country鈥檚 homophobic culture to a greater degree because of the greater number of people likely enrolled in school. Interestingly, the education variable only becomes significant with the addition of the GDP per capita and religiosity variables. It seems conceivable that education could reinforce homophobia in both more religious settings and economically less developed countries.

听听听听听听听听听听听 As described by previous studies (Asal 2013; Han and O鈥橫ahoney 2014; Ireland 2013; Sanders 2009), the type of former colonizer has a significant impact on contemporary homophobia. Countries formally ruled by the UK show a positive relationship compared to Portuguese, Spanish, German, Belgian, and Italian colonized countries, which supports hypothesis 7. The variable鈥檚 beta coefficient is fairly large at 0.343 (in the sixth model), showing that type of colonial influence makes quite a difference in the structure of African states. This falls in line with the model of path-dependence, posed by Inglehart, where culturally historic practices are reflected in states鈥� unique paths of development (1999). This suggests that, overall, former colonies of the UK have developed more homophobic institutions, compared to countries with other colonial influences, because the British introduced homophobic colonial laws. An interesting conclusion is that British colonies have actually continued more discriminatory systems partially because their former colonizer originally planted the homophobic seed. In contrast, former colonies of the French show no meaningful difference in the level of institutional homophobia compared to countries colonized by other European powers.

听听听听听听听听听听听 It is interesting to notice that median age of a country was also an important negative predictor in the analysis. The relationship suggests that countries with younger populations on average have more homophobic systems. This could imply that younger populations within Africa are pushing for discriminatory institutions. Importantly, Inglehart鈥檚 theory depends on younger generations to drive changing societal interests during the course of population turnover (1987). With economic development, generational changes in Western societies have made social issues increasingly important. However, in countries where the culture indoctrinates homophobic ideals, population turnover may actually reinforce those ideals, and this seems to be reflected in African states鈥� institutions.

The Cause of Personal Homophobia听

听听听听听听听听听听听 Table 4 illustrates the regression models developed for the analysis of personal homophobia. Interestingly, the initial results show a marked difference from the analysis of institutional homophobia. With the addition of controls, only the type of colonial rule and the average age of the country show significance. In model 5, religiosity approaches the level significant with a P>|t | score of 0.053, but this relationship doesn鈥檛 hold up in the subsequent models. Importantly for this analysis, Inglehart鈥檚 theory does not seem to hold weight since GDP and economic development do not have a relationship with personal homophobia.听 However, upon closer inspection of model 1, there appears to be an outlier country, which has a significant effect on the results (Figure 2). The models seem to be skewed by Gabon whose GDP per capita is one of the highest in Sub-Saharan Africa. The country has a small population of slightly over 1.5 million (鈥淧opulation, total鈥� 2014), and derives its wealth primarily from oil exports. As with most oil exporters, Gabon鈥檚 substantial petroleum revenue appears to be concentrated in the hands of a select few, which makes its GDP per capita artificially high, especially for the underdeveloped region (see Table A2 for information on the regions GDPs per capita). Overall, the country鈥檚 population remains extremely poor and lives in economically difficult conditions. Consistent with Inglehart鈥檚 theory the majority of the population seems to have very negative views of homosexuals, but the inflated GDP per capita impedes an investigation into a possible relationship. The country was therefore omitted and the models were reanalyzed (Figure 3).

听

Table 4: Regression Models for Personal Homophobia (Standard Error in Parenthesis)

Factor	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
GDP Per Capita, PPP (in 000鈥檚 of US$)	-0.876** (0.285)	--------	--------	-0.518 (0.377)	-0.456, (0.367)	-0.553 (0.356)
Education (in Years)	--------	-1.452* (0.606)	--------	-0.012听 (0.775)	0.750听听 (0.849)	0.872 (1.018)
Religiosity	--------	--------	0.808** (0.234)	0.577听听听听听听 (0.286)	0.321听听 (.302)	0.236 (0.292)
Polity Index	--------	--------	--------	--------	-0.081 (0.290)	-0.051 (0.301)
Median Age (in Years)	--------	--------	--------	--------	-1.649听听听听 (0.842)	-1.919* (0.788)
Ethnic Fractionalization	--------	--------	--------	--------	4.560 (6.752)	1.467 (6.399)
UK	--------	--------	--------	--------	--------	4.579 (4.186)
France	--------	--------	--------	--------	--------	7.572* (3.087)
Other Colonizers	--------	--------	--------	--------	--------	Baseline
Constant	91.911*** (1.589)	95.407*** (3.039)	11.502 (22.458)	35.551 (28.901)	83.856* (36.511)	93.804* (34.923)
N	32	32	32	32	32	32
R-Squared	0.240	0.161	0.284	0.346	0.451	0.569

*P-Value < 0.05, **P-Value < 0.01, ***P-Value < 0.001

听听听听听听听听听听听 Table 5 depicts this revised analysis with the dropped variable, and the results are surprising. With one less country in the analysis, GDP per capita becomes highly significant in every model (with a large beta coefficient of -0.771 compared to all other variables in the sixth model). The variable shows a negative correlation with personal homophobia, in support of hypothesis 2 (Figure 3), which means that higher GDP per capita African countries tend to have lower levels of discriminatory views within their populations. This falls squarely in line with the post-materialist explanation by supporting the notion that economic development is at the backbone of homophobic views. Dropping Gabon allows GDP per capita to explain far more in the model than it had before. In the previous analysis, GDP per capita explained about a quarter of the variance in personal homophobia. With the dropped variable, GDP per capita can explain almost two-thirds of the variance. The vast difference in the two analyses is due to the small number of countries included. With an N of only 32 to begin with, the model can be radically changed with the omission of one country. For this reason, an individual-level of analysis of the fifth and sixth wave of the World Values Survey was performed later in this study, to corroborate the results of state-level analysis. 听

Table 5: Regression Models for Personal Homophobia with the Dropped Variable (Standard Error in Parenthesis)

Factor	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
GDP Per Capita, PPP (in 000鈥檚 of US$)	-1.823*** (0.272)	--------	--------	-1.948*** (0.323)	-1.792** (0.471)	-1.803*** (0.415)
Education (in years)	--------	-1.674* (0.612)	--------	0.626 (0.608)	0.843 (0.693)	0.184 (0.807)
Religiosity	--------	--------	0.820** (0.234)	0.145听听听 (0.238)	0.074听听听 (0.255)	-0.098 (0.241)
Polity Index	--------	--------	--------	--------	-0.046听听听 (0.237)	-0.182 (0.235)
Median Age (in Years)	--------	--------	--------	--------	-0.601听 (0.744)	-0.699 (0.681)
Ethnic Fractionalization	--------	--------	--------	--------	3.269 (5.523)	-0.178 (4.977)
UK	--------	--------	--------	--------	--------	8.155* (3.363)
France	--------	--------	--------	--------	--------	6.952** (2.398)
Other Colonizers	--------	--------	--------	--------	--------	Baseline
Constant	94.125*** (1.222)	96.062*** (3.008)	10.209 (22.424)	77.870** (23.912)	92.123** (29.123)	117.128** (27.370)
N	31	31	31	31	31	31
R-Squared	0.607	0.205	0.297	0.625	0.641	0.747

*P-Value < 0.05, **P-Value < 0.01, ***P-Value < 0.001

听

Figure 2: Regression of Personal Homophobia vs. GDP Per Capita, PPP

Figure 3: Regression of Personal Homophobia vs. GDP Per Capita, PPP with Dropped Variable

听听听听听听听听听听听听

听听听听听听听听听听听 It is interesting that neither education nor religiosity shows a robust relationship in this analysis. In the bivariate regression models, both education and religiosity have a significant relationship to personal homophobia, but these correlations fades in subsequent models. It is interesting to notice the near significance of religiosity in the first analysis of personal homophobia, and its complete absence in the revised one. Moreover, Inglehart鈥檚 theory would lead one to believe that religiosity should play a major role in the development of materialist beliefs in underdeveloped societies (1999). Economic development鈥檚 overwhelmingly strong relationship raises the possibility that smaller relationships are hiding in the analysis. With the information given thus far, neither hypothesis 4 or 6 can be accepted or denied, but an individual-level investigation of the World Values Survey is provided to corroborate these results.

听听听听听听听听听听听 The difference between the original and subsequent analyses is also intriguing. Originally, median age showed an important negative relationship, which would suggest that countries with younger populations tend to have populations with more negative views of gays. This would seem to reflect the logic put forth by Inglehart whereby the younger generation is driving societal interests in Africa in a more conservative direction and would also support the results found in the analysis of institutional homophobia. In African societies, where homophobic views are the norm, it would be expected that populations composed mainly of a young, impressionable generation would reflect the homophobic norms that they grew up with. The first analysis on personal homophobia seems to support this notion with a significant negative relationship. However, the relationship disappears in the second analysis, which makes the true relationship less clear.听

听听听听听听听听听听听 The first analysis also shows a positive relationship between colonies of the French and personal homophobia compared to other colonizers, which would suggest that African countries with French colonial influence actually have more homophobic views overall. This appears to contradict hypothesis 8, as well as discredit the notion that homophobic views in Africa originated from colonial laws introduced by the British. However, in the updated analysis, former colonies of the British and French have positive correlations with personal homophobia. The coefficients of both variables are surprisingly large with a beta coefficient of 0.511 for the British variable and 0.441 for the French variable coefficient (in the sixth model). In view of this, influence from both colonizers may have had a lasting impact on the development of homophobic views to the present day.

Individual-Level Results

Individual-Level Methods

听听听听听听听听听听听 To corroborate the results of the state-level analysis, and to shed light on the true causal relationship of various factors on personal homophobia in Sub-Saharan Africa, an individual-level investigation was done using the fifth and sixth waves of the World Values Survey. The fifth wave of the survey was done during the period of 2005 to 2009 and the sixth wave was conducted from 2010 to 2014. This report used only countries in the two waves of the World Values Surveys from the region of Sub-Saharan Africa. The summary statistics for these seven countries and the variables used are available in tables B1 and B2 (see appendix B). The purpose of this additional analysis was to see if relationships posed at the state-level hold up at the individual-level. This aims to avoid the individualistic fallacy, whereby state-level phenomenon are inferred to be the result of aggregated individual actions or interests even when a causal link is not made at the micro-level (Seligson 2002). This is especially important when using Inglehart鈥檚 theory, which justifies changes in populations鈥� interests from material to social on changing lifestyles of individuals. It was recognized in this study that a spurious relationship was possible when casual relationships between state-level variables and individual homophobic views are assumed to be true without individual-level data. Therefore the World Values Survey data was used to back up the claims made at the state-level.

听听听听听听听听听听听 Personal homophobia was measured using questions from the surveys that ranked the respondent鈥檚 view of homosexuality on a 1-10 scale, with 10 denoting homosexuality as always justifiable and 1 as never justifiable. This scale gave a measure of tolerance, which is the inverse of the dependent variables used thus far. To fit the format of the rest of this study, all answers were multiplied by -1 in order to give a measure of homophobia, which serves as the dependent variable.

听听听听听听听听听听听 Several individual-level variables are incorporated from questions in the World Value Survey, which reflect independent variables used in the previous analysis. Income was measured using questions that asked respondents to place their household revenue on a scale from 1-10, with 10 being the highest in society and 1 the lowest.[10] The educational level was quantified using a survey scale from 1-9, with each subsequent point representing a higher educational level attained by the respondent.[11] Religiosity was approximated using questions that asked respondents how important God was in their lives on a scale from 1-10, with 10 being highest and 1 lowest.[12]

听听听听听听听听听听听 To test Inglehart鈥檚 theory of state-level economic development GDP per capita, PPP was utilized. Each respondent was assigned a GDP per capita value corresponding to the year his or her country was surveyed. All GDP per capita values were taken from the World Bank. This study clustered the data around country type in order to standardize individual responses by the country that they originated from. Since only seven countries were represented in the surveys, not many state-level variables could be included. For this reason, neither polity score nor ethnic fractionalization were utilized in the study. In addition, all but two of the countries in the surveys were former British colonies, and none were French, so dummy variables for former colonizer were not included.

听听听听听听听听听听听 Age[13] and gender[14] of the respondent, along with religious denomination, were used as controls, which mirrors similar individual-level studies (Anderson and Fetner 2008; Henshaw 2014; Stulhofer and Rimac 2009). Dummy variables for religious denomination were created to see if one type of religious group mattered more than another.[15] Separate dummy variables were created for Protestants, Roman Catholics, Orthodox, Muslims, and other religions (as a baseline). These denominations were selected because they represent the largest religious groups in the Sub-Saharan region (Cooperman et. al. 2010; Hackett et. al. 2012, 2011).

听Individual-Level Findings

听听听听听听听听听听听 Table 6 shows the results of the individual-level analysis of the Sub-Saharan countries within the fifth and sixth wave of the World Values survey. Interestingly, the results present nearly an identical picture to the state-level analysis done previously. Even while controlling for the socioeconomic status of individuals, the GDP per capita of the country in which the person resides, has the ability to predict homophobic views. The negative correlation gives evidence that individuals residing in more economically developed countries will have less homophobic views compared to others in similar socioeconomic conditions within less developed countries. In light of this, economic underdevelopment is likely an important driver of the larger homophobic cultural views within African states. This seems to fit into a post-materialistic framework where economic development influences cultural norms, which impinge negative views of homosexuality onto the population.

听听听听听听听听听听听 The analyses also show that income, religiosity, as well as religious denomination seem to have standalone effects on people鈥檚 views outside of the states economic development, although their r-squared values seems to imply that these effects are less significant. Income鈥檚 negative correlation illustrates that poorer people tend to hold more discriminatory attitudes towards homosexuals. It would appear that while statewide economic development can shift society鈥檚 view of homosexuality, socio-economic status can also change the views of individuals within Africa. This matches the results of other studies that have investigated the connection between income and homophobia (Anderson and Fetner 2008).

Table 6: World Value Survey Individual-Level Regressions for Personal Homophobia (Standard Error in Parenthesis)[16]

Factor	Model 1	Model 2	Model 3	Model 4
GDP Per Capita, PPP (000's of US$)	--------	-0.213*** (0.012)	-0.217*** (0.0125)	-0.209*** (0.015)
Income	-0.11* (0.041)	-0.128* (0.048)	-0.128* (0.048)	-0.129* (0.049)
Education	-0.141 (0.062)	0.002 (0.019)	0.010 (0.021)	0.011 (0.020)
Religiosity	0.244 (0.112)	0.153* (0.056)	0.152* (0.055)	0.142* (0.053)
Age (in Years)	--------	--------	0.006 (0.004)	0.007 (0.003)
Sex	--------	--------	-0.085 (0.050)	-0.091 (0.054)
Protestant	--------	--------	--------	0.242* (0.073)
Catholic	--------	--------	--------	0.027 (0.080)
Orthodox	--------	--------	--------	0.260 (0.127)
Muslim	--------	--------	--------	0.426 (0.254)
Other Religion	--------	--------	--------	Baseline
N	17773	17773	17772	17735
Constant	-3.409* (1.110)	-2.034** (0.377)	-2.125** (0.379)	-2.232** (0.362)
R-Squared	0.075	0.217	0.218	0.222

*P-Value < 0.05, **P-Value < 0.01, ***P-Value < 0.001听听听听听

听听听听听听听听听听听 An investigation of religiosity was also interesting because the results from the previous analysis of personal homophobia were inconclusive. While the term was on the level of significance in the original models of personal homophobia, the updated analysis no longer showed a role for religiosity. Religiosity鈥檚 significance was likely not apparent in the previous analysis because of the dominant role of the GDP per capita variable, or because of the small sample size. Additionally, the important role of religiosity in the institutional homophobia models would make it surprising that the variable plays no role in personal homophobia. The individual-level analysis is useful to look at relationships that were not apparent at the state-level. Intriguingly, religiosity shows significance in almost all models with a positive correlation, further demonstrating that there is a relationship between religiosity and homophobic views. It is also interesting that religiosity only becomes significant in the models after the addition of GDP per capita, which might suggest that the relationship between homophobic views and religiosity is different at various levels of economic development. This would make sense under Inglehart鈥檚 theory where religious cultural norms change as states go through stages of development (1999).

听听听听听听听听听听听 Another interesting result coming from table 6 is the significance of the Protestant dummy variable compared to the other religions. The relationship to homophobic views is positive, meaning that Protestants tend to be more homophobic compared to other denominations. Several studies have noted that Protestants tend to be more homophobic overall compared to other religious denominations (Adamczyk and Pitt 2009; Ogland and Verona 2014; Olson, Cadge, and Harrison 2006; Whitehead 2010). The results from this study seem to support this notion in African states. In contrast, neither Catholic nor Muslim individuals showed any difference in homophobic views compared to other religious denominations.

听听听听听听听听听听听 Several of the variables, including educational level, age, and gender appear to play no significant role. The lack of robust correlation between education and homophobic views supports the state-level analysis of personal homophobia, which seems like sufficient evidence to reject hypothesis 4. The age and gender variables lack of influence in the models was somewhat surprising since previous studies have found them to be important factors to predict homophobic views (Anderson and Fetner 2008; Stulhofer and Rimac 2009). It seems that the previously established relationships with both age and gender in developed countries may not be true in Sub-Saharan Africa.

Discussion

听听听听听听听听听听听 This paper is noteworthy because, to my knowledge, it is the first time the post-materialist argument has specifically been used to explain variations in African homophobia. While past studies have certainly used similar variables to explain why Africa is so homophobic compared to the rest of the world, this study is unique in how it places the results within Inglehart鈥檚 framework. Viewing African homophobia in this way could have important implications for how future research both analyzes African homophobia, and how the West chooses to respond to the discrimination of sexual minorities in the region.

听听听听听听听听听听听 Primarily, this paper demonstrates the importance of economics in the development of homophobia in the region. This component may be particularly problematic because the region is so impoverished overall, and under conditions of severe poverty, homophobia may be self-reinforcing. As implied by other authors, it is conceivable that African governments use homophobic policies as a way to obtain political support from a homophobic electorate, which in turn may reaffirm the discriminatory views of the population (Awondo 2012; Biruk 2014; Epprecht 2013; Thoreson 2014). In line with this, previous research has postulated that many governments frame homosexuality as a threat and often set up institutions to reflect this mentality for political gain (Henshaw 2014). This is especially relevant in African countries where materialist attitudes predominate because governments can easily institute homophobic laws and issue state sponsored crackdowns on homosexuals. Both can be used to reassure the public of the government鈥檚 ability to effectively govern, and to distract from problems of chronic underdevelopment (Epprecht 2014). For instance in Malawi, when the country went through an economic downturn in 2011, the country鈥檚 president, Bingu wa Mutharika, rallied popular support for his own administration by condemning the country鈥檚 problems on gays, which stoked persecution of the LGBT community (Biruk 2014). In other instances, anti-gay movements and legislation, like ones seen in both Uganda and Senegal, have appeared during times of growing dissatisfaction with their governments鈥� handling of country-wide economic stagnation (Awondo 2012; Thoreson 2014).

听听听听听听听听听听听 The perverse role of religion in this study also suggests large problems for underdeveloped African states because religion remains an integral part of those societies and permeates into the core beliefs of most Africans. Several scholars have addressed how religious cultural ideals, which teach heterosexuality as the only acceptable form of sexual orientation, are an integral part of most African鈥檚 upbringings (Johnson 2007; 鈥淢aking Love a Crime鈥� 2013). It is also of interest that African Protestants seem to show more homophobic views compared to other religious denominations. These results back up studies from other regions in the world (Adamczyk and Pitt 2009; Ogland and Verona 2014; Olson, Cadge, and Harrison 2006; Whitehead 2010). This is particularly important for Sub-Saharan Africa were the largest religious group is Protestant (Cooperman et. al. 2010; Hackett et. al. 2012, 2011), and many missionary groups from the West are of a Protestant denomination. Many blame Western missionaries for Africa鈥檚 extreme homophobia (Epprecht 2013; Johnson 2007; 鈥淢aking Love a Crime鈥� 2014). This study provides tentative evidence that Protestant missionary groups within African states are worsening homophobic sentiments compared to other religious denominations.

听听听听听听听听听听听 The structure of homophobia found in the context of this paper raises further questions. For instance, if economic development underlies African homophobia, should the West increase African aid and development projects to curb discrimination against sexual minorities in the region? If religious ideology has been creating a hostile environment for gays, how should the United States view the actions of domestic religious organizations working in the Sub-Saharan region? In addition, what other factors have created widespread homophobia within Africa? Future research should attempt to address these issues, in order to find the best ways to help the many persecuted individuals living within homophobic states.

听听听听听听听听听听听 Some limitations of this study should also be discussed. The biggest of which comes from doing a study at the state-level of analysis. The possibility of a spurious relationship is always a drawback of a state-level study because the research infers micro-level mechanisms from macro-level data. This could be especially problematic for institutional homophobia, as an individual-level analysis could not be done to back up the state-level data, which was the case for personal homophobia. Therefore, future research should verify the state-level results found in this study. Having said this, the fact that both the analysis of institutional and personal homophobia as well as the individual-level analysis had similar results, suggests that this study鈥檚 findings are valid.

Conclusion

听听听听听听听听听听听 This study sought to explain variations in Sub-Saharan homophobia through the lens of Inglehart鈥檚 post-materialist theory, which suggests that societal interests change from material to social, and survivalist to individualist, as countries economically develop (1987, 1990, 1997, 1999). The results of this study strongly suggest that economic development, religiosity, and colonialism play important roles in the development of African homophobia. Further, the relationships follow the predictions made by Inglehart鈥檚 theory of post-materialism, which suggest that variations in both institutional and personal homophobia can be explained by post-materialism in remarkably similar ways.

听听听听听听听听听听听 This study found that both the economic developmental of a state and a greater individual income appear to counter the effects of homophobia within African nations. Just as increases in standard of living improve individual tolerance of homosexuality at the micro-level, economic development at the macro-level increases whole population鈥檚 tolerance of gays. The religious culture in the region appears to works in conjunction with economic development, and negatively impacts homophobic state structures and views within African states. However, the two types of homophobia seem to diverge in how they were affected by colonial influence and the educational level of society. Institutional homophobia was found to have British influence, which would make sense since the British enacted homophobic laws into all of its colonial landholdings. Unexpectedly, personal homophobia seemed to be greater in both French and British former landholdings. This makes sense in former colonies of the British, where homophobic laws would naturally influence personal views over time, but the same does not hold for former colonies of the French, which largely avoided sodomy laws. It could be that while the French did not put homophobic laws on the books, they enforced discriminatory policies in reality.

听听听听听听听听听听听 Also surprisingly, while education seemed to have a strong positive relationship with institutional homophobia, the variable failed to produce a correlation with personal homophobia. Thus, it would seem that education is not significantly making African individuals more homophobic at the same time that it is bolstering homophobic state structures. This seems improbable since education would be expected to reinforce homophobic state structures through individuals. Possibly homophobia is being taught to individuals in African countries outside of the school systems, which could indicate that simply growing up in certain cultures is sufficient to instill widespread homophobic views in Sub-Saharan populations. Regardless, since the African educational system is either reinforcing the existing homophobic culture or playing no significant effect, it is certainly not acting as a counter-force to homophobia in the region.

听听听听听听听听听听听 With the information provided by this study, a post-materialist explanation for the development of homophobia within Sub-Saharan Africa can be conceived. The findings suggest that some of the first homophobic practices came from the former colonizers of Africa and have influenced a homophobic culture to the present day in many states. At the same time, homophobia has been allowed to spread across the continent as a result of economic underdevelopment, which has provided an environment for a conservative religious culture to prosper. Therefore, economic development and religion seem to share an intimate relationship in the maturation of African homophobia. It appears that African populations stuck in a survivalist mentality tend to uniformly adopt religiously inspired homophobic norms. At the same time, African governments conform to both the interests of the population and to culture鈥檚 religious norms. Interestingly, this study found that even with the underdeveloped Sub-Saharan region, the wealthier a state, the less homophobic the country is overall. It seems that economic development has allowed cultural norms to be secularized in some Sub-Saharan sates, which has helped subdue the homophobic culture somewhat in those places.

听听听听听听听听听听听 If this explanation is correct, there could be important implications for how Western governments respond to African homophobia. Since the phenomenon seems to be driven by a post-materialist framework, one policy recommendation for the West would be to allow improving economic conditions counter homophobia in the region naturally. In the last several years, most African states have been experiencing steady growth, which would be expected to increase societal tolerance of homosexuals over time. As of now, the homophobic actions of African governments have been tied to foreign aid, which has actually increased both anti-Western and ant-homosexual sentiments in many places. In the current economic climate, where African countries continue to show improvements in terms of economic development, Western countries may want to take a step back and allow local gay rights organizations to promote tolerance on their own.

听

Acknowledgments

听听听听听听听听听听听 First, I would like to thank my primary advisor, Dr. Andy Baker, for his continual patience and time. I would not have been able to complete this project without his advice and expertise. I am also very grateful to Dr. Scott Adler for encouraging a discussion about my topic among my peers, which helped mature this thesis. Furthermore, I am very appreciative of the Undergraduate Research Opportunities Program for providing the funding for this research. From the beginning, Dr. Rolf Norgaard helped me develop and bring this project into reality, for which I express the utmost gratitude. Additionally, I want to thank both Dr. Norgaard and Dr. Janet Donavan for participating on my honors committee. Finally, I would like to give my highest praises to Cali Greksa and Cameron Lees for the endless hours they spent proofreading my work.

听

References

Adamczyk, Amy, Cassady Pitt. 鈥淪haping Attitudes about Homosexuality: The Role of Religion and Cultural Context.鈥� Social Science Research 38, (2009): 338-351.

Albright, Madeleine K., Andrew Kohut, Mary McIntosh, Bruce Stokes, Elizabeth Gross, Carroll Doherty, and Nicole Speulda 鈥淰iews of a Changing World.鈥� Pew Global Attitudes Project, (2003):1-136.

Alesina, Alberto, Arnaud Devleeschauwer, William Easterly et. al. 鈥淔ractionalization.鈥� Journal of Economic Growth 8, (2003): 155-194.

Anderson, Robert, Tina Fetner. 鈥淓conomic Inequality and Intolerance: Attitudes toward Homosexuality in 35 Democracies.鈥� American Journal of Political Science 52, no. 4 (2008): 942-958.

Arreola, Sonya, Glenn-Milo Santos, Jack Beck, Mohan Sundararaj, Patrick A. Wilson, Pato Hebert, Keletso Makofane, Tri D. Do, and George Ayala. "Sexual Stigma, Criminalization, Investment, and Access to HIV Services Among Men Who Have Sex with Men Worldwide." AIDS and Behavior (2014): 1-8.Asal, Victor, Udi Sommer, and Paul G. Harwood. "Original Sin A Cross-National Study of the Legality of Homosexual Acts." Comparative Political Studies 46, no. 3 (2013): 320-351.

Awondo, Patrick, Peter Geschiere, and Graeme Reid. "Homophobic Africa? Toward A More Nuanced View." African Studies Review 55.3 (2012): 145-68.

Biruk, Crystal. "鈥楢id for Gays鈥�: The Moral and the Material in 鈥楢frican Homophobia鈥� in Post-2009 Malawi." The Journal of Modern African Studies 52, no. 3 (2014): 447-73.

Blumenfeld, Warren J., ed.听Homophobia: How we all pay the price. Beacon Press, 1992.

Bobo, Lawrence, and Frederick C. Licari. "Education and Political Tolerance: Testing the Effects of Cognitive Sophistication and Target Group Affect." Public Opinion Quarterly 53, no. 3 (1989): 285. Print.

Cooperman, Alan, Gregory Smith, Brian Grim, Allison Pond, Neha Sahgai, Scott Clement, Conrad Hackett, Jacqueline Wenger, Sahar Chaudhry, and Elizabeth Podrebarac. 鈥淭olerance and Tension: Islam and Christianity in Sub-Saharan Africa.鈥� Pew Forum on Religion & Public Life, (2010): 1-133.

鈥淐ountry Reports on Human Rights Practices for 2013.鈥� U.S. State Department, (2013).

Denny, Kevin.听鈥淐ivic returns to education: its effect on homophobia.鈥� UCD Centre for Economic Research 11, no. 8 (2011): 1-22.

Epprecht, Marc. "Africa鈥檚 New Political Homophobia." Current History, (2014): 203-04.

Epprecht, Marc. Sexuality and Social Justice in Africa: Rethinking Homophobia and Forging Resistance. London, UK: Zed books, 2013.

Fiske, Susan T. "Effects of power on bias: Power explains and maintains individual, group, and societal disparities." Psychology Press, (2001): 181-193.

鈥淕DP per capita, PPP (current international $).鈥� The World Bank, 2014. .

鈥淐rabtree, Steve. 鈥淩eligiosity Highest in World鈥檚 Poorest Nations,鈥� Gallup, Aug. 31. 2010. .

Hackett, Conrad, Brian J. Grim, Alan Cooperman, Juan C. E. Ochoa, Chris Gao, Philip Connor and Anna F. Shi, and Noble Kuriakose, Elizabeth A. Lawton, and Elizabeth Prodrebarac. 鈥淕lobal Christianity: A Report on the Size and Distribution of the World鈥檚 Christian Population.鈥� The Pew Forum on Religion & Public Life, (2011): 1-130.

Hackett, Conrad, Brian J. Grim, Alan Cooperman, Juan C. E. Ochoa, Chris Gao, Philip Connor and Anna F. Shi, and Noble Kuriakose. 鈥淭he Global Religious Landscape: A Report on the Size and Distribution of the World鈥檚 Major Religious Groups as of 2010.鈥� The Pew Forum on Religion & Public Life, (2012): 1-82.

Han, Enze, and Joseph O鈥橫ahoney. 鈥淏ritish Colonialism and the Criminalization of Homosexuality.鈥� Cambridge Review of International Affairs 27, no. 2 (2014): 268-288. 听

Henshaw, Alexis L. 鈥淕eographies of Tolerance: Human Development, Heteronormativity, and Religion.鈥� Sexuality and Culture 18, (2014): 959-976.

Herek, Gregory M. "Heterosexism and homophobia." (1996).

Herek, Gregory M. 鈥淪tigma, Prejudice, and Violence Against Lesbians and Gay Men,鈥� Homosexuality: Research Implications for Public Policy, (1991): 60-80.

Horowitz, Juliana M., Katie Simmons, Jacob Poushter, Aaron Ponce, Cathy Barker, and Kat Devlin. 鈥淭he Global Divide on Homosexuality.鈥� Pew Global Attitudes Project report, (2013): 1-26.

Hunter-Gault, Charlayne. 鈥淰iolated Hopes: A Nation Confronts a Tide of Sexual Violence.鈥� The New Yorker. May 28 2012. .

Inglehart, Ronald. 鈥淰alue Change in Industrial Societies.鈥� American Political Science Review 81, no. 4 (1987): 1289鈥�1303.

Inglehart, Ronald. Culture Shift in Advanced Industrial Society. Princeton, NJ: Princeton University Press, 1990.

Inglehart, Ronald. Modernization and Postmodernization: Cultural, Economic, and Political Change in 43 Societies. Princeton, NJ: Princeton University Press, 1997.

Inglehart, Ronald, and Paul R. Abramson. 鈥淢easuring Post-materialism.鈥� American Political Science Review 93, no. 3 (1999): 665鈥�77.

Ireland, Patrick R. "A Macro-Level Analysis of the Scope, Causes, and Consequences of Homophobia in Africa." African Studies Review 56, no. 2 (2013): 47-66.

Itaborahy, L., and Jingshu Zhu. 鈥淪tate-Sponsored Homophobia: a World Survey of Laws: Criminalization, Protection and Recognition of Same-Sex Love.鈥� ILGA, (2013): 1-111.

Gupta, Alok. 鈥淭his Alien Legacy: The Origins of "Sodomy" Laws in British Colonialism,鈥� Human Rights Watch, (2008): 1-70.

Johnson, Cary Alan. "Off the map. How HIV/AIDS Programming is Failing Same-Sex Practicing People in Africa." IGLHRC, (2007): 1-138.

鈥淢aking Love a Crime: Criminalization of Same-Sex Conduct in Sub-Saharan Africa,鈥� Amnesty International, (2013): 1-125.

Marshall, Monty G. 鈥淧olity IV Individual Country Regime Trends, 1946-2013.鈥� Polity IV Project, 2014.

鈥淢edian Age.鈥� The CIA World Factbook, 2014. .

McCarthy, Justin. "Nearly 3 in 10 Worldwide See Their Areas as Good for Gays,鈥� Gallup, Aug. 27, 2014. .

鈥淢orality Interactive Topline Results.鈥� Pew Research Center, (2013).

Naurath, Nicole. 鈥淧erceived Acceptance of Homosexuals Differs Around Globe.鈥� Gallup, Nov. 1, 2007. .

鈥淣et ODA Received Per Capita (Current US$).鈥� The World Bank, 2014.

Ogland, Curtis P., and Ana Paula Verona. 鈥淩eligion and the Rainbow Struggle: Does Religion Factor Into Attitudes Towards Homosexuality and Same-Sex Civil Unions in Brazil?鈥� Journal of Homosexuality 61, no. 9 (2014): 1334-1349.

Olson, Laura R., Wendy Cadge, James T. Harrison. 鈥淩eligion and Public Opinion About Same-Sex Marriage.鈥� Social Science Quarterly 87, no. 2 (2006): 340-360.

Petersen, Larry R., and Gregory V. Donnenwerth. "Secularization and the influence of religion on beliefs about premarital sex."听Social Forces听75, no. 3 (1997): 1071-1088.

鈥淧opulation, total.鈥� The World Bank, 2014. .

Raja, Sheela, and Joseph P. Stokes. "Assessing attitudes toward lesbians and gay men: The modern homophobia scale."听International Journal of Sexuality and Gender Studies听3, no. 2 (1998): 113-134.

Seligson, Michell. 鈥淭he Renaissance of Political Culture or the Renaissance of the Ecological Fallacy?鈥� Comparative Politics 34, no. 3 (2002): 273-292.

Shoko, Tabona. "鈥橶orse than Dogs and Pigs?鈥� Attitudes Toward Homosexual Practice in Zimbabwe." Journal of Homosexuality 57, no. 5 (2010): 634-49.

Strudwick, Patrick. 鈥淐risis in South Africa: The Shocking Practice of 'Corrective Rape'鈥擜imed at 'Curing' Lesbians.鈥� The Independent. Jan. 4 2014. .

Stulhofer, Aleksander, Ivan Rimac. 鈥淒eterminants of Homonegativity in Europe.鈥� Journal of Sex Research 46, no. 1 (2009): 24-32.

Thoreson, Ryan R. "Troubling the Waters of a 'Wave of Homophobia': Political Economies of Anti-Queer Animus in Sub-Saharan Africa." Sexualities 17, no. 23 (2014): 23-42.

鈥淭rust and Non-Self-Governing Territories (1945-1999).鈥� The United Nations. 1999. .

Van Klinken, Adriaan S. 鈥淕ay Rights, the Devil and the End Times: Public Religion and the Enchantment of Homosexuality Debate in Zambia.鈥� Religion 43, no. 4 (2013): 519-540.

鈥淲ave 5 2005-2009.鈥� World Values Survey Association, (2009).

鈥淲ave 6 2010-2014.鈥� World Values Survey Association, (2014).

Whitehead, Andrew L. 鈥淪acred Rites and Civil Rights: Religion鈥檚 Effect on Attitudes Toward Same-Sex Unions and Perceived Cause of Homosexuality.鈥� Social Science Quarterly 91, no. 1 (2010): 63-79.

Wike, Richard, and Juliana M. Horowitz. 鈥淲orld Publics Welcome Global Trade 鈥� But Not Immigration.鈥� Pew Global Attitudes Project report, (2007): 1-140.

听

Appendix A: Summary Statistics

Table A1: Independent Variable Statistics Summary

Variable	Description	Predicted Effect on Dependent Variables	Mean	Std. Dev.	Max	Min
GDP Per Capita, PPP (in 000鈥檚 of US$)	GDP per capita divided by the number of people in the population and standardized for purchasing power parity, which is the standardized price within the domestic market. The measure was further divided by 1000 for ease of analysis. Source: The World Bank.	Negatively correlative with both dependent variables.	4.818	6.731	33.72	0.604
Education (in years)	The average number of years of schooling a person within the population of country receives. Source: a component of the Human Development Index.	Positively correlative with both dependent variables.	4.609	2.112	9.9	1.3
Religiosity	100 minus the percent of a country鈥檚 population who defines their religious following as unaffiliated. Source: the Pew Research Center.	Positively correlative with both dependent variables.	3.902	4.788	20.6	0
Polity Score	Quantifies a country鈥檚 regime type. A score of 6 to 10 is defined as a democracy, 5 to -5 as an anocracy, and -6 to -10 as an autocracy. Source: Polity IV Project.	None	2.596	4.972	10	-9
Age (Years)	The median age of the entire population of a country. Source: the CIA鈥檚 World Factbook.	None	19.633	3.704	33.9	15.1
Ethnic Fractionalization	Measures ethnic heterogeneity within a country鈥檚 population. Source: Alberto Alesina et. al.	None	0.645	0.236	0.93	0
UK	Describes countries formerly controlled by the British directly before colonization. Source: the United Nations.	Positively correlative 听with both dependent variables.	0.408	0.497	1	0
France	Describes countries formerly controlled by the French directly before colonization. Source: the United Nations.	None	0.347	0.481	1	0
Other Colonizers	Describes countries formerly controlled by other European countries, excluding the British and French, directly before colonization. Source: the United Nations.	None	0.245	0.434	1	0

听

Table A2: Summary of Dependent and Independent Variable Statistics Grouped by Country

Country	Institutional Homophobia听 Score	Personal Homophobia Score	GDP Per Capita, PPP (US$)	Education (in years)	Religiosity
Angola	7	74.098	7538	4.7	94.9
Benin	6	87.776	1791	3.2	95
Botswana	7	72.727	15675	8.8	79.4
Burkina Faso	6	88.679	1634	1.3	99.6
Burundi	9	82.716	771	2.7	100
Cameroon	11	91.806	2711	5.9	94.7
Cape Verde	2	----------	6412	3.5	90.9
Central African Republic	7	----------	604	3.5	99
Chad	7	90.588	2081	1.5	97.5
Comoros	11	----------	1559	2.8	99.9
Cote d'Ivoire	7	89.864	3012	4.3	92
Democratic Republic of the Congo	8	87.356	747	3.1	98.2
Djibouti	6	----------	2998	3.8	99.8
Equatorial Guinea	6	----------	33720	5.4	95
Eritrea	11	----------	1195	3.4	99.9
Ethiopia	12	90.662	1354	2.4	100
Gabon	6	95	19260	7.4	94.4
Gambia	10	----------	1666	2.8	100
Ghana	10	91.652	3974	7	95.8
Guinea	9	93.684	1255	1.6	98.2
Guinea-Bissau	5	----------	1242	2.3	95.7
Kenya	12	94.562	2265	6.3	97.5
Lesotho	6	----------	2586	5.9	96.9
Liberia	10	93.671	878	3.9	98.6
Madagascar	9	91.930	1395	5.2	93.1
Malawi	8	94.274	780	4.2	97.5
Mali	7	89.162	1641	2	97.3
Mauritania	12	95.355	3042	3.7	99.9
Mauritius	4	----------	17200	8.5	99.4
Mozambique	6	91.398	1045	3.2	82.1
Namibia	7	----------	9685	6.2	98.1
Niger	9	94.372	913	1.4	99.3
Nigeria	14	91.330	5601	5.2	99.6
Republic of the Congo	7	84.375	5867	6.1	91
Rwanda	5	82.941	1452	3.3	96.4
Sao Tome and Principe	5	----------	2970	4.7	89.9
Senegal	11	96.570	2296	4.5	100
Seychelles	8	----------	24189	8.4	97.9
Sierra Leone	9	87.802	1927	2.9	99.9
Somalia	14	----------	----------	----------	100
South Africa	1	58.971	12504	9.9	85.1
South Sudan	10	----------	2330	----------	99.5
Sudan	14	----------	3372	3.1	99
Swaziland	9	----------	6683	7.1	87.4
Tanzania	13	93.354	1775	5.1	98.6
Togo	11	90.217	1390	5.3	93.8
Uganda	13	96.592	1410	5.4	99.5
Zambia	12	84.982	3181	6.5	99.5
Zimbabwe	11	93.765	1700	7.2	92.1

听

Appendix B: Individual-Level Data

Table B1: Individual-Level Statistics Summary

Variable	Description	Mean	Std. Dev.	Max	Min	N
Personal Homophobia	A score from 1-10 denoting how respondents felt about homosexuality. 1 denoted never justifiable and 10 denoted always justifiable. The scores were multiplied by -1 so that the highest scores would denote dislike of homosexuality.	-2.456	2.368	-10	-1	18583
GDP Per Capita, PPP (000's of US$)	GDP per capita divided by the number of people in the population and standardized for price within the domestic market, and further divided by 1000 for ease of analysis. Source: The World Bank.	5.525	4.487	12.507	0.813	18898
Income	A scale of income from 1-10 denoting how a respondent鈥檚 household income compared to others in society. 10 denoted the highest income and 1 denoted the lowest.	4.920	2.218	10	1	18183
Education	The highest level of education attained by the respondent using nine possible groups. 1 point denoted the lowest level of education (no formal education) and 10 denoted the highest (full university education).	4.881	2.203	9	1	18823
Religiosity	A score from 1-10 denoting how important God was in the respondent鈥檚 life. 10 denoted the highest importance and 1 the least.	8.988	1.758	10	1	18783
Age	The respondent鈥檚 age in years.	34.097	13.862	98	16	18897
Gender	The respondent鈥檚 gender. (Male=1 and Female=2).	1.501	0.500	2	1	18898
Protestant	A dummy variable denoting a respondent鈥檚 religious denomination. A 1 was assigned to every respondent who said they belonged to a Protestant religious denomination. This included people who identified as being a Protestant, Evangelical, South African Evangelical, and Pentecostal. All other respondent鈥檚 were given a 0.	0.386	0.487	1	0	18844
Catholic	A dummy variable denoting a respondent鈥檚 religious denomination. A 1 was assigned to every respondent who said identified as a Roman Catholic. All other respondent鈥檚 were given a 0.	0.226	0.418	1	0	18844
Orthodox	A dummy variable denoting a respondent鈥檚 religious denomination. A 1 was assigned to every respondent who said they identified as Orthodox. All other respondent鈥檚 were given a 0.	0.068	0.252	1	0	18844
Muslim	A dummy variable denoting a respondent鈥檚 religious denomination. A 1 was assigned to every respondent who said they identified as a Muslim. All other respondent鈥檚 were given a 0.	0.090	0.287	1	0	18844
Other Religion	A dummy variable denoting a respondent鈥檚 religious denomination. A 1 was assigned to every respondent who said they identified anything other than a Protestant, Evangelical, South African Evangelical, Pentecostal, Roman Catholic, Orthodox, or Muslim. All other respondent鈥檚 were given a 0.	0.223	0.416	1	0	18844

听

Table B2: Summary Statistics for Countries in the Fifth and Sixth World Values Surveys

Country	Average Personal Homophobia Score	GDP Per Capita, PPP in US$ (Year)	Average Income Score	Average Education Score	Average Religiosity Score	Average Age
Ethiopia	-1.509	813.1 (2007)	5.158	4.421	9.212	29.933
Ghana	-1.583	2,554 (2007) and 3,446 (2011)	4.659	3.924	9.738	32.384
Nigeria	-2.013	5,217 (2011)	5.179	5.195	9.031	31.224
Rwanda	-1.515	1,024 (2007) and 1,426 (2012)	4.395	3.894	8.571	34.209
South Africa	-3.856	10,149 (2007) and 12,507 (2013)	5.087	5.787	8.578	37.656
Zimbabwe	-1.813	1,626 (2011)	4.859	5.121	9.605	33.772
Zambia	-2.622	2,733 (2007)	5.364	4.776	9.182	29.787

听

听听听听听听听听听听听 [1] In this paper I use the words homosexual, gay, LGBT, queer, and sexual minority interchangeably to mean people who have sex with others of the same gender. Although there are slight differences between the terms, they are all used in the literature to describe men who have sex with men and women who have sex with women. LGBT includes bisexuals and transgender individuals, but the way the term is used in the course of this paper does not significantly distort or change its meaning.

[2] In terms of scoring for laws regarding homosexuality, one point was awarded for the given level of severity of the law and one point was also given for each level of severity beneath it. For example, countries with the death penalty for homosexual acts got six points, one for every 鈥渉omosexual legality鈥� section, and countries with life imprisonment got five points.

听听听听听听听听听听听 [3] 1950 was chosen because it represents at least a generation away from modern policy- making in the region. Homophobic laws passed within the more modern era should reflect a more discriminatory cultural mentality.听听听

听听听听听听听听听听听 [4] These questions, regarding homophobic laws, were answered using Itaborahy and Zhu鈥檚 report on the criminalization of homosexuality.

[5] To answer these questions the U.S. State Department鈥檚 Country Reports on Human Rights Practices for 2013 were consulted for each individual country.

听听听听听听听听听听听 [6] To answer this questions the academic database Lexis Nexis (say who this is) was consulted. For each country the commands 鈥渢he name of the country OR the name of the country鈥檚 leader AND homosexuality OR gay鈥� was entered into the search engine. This produced all newspaper, magazine, and journal articles from the period of 2000-2014 that mentioned both homosexuality and the African country or the state鈥檚 leader in the same article. Public statements by elites or country leaders were then used to gauge opinions of homosexuality.

[7] The two Gallup surveys asked, 鈥淚s the city or area where you live a good place or not a good place to live for gay or lesbian people?鈥� (McCarthy 2013; Naurath 2007). The Pew Research Global Attitudes Project survey asked, 鈥淪hould society accept homosexuality?鈥� (Horowitz 2013). In another survey, the Pew Research Center鈥檚 Global Attitudes and Trends project asked, 鈥淒o you personally believe that homosexuality is morally acceptable, morally unacceptable, or is it not a moral issue?鈥� (鈥淢orality Interactive鈥� 2013). A Pew Forum on Religion and Public Life survey told respondents, 鈥淣ext, I鈥檓 going to read some behaviors. For each, please tell me whether you personally believe that it is morally acceptable, morally wrong, or is it not a moral issue?鈥擧omosexual behavior鈥� (Cooperman et. al. 2013). This survey was not included because the answers were broken up into four categories, which did not match the scoring framework used for the rest of the reports. The Pew Global Project Attitudes Survey asked respondents to complete the statement, 鈥淗omosexuality should be鈥ccepted鈥攔ejected鈥� (Wike and Horowitz 2007). An earlier Pew Research Global Attitudes Project stated, 鈥淗omosexuality should be accepted by society鈥� to respondents who were asked to answer, 鈥淵es or no鈥� (Albright et. al. 2003). The fifth and sixth World Values Surveys asked, 鈥淥n this list are various groups of people. Could you please mention any that you would not like to have as neighbors? 鈥揌omosexuals鈥� (鈥淲ave 6鈥� 2014; 鈥淲ave 5鈥� 2009).

听听听听听听听听听听听 [8] a is the percent of total respondents showing dislike for homosexuality, and b is the total percent of respondents showing positive views of homosexuality.

[9] In the analysis 2013 polity scores were used. Where data was not available for 2013, scores from 2012 were employed.

听听听听听听听听听听听 [10] Question V253, from the fifth wave, and V239, from the sixth wave, stated, 鈥淥n this card is an income scale on which 1 indicates the lowest income group and 10 the highest income group in your country. We would like to know in what group your household is. Please, specify the appropriate number, counting all wages, salaries, pensions and other incomes that come in.鈥�

听听听听听听听听听听听 [11] Question V238, in the fifth wave, and V248, in the sixth wave, asked, 鈥淲hat is the highest educational level that you have attained?鈥� The nine levels, from lowest to highest on the scale, were no formal education, incomplete primary school, complete primary school, incomplete technical/vocational training, complete technical/vocational training, incomplete secondary school (university preparatory), complete secondary school (university preparatory), some university level education (without degree), and finally university level education (with degree).

听听听听听听听听听听听 [12] Questions V192, in the fifth wave, and V154, in the sixth wave, asked, 鈥淗ow important is God in your life? Please use this scale to indicate. 10 means 鈥渧ery important鈥� and 1 means 鈥渘ot at all important.鈥�

听听听听听听听听听听听 [13] Age, in years, of the respondent was determined using questions V237 and V242, from the fifth and sixth waves respectively, which stated, 鈥淭his means you are ____ years old.鈥�

听听听听听听听听听听听 [14] The respondent鈥檚 gender was determined using observation by the survey administrator in questions V235, in the fifth wave and 240, in the sixth wave.

听听听听听听听听听听听 [15] QuestionsV185, in the fifth wave, and V144, in the sixth wave, asked, 鈥淒o you belong to a religion or religious denomination? If yes, which one?鈥�

听

听听听听听听听听听听听 [16] A dummy variable controlling for the specific wave of the World Values Survey was included in a final model (not published) but was not significant. This was done to demonstrate that the results are not due to changes in attitudes in the surveys from the fifth to the sixth wave.

Miracle on the Han River, Gookjin Jeong

(Back to Top)

Miracle on the Han River: A Regression Analysis of the Effect of Chaebol Dominance on South Korea鈥檚 Economic Growth

Table of Contents

Abstract听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 1

List of Tables and Figures听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 2

Introduction听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 3
Background: Chaebol Origins and Characteristics听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 8
Research Question听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 11
Theory听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 15
Analysis听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 18
1. Methodology听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 18
2. Data Sources, Coding, and Coverage听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 21
3. Model听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 22
4. Descriptive Statistics听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 23
5. Results听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 27
6. Qualitative Discussion: Diversification and Vertical Integration听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 33
Conclusion听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 40

Bibliography听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 43

听

听
听

Appendix听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 45

听

Abstract

听

How did South Korea, one of the poorest countries in the world in the 1960s, become a model of success for developing economies just five decades later? This paper analyzes the role that chaebols, large Korean conglomerates, played in bringing about robust economic growth to the country since the conclusion of the Korean War. The broad research question was: What is the effect of chaebols鈥� dominance in the Korean economy on Korea鈥檚 growth rate? I argue that chaebols have had a positive impact on Korea鈥檚 growth rate by being the main drivers of Korea鈥檚 export-oriented industrialization, achieved through economies of scale. I also discuss two most prominent features of chaebols鈥攄iversification and vertical integration鈥攁nd how they enabled chaebols to achieve overall efficiency in production. Using cross-sectional and time- series data, I perform regression analysis to examine the effect of chaebols鈥� dominance on Korea鈥檚 growth rate, compared with the average growth rate of the rest of the world and with Korea鈥檚 own long-term average growth rate. The results show that, holding constant exports, education, savings rate, population growth, regime type, and natural resources, the effect of chaebols鈥� dominance on Korea鈥檚 average growth rate was statistically significantly positive when compared with the average growth rate of the rest of the world, but statistically insignificant when compared with Korea鈥檚 own average growth rate. The findings of this study present a new perspective on the isolated effect of chaebol dominance on Korea鈥檚 economic growth, and lay a foundation for several avenues for future research.

听

List of Tables and听 Figures

Table 1: Operationalization of Variables听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 19

Table 2: Top 10 Chaebols鈥� Share of Korean GNP and Korea鈥檚 GDP Growth听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 24

Table 3: Summary Statistics听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 26

Table 4: Multivariate Regression 鈥� No Dummy Variables听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 29

Table 5: Multivariate Regression 鈥� Year Dummy Variables Only听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 30

Table 6: Multivariate Regression 鈥� Country Dummy Variables Only听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 31

Table 7: Multivariate Regression 鈥� Year and Country Dummy Variables听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 32

Table 8: Diversification Patterns of Top 30 Chaebols听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 35

听

听
听

Table 9: The Pattern of Chaebol Expansion听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 36

听

听
听

Figure 1: Top 10 Chaebols鈥� Share of Korea鈥檚 GNP vs. Korea鈥檚 GDP Growth, 1973-1995听听听听听听听听 24

听

Figure 2: Chaebols鈥� Share of GNP and GDP Growth (Entire Sample)听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 27

听

听
听

Figure 3: Chaebols鈥� Share of GNP and GDP Growth (Korea Only)听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 28

听

I.Introduction

听

After a devastating civil war that left the country in ruins, South Korea showed remarkable economic recovery in the following decades. Indeed, from the early 1960s to the mid-1990s, the Korean economy experienced the highest average rate of growth in the world鈥� roughly 9%.1 Scholars later dubbed this phenomenon the 鈥淢iracle on the Han River,鈥� referring to Korea鈥檚 rapid postwar export-fueled growth, accompanied by industrialization, technological

breakthroughs, an education boom, and a large increase in living standards all across the country. One of the poorest countries in the world at the beginning of the 1960s, Korea joined the ranks of industrial democracies within a single generation, becoming a member of the Organization for Economic Cooperation and Development (OECD) in 1996.2 Today, Korea is the 15th largest economy in the world, is a G-20 member economy, has been appropriately called an 鈥淎sian tiger鈥�

nation along with three others, and boasts large automobile and electronics industries, some of which (e.g. Samsung, LG, Hyundai) have become household names in the United States. What explains this complete transformation? Some have argued that it primarily comes down to the export-led growth strategy adopted by Korea, often grouping Korea with several other Asian economies that also grew robustly by using the same strategy. However, this overly general explanation is neither sufficient nor satisfactory. Indeed, one scholar aptly emphasizes that development strategy for any country is a complex, multidimensional problem involving wide- ranging areas such as the establishment of long-term targets for growth and structural change, and that development strategy should not be reduced to the dichotomy between export-oriented

growth and import-substitution industrialization.3听 The general consensus in the current literature

听

听
听

听

1 Peter M. Beck, 鈥淩evitalizing Korea鈥檚 Chaebol,鈥� Asian Survey 38 (1998): 1018.

2 Wonhyuk Lim, 鈥�Chaebol and Industrial Policy in Korea,鈥� Asian Economic Policy Review 7 (2012): 69.

3 Ha-Joon Chang, 鈥淭he Political Economy of Industrial Policy in Korea,鈥� Cambridge Journal of Economics 17 (1993): 153.

听

is that a form of organizing the structure of companies unique to South Korea鈥攃alled the chaebol鈥攑layed an instrumental role in developing a highly versatile economy that was able to adapt to the changing international demands over time.

There has been ample literature dealing with different factors that served as impetus for Korea鈥檚 economic growth. Much of it has focused on Korea鈥檚 exports-driven growth strategy, while some have also examined the effect of other variables, such as democratization, emphasis on education, and population growth. The linkage between export performance and economic growth has been studied widely and there is a clear positive relationship between the two.

Scholars explain that, both in general and specifically for Korea, an export-oriented industrialization strategy tends to exert a positive effect on growth through increases in saving and investment, technology, and possibility of structural change stemming from opening industries to world markets.4 There also appears to be a mutually reinforcing relationship between democracy and economic growth, as suggested by economist John F. Helliwell. He

asserts that income has a positive effect on democracy because increases in income levels are likely to increase people鈥檚 demands for political and civil freedoms, and that democracy also has a positive effect on growth via education and investment.5 Population growth, too, has a non- negligible positive effect on growth by encouraging the development of industries that are labor- intensive.6

听

听
听

While, as mentioned, there is a consensus in the literature that chaebols had an important role in growing and developing the South Korean economy, most lack a definitive focus of specific aspects of chaebols that enabled them to have a positive effect on growth. For instance,

4 Rudiger Dornbusch et al., 鈥淜orean Growth Policy,鈥� Brookings Paper on Economic Activity 2 (1987): 404.

5 John F. Helliwell, 鈥淓mpirical Linkages Between Democracy and Economic Growth,鈥� British Journal of Political Science 24 (1994), 225.

6 Wong Hock Tsen and Fumitaka Furuoka, 鈥淭he Relationship between Population and Economic Growth in Asian Economies,鈥� ASEAN Economic Bulletin 22 (2005): 314-330.

听

many scholars have placed a lot of emphasis on the emergence and development of chaebols, tracing their evolution from the 1960s to the 1990s.7 They focus on the role of state intervention and the state-business cooperation, mentioning such government policies as the Five-Year Economic Development Plans.8 Some even go into the cultural and political foundations for chaebol, discussing such factors as characteristics of Korean employees and behavioral patterns of Korean people in general.9 While these authors certainly do examine the unique features of chaebols, such as family ownership and management, diversification into different industries, and vertical integration, they do not specifically delve into the effect of these features on Korea鈥檚 economic growth.

Given this review of the current literature, the central question of my research is as follows: What is the effect of chaebols鈥� dominance in the Korean economy on Korea鈥檚 growth rate? The purpose of this paper is to examine the link between chaebol dominance in Korea and Korea鈥檚 economic growth. The aforementioned features of chaebols enabled them to grow quickly and establish a dominant position in the Korean economy. I contend that this dominant presence of chaebols, or the chaebol-centered growth strategy, played a crucial role in promoting Korea鈥檚 growth between the 1960s and the 1990s. In an attempt to establish the statistical significance of this relationship, I conduct multivariate regression analysis. Separately, I also offer insights into the key characteristics of chaebols, and how these characteristics contributed to their dominance in Korea鈥檚 economy. This combination of regressions analysis and qualitative discussion cannot easily be found in the current literature. Thus, by discussing chaebol dominance as a specific variable that contributed to Korea鈥檚 growth, I hope to fill in the gap in the existing set of explanations.

听

7 Haggard, Lim, and Kim 2003; Chang 2003; Kim 1997; Chang and Chang 1994; Jwa 2002.

8 Jwa 2002; Kim 1997; Chang 2003.

9 Chang and Chang 1994.

听

This paper explains that chaebols鈥� dominance has had a positive impact on growth through their superior operation of economies of scale. After their creation and initial development in the 1960s, chaebols quickly and firmly established their dominant position in Korea鈥檚 economy with stellar export performance. In the process, chaebols gradually became what they are today鈥攈uge multi-industry conglomerates run by specific families. Government policy at the time helped form chaebols鈥� special structural features, the two most prominent of which are vertical integration and diversification. The former essentially refers to a process of indigenizing inputs for different stages of production, and the latter to the expansion of production into different, sometimes unrelated, industries. After chaebols became the cornerstone of Korea鈥檚 export-driven growth strategy by the mid- to late-1960s, they were considered, both by the government and private actors like banks, to be in an advantageous position to operate economies of scale to maximize profits. Faced with the increasing demands from the growing world markets, the Korean government pushed chaebols, who had the necessary capital base, into new industries, rather than subsidizing non-chaebol firms specializing in such industries. While this further widened the gap between chaebol firms and non-chaebol firms, this system of exclusive support for the large, established chaebol firms helped attain economies of scale more efficiently, thereby maximizing profits and promoting economic growth.

In addition to an investigation of the mechanism through which chaebol dominance has affected Korea鈥檚 economic growth, I present findings from regression analysis to confirm the statistical significance of the effect. Using time-series/cross-sectional data obtained from extant literature, the World Bank鈥檚 World Development Indicators and the Quality of Government Institute, I perform a series of multivariate regressions. The results suggest that, holding constant

听

export performance, education, savings rate, population growth, regime type, and availability of natural resources, chaebol dominance, measured by top 10 chaebols鈥� share of Korea鈥檚 GNP, has a statistically significant effect on Korea鈥檚 GDP growth.

Finally, I explore how chaebol features such as diversification and high degree of vertical integration helped chaebols operate economies of scale and overall efficiency of production. To this end, I look closely at how and why chaebols decided to diversify and vertically integrate in the first place, citing specific examples of chaebols engaging in this practice.

听

II.Background: Chaebol Origins and Characteristics

听

What, then, is the chaebol? Many scholars define it in different ways. Most simply, a chaebol is a Korean business group that encompasses many subsidiary firms under the same name.10 There are many Korean companies with more than two subsidiaries that are controlled by one family, but the Korean government and media typically use 鈥渃haebols鈥� to denote thirty largest business groups.11 Every year, the Korean government identifies the thirty largest business groups and publishes a listing of their affiliates under the 鈥淢onopoly Regulation and Fair Trade Act鈥� to block any anticompetitive behaviors. The act defines chaebols鈥� affiliates as those for which 鈥渆ither more than 30% of whose issued shares are owned by one person, his relatives, or a company controlled by him, or whose management such as appointing its officers is substantially affected.鈥�12 There are numerous chaebol groups in Korea, the biggest and most prominent of which include companies like Samsung, Hyundai Motor, SK, and LG. It is important to note a few key characteristics that describe how chaebols are structured and designed to operate. First, many chaebols are family-run; often, the chairmanship is inherited from father to son. The current chairmen of Samsung and Hyundai, for example, have their respective sons in the company鈥檚 vice president positions, and they are expected to become chairmen after their fathers retire. Similarly, the current leaders of SK Group and LG also inherited their positions from family-member founders. Moreover, the rest of the founding family members virtually control chaebol firms through cross-holding of equity.13

听

听
听

听

10 Sea-Jin Chang, Financial Crisis and Transformation of Korean Business Groups: The Rise and Fall of Chaebols

(Cambridge: Cambridge University Press, 2003), 9.

11听 Ibid., 10.

12听 Ibid.

13 Seung-Rok Park and Ky-hyang Yuhn, 鈥淗as the Korean chaebol model succeeded?鈥� Journal of Economic Studies

39 (2012): 261.

听

Second, chaebols typically consist of a large number of affiliate companies that are engaged in different businesses, all operating under a single corporate group. This is called diversification, which essentially means that big chaebols are not simply engaged in one type of business, but instead in numerous different types of businesses that are often unrelated. For instance, the general American public perception of Samsung, the biggest chaebol in Korea, seems to be that it is a company focused primarily on producing electronic devices such as TVs, computers and mobile phones, but its corporate reach goes far beyond consumer electronics.

Other industries in which Samsung subsidiaries are engaged include shipbuilding, construction, life insurance, surveillance and defense, and advertising. Although not as extensive as Samsung, other big chaebols in Korea are engaged in businesses they are not primarily known for. The average number of subsidiaries of the top 30 chaebols hovered around 22 for the 1990s,14 up from around 15 for the previous decade.

Third, chaebol firms exhibit a high degree of vertical integration. Highly vertically integrated firms seek to 鈥渋ndigenize intermediate inputs imported from foreign upstream industries,鈥� and Korean firms that would later emerge as big chaebols showed early efforts in the 1960s by trying to vertically integrate through technology acquisition, human resource development, and construction of optimal-scale plants aimed for the global market.15 Samsung

Electronics, for instance, makes most of the components that go into its own phones. This is in

听

stark contrast with Apple, which does its own operating system and designs but contracts out all of the underlying hardware. The result is that it takes significantly less time for Samsung to go from contriving an idea to bringing a final product to markets, because it does most of the production process itself.

听

14 Sung Hee Jwa, The Evolution of Large Corporations in Korea: A New Institutional Economics Perspective of the Chaebol (Cheltenham: Edward Elgar Publishing, Ltd., 2002), 45.

15听 Lim, 74.

听

Given their sizes and the extent of their diversification and vertical integration, the chaebol firms鈥� contribution to Korean economy has recently been more significant than ever before. In 2011, the sales of Korea鈥檚 ten largest companies were equal to about 80% of the country鈥檚 GDP that year.16 Even more surprising is the fact that among all the chaebols that exist in Korea, the top firms are by far the biggest contributors. According to the Korea Fair Trade

Commission, in 2013, South Korea鈥檚 top four conglomerates generated some 90% of the total net profit earned by the top 30 conglomerates.17 In the following sections, I discuss how chaebols became such big conglomerates, as well as how they established a dominant presence in Korea鈥檚 economy.

听

听
听

听

16 Eun-jung Kwon, 鈥淭op ten chaebol now almost 80% of Korean economy,鈥� The Hankyoreh, August 28, 2012, accessed December 9, 2014,

17 Koichi Kato, 鈥淭op four chaebol generate 90% of South Korean conglomerate profits,鈥� Nikkei Asian Review, April

10, 2014, accessed December 9, 2014, four-chaebol-generate-90-of-South-Korean-conglomerate-profits.

听

III.Research Question

听

As mentioned earlier, there has been plenty of literature that explores sources for Korea鈥檚 economic miracle. While much of it has examined Korea鈥檚 export promotion and its effect on growth, some have also analyzed the effect of other variables, including regime type, emphasis on education, and population growth.

Exports, for one, are a key component of gross domestic product, so an increase in exports is an important source for GDP growth. Unlike many Latin American countries that adopted import substitution industrialization (ISI) as their principal method of achieving economic growth throughout the post-WWII era,18 Korea quickly adopted an export-driven growth strategy. Korea started with an import substitution strategy in the early 1950s, with foreign aid financing the trade gap.19 By around 1960, however, it had 鈥渧irtually exhausted the possibility of rapid growth through import substitution of nondurable consumer goods and intermediate inputs.鈥�20 Additional import substitution of machinery, consumer durables, and their intermediate inputs was rejected because the domestic market was too small.21 Thus, faced with implied negative consequences of a continued ISI strategy, the Korean government intervened in order to promote exports-oriented industrialization. Intervention in the form of trade restrictions, subsidies, and credit allocation was pervasive.22 Scholars have noted that this type of outward-oriented strategy is supported by the efficiency of freer trade over a restrictive trade regime.23 Because export promotion is more closely related to free trade than is import substitution, Korea鈥檚 export-driven growth strategy has had a positive effect on saving and

听

听
听

听

18 Werner Baer, 鈥淚mport Substitution and Industrialization in Latin America: Experiences and Interpretations,鈥� Latin American Research Review 7 (1972), 95.

19 Dornbusch et al., 405.

20听 Ibid.

21听 Ibid.

22听 Ibid., 403.

23听 Ibid., 404.

听

investment, technology, and the possibility of structural change coming from opening industries to world markets.24 In summary, the high positive linkage between export performance and economic growth is 鈥渁n accepted fact in development economics.鈥�25

On the relationship between regime type and economic growth, Helliwell explains that there is a two-way reinforcing linkage between democracy and economic growth.26听 The effects of income on democracy, he notes, are found to be robust and positive. This positive effect is also theoretically supported because increasing levels of education and income are 鈥渓ikely to increase citizen demands for many things, including the range of political and civil freedoms that characterizes democratic systems.鈥�27 Democracy also exerts a positive effect on growth via investment and education.28听听 This two-way strengthening relationship has held true for Korea.

One scholar notes that the improvements in the quality of democracy contributed to sustaining high economic growth after the democratic transition in Korea, arguing essentially that an authoritarian Korea would likely have 鈥渇ailed to adjust to the changes in the economic environment caused by the end of the Cold War, the IT revolution, and the spread of globalization.鈥�29

Education, too, has served as a key determinant of industrialization鈥攁nd in turn of

听

economic growth鈥攊n Korea. Amsden asserts that while the role played by education in economic development in Korea ought not to be deified, a well-educated population in general

听

24听 Ibid.

25听 Ibid.

26听听 Helliwell, 225-248.

27听 Ibid., 225.

28听 Ibid.

29 Hyung Baeg Im, 鈥淏etter democracy, better economic growth? South Korea,鈥� International Political Science Review 32 (2011): 596.

听

and a plentiful supply of trained engineers in particular have been critical inputs in the industrialization process.30

Population growth also has a non-negligible effect on growth. Tsen and Furuoka, through statistical analysis, have established bidirectional Granger causality between population and economic growth for Korea.31 They explain the rationale with a discussion about minimum wage:

The issue of population and economic growth is [鈥 closely related to the issue of minimum wage. Population growth enlarges labour force and, therefore, will push wage down. The standard economic labour demand model predicts that low wage will raise the demand for labour. As a result, the welfare of the economy is likely to increase.

Moreover, low wage would encourage industries that are labour intensive. Low wage is said to be an important factor that has contributed to the industrialization of Asian newly industrialized economies (NIEs), namely Korea, Hong Kong, Taiwan, and Singapore.32

听

While, as mentioned, there is plenty of literature examining the emergence, development and success of chaebols, most neglect a discussion about the specific aspects of chaebols that enabled them to have a positive effect on growth. For instance, many scholars have placed a lot of emphasis on the creation of chaebols and have traced their evolution from the 1960s to the 1990s.33 They focus on the role of state intervention and the state-business cooperation, mentioning such government policies as the Five-Year Economic Development Plans.34 Some even go into the cultural and political foundations for chaebol, discussing such factors as characteristics of Korean employees and behavioral patterns of Korean people in general.35 While these authors certainly do examine the unique features of chaebols, such as family ownership and management, diversification into different industries, and vertical integration,

听

30 Alice H. Amsden, Asia鈥檚 Next Giant: South Korea and Late Industrialization (New York: Oxford University Press, Inc., 1989), 238-9.

31 Tsen and Furuoka, 314-330.

32听 Ibid., 315.

33 Haggard, Lim, and Kim 2003; Chang 2003; Kim 1997; Chang and Chang 1994; Jwa 2002.

34 Jwa 2002; Kim 1997; Chang 2003.

35 Chang and Chang 1994.

听

they fail to delve into the effects of these features on Korea鈥檚 economic growth. Given this gap in the current literature, the central question of my research is as follows: What is the effect of chaebols鈥� dominance in the Korean economy on Korea鈥檚 growth rate?

Through this paper I hope to fill in an existing gap in the literature about the effects of chaebols鈥� dominance鈥攐r growing importance and influence鈥攐n Korea鈥檚 economic growth. But while my hypothesis links chaebol dominance with growth, I also consider what caused chaebols to dominate the Korean economy in the first place. In this paper I present multivariate regression analysis and comparative perspective, neither of which can easily be seen from the current literature. In addition to the regression analysis, I offer a qualitative discussion, explaining how and why top chaebol firms diversified and integrated vertically, the effect of diversification and vertical integration on efficiency and, by extension, on growth. Examining the pattern of chaebol expansion is important, as it helps further our understanding of the regression analysis by providing insight into the extended mechanism by which chaebols鈥� share of Korea鈥檚 output increased so significantly, thus complementing the results of regressions.

听

IV.Theory

听

States in which government-business collusion prevails can be effective in promoting domestic economic growth. The mechanism involved starts with a discussion about identifying the interests of the actors involved鈥攏amely, national government and corporations. A chief interest of government officials is to stay in power. The means to achieve maintenance of power can vary depending on the type of government, but a common way to secure power seems to be increasing prosperity and delivering economic growth. On the other hand, it is well known that all firms have a core interest in maximizing profit. To this end, firms strive to operate economies of scale to bring down production costs, and the government seeks out a strategy that can help the country grow. The result of interaction between growth-pursuing government and profit- maximizing firms is a collaborative policy that can serve to further both ends.

On the basis of the above logic, I generate my two-part hypothesis for this paper. The proposed argument expects an increase in chaebols鈥� dominance in Korea to increase Korea鈥檚 economic growth rate, (a) relative to the world鈥檚 average growth rate and (b) relative to its own growth rate over time. The primary mechanism through which chaebols鈥� dominance affects Korea鈥檚 economic growth is economies of scale and export promotion through government support and incentives. Chaebols, like all firms, have a core objective of maximizing profit. The Korean government鈥檚 key concern in the late 1950s and early 1960s was delivering economic growth. Thus, the Korean government, in order to deliver economic growth to its people, selected a few firms based on family relations and essentially made them more productive by providing various types of economic support, including subsidies and tax breaks contingent on strong performance. In the 1960s, Korea also chose to promote growth through export-oriented industrialization, given its lack of natural resources and its relatively small

听

domestic market. With this directive, the government provided subsidies and tax breaks to chaebol firms for export promotion, helping them grow in size, market share, and overall contribution to the economy. Further, exporters have a productivity advantage even before they start exporting;36 on balance, exporters are more productive, not as a result of exporting, but because only the most productive firms are able to overcome the costs of entering export markets.37 In the case of Korea, the most productive firms have been chosen and nurtured by the government.

Initially, a set of policies by the Korean government designed to identify and support certain firms based on family relations led to chaebol formation. Because of the incentives鈥攕uch as tax breaks, subsidies and allocation of foreign capital at low interest rates鈥攑rovided by the government, chaebols began to diversify into various, sometimes unrelated, industries. The Park regime鈥檚 astute analysis of international economic demands during its reign enabled Korea to focus on key strategic industries that changed from decade to decade. Beginning with the advent of the heavy and chemical industrial (HCI) drive in the 1970s, and fueled by continued support from the government, chaebols were able to achieve economies of scale and increase profits.

This, coupled with robust export markets provided by Japan and the United States, ultimately helped Korea produce export-led economic growth. Jae-Seung Shim and Moosung Lee explain that the government determined that chaebols were more suited to taking advantage of economies of scale:

Big businesses were seen by government to be in a better position than small firms in the efficient use of resources and in strengthening competitiveness through economies of scale.38 [Government] measures resulted in a rapid increase in production and export capacity since business success for firms depended to a large extent on their export

听

听
听

听

36 Bernard, et.al, 鈥淔irms in International Trade,鈥� Journal of Economic Perspectives 21 (2007): 106.

37听 Ibid.

38 Jae-Seung Shim and Moosung Lee, The Korean Economic System: Governments, Big Business and Financial Institutions (Hampshire: Ashgate Publishing, Ltd., 2008), 50.

听

performance. Chaebol were in a much better position than small firms in benefiting from economies of scale as well as in expanding production capacity. The government, too, favored a small number of large firms over a large number of small firms, since it was convenient for them to focus government attention in this way. Moreover, in allocating loans, banks preferred chaebol to small firms.39

Thanks to the size they have achieved by the start of the HCI drive in the early 1970s, chaebols were in an advantageous position to attain economies of scale through large-scale production. This, along with Korea鈥檚 export-driven growth strategy, caused the importance of exports in Korea鈥檚 economy to increase dramatically throughout the decade. Indeed, to achieve scale economies, the government allocated export licenses to only a few companies, furthering the rise of chaebols. For instance, it chose Hyundai and Daewoo to develop power plant facilities and Hyundai, Samsung, and Daewoo to build ships.40 As a result, from 1971 to 1979, the proportion of exports to GNP increased from 16% to 36%.41 Thus, we can infer that chaebols鈥� attainment of scale economies鈥攁long with continued government support in the form of subsidies and tax breaks鈥攈elped them achieve better export performance, which in turn contributed directly to economic growth.

听

39听 Ibid., 82.

40 Sea-Jin Chang, Financial Crisis and Transformation of Korean Business Groups: The Rise and Fall of Chaebols

(Cambridge: Cambridge University Press, 2003), 54.

41听 Ibid.

听

V.Analysis

听

Methodology

Conducting regression analysis for the effect of chaebol dominance on Korea鈥檚 economic growth first requires operationalizing the independent and dependent variables. The independent variable, chaebol dominance, can be appropriately measured by looking at the value added share of Korean GNP by top chaebol firms. I chose the chaebols鈥� value added share of GNP rather than sales share of GNP, since the latter gives 鈥渁n inflated estimate of the importance of the chaebol [鈥 because transactions of semi-finished goods between firms within a group are included.鈥�42 On the other hand, value added is defined as 鈥渘ew value created by the producer during the production process,鈥� and since GNP represents the sum of value added produced within a country from the production activity of all economic actors (plus overseas income earned by domestic residents minus domestic income earned by overseas residents), the total value added for a company can be considered a strong indicator of the importance that company holds in the national economy.43 In this sense, the value added measure is a much more reasonable and accurate measure of chaebols鈥� influence. This measure also makes intuitive sense because as chaebols become a more dominant influence in the Korean economy, their share of the country鈥檚 total output also increase鈥攁nd vice versa. The dependent variable, economic growth, is expressed using the standard measure of annual GDP growth rate.

听

In an attempt to reduce omitted variable bias, or bias produced by leaving out variables that also have an effect on economic growth, I have included in my regressions a list of variables that scholars agree affect economic growth: savings rate, population growth, exports, education,

听

42 Robert C. Feenstra, Tzu-Han Yang and Gary G. Hamilton, 鈥淏usiness Groups and Product Variety in Trade: Evidence from South Korea, Taiwan and Japan,鈥� Journal of International Economics 48 (1999): 74.

43 Jung-ae Lee, 鈥淭op four chaebol鈥檚 value added accounts for nearly 10% of GDP,鈥� The Hankyoreh, September 18,

2014,

听

natural resources, and regime type. Although this is not a comprehensive list, including these variables in the regressions helps to isolate the effect of chaebol dominance. Table 1 below summarizes the operationalization of the variables.

听

Table 1. Operationalization of Variables

听

听
听

听

Variables听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 Measures

听

听
听

听

Chaebol dominance听听听听听听听听听听听听听听 Top 10 chaebols鈥� share of Korean GNP

听

Economic growth听听听听听听听听听听听听听听听听听 GDP growth (annual %)

听

Exports听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 Exports of goods and services (% of GDP)

听

Education听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 Gross enrollment ratio for tertiary (ISCED 5 and 6)

听

Savings rate听听听听听听听听听听听听听听听听听听听听听听听听听听 Gross savings (% of GDP) Population growth听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 Population growth (annual %)

Regime type听听听听听听听听听听听听听听听听听听听听听听听听听 Democracy (dichotomous; 1 if democracy, 0 otherwise) Natural resources听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 Total natural resources rents (% of GDP)

Source: World Bank World Development Indicators; Quality of Government Institute; Lim (2003); Hattori (1997); Jwa (2002).

听

Three points are worth clarifying with regard to the operationalization of the variables listed in Table 1. First, in operationalizing the education variable, I selected gross enrollment for tertiary education from a number of variables that could also serve as viable measures of education because the enrollment ratio variable most widely covered the countries and years included in my sample. Second, again for the education variable鈥檚 measure, ISCED refers to the International Standard Classification of Education, constructed by the United Nations as a result

听

of international agreement and adopted formally by UNESCO member states.44 The U.S. equivalents of ISCED level 5 (short-cycle tertiary education) and level 6 (Bachelor鈥檚 or equivalent level) are completion of an Associate鈥檚 degree and completion of a Bachelor鈥檚 degree, respectively.45 Third, total natural resources rents as a share of GDP can be calculated by taking the difference between the price at which the output from natural resources is sold and the costs

of extraction and production. An increasing share of natural resources rents thus essentially represents a higher contribution of natural resources to GDP. Therefore, the resource rents share variable is hereafter referred to as the contribution of natural resources to GDP.

Because of the limited availability of data for chaebols鈥� share of GNP for Korea, the initial table of regressions without sample restrictions showed a decrease in the number of observations each time a new control variable was added. Thus, in order to keep the sample consistent, the final multivariate regression with all of the controls included, which produced the smallest number of observations, effectively determined the sample for the entire regression analysis.

First, using a T-test, I establish that the difference in the average GDP growth rate between Korea and other countries in the sample is statistically significant鈥攖hat is, the growth rate differential observed between Korea and the other countries is consistent and not due to chance.

Second, I run a simple bivariate regression鈥攊.e. a regression of growth on chaebol dominance鈥攊n order to see if there is any association between the two variables. A statistically significant result here would then provide the basis for further multivariate regression analysis.

听

听
听

听

44 UNESCO Institute for Statistics, International Standard Classification of Education: ISCED 2011, UIS/2012/INS/10/REV (2012),

45 National Center for Education Statistics, Comparative Indicators of Education in the United States and Other G-8

Countries: 2011, NCES 2012-007 (October 2011),

听

Third, provided that the bivariate regression yields statistically significant results, I proceed to run a series of multivariate regression, including the other aforementioned variables to reduce omitted variable bias. If the regression coefficient on the chaebol share of GNP is still statistically significant after controlling for omitted variables, then we can conclude that chaebols鈥� dominance in the Korean economy is a factor that causally contributed to Korea鈥檚 rapid

economic growth in the decades following the conclusion of the Korean War.

听

Finally, to supplement the empirical findings for the regression analysis, I incorporate a qualitative discussion on the mechanism by which chaebols were able to attain a dominant position in the Korean economy, explaining how and why chaebols diversified and integrated vertically.

听

Data Sources, Coding, and Coverage

Data for the main explanatory variable, top 10 chaebols鈥� value added share of GNP, were procured from three different authors. Lim (2003) and Hattori (1997) provide data points for the years 1973-197846 and 198347, respectively, and Jwa (2002) gives data points for 1985-1995.48 Due to the lack of availability of data for this measure for 1979-1982 and 1984, in coding this variable I repeated the value for 1978 for 1979-1982 and the value for 1983 for 1984. Data for one of the control variables, regime type, were collected from the Quality of Government Institute database. Data for the main dependent variable, annual GDP growth rate, were obtained from the World Bank collection of development indicators. Data for the dependent variable,

听

46 Stephan Haggard, Wonhyuk Lim and Euysung Kim, Economic Crisis and Corporate Restructuring in Korea: Reforming the Chaebol (New York: Cambridge University Press, 2003), 47.

47 Tamio Hattori, 鈥�Chaebol-style Enterprise Development in Korea,鈥� The Developing Economies 35 (1997): 466.

48听 Jwa, 33.

听

annual GDP growth rate, as well as remaining control variables鈥攅xports as share of GDP, tertiary enrollment ratio, savings rate, population growth, and contribution of natural resources to GDP鈥攈ave all been obtained from World Bank鈥檚 World Development Indicators online database. The final dataset includes 2,999 observations stretched across 163 countries for the years 1970-2008. A table listing all the countries and the years in the restricted sample for regressions is included in the appendix.

听

Model

听

My hypothesis stated that higher chaebol dominance, as measured by top 10 chaebols鈥� share of Korea鈥檚 GNP, would lead to higher economic growth. The basic linear bivariate regression model therefore included economic growth as the dependent variable and chaebol dominance as the independent variable. In order to account for the effects of other factors on economic growth, I controlled for six additional variables: 1) exports, measured by the total share of exports in GDP; 2) education, measured by the gross enrollment ratio for tertiary education; 3) savings rate, measured by the gross savings as a share of GDP; 4) population growth measured in annual percentage; 5) regime type, a dichotomous variable that indicates whether a country is democratic or autocratic; and 6) natural resources, measured by the contribution natural resources to GDP.

Model: Change in GDP growth = 伪 + 尾1 (Top 10 chaebols鈥� share of Korea鈥檚 GNP) + 尾2 (Exports as share of GDP) + 尾3 (Gross tertiary enrollment ratio) + 尾4 (Savings rate as share of GDP) + 尾5 (Population growth rate) + 尾6 (Regime type) + 尾7 (Contribution of natural resources to GDP) + 蔚

听

The tables of regressions in the results section show that one 鈥渇ull鈥� multivarirate regression鈥攔egression of GDP growth on chaebol dominance with all six control variables included鈥攚as run for each table. The model above describes the initial regression with no dummy variables. Subsequent regressions included year dummy variables to control for year- specific idiosyncratic effects, country dummy variables to control for country-specific idiosyncratic effects, and both country and year dummy variables to control for both effects.

听

Descriptive Statistics

After their creation in the early 1960s, chaebols quickly became an important influence on the state of the Korean economy. Supported by the government鈥檚 trade policies and financial resources, such as subsidies and tax breaks, select chaebols grew especially rapidly and were essentially prevented from ever failing. Indeed, active government intervention can be found in every stage of the chaebol creation process. Entry barriers, erected by the government, allowed for only a handful of firms to dominate certain industries, and policies designed to support these firms allowed them to become chaebols. If these firms found themselves in trouble, government- led industrial and corporate restructuring awaited. Thus, this cycle created the perfect environment for chaebol creation and retention.49

听

听
听	听

49 Sung-Hee Jwa, The Evolution of Large Corporations in Korea: A New Institutional Economics Perspective of the Chaebol (Cheltenham: Edward Elgar Publishing, Ltd., 2002), 17.

听

Berlin Revisited, Daniel Naftel

(Back to Top)

听

Existing research suggests that, in most cases, foreign elites hold influence over the domestic media coverage of international crises only when they are considered to be hostile. This is unsurprising, as it is often their hostile rhetoric that drives the events of crises. Thus far, however, there has been no distinction made between the various types of rhetoric these hostile leaders engage in, or how different categories of statements might have different degrees of influence on domestic media coverage. For example, might these foreign leaders鈥� more benign public utterances, that do not change the status quo of a crisis, still have an impact on how salient the press views the issue to be? This study attempts to fill this gap. Specifically, I study the influence of the non-status-quo-changing statements of Soviet Premier Nikita S. Khrushchev on American media coverage of the Berlin Crisis of 1958-1960. This period was selected as a case study for the following reasons: (1) U.S. President Dwight D. Eisenhower displayed a strong reticence to publicly commit to a specific course of action, largely forgoing his role as president to guide policy discourse at the outset of the crisis, (2) Premier Khrushchev was viewed in the American media as being the ultimate voice in the formation of Soviet foreign policy, giving him outsized influence on the direction of the crisis, and (3) the historical memory of the Berlin blockade of 1948-1949 predisposed the American media to pay attention to the crisis and view it as urgent. An analysis of New York Times articles was conducted and suggests that Khrushchev鈥檚 non-status-quo-changing 鈥減ropaganda鈥� did have some influence on how salient the crisis was in the news; however, this influence was only seen during periods of high tension, suggesting that these effects are conditional on the media鈥檚 preexisting levels of interest.

听

Table 听of Contents:

听

Abstract:.............................................................................................................................. 2

Introduction:........................................................................................................................ 4

Literature Review:............................................................................................................... 6

Historical Background:...................................................................................................... 13

Methods and Data Collection:........................................................................................... 30

Findings and Discussion:................................................................................................... 35

Conclusion:........................................................................................................................ 41

Sources Cited:.................................................................................................................... 43

Appendix:.......................................................................................................................... 49

听

Introduction:

听

Heads of state face a two front war during international crises. Not only must they define and promote the national interest, cognizant of the pressures of the international system of power and alliances, but they must do so in a way that takes into account their own domestic political incentives (Putnam 1988). Their rhetoric often demonstrates the existence of this dual audience, signaling intent and resolve to the opposition as well as framing the crisis in a way that can be 鈥渟old鈥� to their domestic constituents. Rhetoric is not simply a product of the strategic situation, however. Leaders鈥� words can also actively alter their incentives and shape the course of events; in crisis diplomacy, words are often synonymous with actions. 听听

Audience cost theory suggests one such possible use of crisis rhetoric: augmenting domestic political risks in order to strengthen bargaining positions. When leaders publicly escalate international crises they create the risk of political ramifications suffered in the case of reneging on their stated commitments. More importantly, the potential new consequences that these leaders would experience in the case of backing down from their threats are greater than those they would have received in the case of immediately capitulating to their opponent (Fearon 1994). When a leaders makes a public promise to stand firm on an issue and then are forced to backtrack, they appear weak and incompetent. This reality implies that public escalationng of a crisis is analogous to the idea of burning bridges in game theory, whereby a leader effectively cuts off the possibility of an action (in this case backing down), to lend credibility to the leader鈥檚ir threats. Despite the perceived benefits of garnering potential audience costs, there are many cases in which a leader would instead choose to avoid actively escalating a crisis, opting to keep the situation out of the public eye as much as possible. Baum (2004, 612) notes that in this situation, it would be advantageous for an opponent to co-opt the other leader鈥檚 bargaining position by drawing attention to the delicate point of contention.

Is this possible, however? Even if unintentional, the ability to manipulate an opponent鈥檚 domestic political payoffs would have significant implications for how crisis diplomacy is conducted. It seems evident that an opposing leader would attract press attention by making statements that threaten the status quo or provide new developments in a crisis鈥攂ut what about more routine, non-status-quo changing statements that could be termed as 鈥減ropaganda,鈥� and that are not necessarily aimed at an international audience?

In order to answer this question, I will examine the Berlin Crisis of 1958-1961, a period in which U.S. President Dwight D. Eisenhower鈥檚 public rhetoric displays an avoidance of undue escalation and press attention in the face of domestic and international pressures. Using media content analysis of New York Times articles from this period, I will test to see if Soviet Premier Nikita S. Khrushchev鈥檚 own public rhetoric (which was much more hostile in tone than Eisenhower鈥檚) had any influence on the salience of the crisis in the American media. Part one of this paper builds a theoretical framework that explores the relationship between public opinion, presidential rhetoric, media coverage, and foreign elites. The second part explores the background of the crisis, arguing that the strategic conditions of the time made the American press susceptible to foreign influences. Finally, part three analyzes the ways in which the media framed the crisis, and how Khrushchev (inadvertently or not) influenced their coverage of the situation.

Literature Review:

听

The Public and the Shaping of Foreign Policy Consensus:

According to traditional international relations scholarship, the general public鈥檚 views on foreign policy were ill-informed, irrational, and inconsistent. This 鈥淎lmond-Lippmann鈥� consensus, based on realist models of foreign policy decision-making, did not face significant challenges until relatively recently when a series of studies suggesting that the public holds both rational and consistent views, brought civil society back into the conversation (Holsti 1992; Page & Shapiro 1988). Not only did the public appear rational, they also had a marked influence on how politicians chose and implemented their grand strategies. Attempting to bridge the gap between realist views of an international system that places certain demands on leaders, and the realities of domestic politics, Putnam (1988) proposed a two-level game analysis of heads of state, wherein these leaders must balance the demands of the electorate with developments and realities of the international system.[1] In this model, 鈥渄iplomatic strategies and tactics are constrained both by what other states will accept and by what domestic constituents will ratify鈥� (Evans, Jacobson, & Putnam [Moravcsik] 1993, 15). This focus on the domestic as a 鈥渓imiting鈥� force on international behavior has been particularly popular in research into democracies and how domestic political restrictions could be a possible explanation for the theory of the democratic peace.

However, the influence of the public on foreign policy is not always consistent (Monroe 1979). Explanations for this variation include different institutional structures (Soroka & Wlezien 2004; Prins & Sprecher 1999; Trumbore 1998; Risse-Kappen 1991), the nature of media coverage (Baum & Potter 2008), or even presidential beliefs on the role of politicians as democratic agents (Foyle 1999, 9). Despite the breadth of the extant literature, it is generally agreed upon that the influence of public opinion on the foreign policy agenda is contingent on the attention paid by the public to specific issues, often called issue salience (Soroka 2003; Page & Shapiro 1983).[2]

Paying Attention: What makes the News?

The study of issue salience, particularly what makes some issues more salient to the public than others, provides an important link between the demands of the public and the incentives of those with decision-making power. This process of determining salience, better termed agenda setting, has been studied extensively both in communications literature (Scheufele & Tewksbury 2007; Dearing & Rogers 1996; McCombs & Shaw 1993) and by political scientists (Wood & Peake 1998; J. Cohen 1995; McCombs & Shaw 1972). One of the most marked findings of these studies is the link between media coverage and what the public considers salient (Soroka 2003; Wanta & Hu 1993; Semetko et al. 1992). In an oft-quoted line, Cohen (1963, 13) notes that the media, 鈥渕ay not be successful much of the time in telling people what to think, but it is stunningly successful in telling its readers what to think about.鈥� The average American is more separated from and less directly influenced by foreign affairs than domestic ones, making them even more reliant on media cues (Baum & Potter 2008; Auerbach & Bloch-Elkon 2005).

The central role the media plays as an agenda-setter, and the implications that this may hold for foreign policy decision-making, begs a pertinent question: who or what determines what the press covers? The media acts as a key interlocutor between the public and political elites (Baum & Potter 2008; Auerbach & Bloch-Elkon 2005; Zaller 1992; Page & Shapiro 1988; Rogers & Dearing 1988), but exactly how each of these three actors influence the act of news-making appears to be complex, conditional, and multifaceted. 听This interrelatedness has generated a wide range of findings, from studies suggesting that the media is mostly responsible for determining what the public and democratic leadership consider significant (Cook et al. 1983; McCombs & Shaw 1972) to other models suggesting that the leadership merely reacts to media coverage of events (Zhang & Meadows 2012; Edwards & Wood 1999) or even that elite discourse is the main driver of the political agenda (J. Cohen 1997; Dearing & Rogers 1996). Walgrave and& Van Aelst (2006) deal with these theoretical difficulties by arguing that the media does hold significant agenda setting power, but that this power is contingent on the political climate.

Behr and& Iyengar (1985, 53) find that televised journalism has an influence on the public agenda, but that 鈥渞eal-word events鈥� and the issues that the president addresses also determine what voters consider significant. In fact, it appears that the president may have an outsized influence as a setter of the public agenda (Villalobos & Sirin 2012; Edwards & Wood 1999; J. Cohen 1997, 1995; Dearing & Rogers 1996; Kingdon 1995; Light 1991; Bond & Fleisher 1990; Downs 1972), and more specifically in the realm of foreign policy (Lawrence 2004; J. Cohen 1995; Bennett 1990). Perhaps the best example of this elite influence on the media is in the observation that press coverage is closely 鈥渋ndexed鈥� to elite discourse in both terms of content and media frames (Zaller & Chiu 1996; Bennett 1990). Whether due to an overemphasized respect for official versions of stories or merely journalistic laziness, the result is that ruling elites鈥� policy debates bound the national conversation and topics that fall outside of these constraints are usually passed over. In the realm of foreign policy, presidents are central to the agenda setting process, and their public addresses shape the topics that are discussed in the news (Druckman & Holmes 2004; Iyengar & Kinder 1987; Bennett 1990; Behr & Iyengar 1985). As the media seems to give a great deal of weight to presidents鈥� words, there seem to be certain strategic benefits to this influence during international crises, particularly when the relationship between the media and the public is considered. In their study of the relationship between public opinion, the media, and elite policies in shaping public views of foreign countries, Zhang and& Meadows (2012, 89) argued that, 鈥淸their] findings support the postulate that the news media is a diplomatic device as powerful as diplomatic statements.鈥� According to some lines of reasoning, the two might be more connected than these authors acknowledge.

Audience Costs: Leveraging Deference to the Executive

Audience cost theory has recently helped to operationalize presidential influence over agenda setting within the strategic framework of crisis diplomacy. According to this model of international crisis bargaining first proposed by Fearon (1994), every public statement that a head of state makes in an international crisis bargaining situation is expensive鈥攖hat is to say, it generates some sort of risk for the speaker鈥攁nd so is an effective signal of resolve that weeds out 鈥渃heap talk.鈥� Fearon and others (Tomz 2007; Slantchev 2006; Schultz 2001; Fearon 1997) have found that leaders who decide to verbally escalate an international conflict bestow upon themselves the risk of political ramifications that would hurt them if they were to eventually capitulate. Interestingly, these consequences are much greater than if they were to relent at the beginning of the conflict. This observation suggests that leaders are able successfully to leverage their domestic political capital to engage in bridge burning and thus strengthen their bargaining position. The generation of audience costs changes these leaders鈥� payoffs, making further conflict escalation more attractive for them after a 鈥減ublic battle of wills and honor鈥� has begun. This change in personal incentives signals resolve much more effectively than sunk costs, such as military build-ups, do (Fearon 1997). According to this theory, leaders who are better at generating audience costs should prevail over their rivals in international crises as their political futures depend on the bargaining outcome (Fearon 1994). In their analysis of Militarized Interstate Dispute (MID) data, Partell and Palmer (1999) agree that there is some empirical evidence to support this claim.[3]

Despite these benefits, there are many cases in which heads of state choose instead to minimize their public exposure during crises (Baum 2004). This act of 鈥済oing private,鈥� however, can run into various problems. While leaders can choose not to escalate crises themselves, they are not always able to keep them out of the news, as more often than not, crises take on a life of their own in the media (Knecht 2010; Wood & Peake 1998). In cases where the president is in a particularly perilous diplomatic position, this public attention can negatively impact them given that greater public exposure often reduces a president鈥檚 ability to diplomatically maneuver (Page & Shapiro 1983). Following this line of reasoning, Baum (2004, 612) postulated that an enemy leader would have the incentive to bring greater attention to a crisis in which their opponent was perceivably weak in order to force their hand.听

The Public Eye and the Foreign Statesman

Is this sort of influence over media content even possible? Various studies of foreign voices in the news suggest that the opinions of foreigners count for very little in the eyes of the American public (Entman 2004; Page, Shapiro, & Dempsy 1987). While Hayes and Guardino (2013) argue that media sources and politically active members of the public can rely on foreign opinions in the face of a unified elite that acts against the wishes of the public, their study deals with foreign elites who are considered friendly and not acting against the best interests of the United States. For example, it is doubtful that an adversarial leader, such as Vladimir Putin, would be able to change the hearts and minds of the American public with claims that he is merely out to protect Russian citizens from Ukrainian oppression. After all, his expansive views of the Russian people were described in the American press as a 鈥減erverted version of the Western-developed and United Nations-sanctified humanitarian doctrine of the 鈥榬esponsibility to protect鈥欌€� (Ash 2014, New York Times). Is it possible, however, for enemy leaders to alter Americans鈥� views of issue salience by keeping sensitive issues in the news?

The difficulty in answering this question is highly rhetorical in nature. The words of foreign leaders do not have a direct connection to the public; instead they must first pass through the interpretive and conceptual frameworks of members of the media and domestic elites. One aspect of this information processing is known as framing.[4] By Entman鈥檚 (1993, 52) definition,: 鈥淸t]To frame is to select some aspects of a perceived reality and make them more salient in a communicating text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation鈥� (as quoted by Scheufele 1999). Unlike agenda setting, which deals with the salience of individual issues and events, framing can be seen as the aspects of those events that are pulled out by the press in order to construct a narrative (Scheufele 1999). Consider news headlines: the press engages in the act of reducing an entire event or situation into a short phrase necessitates that a simplified narrative of the story be created. In the case of public diplomacy, the way in which the media frames elites鈥� statements creates new realities in the rhetorical landscape by placing the emphasis on information that fits into their preconceived notions of what is, or rather should be, considered important. These fames have a significant impact on the national discourse and on how people view an event (Terkildsen & Schnell 1997; Iyengar 1991). In the case of my study, the categorization of a statement being status-quo-changing or not largely depends on media frames; leaders sometimes have little control over how their statements are received by foreign audiences. Ultimately, framing theory provides a useful methodological scaffold to approach this rhetorical ambiguity and explore how Khrushchev influenced the American media.

Historical Background:

听

听听听听听听听听听听听 There are several reasons that the Berlin Crisis, specifically the period between November 10, 1958 and May 1960, presents an ideal case study for examining the influences of foreign elite rhetoric on the salience of foreign policy issues. The legacy of the Berlin blockade a decade earlier left the American media and public predisposed to pay attention to the situation and consider it to be urgent. In contrast with these naturally raised levels of interest, U.S. President Eisenhower displayed many of the signs of 鈥済oing private鈥� as modeled by Baum (2004), due to a precarious diplomatic situation in which the United States had staked too much of its prestige on the isolated city鈥檚 fate. Lastly, a view in the American media perpetuated by certain American government officials that Premier Khrushchev was the sole voice of import in Soviet foreign policy led the media to pay more attention to his words than they might have otherwise.

Berlin and the Legacy of the Blockade: Defining the Crisis

听听听听听听听听听听听 As the defeat of the Third Reich appeared imminent, Allied leaders met at Potsdam, Germany in July 1945 to discuss the future of Europe, and perhaps most important, the future of Germany. On August 1, it was decided that the defeated state would be divided into four occupation zones, run by the US, USSR, UK[AH1]听, and France. Although it was understood that 鈥淕ermany [would] be treated as an economic unit,鈥� it was to be politically decentralized with the goal of 鈥渢he complete disarmament and demilitarization of Germany,鈥� 鈥渢o destroy the National Socialist Party,鈥� and 鈥渢o prepare for the eventual reconstruction of German political life on a democratic basis鈥� (Art. 2, sec. A). It is not surprising, however, that in the context of the burgeoning Cold War, these accords were never implemented within the confines of strict legalism.

Complicating matters was the 1949 consolidation of the three Western zones into the Federal Republic of Germany (FRG), and the creation of the communist German Democratic Republic (GDR) out of the Soviet zone. While these two Germanys ostensibly became sovereign nations, the GDR was not recognized by any of the Western powers鈥攊nstead, they adopted a unified stance, asserting that the formation of the FRG fulfilled the original goal of reunifying Germany, and that the GDR, or 鈥淪oviet occupation zone鈥� should join with the others through democratic elections, a proposition the Soviets would never have agreed to (Foreign Relations of the United States 1949, vol. III doc. 131; hereafter FRUS). Situated a hundred miles behind the border of the GDR, and increasingly at the center of a volatile geopolitical situation, Berlin existed under a unique and contested legal regime. According to the Protocol on the Zones of Occupation in Germany and the Administration of Greater Berlin of September 1944, Berlin was considered an entity separate from the other occupation zones. Despite the fact that West Germany considered the city to be one of its states, the GDR claimed it as its capital and the Western powers considered it to be a separate entity occupied under the auspices of international law (Central Intelligence Agency, 1961).

听听听听听听听听听听听 Conflict regarding the city鈥檚 status first erupted when the Soviet Union blocked the Western powers鈥� overland access to the city on June 24, 1948, in reaction to proposed currency reforms aimed at improving the West German economy (Schick 1971). Unwilling to risk armed conflict over Berlin, the Western powers instead airlifted supplies to the city for nearly a year. Ultimately an end to the blockade was successfully negotiated in May 1949 with the promise of a foreign minister鈥檚 meeting to discuss issues surrounding the city (FRUS 1949, vol. III doc. 374). Despite this lull in tensions and increased East-West dialogue, there was now a precedent for using the isolated city as a political lever, a possibility that did not escape the American government. In May 1955, a disagreement over increased tolls on the German Autobahn leading into West Berlin led U.S. State Department officials to privately consider contingency plans in the case of 鈥渇urther deterioration of the situation,鈥� and the possibility that 鈥渨e might find ourselves again in a blockade鈥� (Digital National Security Archive, May 10 1955; hereafter DNSA).

The continued existence of an occupied Berlin within the GDR鈥檚 territory, the refusal of America and its allies to recognize the GDR, and the increasing involvement of the FRG in the economic and military structures of Western Europe all became a significant source of contention with the Soviet Union. In 1957, the USSR sent a strongly worded diplomatic note to the FRG loudly protesting an American plan to store nuclear weapons on its soil, warning against any attempt by the West Germans to obtain such weapons for themselves (Documents on Germany, 496; hereafter DG). The Soviet Politburo had considered preventing a nuclearized FRG as a top priority since 1955 (Zubok 1993) and West German Chancellor Konrad Adenauer鈥檚 seeming refusal to give up on either the idea of a nuclear arsenal or on his claims on German territory ceded to Poland continued to provoke hostilities. In early 1958, a Polish plan to create a denuclearized zone in Central Europe was rejected by the Americans (DG 517, 528) and the issues surrounding armaments and the FRG鈥檚 policies remained unsolved. Against this diplomatically strained backdrop, the standards of living between West Berlin and the surrounding GDR were becoming much more pronounced, as the West began to experience rapid economic growth while the GDR stagnated (Harrison 1993).

Speaking on November 10, 1958 at a Soviet-Polish friendship meeting in Moscow, Soviet Premier Nikita Khrushchev finally went on the offensive, unleashing one of the longest-lasting crises of the Cold War (Schick 1971). Accusing the United States and other Western powers of allowing Bonn, the FRG鈥檚 capital, to become a hotbed of growing Nazi militarism and of using their position in Berlin to engage in 鈥渟ubversive activities鈥� against the community of Communist states, Khrushchev declared that the West had dishonored the spirit of the Potsdam Accord. According to the Premier, the agreements guaranteeing Western rights in Berlin were now invalidated. He suggested that the Soviet Union replace its agents stationed at checkpoints with East German officials, and to 鈥渓et the United States, France and Britain themselves build their relations with the German Democratic Republic鈥f they are interested in any questions concerning Berlin鈥� (DG 542). Khrushchev鈥檚 proposal/declaration seemed to do several things at once; in addition to attempting to force de facto recognition of the GDR by the West, the Premier鈥檚 words were seen as a direct threat against the Allies鈥� occupied status of Berlin.

This surprise statement triggered a flurry of internal telegrams and notes among U.S. State Department officials as ambassadors, analysts, and decision-makers attempted to decipher what exactly Khrushchev had meant by his speech. While most of his threats were veiled, they broached exceedingly sensitive points of East-West contention with high margins for escalation. The idea that this was mere propaganda was quickly dismissed. U.S. ambassador to the Soviet Union Llewellyn Thompson called it 鈥渁 most dangerous move,鈥� and posited that Khrushchev was using the vulnerable position of Berlin to force Western recognition of the East German regime (FRUS 1958-1960, vol. VIII doc. 25). Political Officer at the Embassy in Berlin Findley Burns agreed with this appraisal, arguing that Khrushchev鈥檚 mentions of Berlin were 鈥渋ntended to include access thereto鈥� (FRUS 1958-1960, vol. VIII doc. 27).

Despite these initial worries, internal correspondence within the State Department suggests that the majority of American government officials did not see a new blockade as a strong possibility. In a November 13 telegram sent to various American embassies, Under Secretary of State Christian A. Herter expressed the belief that the Soviets would not attempt anything as blatant as a full blockade again, due to their failure in 1948 and the potential consequences such a reprisal would have on world public opinion. Furthermore, he stated that 鈥渋n discussing [the] subject addressees should not convey impression of excitement or undue anxiety鈥� and that they 鈥渟hould point out Communist threats against Berlin [were] frequently made but not often implemented and [that] whether and how [the] latest threat [would] be implemented [would] depend greatly on [the] firmness of [the] Western reaction鈥� (FRUS 1958-1960, vol. VIII doc. 31).

In spite of this endeavor by the American government to appear calm and to make the explicit argument that Khrushchev鈥檚 speech did not foretell 鈥渁 reimposition of the Berlin blockade of 1948-49鈥� (Kenworthy 1958, 11 New York Times; hereafter NYT), the American press quickly drew parallels to the previous crisis. Interestingly, the connection was often used to suggest that 鈥淢r. Khrushchev [had] presented the Western Powers with a problem potentially far more difficult for them than the Berlin blockade of 1948-49鈥� (Frankel 1958, 1 NYT). This return to the language and imagery of the blockade also included the implication of conflict. In a private telephone conversation with Herter, days after the Khrushchev speech, Secretary of State John Foster Dulles expressed his displeasure at the reports in the news that there were 600 planes at the ready, prepared to resume an airlift at any moment (Log, Nov. 13 1958 DNSA).

U.S. officials continued to assert that this was not a crisis鈥攁t a news conference Dulles even quipped that he was 鈥渘ot surprised that this Berlin probe took place,鈥� claiming to have thought it a probable scenario (DG, 549). Media figures and certain foreign allies had adopted a tone very much in opposition to this view, and one that carried a greater degree of influence in the pubic discourse of the time. West German Chancellor Adenauer immediately donned a mantel of defiance against the Soviet threat, publicly warning them that 鈥渁n extremely dangerous situation had been created,鈥� and that any unilateral actions would be met with a strong Western response (NTY, Nov. 16 1958, 29). A tenser situation would have benefited the leader of the FRG; an imminent threat to Berlin would force the public implementation of American security guarantees, whereas bureaucratic disagreements and back-channel deals would give the U.S. the option to quietly compromise without losing face. West Berlin鈥檚 Mayor Willy Brandt also contributed to the public鈥檚 sense of urgency; an article at the time noted that when the Mayor 鈥渁sserts鈥hat 鈥楤erlin belongs to the free world,鈥� the city stands as a new kind of symbol鈥� (NYT Nov. 23 1958).

Viewing Berlin as a symbol was certainly not a new construct鈥攂oth the press and the West Germans were drawing on the legacy of the first crisis. Arguably, the Berlin blockade of 1948 shifted the perception of the city from an ex-Nazi capital that needed to be controlled to a symbol of budding democracy that needed to be protected. During the blockade, then-President Harry S. Truman hailed the 鈥渃ourage that [its inhabitants were] displaying in their beleaguered outpost,鈥� calling it 鈥減roof to the world of the strength of the democratic spirit鈥� (Public Papers of the Presidents March 1948; hereafter PPP). The success of the airlift in outlasting the Soviet blockade led to an association between the city and the fight against communism. Then-Secretary of State Dean Acheson claimed that the city was a 鈥渟ymbol of courage,鈥� and that issues surrounding Berlin were due to Soviet attempts to 鈥渋nterfere with the whole environment of freedom and liberty鈥� (Middleton 1949, 1 NYT). Thus, an attack on Berlin became an attack on the West and its democratic principles.

This symbolic language, combined with repeated and emphatic guarantees to protect the city, created a schema of the city鈥檚 importance within the media that effectively tied the hands of future leaders. The United States would not be able to back down on issues surrounding the city without serious political consequences (DNSA, May 10 1955). As one journalist at the time noted, despite the fact that many points of contention during the crisis were 鈥渁bout seemingly trivial details,鈥� the situation, and the city, had 鈥渂ecome a test of the value of the United States鈥� solemn word鈥� (Lewis 1959, SM14 NYT). A separate opinion piece aptly claimed, 鈥淲e Painted Ourselves Into the Corner鈥� (Krock 1958, 32 NYT). These high stakes make Eisenhower鈥檚 relative silence at the time all the more striking.

Eisenhower and Berlin:

听听听听听听听听听听听 President Eisenhower鈥檚 role at the beginning of the Berlin Crisis has been given surprisingly little attention considering its unusual nature in comparison to other Cold War crises. For several weeks following Khrushchev鈥檚 November 10 speech, the President maintained total silence on the issue. Declassified documents suggest that the President was well informed on the pronouncement (Memorandum, Nov. 18 1958 DNSA); however, the vast majority of internal communication ran through Secretary Dulles and Under Secretary Herter. In addition to staying out of the public eye, Eisenhower ensured that other members of his Administration were not engaging in any behavior that could be interpreted as escalatory. At a meeting at which a potential visit by Vice President Nixon to Berlin was floated, Eisenhower insisted that no speech on the issue be made, lest it be interrupted as a provocation. Tellingly, a party to the conversation noted that the visit would be 鈥減sychologically a good gesture,鈥� but that there was a risk that, 鈥渢hings will happen that will make [the President] a prisoner later on as to what to do鈥� (Memorandum, Nov. 26 1958 DNSA).

听听听听听听听听听听听 This provides one explanation for Eisenhower鈥檚 preliminary caution: at the time, his Administration鈥檚 policy was not entirely clear. Considering the vague nature of Khrushchev鈥檚 November 10 announcement and the need for the U.S. to appear to be in total agreement with the FRG, Britain, and France, this is unsurprising. Trachtenberg (1991) argues that Eisenhower eventually adopted a very flexible stance on Berlin, pointing out that the President was later willing to consider the Soviet鈥檚鈥渇ree city鈥� proposal. However, this would have put him in direct odds with Adenauer and other elements of the West German government who quickly adopted a hardline stance on issues of both reunification and Berlin鈥檚 status (DG, 577). In balancing the diverse interests of his allies and attempting to take accurate stock of the situation, Eisenhower evidently did not want to tie his hands by making any sort of public statement that could be seen as a commitment to a specific course of action, particularly in the context of Berlin鈥檚 emblematic importance in the fight against communism.

The Administration鈥檚 avoidance of explicit commitments appears to be justified in light of the public reaction to Dulles鈥檚 press conference on November 26. During the conference, the Secretary of State suggested that the United States might treat East German officials as 鈥渁gents鈥� of the USSR. This undermined the previous American policy of no-tolerance regarding the transfer of functions from the Soviet Union to the GDR, and therefore made American threats seem less credible. The comment generated a great deal of bad press for the Secretary. Accusations of tacit recognition of the GDR and of abandoning the West Germans were widespread and enflamed by the ridicule of French and other American officials (Gruson 1958, NYT). Interestingly, the President was never connected with the incident, but it did demonstrate the dangers of needing to backpedal, and it was the first in a series of episodes that made the West look divided (an issue that never seemed to exhaust the press鈥� imagination over the ensuing months).[5]

Eisenhower and Dulles may have appeared reticent at the beginning of the crisis, but their general avoidance of the public eye is a poor indicator of the political importance that they placed on Berlin. The city, defeated German capital turned island of democracy, came to represent many of the geopolitical, ideological, and class struggles of the Cold War, but, more importantly, it was a test of American and NATO security guarantees (trust in which was considered so vital that war was preferable to reneging on them) (DG, 612). Imagery drawn from British Prime Minister Neville Chamberlain鈥檚 visit to Munich on the eve of the Second World War fueled a mantra that any form of appeasement would increase the risk of war, and that war was preferable to capitulation. Privately, Eisenhower made it clear that he was willing to wage nuclear war to protect Berlin if necessary, asserting that using this threat was the best way to force the Soviets to be 鈥渞easonable鈥� (DNSA, March 31 1959). Burr (1994, 179) argues that this extreme stance was responsible for the President鈥檚 desire to keep press exposure on the crisis to a minimum. While threatening military action over the relatively minor incident of the transfer of functions to the East Germans might be considered a good coercive tool, it would have been 鈥減olitically disastrous鈥� if made public. Another confounding factor was the (later confirmed) presumption 鈥渢hat turn over to [the] GDR would be implemented gradually,鈥� and owing to the fact that 鈥淕DR pressure would be applied in stages,鈥� communist forces would consciously 鈥渁void any action which鈥ould bring into force [the American] security guarantee鈥� (FRUS 1958-1960, vol. VIII doc. 27).

This begs the question: how would one decide which minor infraction in a series of many would justify the use of military force? In this scenario, any retaliation would seem too disproportionate and it would become impossible to garner domestic or international support. However, loss of the city was not considered an option under any circumstance. Later on in the crisis, Eisenhower argued in conversations with congressional leaders that it was necessary for the U.S. to be willing to 鈥渆ngage in general war鈥� to preserve access to Berlin because appeasement would lead to total defeat (Declassified Documents Collection, Mar. 6, 1959, 597 [cited by Trachtenberg 1991]). This deterrence was both necessary and risky: how long would American allies and the public accept a strategy that could lead to total annihilation over the political status of a single, isolated city?

This scenario is much like the one described in Baum鈥檚 (2004) model of audience costs, in which leaders often have an incentive to 鈥済o private鈥� and engage in less public forms of diplomacy to minimize the salience that the public attributes to a crisis. Internal and external pressures, as well as the vulnerable geographic location of Berlin, made escalation of the crisis dangerous. In a private conversation with Eisenhower, Dulles admitted that they were 鈥渂oxed in鈥� to the point that the situation in Berlin had become 鈥渦ntenable鈥� (DNSA, Jan. 7 1959). Despite this precarious position, there is no conceivable way that Eisenhower could have smothered discussion of the city. The ability of presidents to act as gatekeepers of public attention is largely limited to minor events鈥攃rises tend to take on a life of their own in the media, the political landscape, and the public imagination (Knecht 2010, 42). Although Eisenhower waited for a long period of time to publicly discuss the Berlin situation, remaining silent for too long would have made him appear weak and he would have forfeited his unique ability as president to mold public opinion and the general discussion of the crisis (Baum & Potter, 2008). As such, he was forced to respond, trying to set a tone that maneuvered the treacherous backdrop of alliances, the Soviet Union, the press, and the public.

Eisenhower鈥檚 crisis rhetoric reflects this need for balance. Many of the President鈥檚 statements followed the same rhetorical line, and displayed consistency to the point of redundancy as time progressed. His strategy seemed to be two-fold: to appear firm and resolute (but not hostile), and to justify his actions by emphasizing the legal and moral responsibility that the United States had toward the people of West Berlin. During his first press conference after the crisis began, Eisenhower stated that:

听

There is no attempt on the part of the United States in this position to be arbitrary in the sense of trying to irritate or anger anybody else鈥 want to be clear that we are not doing anything for ourselves. We are doing it for a free people that have a right to this kind of cooperation, promised them in the pledged word of a number of governments (PPP, Dec. 10, 1958).

听

Several months later, he maintained this argument that the US鈥檚 goals were benign and selfless, and that he was compelled by a solemn 鈥渞esponsibility鈥� not to give in or compromise. In response to more pointed questions, Eisenhower remained cagey, unwilling to lay out firm plans or promises other than the general commitment to maintain access to Berlin and to preserve the freedom of the people living there. The following interaction is characteristic:

听

Reporter: Is there an agreement to use force, if necessary, to defend our rights in Berlin?

Eisenhower: We have not said we are using force. We are saying we do not abandon our responsibilities. We will continue to carry them out, and it will be the responsibility of the other side if there is going to be any force; they have to use that to block our carrying out our responsibilities (PPP, Feb. 18, 1959).

听

Fortunately for the President, the press of the 1950鈥檚 was more deferential to the executive on matters of foreign policy than it is today, and did not often pick apart the rhetorical pretzel into which the President had to twist himself. The President also had the advantage of muted public awareness of the details of the crisis despite high levels of press coverage. A survey done by the New York Times found that most of the public was 鈥渋ll-informed but trustful of the Administration and particularly of the President, nonpartisan in their approach and confident that somehow this crisis will pass鈥� (Reston, 1959, 1 NYT). The public鈥檚 certainty of a successful outcome most likely owed to the tone of calm resoluteness that Eisenhower adopted, thereby avoiding hysteria or a sustained questioning of his policies. The various turns, dead-ends, and tensions that were created during the crisis were largely handled by the State Department through diplomatic channels, allowing the President to largely remain aloof, and left usually only to comment on the situation when directly questioned at press conferences.

听听听听听听听听听听听 There were notable exceptions to this rhetorical approach, however. Following Khrushchev鈥檚 rejection of a foreign minister鈥檚 conference and renewed threats of a separate treaty with the East Germans in early March 1959, Eisenhower gave one of his most public addresses on the state of the Berlin situation. Broadcast on television and the radio, and announced several days prior, the speech was clearly meant to bring attention to his Berlin policies. However, an examination of the text and context of the speech demonstrates that it represents more continuity than a transformation of tactics.

听听听听听听听听听听听 Going in the opposite direction of 鈥減rivate,鈥� Eisenhower, in his address, cultivated a sense of urgency surrounding Berlin, presenting it as being besieged by a Soviet Union that wanted 鈥渘o less than world domination, to be achieved by gaining political power successively in each of the many areas which had been afflicted by war, so that in the end the United States鈥� would be isolated and closely encircled.鈥� Themes concerning morality sought to label the USSR as an aggressor, while his use of historical precedent established the legal defense of the Western position. Eisenhower denounced the Soviet鈥檚鈥渇ree city鈥� proposal, famously saying: 鈥淚t is by no means clear what West Berlin would be free from, except perhaps freedom itself.鈥� The President, after presenting the crisis as an urgent situation in which the United States had a vital and justified role to play, laid out three possible courses of action:听 acquiescence to the Soviets鈥� wishes, which would threaten 鈥渢he mutual confidence upon which our entire system of collective security is founded鈥�; 鈥渢he possibility of war鈥�; or negotiation given that the Soviets demands were reasonable. Naturally, the President advocated the third option (DG, 612).

听听听听听听听听听听听 Despite the harsh words, escalatory language, and list of reasonably definitive boundaries that he would not cross, Eisenhower鈥檚 speech did not say anything new in the way of policy proposals. Perhaps owing to this, the report received minimal press coverage, generating a few articles at most in the New York Times. Whether intentional or not, its purpose was received as informational rather than as a diplomatic exchange with the USSR. While the President generally used his rhetoric to frame the crisis as a struggle to protect freedom and uphold the rule of law, he rarely acted in a proactive way, instead reacting to new developments and depending on officials at the State Department to conduct diplomacy. Unlike Khrushchev, Eisenhower鈥檚 rhetoric rarely presaged new crisis developments, and so did not capture the press鈥� attention; instead the media turned to other sources of information, including the Soviet Premier.

Khrushchev and the Russian Monolith

听

During the crisis, Khrushchev was much more diplomatically active than his American counterpart. While the exact purposes the Premier had in mind when he began the crisis are contested (Fursenko & Naftali 2006; Richter 1994; Zubok 1993; Harrison 1993), evidence suggests that his decision represented a clear break from previous Soviet policies. In the months leading up to the crisis, the Soviet Ministry of Foreign Affairs had adopted a noncommittal, and relatively uninteresting, policy of non-action toward Berlin and the German question (Archive of the President, Russian Federation Nov. 3, 1958, 185 [cited by Fursenko & Naftali]). Tellingly, a peace treaty with the GDR had not yet been drafted at the time of Khrushchev鈥檚 speech (Zubok 1993). Furensko & Naftali (2006) go as far as to argue that the November 10 speech caused an internal policy struggle between an aggressive faction led by Khrushchev and a more conciliatory one led by Deputy Premier Anastas I. Mikoyan that worried about how far the West would go to prevent unilateral actions being taken against Berlin.

听听听听听听听听听听听 Whatever his specific goals were, Khrushchev remained relatively quiet early on in the crisis, waiting on the result of a visit by Mikoyan to the United States in early 1959. The visit by a senior Soviet official, a rare occurrence at the time, was understood by Acting Secretary Herter as being primarily fact-finding in nature, and not an opportunity for any substantive policy decisions to be made (Memorandum to President, Jan. 2 1959 DNSA). In the press, it was claimed that Mr. Mikoyan 鈥渨as acting as Mr. Khrushchev鈥檚 eyes and ears鈥� (Salisbury 1959, 1 NYT). Following the visit, this position was reinforced by a State Department report, remarking that 鈥淢ikoyan introduced several modifications, but no basic change in the Soviet position on Berlin.鈥� Perhaps most tellingly, it was noted:

听

There is another possible explanation of the Mikoyan trip, namely, that it is related to possible difference within the Kremlin leadership on foreign policy issues. According to this hypothesis, Mikoyan came to the US (1) to resolve differences within the Kremlin on the proper course of intention (with Mikoyan favoring a go-slow approach), or (2) to build a case for a renewed, concentrated drive for an East-West d茅tente, in opposition to those in the Kremlin favoring a generally harder line. The proceedings of the Party Congress do not bear out the latter point (which, of course, is contrary to the general line of analysis presented above) but there is some inconclusive evidence tending to support the first point (DSNA, Jan. 1959).

听

It was generally agreed that if internal Soviet schisms over the Berlin question did exist, they were not major enough to warrant extensive consideration; the theme of Soviet divisions was rarely mentioned throughout the crisis, and there was a general deference to Khrushchev鈥檚 statements over those of other Soviet officials. Reading a report by Ambassador Thompson of a luncheon he had attended to discuss Berlin policy with top Soviet officials including Khrushchev, Mikoyan, Frol R. Kozlov, and Foreign Minister Andrei Gromyko, it would be easy to assume no one had been there save for the Premier (DNSA, June 26 1959).

This view was held by other Western powers as well. At a tea party attended by Vice President Nixon and several U.S. Congressmen, it was reported that British Prime Minister Harold Macmillan claimed:

听

Of all the Soviet officials, the West can do business only with Premier Khrushchev. The attitude of a Soviet Foreign minister, cooperative or obstructive, is no chart at all to Khrushchev鈥檚 attitude. Khrushchev alone can speak the ultimate word on act and policy. He will not relay it through a foreign minister, and he is determined to meet with President Eisenhower in person as an equal for such momentous purpose. (Krock 1959, 30 NYT).

听

In the analysis of Khrushchev鈥檚 public statements on the crisis, there was a consistent tendency to view his rhetoric as significant and holding important policy implications, as opposed to being mere politicking or propagandizing (Thompson, July 30 1959 DNSA). Khrushchev himself helped to promote this view in his insistence on a summit of heads of state鈥擡isenhower privately complained that the Premier held the view that 鈥淔oreign Ministers [were] only errand boys and incapable of substantive agreement鈥� (Dillon, June 14 1959 DNSA).

听听听听听听听听听听听 The press, unsurprisingly, took this image of Khrushchev a step further. A series of articles appearing in the New York Times titled 鈥淜hrushchev鈥檚 Russia鈥� argued, in a generally accepted vein at the time, that the Premier had consolidated political power to the point of being the undisputed leader of the USSR: 鈥淗e alone has the power to make Soviet decisions. He also can commit his country to a course of peace or a path that leads to disaster鈥� (Salisbury 1959, 1 NYT). This belief that Khrushchev was the only voice of import also led to a tendency to over-interpret his words; when the Premier spoke, the American media generally listened.

After a pair of speeches given by the Premier to the 21^st Party Congress, the press noted that he had failed to mention his six-month ultimatum deadline, positing that this meant a change in policy. Once the headlines had ruan 鈥淲estern diplomats who [had] had talks with Premier Nikita S. Khrushchev warned鈥gainst reading a change of policy into his failure to mention鈥is original six-month deadline on Berlin鈥� (NYT Feb. 9 1959, 3). This type of over-interpretation by the media was a relatively common occurrence at the time. In a short New York Times editorial piece appearing December 13 1958, the author linked together Soviet demands for a summit meeting, Khrushchev鈥檚 recent statements on Soviet nuclear capabilities, and the content of Soviet news coverage, arguing that it was all part of a unified strategy, 鈥渙bviously designed to influence the Western answer to [the Berlin] ultimatum, and, if possible, to sow dissention among the Allies.鈥� The article continued on to state that 鈥渢he free world must prepare itself to take these threats more seriously than heretofore,鈥� in the case that the Soviets become 鈥減risoners of their own propaganda, which could close their avenues of retreat鈥� (NYT Dec. 13 1958). 听

听听听听听听听听听听听 While arguably there were many figures from the Soviet bloc weighing in on the crisis, Khrushchev was taken far more seriously than the others. East German leader Walter Ulbricht was considered by many in the West to be Khrushchev鈥檚 puppet (NYT Mar. 13 1959, 28), and Foreign Minister Gromyko was usually thought to be merely acting according to marching orders. In such a situation, Khrushchev鈥檚 words inherently had a greater ability to set the agenda than a leader who might not have been considered relevant, or who was buffeted by domestic struggles. The fact that 鈥淪oviet Union,鈥� 鈥淐ommunist states,鈥� and 鈥淜hrushchev鈥� could be used interchangeably speaks to the power the Premier held, and the attention he commanded during the crisis.

听

Methods and Data Collection:

听

In order to determine whether Khrushchev鈥檚 public rhetoric had a measurable effect on the perceived issue salience of the Berlin crisis in the American media, a baseline of salience was constructed using media content analysis of New York Times (NYT) articles. Next, all Khrushchev鈥檚 statements appearing in the NYT concerning the crisis were catalogued and coded as status-quo-changing or non-status-quo-changing. Finally, statistical analyses of this time series data were performed.

Patterns of media attention during the crisis

听听听听听听听听听听听 Issue salience here is defined as the degree to which the press reported on the crisis. Numerous studies (Wanta & Hu 1993; Semetko et al. 1992; Soroka 2003; McCombs & Shaw 1972; Cohen 1963) have found that there is a strong level of correlation between what the public and the media consider salient. In the absence of comprehensive polling data during my period of study, press coverage remains the best metric to determine how conscious the American public was of the Berlin crisis. It seems reasonable to assume that the most politically active segments of the public tended to have more exposure to the news and current events, making it even more likely that press coverage of the crisis would create inherent political pressure for President Eisenhower. While some research suggests that the agenda setting power of the media is limited in certain cases (Dearing & Rogers 1996), the literature has frequently affirmed strong media agenda setting on matters of foreign policy (Walgrave & Van Aelst 2006).

Press coverage of the crisis is defined as the number of mentions of the crisis in NYT articles published during the study period. The New York Times has been used in a number of media content studies to determine media agendas and frames (Soroka 2003; Walker 1977; McCombs & Shaw 1972) and is often considered to be a form of 鈥渆lite media鈥� most likely to be read by those in power (Auerbach & Bloch-Elkon 2005). As Dearing and& Rogers (1996, p#[AH2]听) note: 鈥淭he New York Times is generally regarded as the most respected U.S. news medium. When the Times indicates that an issue is newsworthy, other U.S. news organizations take note.鈥� While coverage of foreign affairs tends to be greater in the NYT than in other newspapers (Soroka 2003), and focuses more on traditional security issues (Auerbach & Bloch-Elkon 2005), it remains an often-referenced news source in agenda setting research.

听听听听听听听听听听听 Instead of looking at front-page news (McCombs & Shaw 1972) or editorial articles (Auerbach & Bloch-Elkon 2005), I catalogued all mentions of the Berlin crisis using key words, and then read each article to verify that it was relevant to the crisis situation.^{^[6]} Not all articles were primarily about the crisis, but using the all-mentions metric gives a good measure of how active the crisis was in the press鈥� (and thus the public鈥檚) consciousness. During the process of data collection, I noticed that these more offhand mentions of the crisis increased substantially after a major, status-quo-changing event. This suggests that there is a certain residual attention that other metrics to not capture well.

听听听听听听听听听听听 The data on issue salience (SAL) was used to construct a time-series graphic (Figure I) broken up into one-week segments, which, as noted by Edwards and Wood (1999), gives a reasonably detailed picture of events. The collection period spans from the week of November 10, 1958 when Khrushchev began the crisis, until the week of May 20, 1960, when Khrushchev cut off all negotiations with the Eisenhower Administration in the midst of the U-2 Incident. After this date, all further talks on Berlin were postponed until the American presidential elections later that year, when Khrushchev hoped the next U.S. executive would be more amenable to negotiating on his terms. Examining only this period allowed me to focus on the Khrushchev-Eisenhower dynamic, as the crisis began to take on different characteristics under John F. Kennedy鈥檚 Administration.

Figure I:

This time series provides a map of the crisis from the perspective of the NYT. A total of 1,478 articles were coded with an average of 18.5 mentions per week, the lowest being 2 articles occurring the weeks of Aug. 18, 1959 and January 1, 1960, and the highest being 58 the week of November 23, 1958. A table of all values can be found in Appendix I. Most of the spikes in press coverage occurred, predictably, in conjunction with major status-quo-changing events, such as Khrushchev鈥檚 Berlin ultimatum. The large peak in early spring 1958 (around Week 19) corresponds to a combination of events, including Khrushchev鈥檚 high-profile meetings with British Prime Minister Harold Macmillan, his visit to East Berlin, and his blunt rejection of a proposed foreign minister鈥檚 conference.

Mapping Khrushchev鈥檚 rhetoric

In order to determine what influence Khrushchev鈥檚 rhetoric had on press coverage during the crisis, all statements made by the Premier that appeared in the NYT were collected and coded as either status-quo-changing (SQC) or non-status-quo-changing (NSQC). 鈥淪tatement鈥� is defined as any public or private utterance, written or spoken, that mentioned the Berlin crisis explicitly, or the issues surrounding it (e.g. public speeches, letters to heads of state, communiqu茅s, etc.).

SQC is defined as any statement that could be considered a development in the diplomatic situation, such as new proposals or threats that diverged from the prior crisis situation. The NSQC set includes all other statements. It is important to note that NSQC statements do not necessarily need to be mild or off-hand鈥攁ggressive statements and threats were included, as long as these threats were not new ones. There were several instances where a statement that might have been coded as NSQC was instead listed as SQC, if it represented a major shift in tone from other statements around the same time period.

While the coding of these statements was usually straightforward, there were several instances where the pPremier鈥檚 meaning might have not been to change the status quo, but it was a framed as such in the NYT. In these cases, the coding followed the general consensus of the news coverage, as perceived reality is much more germane for the purposes of this study. One of the clearest examples of this schism between interpretation and intent can be seen during a private reception at the Albanian embassy in Moscow, November 29, 1958. It was reported that Khrushchev had made several statements concerning Berlin, yet the exact details of his talk were uncorroborated. The following day, the NYT ran a front-page headline reading 鈥淜hrushchev Hints Talk Might Delay Action on Berlin.鈥� The article gave two accounts of the event, one in which Khrushchev 鈥渋ndicated the Soviet Union would not act unilaterally to alter Berlin鈥檚 status if the West agreed to start negotiations within six months,鈥� and the other which 鈥渉ad [Khrushchev] repeating the essence of the original announcement.鈥� (Frankel 1958, 1 NYT). The headline and structure of the piece very clearly favored the first interpretation, although it seems clear that the pPremier was not trying publicly to change his bargaining stance at this time.

Khrushchev鈥檚 statements were used to construct a time-series spanning the same period as the issue salience data, broken into one-week segments. Weeks where no statements were made (or at least not covered by the NYT) are included in the table in Appendix I. A total of 82 statements were recorded over this period, 13 of which were coded as SQC, the other 69 being NSQC. This averages to roughly one statement per week.听

听

Findings and Discussion:

Granger F-Test

Using the data collected for media issue salience (SAL), and SQC and NSQC statements, I initially ran a Granger (1969) Test of probable causality, to determine whether or not SQC and NSQC statements could be used to predict SAL variations. A multivariate version of this test, VAR, has been used successfully in multiple agenda setting studies (Edwards & Wood 1999; Peake, 2001), and is a useful tool within econometrics. Significant Granger causality was not found for either SQC or NSQC statements (Table I), so the null hypothesis () of no influence could not be rejected. This does not provide a definitive answer to the research question, however.

The bivariate nature of this test fails to take into account other factors known to influence the media agenda, such as domestic elite discourse (Dearing & Rogers 1996; J. Cohen 1997) and 鈥渞eal-world events鈥� (Behr & Iyengar 1985). I hypothesized that President Eisenhower鈥檚 lack of aggressive, SQC rhetoric would mitigate these effects, but it is possible that in the absence of a strong American voice, the NYT received its cues for issue salience from the rhetoric of other allied leaders, such as West Berlin Mayor Willy Brandt. Furthermore, as Khrushchev鈥檚 SQC rhetoric represented some of the most impactful events of the crisis, such as his six-month ultimatum deadline, I believed that other international events would have minimal effect. While it is still possible that NQCS rhetoric has predictive power over SAL levels, it appears that media attention was too multifaceted in this case鈥攖he aforementioned variables should be considered in future VAR analyses.

听

MA Models

听

听听听听听听听听听听听 Next, I constructed an MA (Moving Average) Model of the SAL data to determine if it had a general trend (Figure II). Expectedly, there was an overall downward slope in press attention as the crisis progressed, in the face of issue fatigue and the absence of major developments. The same was done with Khrushchev鈥檚 NSQC statements, which had a very slight downward slope, but not enough to be considered significant (Figure III). To run a cross correlation, the data was divided into two periods (P.1 and P.2) at the Week 34 mark (June 28, 1959). This was done for both statistical and historical reasons. Dividing the data into two sets yielded stationary results to control for spurious relationships due to general trends in the data (Figures IV & V).

Within the crisis itself, this period in late June marked a significant shift. Shortly before this date on June 19, 1959, Khrushchev proposed a new, six-month deadline for negotiations on Berlin following a failure of Soviet proposals at a foreign minister鈥檚 summit conference in Geneva. After this period, it was remarked in the NYT that Khrushchev鈥檚 rhetoric and demeanor were much more subdued in the lead up to his visit to the United States (Rosenthal 1959, 2 NYT). Richter (1994, 119) argues that it is possible that Khrushchev 鈥渁llowed鈥� the conference of foreign ministers to fail in order to expedite a summit with the other heads of state, where he would be able to negotiate personally with the Western powers. This period of time signaled the end of the first major round of negotiations, and the beginning of a movement toward the next that would directly involve the heads of state rather than proxies.

听

听听听听 Figure II.听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 Figure III.

听

Figure IV.听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听 Figure V.听听听听听听听

听听听听听听听听听听听

During the first period (P.1), there was a moderate (0.490) and significant correlation between NSQC statements and SAL at lag 0, and a weaker (0.424), but still significant correlation for lag 1.* As the data is broken up by week, the lag 1 result is likely due to statements that were made near the end of one week, and covered at the beginning of the next. The correlation between SQC statements and SAL was strong and significant at lag 0 (0.64) and at lag1 (0.593), and was moderate and significant at lag 2 (0.411). During the second period (P.2.), there was no significant correlation between NSQC statements and SAL, save for a spurious weak relationship (0.309) at lag -9.

Discussion

The difference in results between the first and second half of the crisis was not expected, and points to several interesting possibilities. With an average of 28.5 articles appearing weekly on the crisis during P.1, the crisis in this first phase was undoubtedly more volatile and salient than in P.2, with an average of only 11 articles. Khrushchev鈥檚 SQC rhetoric was also much more prevalent during P.1, with and average of 0.26 statements per week as opposed to 0.08 during P.2. It is also important to note that his SQC statements were not homogenous throughout the entire study period. While these statements were defined as any that actively altered the bargaining situation and could be considered a new development, some had a much greater impact on the situation than others. For example, the Berlin ultimatum near the beginning of the crisis was received as a highly escalatory act that represented a definitive break from past policy and a lapse into unilateralism, whereas other statements, such as Khrushchev鈥檚 agreement to a foreign minister鈥檚 conference, was an important development, but was not escalatory and followed as a natural part of the lifecycle of the crisis. In general, the most acutely SQC statements occurred during P.1, and while the Premier continued to make hostile threats in P.2, they were nearly always repetitions of prior ones. After half a year, the opposing sides had fallen comfortably into their entrenched arguments, and little progress was being made鈥攚hen Berlin was no longer being immediately threatened, most laypeople probably lost interest in the convoluted diplomatic proceedings.

There are three main explanations as to why the association between NSQC statements was strong and significant in P.1 and not in P.2: first that Khrushchev鈥檚 NQSC rhetoric was assigned greater importance by the press during periods of high tension, exacerbating the conflict when his words were hostile or not completely clear, second that Khrushchev鈥檚 NSQC statements tended to be grouped with his SQC ones, or third that the NYT only reported on Khrushchev鈥檚 NSQC statements during periods that were already tense. This last hypothesis is unlikely; the number of Khrushchev鈥檚 NSQC statements that made it to press decreased only slightly over the total period of the crisis (Figure III), unlike the overall salience of the crisis, which experienced significant decreases after P.1 (Figure II). Additionally, the finding that media attention correlated with NSQC statements at several lags (but not in the reverse) is highly suggestive that the relationship did not work as a function of selection bias.

Likely, the answer is a combination of hypotheses 1 and 2. While there were instances, such as Prime Minister Macmillan鈥檚 visit to Moscow, where Khrushchev made a series of SQC and NSQC statements in quick suggestion, this was not often the case. Historical evidence, particularly the ways in which the media framed Khrushchev and his speeches, points to the fact that the pPremier鈥檚 words were often misinterpreted, and this speculation increased dramatically during already tense periods. As was discussed earlier in this study a high amount of importance and meaning was assigned to the pPremier鈥檚 words. There were many instances of near egregious over interpretation of the pPremiers鈥� words鈥攏early all of these instances occurred early on during the crisis.

听These results suggest that while NSQC statements can contribute to salience during periods that are already highly salient, they have relatively little effect during other stages of crises.

Conclusion:

听

听听听听听听听听听听听 The Berlin Crisis promised to be a fertile ground for hostile foreign elite influence on the American media for several reasons: President Eisenhower鈥檚 avoidance of crisis escalation and the public eye, the media鈥檚 view that Premier Khrushchev controlled all matters of Soviet foreign policy, and American memory of the Berlin Blockade a decade earlier. The results of this study suggest that there was a moderate influence of non-status-quo-changing statements made by the Premier on the salience of the Berlin crisis in the American media; however, this influence was contingent on an already-tense and volatile situation. Given the politically vulnerable position of Eisenhower, this might have created political consequences for him, but it is highly doubtful that they were intentionally manufactured by Khrushchev. Heads of state must attend the needs and desires of various audiences, and Khrushchev鈥檚 rhetorical behavior likely was an attempt to balance the many interests of his allies, the Soviet leadership, and long-term goals. Given the unique situation of the Berlin crisis, these influences of non-status-quo-changing rhetoric are likely to be much weaker in most other cases. Further research on the influence of foreign rhetoric and propaganda is needed, but for the time being, hostile foreign influence appears to be limited and conditional. Crisis rhetoric is an extremely powerful tool, but ultimately it is a blunt instrument whose efficacy is subject to the interpretations and value judgments of other actors鈥攚hen foreign leaders are not producing new and interesting developments, they are largely ignored.

听

Sources Cited:

听

Auerbach, Yehudith and Yaeli Bloch-Elkon. 2005. 鈥淢edia Framing and Foreign Policy: The Elite Press vis-脿-vis US Policy in Bosnia, 1992-95.鈥� Journal of Peace Research 42 (January): 83-99.

Baum, Matthew A. 2004. 鈥淕oing Private: Public Opinion, Presidential Rhetoric, and the Domestic Politics of Audience Costs in US Foreign Policy Crises.鈥� Journal of Conflict Resolution 48 (October): 603-631.

Baum, Matthew A. and Philip B. K. Potter. 2008. 鈥淭he Relationships Between Mass Media, Public Opinion, and Foreign Policy: Toward a Theoretical Synthesis.鈥� Annual Review of Political Science 11 (June): 39-65.

Behr, Roy L. and Shanto Igengar. 1985. 鈥淭elevision News, Real-World Cues, and Changes in the Public Agenda.鈥� Public Opinion Quarterly 49 (1): 38-57.

Bennett, Lance W. 1990. 鈥淭oward a Theory of Press-State Relations in the United States.鈥� Journal of Communication 40 (June): 103-127.

Bond, Jon R. and Richard Fleisher. 1990. The President in the Legislative Arena. Chicago: The University of Chicago Press.

Burr, William. 1994. 鈥淎voiding the Slippery Slope: The Eisenhower Administration and the Berlin Crisis, November 1958-January 1959.鈥� Diplomatic History 18 (April): 177-205.

Cohen, Bernard C. 1963. The Press and Foreign Policy. Princeton: Princeton University Press.

Cohen, Jeffery E. 1995. 鈥淧residential Rhetoric and the Public Agenda.鈥� American Journal of Political Science 39 (February): 89-107.

Cohen, Jeffery E. 1997. Presidential Responsiveness and Public Policy-Making: The Public and the Policies that Presidents Choose. Ann Arbor, MI: University of Michigan Press.

Cook, Fay L. et al. 1983. 鈥淢edia and Agenda-Setting: Effects on the Public, Interest Group Leaders, Policy Makers, and Policy.鈥� Public Opinion Quarterly 47 (1): 16-35.

Dearing, James W. and Everett M. Rogers. 1996. Agenda Setting. Thousand Oaks, CA: Sage Publications.

Downs, Anthony. 1972. 鈥淯p and Down with Ecology: The Issue Attention Cycle.鈥� Public Interest 28: 38-50

Druckman, James N. and Justin W. Holmes. 2004. 鈥淒oes Presidential Rhetoric Matter? Priming and Presidential Approval.鈥� Presidential Studies Quarterly 34 (December): 755-778.

Entman, Robert M. 1993. 鈥淔raming: Toward Clarification of a Fractured Paradigm.鈥� Journal of Communication 43 (December): 51-58.

Entman, Robert M. 2004. Projections of Power: Framing News, Public Opinion, and U.S. Foreign Policy. Chicago: University of Chicago Press.

Edwards, George C. and Dan B. Wood. 1999. 鈥淲ho Influences Whom? The President, Congress, and the Media.鈥� The American Political Science Review 93 (January): 327-344.

Evans, Peter B., Harold K. Jacobson, and Robert D. Putnam, eds. 1993. Double-Edged Diplomacy: International Bargaining and Domestic Politics. Los Angeles: University of California Press.

Fearon, James D. 1994. 鈥淒omestic Political Audiences and the Escalation of International Disputes.鈥� The American Political Science Review 88 (September): 577-592.

Fearon, James D. 1997. 鈥淪ignaling Foreign Policy Interests: Tying Hands versus Sinking Costs.鈥� Journal of Conflict Resolution 41 (February): 68-90.

Foyle, Douglas C. 1999. Counting the Public In: Presidents, Public Opinion, and Foreign Policy. New York: Columbia University Press.听

Fursenko, Aleksandr and Timothy Naftali. 2006. Khrushchev鈥檚 Cold War: The Inside Story of an American Adversary. New York: W.W. Norton & Company.

Gowa, Joanne. 1998. 鈥淧olitics at the Water鈥檚 Edge: Parties, Voters, and the Use of Force Abroad.鈥� International Organization 52 (Spring): 307-324.

Granger, C. W. J. 1969. 鈥淚nvestigating Causal Relationships by Econometric Models and Cross-spectral Methods.鈥� Econometrica 37 (August): 424-438.

Harrison, Hope M. 1993. 鈥淯lbricht and the Concrete 鈥楻ose鈥�: New Archival Evidence on the Dynamics of Soviet-East German Relations and the Berlin Crisis, 1958-1961.鈥� Cold War International History Project, working paper no. 5. Washington, DC: Woodrow Wilson Center

Hayes, Danny and Matt Guardino. 2013. Influence from Abroad: Foreign Voices, the Media, and U.S. Public Opinion. New York: Cambridge University Press.

Holsti, Ole R. 1992. 鈥淧ublic Opinion and Foreign Policy: Challenges to the Almond-Lippmann Consensus, Mershon Series: Research Programs and Debates.鈥� International Studies Quarterly 36 (December): 439-466.

Iyengar, Shanto and Donald R. Kinder. 1987. News that Matters: Television and American Opinion. Chicago: University of Chicago Press.

Iyengar, Shanto. 1991. Is Anyone Responsible? How Television Frames Political Issues. Chicago: University of Chicago Press.

Jennings, Will and Christopher Wlezein. 2011. 鈥淒istinguishing Between Most Important Problems and Issues?鈥� Public Opinion Quarterly 75 (September): 545-555.

Kingdon, John W. 1995. Agendas, Alternatives, and Public Policies. New York: Harper Collins College Publishers.

Knecht, Thomas. 2010. Paying Attention to Foreign Affairs: How Public Opinion Affects Presidential Decision Making. University Park, PA: Pennsylvania State University Press.

Lawrence, Adam B. 2004. 鈥淒oes it Matter What Presidents Say? The Influence of Presidential Rhetoric on the Public Agenda, 1946-2003.鈥� University of Pittsburgh, unpublished doctoral dissertation.

Light, Paul.1991. The President鈥檚 Agenda. Baltimore, MD: Johns Hopkins University Press.

McCombs, Maxwell E. and Donald L. Shaw. 1972. 鈥淭he Agenda-Setting Function of Mass Media.鈥� Public Opinion Quarterly 36 (2): 176-187.

McCombs, Maxwell E. and Donald L. Shaw. 1993. 鈥淭he Evolution of Agenda-Setting Research: Twenty-Five Years in the Marketplace of Ideas.鈥� Journal of Communication 43 (2): 58-67.

Monroe, Alan D. 1979. 鈥淭he Consistency Between Public Preferences and National Policy Decisions.鈥� American Politics Quarterly 7 (January): 3-19.

Page, Benjamin I. and Robert Y. Shapiro. 1983. 鈥淓ffects of Public Opinion on Policy.鈥� The American Political Science Review 77 (March): 175-190.

Page, Benjamin I., Robert Y Shapiro, and Glenn R. Dempsey. 1987. 鈥淲hat Moves Public Opinion?鈥� The American Political Science Review 81 (March): 23-44.

Partell, Peter J. and Glenn Palmer. 1999. 鈥淎udience Costs and Interstate Crises: An Empirical Assessment of Fearon鈥檚 Model of Dispute Outcomes.鈥� International Studies Quarterly 43 (June): 389-405.

Prins, Brandon C. and Christopher Sprecher. 1999. 鈥淚nstitutional Constraints, Political Opposition, and Interstate Dispute Escalation: Evidence from Parliamentary Systems, 1946-89.鈥� Journal of Peace Research 36 (May): 271-287.

Putnam, Robert D. 1988. 鈥淒iplomacy and Domestic Politics: the logic of two-level games.鈥� International Organization 42 (3): 427-460.

Richter, James G. 1994. Khrushchev鈥檚 Double Bind: International Pressures and Domestic Coalition Politics. Baltimore: John鈥檚 Hopkins University Press.

Risse-Kappen, Thomas. 1991. 鈥淧ublic Opinion, Domestic Structure, and Foreign Policy in Liberal Democracies.鈥� World Politics 43 (July): 479-512.

Semetko, Holli A. et al. 1992. 鈥淭V News and U.S. Public Opinion About Foreign Countries: The Impact of Exposure and Attention.鈥� International Journal of Public Opinion Research 4 (1): 18-36.

Scheufele, Dietram A. 1999. 鈥淔raming as a Theory of Media Effects.鈥� Journal of Communication 49 (March): 103-122.听

Scheufele, Dietram A. and David Tewksbury. 2007. 鈥淔raming, Agenda Setting, and Priming: The Evolution of Three Media Effects Models.鈥� Journal of Communication 57 (March): 9-20.

Schick, Jack M. 1971. The Berlin Crisis 1958-1962. Philadelphia: University of Pennsylvania Press.

Schultz, Kenneth A. 2001. 鈥淟ooking for Audience Costs.鈥� Journal of Conflict Resolution 45 (February): 32-60.

Schultz, Kenneth A. 2012. 鈥淲hy We Needed Audience Costs and What We Need Now.鈥� Security Studies 21 (3): 369-375.

Shapiro, Robert Y. and Benjamin I. Page. 1988. 鈥淔oreign Policy and the Rational Public.鈥� The Journal of Conflict Resolution 32 (June): 211-247.

Slantchev, Branislav L. 2006. 鈥淧oliticians, the Media, and Domestic Audience Costs.鈥� International Studies Quarterly 50 (June): 445-477.

Snyder, Jack and Erica D. Borghard. 2011. 鈥淭he Cost of Empty Threats: A Penny, Not a Pound.鈥� The American Political Science Review 105 (August): 437-456.

Soroka, Stuart N. 2003. 鈥淢edia, Public Opinion, and Foreign Policy.鈥� The International Journal of Press/Politics 8 (January): 27-48.

Soroka, Stuart N. and Christopher Wlezien. 2004. 鈥淥pinion Representation and Policy Feedback: Canada in Comparative Perspective.鈥� Canadian Journal of Political Science 37 (September): 531-559.

Terkildsen, Nayda and Frauke Schnell. 1997. 鈥淗ow Media Frames Move Public Opinion: An Analysis of the Women鈥檚 Movement.鈥� Political Research Quarterly 50 (December): 879-900.

Tomz, Michael. 2007. 鈥淒omestic Audience Costs in International Relations: An Experimental Approach.鈥� International Organization 61 (Fall): 821-840.

Trachtenberg, Marc. 2012. 鈥淎udience Costs: An Historical Analysis.鈥� Security Studies 21 (1): 3-42.

Trachtenberg, Marc. 1991. History and Strategy. Princeton, NJ: Princeton University Press.

Trumbore, Peter, F. 1998. 鈥淧ublic Opinion as a Domestic Constraint in International Negotiations: Two-Level Games in the Anglo-Irish Peace Process.鈥� International Studies Quarterly 42 (3): 545-565.

Villalobos, Jose D. and Cigdem V. Sirin. 2012. 鈥淎genda Setting from the Oval Office: An Experimental Examination of Presidential Influence over the Public Agenda.鈥� International Journal of Public Opinion Research 24 (1): 21-41.

Walgrave, Stefaan and Peter Van Aelst. 2006. 鈥淭he Contingency of the Mass Media鈥檚 Political Agenda Setting Power: Toward a Preliminary Theory.鈥� Journal of Communication 56 (1): 88-109.

Wanta, Wayne, and Yu-Wei Hu. 1993. 鈥淭he Agenda-Setting Effects of International News Coverage: An Examination of Differing News Frames.鈥� International Journal of Public Opinion Research 5 (3): 250-264.

Weeks, Jessica L. 2008. 鈥淎utocratic Audience Costs: Regime Type and Signaling Resolve.鈥� International Organization 62 (Winter): 35-64.

Wlezien, Christopher. 2005. 鈥淥n the Salience of Political Issues: The Problem With 鈥楳ost Important Problem.鈥欌€� Electoral Studies 24 (December): 555-579.

Wood, Dan B. and Jeffery S. Peake. 1998. 鈥淭he Dynamics of Foreign Policy Agenda Setting.鈥� The American Political Science Review 92 (March): 173-184.

Zaller, John R. 1992. The Nature and Origins of Mass Opinion. Cambridge: Cambridge University Press. 听

Zaller, John R. and Dennis Chiu. 1996. 鈥淕overnment鈥檚 Little Helper: U.S. Press Coverage of Foreign Policy Crises, 1945-1991.鈥� Political Communication 13: 1996.

Zhang, Cui and Charles W. Meadows III. 2012. 鈥淚nternational Coverage, Foreign Policy, and National Image: Exploring the Complexities of Media Coverage, Public Opinion, and Presidential Agenda.鈥� International Journal of Communication 6: 76-95.

Zubok, Vladislav M. 1993. 鈥淜hrushchev and the Berlin Crisis 1958-1962.鈥� Cold War International History Project, working paper no. 6. Washington, DC: Woodrow Wilson Center.

听

Historical Documents:

听

U.S. Department of State. Office of the Historian, Bureau of Public Affairs. 1986. Documents on Germany, 1944-85. Washington D.C.: Government Printing Office.听

U.S. Department of State. Office of the Historian, Bureau of Public Affairs. 1993. Foreign Relations of the United States, 1958-1960, Vol. VIII, Berlin Crisis, 1958-1959. Washington DC: Government Printing Office.

Central Intelligence Agency. Office of Current Intelligence.1961. Berlin Handbook. Released Oct. 2005.

Truman, Harry S. Office of the Press Secretary. Office of the Federal Register, National Archives and Records Administration. 1964. Public Papers of the Presidents. Washington D.C.: U.S. Government Publication Office.

Eisenhower, Dwight D. Office of the Press Secretary. Office of the Federal Register, National Archives and Records Administration. 1960. Public Papers of the Presidents. Washington D.C.: U.S. Government Publication Office.

听

Articles in the New York Times:

听

Ash, Timothy Garton. 2014. 鈥淧utin鈥檚 Deadly Doctrine: 鈥楶rotecting鈥� Russians in Ukraine Has Fatal Consequences.鈥� New York Times, July 18, 2014, web.

Frankel, Max. 1958. 鈥淜hrushchev Bids Four Powers End Berlin Control: Says at Polish Fete Soviet is Ready to Yield Role to East Germany.鈥� New York Times, Nov. 11, p. 1.

Frankel, Max. 1958. 鈥淜hrushchev Hints Talk Might Delay Action on Berlin.鈥� New York Times. Nov. 30, p. 1.

Gruson, Sydney. 1958. 鈥淒ulles Staggers Berlin and Allies: Intimation of Readiness to Deal with East Germans Causes Consternation.鈥� New York Times. Nov. 27, p. 19.

Kenworthy, E.W. 1958. 鈥淲ashington Calm on German Issue: U.S. Officials Recall Other Soviet Moves to Weaken Western Tie to Berlin.鈥� New York Times, Nov. 12, p.11.听

Krock, Arthur. 1958. 鈥淚n the Nation: We Painted Ourselves Into the Corner.鈥� New York Times. Nov. 13, p. 32.

Krock, Arthur. 1959. 鈥淚n the Nation: A Very Important Tea Party Indeed.鈥� New York Times, Mar. 26, p. 30.

Lewis, Flora. 1959. 鈥淎 Symbol, a Weapon, a Refuge.鈥� New York Times, Nov. 22, p. SM14.

Middleton, Drew. 1949. 鈥淎cheson in Berlin Pledges Countering of Soviet Moves.鈥� New York Times, Nov. 15, p. 1.

Middleton, Drew. 1959. 鈥淲estern Accord on Berlin Urged: Diplomats Abroad Forecast Soviet-East German Pact if Disunity Persists.鈥� New York Times, Nov. 22, p. 39.

New York Times. 1958. 鈥淏erlin: A Symbol.鈥� Nov. 23, p. E8.

New York Times. 1958. 鈥淪oviet Campaign.鈥� Dec. 13, p. B2.

New York Times. 1959. 鈥淏erlin Shift Doubted: Khrushchev Holds to Deadline, Western Envoys Believe.鈥� Feb. 9, p. 3.

New York Times. 1959. 鈥淜hrushchev and His Puppets.鈥� Mar. 13, p. 28.

Reston, James. 1959. 鈥淎 Spot Check: Public Firm on Berlin, but Ill-Informed.鈥� New York Times. Mar. 23, p. 1.

Rosenthal, A. M. 1959. 鈥淜hrushchev Sees Scandinavia Risk: Tells Poles West is Plotting to Make Area Decoy for Soviet Bombs in War.鈥� New York Times. July 18, p. 2.

Salisbury, Harrison E. 1959. 鈥淜hrushchev鈥檚 Russia Part II: 鈥楥ollective Leadership鈥� is Replaced by Premier鈥檚 Own Undisputed Rule.鈥� New York Times. Sept. 9, p. 1.

听

Appendix:

听

Date	N News Articles	NSQC Statements	SQC Statements
11/9/58	24	1	1
11/16/58	34	0	0
11/23/58	58	0	2
11/30/58	52	2	0
12/7/58	30	0	0
12/14/58	14	1	0
12/21/58	6	0	0
12/28/58	23	0	0
1/4/59	19	0	0
1/11/59	20	0	0
1/18/59	19	0	0
1/25/59	15	0	0
2/1/59	23	1	0
2/8/59	25	0	0
2/15/59	25	1	0
2/22/59	36	0	1
3/1/59	53	4	1
3/8/59	54	4	1
3/15/59	48	1	1
3/22/59	33	1	0
3/29/59	34	0	0
4/5/59	34	1	0
4/12/59	18	0	0
4/19/59	21	0	0
4/26/59	24	1	0
5/3/59	23	2	0
5/10/59	31	2	0
5/17/59	23	2	0
5/24/59	24	1	1
5/31/59	32	1	0
6/7/59	27	2	0
6/14/59	27	1	1
6/21/59	23	1	0
6/28/59	18	0	0
7/5/59	16	0	0
7/12/59	34	2	0
7/19/59	24	0	0
7/26/59	12	0	0
8/2/59	19	1	0
8/9/59	4	0	0
8/16/59	2	1	0
8/23/59	11	1	0
8/30/59	9	1	0
9/6/59	4	1	0
9/13/59	5	4	0
9/20/59	7	1	0
9/27/59	29	3	1
10/4/59	14	1	0
10/11/59	6	0	0
10/18/59	3	0	0
10/25/59	9	1	0
11/1/59	7	0	0
11/8/59	7	0	0
11/15/59	13	0	0
11/22/59	12	0	0
11/29/59	9	1	0
12/6/59	3	0	0
12/13/59	5	1	0
12/20/59	15	1	1
12/27/59	7	1	0
1/3/60	2	0	0
1/10/69	5	1	0
1/17/60	3	0	0
1/24/60	4	0	0
1/31/60	10	2	0
2/7/60	11	1	0
2/14/60	6	0	0
2/21/60	8	0	0
2/28/60	8	1	0
3/6/60	10	0	0
3/13/60	19	0	0
3/20/60	14	2	0
3/27/60	4	2	0
4/3/60	18	1	0
4/10/60	14	0	0
4/17/60	10	0	0
4/24/60	24	1	0
5/1/60	9	1	0
5/8/60	11	4	0
5/15/60	32	2	2
Total:听	1478	69	13

听

[1] Although I discuss public opinion in the context of democratic states, there are some scholars who argue that the same mechanisms of accountability apply to non-democratic regimes (see: Weeks 2008).

[2] It has been noted that while political scientists often use the concept of salience in reference to policy issues, rarely do scholars define what exactly they mean by the term (Wlezein 2005).听 While Wlezien notes that there may be a need to distinguish between 鈥渋mportance鈥� and 鈥渟tatus as a problem鈥� when discussing salience, later studies suggest that in survey data, different wording does not generate statistically significant differences (Jennings & Wlezien 2011). For the purposes of my study, I will define salience as the degree to which voters consider a political issue to be a significant problem by which they can gauge their leaders鈥� performance.

听

[3] Despite these insights, audience costs have garnered a large amount of criticism due to difficulties in detecting them (Tomz 2007; Schultz 2001), their rarity (Trachtenberg 2012), doubts over the consequences of empty threats (Slatchev 2006; Gowa 1998), and the persistence of private crisis diplomacy (Baum 2004), among others (Snyder & Borghard 2011). While many of these problems might be accounted for by the theory itself (the rarity of occurrence is not surprising as leaders not likely to win would never incur this risk in the first place) or with slight modifications, others are part of an ongoing debate over the nature, or even existence of audience costs. Despite these theoretical schisms, audience costs have proven notably instructive in explaining other aspects of international politics (Schultz 2012), and have provided a newfound importance of public opinion and domestic politics in the formation and conduct of foreign policy.

[4] I would like to thank Dr. Rolf Norgaard of the Program for Writing and Rhetoric for pointing out the usefulness of framing to me as a possible approach to this problem.

[5] For example, in the months leading up to the summit of heads of state in spring, 1960, a slew of news articles were published on the perceived differences in goals of the different Allies going into the conference. While the negotiations and disagreements between the U.S., Britain, France, and West Germany were arguably a natural part of the diplomatic process, many articles presented the issue as if the West was tearing itself apart before it could face the Soviets. For an example, see (Middleton, Nov. 22 1959, 39 NYT).听

[6] Also included in this collection were mentions of the refugee flow of East Germans to the West via Berlin in early 1960. The media often linked this issue with the larger geopolitical crisis, and the situation helped to perpetuate a public consciousness of the city. This refugee flow was a significant source of tension, and arguably one of the main reasons East German leader Walter Ulbricht exerted pressure on the Soviets to press the Berlin issue (Harrison, 1993).

听

* Significance is defined at the 95% level (p<0.05). Lag, in this case, means the shifting of one data set against the other to test for correlations between non-synchronous events. For example, NSQC data from week one is tested for its correlation to SAL from week two.

听[AH1]write out the country names

听[AH2]A page number for this quote is needed

Geothermal Reconnassance, Jennifer Bloom

(Back to Top)

The Atheist Experience, Cali Greksa

(Back to Top)

听

THE ATHEIST EXPERIENCE:

A SOCIOLOGICAL APPROACH TO ATHEIST IDENTITY IN COLLEGE STUDENTS 听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听听

听

ABSTRACT

听听听听听听听听听听听 Atheists in the United States remain a largely invisible minority in a highly religious social climate. College atheists as individuals have also been largely dismissed in sociology and the sociology of religion, especially with regard to their identity formation, identity salience, and experiences of stigma and stigma-coping mechanisms. Based on in-depth, qualitative interviews with 14 self-identified atheist college students, I argue that atheists begin the process of forming their identity during adolescence as a result of numerous background factors and through an ongoing negotiation of a society that they perceive as unaccepting on some level. I also argue that college plays a role in providing a supportive context in which atheists can affirm and crystallize their identities as a result of processes set in motion before college. This thesis seeks to gain a nuanced understanding of how college atheists as individuals have formed their identities and how they negotiate this embattled identity that remains stigmatized by U.S. society based on social constructions of the atheist as immoral. Respondents reported similar trends in their backgrounds and events in their childhoods or young adulthoods that sparked the eventual adoption of an atheist identity. They also reported variations in atheist identity salience as well as perceptions stigma and acceptance. Respondents coped with stigma by redefining friend groups, controlling information, and positing themselves as moral people without believing in God.

听

TABLE OF CONTENTS

CHAPTER 1: INTRODUCTION鈥︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌�.5

听听听听听听听听听听听 I. What is an Atheist?...........................................................................................................5

听听听听听听听听听听听 II. Research Focus and Questions鈥︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€�....7

听

CHAPTER 2: A REVIEW OF THE LITERATURE鈥︹赌︹赌�.鈥︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌�..10

听听听听听听听听听听听 I. Embattled Identity: Stigma and Casting Atheists as the 鈥淥ther鈥濃€︹€︹€︹€︹€︹€︹€︹€�...10

听听听听听听听听听听听 II. Atheist Identity: Formation and Salience鈥︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€�..17

听听听听听听听听听听听 III. College and Atheism鈥︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€�27

听听听听听听听听听听听 IV. The Current Study鈥︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€�28

听

CHAPTER 3: METHODS鈥︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌�..30

听听听听听听听听听听听 I. Setting鈥︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€�.30

听听听听听听听听听听听 II. Recruitment Process鈥︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€�......30

听听听听听听听听听听听 III. Sample鈥︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€�..31

听听听听听听听听听听听 IV. Interviews鈥︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€�...鈥︹€︹€︹€︹€�.32

听

CHAPTER 4: FINDINGS鈥�....鈥︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌�...35

听听听听听听听听听听听 I. Precursors to Atheism: Backgrounds鈥︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€�.35

听听听听听听听听听听听 II. Pathways to Atheism: 鈥淭he Spark鈥濃€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€�...42

听听听听听听听听听听听 III. Experiences with an Atheist Identity鈥︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€�..鈥︹€�.54

听

CHAPTER 5: DISCUSSION鈥︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌�..69

听听听听听听听听听听听 Study Limitations鈥︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€�...69

听听听听听听听听听听听 I. Backgrounds鈥︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€�...鈥︹€︹€︹€︹€︹€︹€︹€︹€︹€�........70

听听听听听听听听听听听 II. Pathways to Atheism鈥︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€�.........................71

听听听听听听听听听听听 III. Experiences with an Atheist Identity鈥︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€�...73

听听听听听听听听听听听 IV. Perceptions of Stigma鈥︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€�..74

听听听听听听听听听听听 V. Individual Coping Mechanisms鈥︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€�........75

听

CHAPTER 6: CONCLUSION鈥︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌�...78

听

REFERENCES鈥︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌︹赌�80

听

APPENDIX A: Recruitment Material鈥︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€�..............85

听

APPENDIX B: Consent Form (HRP-502)鈥︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€�.......87

听

APPENDIX C: Interview Schedule (HRP-503: Protocol)鈥︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€�...91

听

APPENDIX D: Religious Information of Participants鈥︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€︹€�.94

听

CHAPTER 1: INTRODUCTION

听

With only 2.4 percent of the population claiming an atheist identity, to be an atheist in the United States is to belong to a clear minority (Lipka 2013). Indeed, 82.0 percent of the U.S. population report that religion is either 鈥渧ery important鈥� or 鈥渟omewhat important鈥� in their lives, and 78.4 percent of U.S. adults describe themselves as specifically Christian (Carriere-Kretschmer and Holzwart 2008). These figures place the United States as the most religious developed country in the world, ranking closer to countries such as Mexico, Brazil and India where 86, 87, and 89 percent (respectively) of these populations cite religion as 鈥渧ery鈥� or 鈥渟omewhat鈥� important. By contrast, in the developed West, 42 percent of the British population and 37 percent of the French population rate religion as 鈥渧ery鈥� or 鈥渟omewhat鈥� important (Carriere-Kretschmer and Holzwart 2008). Sociologists of religion have recognized how deeply ingrained religion and more specifically, a 鈥渘ormative鈥� belief in God is in U.S. society (Smith 2010:6). This social fact has certain implications for the minority of U.S. citizens who reject a belief in God and who claim an atheist identity.听

I. What is an Atheist?

In a highly religious social climate, to be an atheist in the U.S. is to deviate from the norm of believing in God and to be cast as a social 鈥渙ther鈥� (Smith 2010:15). Until recently, scant research has been conducted on irreligion and nonreligious individuals, but in order to truly understand the religious landscape, both religion and irreligion must be taken into account (Bainbridge 2009:319). Although there is some debate about the true denotation, irreligion is most commonly cited as the absence of religious belief or the indifference to religion. It can also imply negative attitudes toward religion or a rejection of religion (Cline 2015). Campbell explains that although terms such as 鈥渋rreligion,鈥� 鈥渟ecularism,鈥� and 鈥渁theism鈥� are sometimes conflated, there are nuances, and irreligion does not necessarily have to include a feeling of hostility toward religion (1972:22). In fact, as Campbell observes, there exist many 鈥渞eligiously sympathizing unbelievers鈥� inside and outside churches and religious institutions who are raised in a religious culture, but who cannot make themselves believe (Campbell 1972:22). This is important to note because not all non-believers feel the same way toward religion, even though they are often cast as uniformly hostile and rejecting of religion in all of its forms. Among atheists, there exists a more complicated and nuanced spectrum of beliefs than a simple dichotomy between religious and irreligious (1972:25). Ultimately, Campbell urges scholars to take into account the beliefs, actions, attitudes, and experiences of nonbelievers when studying irreligion in the same way these aspects of behavior are taken into account when studying religion and religious believers (1972:23).

It is important to note that my use of the term 鈥渁theist鈥� is intentional. In the United States, the word 鈥渁theist鈥� is sometimes hesitantly used even by atheists themselves due to the stigma the word carries that maligns atheists and often labels them as immoral and untrustworthy (Edgell, Gerteis, and Hartmann 2006:211). However, what does it really mean to be an atheist? Even the atheist community is somewhat split on this issue. Bainbridge observes: 鈥淚n a monotheistic society, an atheist may be someone who simply does not subscribe to a belief in God while in a polytheistic society, an atheist may be one who does not believe in enough gods鈥� (2009:320). Bainbridge also contrasts an implicit atheist with a lower-case 鈥渁鈥� from an explicit Atheist with a capital 鈥淎.鈥� The implicit atheist is one who simply does not subscribe to a belief in God, while the explicit Atheist consciously rejects a belief in God and actively proclaims this (non) belief (2009:320).

For the purposes of my study, I refer to atheists in the sense of those who do not subscribe to a belief in any god(s) and who do not consider themselves a member of any religion. For the purpose of ease and clarity, I will conflate terms such as 鈥渟ecular,鈥� 鈥渋rreligious,鈥� 鈥渦nbeliever,鈥� and 鈥渁postate鈥� with 鈥渁theism鈥� with the understanding that each of these terms is itself nuanced and not necessarily equivalent. To encompass all of these individuals, I refer to all atheists with a lower-case 鈥渁.鈥� Agnosticism will not be a focus of my examination since agnostics often 鈥渄isagree less strenuously with those who possess religious faith鈥�[1] (Bainbridge 2009:320). In my own research, I employed the word 鈥渁theist鈥� on all recruitment material, but allowed interview participants to expand on their preferred terminology, asking them to give their own definition of what being an atheist means to them personally. This is important because views on religion and other personally held beliefs are often very subjective and difficult to define in very strict terms or as a binary (atheist versus religious).[2]听

II. Research Focus and Questions

听听听听听听听听听听听 Important studies have already been conducted on atheist identity formation in recent years with an emphasis on atheists in a social context, studying what LeDrew (2013:1) terms 鈥渁ctive atheists,鈥� and more specifically, atheists that are members of atheist clubs and social groups. Empirical work also points to trends in the ebbs and flows of religious practice across the life-course, with the least religious years (operationalized by a low practice frequency) most typically occurring during an individual鈥檚 college years (Caplovitz and Sherrow 1977; Regnerus and Uecker 2007; Schweitzer 2004). The Pew Research Center reports that young people make up the least religious group in the U.S. with only 48 percent of people between the ages of 18 and 39 citing religion as 鈥渧ery important鈥� compared to 55 percent of adults aged 40 to 59 and 64 percent of people 60 and older (Carriere-Kretschmer and Holzwart 2008). This trend of younger people being the most disinterested in religion is mirrored by the typical demographics of self-identified apostates who tend to be 鈥渓iberal, Democrats, independents, younger, living in the West, students, and those who are living with someone without being married鈥� (Beit-Hallahmi 2011:301). Several studies that have explored the trend of lessened religious practice in the younger demographic have hypothesized that a tendency toward a lessened religious salience or practice may just be a phase, and that individuals will eventually return to religion (Regnerus and Uecker 2007; Schweitzer 2004). However, this thesis will address a gap that seems to appear in the research relating to a disaffiliation with religious belief early in life (not simply religious practice).

听听听听听听听听听听听 This thesis seeks to supplement the extant literature by focusing primarily on the individual, biographical experiences of college atheists who do not necessarily fall inside the group context studied by scholars such as LeDrew (2013) and Smith (2010), or whose identity formation processes are compared to those of other religious college students (Bowman and Small 2010). This research also seeks to fill the gap in the literature regarding the role of college. I argue that college functions as an institution that serves to crystallize an identity-achievement process already set in motion in childhood and adolescence, rather than as an institutional moment in a person鈥檚 life that causes a person to become an atheist, as suggested by Caplovitz and Sherrow (1977). My research will aim to understand the lived experiences of atheists including their backgrounds, how they come to identify as an atheist, their own personal definitions of an atheist identity, how salient this identity is, and the experiences of stigma and of coping with stigma. I therefore ask the following research questions:

听听听听听听听听听听听 1. What are the background factors that lead college students to adopt an atheist 听听 identity?

听听听听听听听听听听听 2. What is the process of adopting an atheist identity?

听听听听听听听听听听听 3. How salient does an atheist identity become?

听听听听听听听听听听听 4. Do people feel that this part of their identity is stigmatized? If so, how do they cope?

听听听听听听听听听听听

听

CHAPTER 2: A REVIEW OF THE LITERATURE

听

听听听听听听听听听 This review of the literature is comprised of three main sections. In the first section, Embattled Identity: Stigma and Casting Atheists as the 鈥淥ther,鈥� the concept of stigma will be discussed, given its importance in casting an atheist identity as an embattled and stigmatized identity. Included in the examination of stigma are subsections discussing commonly held beliefs regarding morality that contribute to the stigma surrounding atheists. The second section, Atheist Identity: Formation and Salience, discusses existing research documenting the various pathways taken by individuals in adopting an atheist identity. This section also discusses other aspects of identity such as salience. The third section, College and Atheism, focuses on the institution of college and its controversial role in the formation of an atheist identity.

I. Embattled Identity: Stigma and Casting Atheists as the 鈥淥ther鈥�

Although levels of stigma against an individual or group can vary in intensity, historically, there has been and continues to be a certain level of stigma and societal unacceptance of atheists. This informs the idea that to be an atheist is to claim a stigmatized and embattled identity. Campbell explains the significant development during the Victorian era that shaped the current societal consciousness toward atheists:

The Victorian caricature of the atheist as a depraved, friendless and tortured soul has left 听 its mark on the collective consciousness of contemporary society (and equally upon 听听听听 contemporary sociology) in the form of a tendency to view his modern counterpart as an aggressively nonconforming and neurotic person who is alienated from his social milieu. 听听听听听听 (1972:39)

听

These ideas of the atheist as 鈥渄epraved,鈥� 鈥渇riendless,鈥� and so on, are important to understand when exploring the atheist identity as embattled and stigmatized. This section will discuss stigma as applied to an atheist identity as well as the causes and effects of the stigmatization of atheists.

According to Erving Goffman, the term stigma refers 鈥渢o bodily signs designed to expose something unusual and bad about the moral status of the signifier鈥� (1968:1). Goffman uses this term to refer to a personal characteristic including race, religion, addiction, homosexuality, and so on that discredits a person and denies them 鈥渇ull social acceptance鈥� (1968: Preface-5). He explains:

By definition, of course, we believe the person with a stigma is not quite human. On this assumption we exercise varieties of discrimination, through which we effectively, if often unthinkingly, reduce his life chances. We construct a stigma-theory, an ideology to explain his inferiority and account for the danger he represents, sometimes rationalizing an animosity based on other differences, such as those of social class. (1968:5)

听

Different types of stigma exist in society and do so to varying degrees of visibility. Goffman explains that the visibility of a stigma is important to a person鈥檚 identity in social situation. If, for example, the stigma is visible or impedes interaction (such as having a large birth mark on one鈥檚 face or speaking with a stammer), the person is stigmatized upon contact with others (1968:48-49). Atheists are an interesting group because they can usually 鈥減ass鈥� or be secretive about their 鈥渇ailings鈥� in order to get along in every day life by being perceived as 鈥渘ormal鈥� (1968:73).[3] 鈥淧assing鈥� can be achieved through the use of a strategy suggested by Goffman where only a controlled amount of information is shared about oneself to a majority of the public. A stigma-bearer might only reveal her true identity to a trusted minority, thereby dividing the world into two groups: a large majority to which little information is shared and a small minority to whom the stigma-bearer 鈥渢ells all鈥� (1968:95). Similarly, in many cases, atheists do not disclose their beliefs or identity to most acquaintances and may only talk freely with close friends and other trusted individuals.

听听听听听听听听听听听 Forming in-group and out-group affiliations is another way people cope with stigma. In-groups are 鈥渇ellow-sufferers鈥� of a particular stigma (1968:112). Goffman suggests that in some cases, in-groups adopt a militant or secessionist ideology where members might actually 鈥渇launt鈥� some of their traits that clearly 鈥渙ther鈥� them from the rest of society or the 鈥渙ut-group鈥� (people who do not share their stigma) (Goffman 1968:113). Through this framework, some atheists take on an 鈥渦s鈥� versus 鈥渢hem鈥� mindset and adopt a more combative approach against the people who stigmatize them. A good example of this in-group affiliation where atheists make themselves more visible is Richard Dawkins鈥檚 online 鈥淥ut Campaign.鈥� This website is a space where atheists can read, share stories, blog, and buy bumper stickers, T-shirts, and other merchandise relating to atheism. Atheists are encouraged to 鈥渃ome out of the closet鈥� in order to publicly affirm their identity and make themselves more visible rather than trying to 鈥減ass鈥� (Dawkins 2013).

听听听听听听听听听听听 Smith explains that forming social atheist groups is yet another way that atheists form in-groups instead of trying to 鈥減ass鈥� (Smith 2013). Some atheists who 鈥渃ome out鈥� publicly manage their stigmatized identities by finding common ground with other atheists through atheist-run charities, social groups, and activism campaigns. These collective group affiliations have become much more common and provide spaces for atheists to reaffirm their identities and mobilize for a common cause (2013:96). Despite forming their own in-groups, atheists are still cast by the majority of U.S. society as the external 鈥渙ut-group鈥� (Smith 2010:14). The importance of the designation of an atheist identity as a deviance from the norm cannot be understated when trying to understand the implications of adopting this embattled identity. Smith explains:

Because of the stigmatized and deviant status of atheism, it can initially be difficult to 听听听听听 claim the identity in a social setting. This results in part from the tension experienced from their knowledge of having rejected the normative views of the larger society, and 听听听听听 feeling disconnected from the rest of American culture鈥�(2010:15)

听

It is important to explore some of the reasons why much of the U.S public casts atheists as the 鈥渙ther.鈥� One of the major sources of othering and stigma is the belief that atheists reject the foundations of morality.听

Causes of Stigma: 鈥淩ejection鈥� of Morality

听听听听听听听听听听听 A passage of the King James Version of the Bible (Psalm 14) states: 鈥淭he fool hath said in his heart, There is no God. They are corrupt, they have done abominable works, there is none that doeth good鈥� (as cited by Bainbridge 2009:324). This passage is an example of a moral argument against atheists that roots itself in religion. Further, words such as 鈥渨ickedness鈥� are often conflated with atheism (2009:324). As a result, it is no surprise that language surrounding atheism in religious texts and ideology is where much hostility against atheists begins. The ideological structure of the U.S. sets a baseline of 鈥済ood鈥� and 鈥渕oral鈥� in the context of believing in God. Thus, anyone who does not adhere to this baseline structure may be viewed as 鈥渨icked.鈥� 听听听听听听听听 An interesting historical note is that in the nineteenth century, 鈥渇ree-thinkers鈥� were seen as 鈥渘aturally immoral and depraved,鈥� and the term 鈥渁theist鈥� was used in a derogatory fashion 鈥渢o denote the immoral鈥� (Campbell 1972:97-98). Campbell explains that the historical conflation of atheism and immorality has changed to an idea that if an unbeliever was raised by Christian parents, then he or she 鈥渁bsorbed a Christian moral ethic鈥� (Campbell 1972:99). However, there exists a fear that this moral ethic will eventually disintegrate with the successive generations of unbelievers, thus a 鈥渢ime-lag鈥� in the acquisition of immorality (Campbell 1972:99). Although Campbell wrote these theories on atheism several decades ago, his findings are still relevant to the current social landscape. The fear that the moral fabric of society is disintegrating is present today where about 74 percent of United States citizens think that 鈥渋t is a bad idea to raise children without any religion鈥� (Edgell, et al. 2006:213). Indeed, 45 percent of U.S. citizens believe that religious beliefs are necessary for morality (as cited by Bainbridge 2009:324). Although not a majority, this is still a significant portion of citizens who agree with the historical view that 鈥渟ince religion fosters morality, irreligion fosters immorality鈥� (Campbell 1972:97). According to Edgell, Gerteis, and Hartmann, 鈥淎mericans construct the atheist as the symbolic representation of one who rejects the basis for moral solidarity and cultural membership in American society altogether鈥� (2009:230). The conflation of religion with morality comes, in many ways, from both an idea of divine policing and a culture based around Christianity as the acceptable ideology, used as a yardstick to judge any other moral views.

In the eyes of some religious people, the suggestion that 鈥淕od is always watching鈥� keeps the faithful on their best behavior, and this idea of constant policing serves as a religious foundation for morality (Gervais and Norenzayan 2012:484). The presumption that morality is rooted in God or in religion proves problematic for atheists who do not subscribe to the idea of a divine being policing their actions, and creates a certain amount of distrust from people who fear that immoral actions may accompany a lack of belief in eternal punishment for wrongdoing. This is the primary explanation for atheists being stigmatized, labeled as deviant, and sometimes confronted with marginalization, discrimination, and exclusion from their communities.

Profound examples of this stigmatization are evident in political life. Although the Constitution of the United States prohibits the use of religious tests in order to hold public office (U.S. Constitution, Article 6, Section 3), several states have laws excluding atheists from holding office. The Arkansas Constitution forbids atheists from holding public office or serving on juries because they are not considered 鈥渃ompetent鈥� (Arkansas Constitution, Article 19, Section 1). Mississippi, Texas, South Carolina, and other states contain similar provisions in their constitutions (Bulger 2013). Although many regard these provisions as irrelevant in the twenty-first century, atheists in public office are far and few between, indicating that legal prohibitions still on the books may play a factor in the election of atheists or the attitudes of the voting population.[4] There are currently no (uncloseted) atheists in the United States Congress, and only one open atheist, California鈥檚 Pete Stark, has ever served in the U.S. Congress (Moreno 2014). Not only are atheists institutionally excluded from political life, but their status as outsiders sometimes results in experiences of discrimination and feelings of otherness from the rest of 鈥渕ainstream鈥� society. Previous research has noted other instances of atheist marginalization including bullying and harassment of atheist children in schools (Hamblin 2013) and the exclusion of atheists from organizations such as the Boy Scouts (Hill and Sprigg 2013).

In addition to the umbrella notion of 鈥渞eligion鈥� indicating a baseline for moral goodness (and as a requisite for holding public office and being able to fully participate as a U.S. citizen), Christianity, specifically, is seen as the baseline of morality with which other sources of morality are judged. Campbell explains:

Measurement of the degree of immorality of various groups in society thus becomes a very slippery exercise in which there is a real danger of projecting the morality of a dominant ethnic or class group on to a nonconforming minority and hence 鈥榙efining鈥� immorality into existence. Even where all groups in society are agreed on the acts which are to be regarded as immoral, they may attach a very different emphasis to the degree of immorality involved. (1972:103)

听

This statement explains how any variations in moral belief outside of Christianity (the dominant religious belief) can be perceived as immoral. For example, since orthodox Christian doctrine forbids premarital sex or homosexual activity, these acts can become 鈥渋mmoral鈥� in the eyes of society as a whole. Thus, it may be seen as problematic when a person has premarital sex or engages in homosexual activities even though the person may not see these actions as morally wrong (Campbell 1972:103-104). This heavily Christian-influenced idea of morality is still seen today in several states that still have anti-sodomy and anti-homosexuality laws on the books (Ayres and Eskridge 2014). Many of the arguments that influence policies banning gay marriage are heavily rooted in religion and Christianity specifically, thus making debates regarding such subjects more about whether or not such acts should be legal from a Christian standpoint. Campbell suggests that even if atheists do not share a specifically Christian or religiously informed morality, this does not mean that they are immoral or have a lessened 鈥渃apacity for self-control,鈥� but rather that there exists a variation in moral judgments (Campbell 1972:104). It is worth noting that several studies have been conducted that have refuted the assumption that atheists are less moral than their religious counterparts (often measured by the willingness to commit crimes). In a study performed in the United States between 1924 and 1934, it was shown that delinquency and criminality were actually lower in people without religious affiliation than in Catholics, Protestants, and Jews, indicating a lack of a correlation between immorality and lack of religious affiliation (as cited by Campbell 1972:101).

Despite a lack of correlation between religious belief and morality, atheists continue to suffer from negative stereotypes that have serious social consequences. In fact, Edgell, Gerteis, and Hartmann cite atheists as one of the most stigmatized groups in the United States (2006:212). The same researchers explain that more people in the U.S. would have their children marry or otherwise associate with a person from a background of just about any other cultural, minority, or religious group than atheist(s). Most U.S. citizens also believe that atheists are the group least likely to share their worldview (2006:216). In a list comparing commonly stigmatized groups including Muslims, homosexuals, immigrants, and African Americans, 39.6 percent of participants cited atheists as least likely to share their vision of American society. This is compared to 26.3 percent of participants who cited Muslims and 22.6 percent of participants who cited homosexuals as the least likely to share their vision. Further, 47.6 percent said they would oppose their children marrying an atheist compared to 33.5 percent and 27.2 percent of people who were opposed to their child marrying a Muslim or an African American, respectively (Edgell, et al. 2006:218).

Not only are atheists commonly mistrusted for not sharing a common 鈥渟ource鈥� of morality, but they are often disliked and maligned by other groups. The Pew Research Center performed a study of how different groups of Americans rated other groups on a 鈥渇eeling thermometer鈥� where a temperature of 100 was the warmest in terms of feeling, and a 0 was the coldest (Cooperman et. al. 2014). According to this study, the American public viewed Jews, Catholics, and Evangelical Christians most warmly on the 鈥渇eeling thermometer.鈥� These groups were given a rating of at least 60 degrees. Atheists and Muslims were on the coolest side of the thermometer with atheists receiving an average of 41 degrees and Muslims an average of 40 degrees. Atheists were rated on the coldest part of the scale by 40 percent of the studied population with Evangelical Christians tending to rate atheists the lowest with an average of 25 degrees (Cooperman et. al. 2014).

As discussed above, atheists encounter a large amount of mistrust in the social world, 听which reinforces the idea of an atheist identity as an embattled identity that is difficult to outwardly affirm without the possibility of consequences such as marginalization and exclusion. Despite the stigmatization of atheists, people continue to achieve this identity, often undergoing a process of identity formation.

II. Atheist Identity: Formation and Salience

Religion is often taken to be an important part of a person鈥檚 identity. Similarly, a lack of a religious identity can become just as important to an individual. Social psychologist George Herbert Mead, introduces the concept of the self, an important entity that comprises a person鈥檚 identities. He suggests that the self is not present at birth, but arises through a gradual process of social interaction with other individuals (Mead 1934:135). Mead鈥檚 conceptualization of the self is instructive when studying identity because it provides a framework for viewing the self and identity as emergent through socialization, experience, and interaction. The self comprises a variety of identities arranged in a hierarchy of importance. Smith articulates the term 鈥渋dentity鈥� as referring to 鈥渢hat which individuals use to describe and define their membership in groups, their relative positions in social life, and the various ascribed and achieved statuses they hold鈥� (2010:3). The idea of atheism as an achieved rather than an ascribed identity is important when looking at the factors involved in the process of becoming an atheist. This section will focus on the identity formation process of atheists as well as the religious disaffiliation of adolescents and young adults. Later sections will deal with identity salience and the factors that play in to the identity salience of atheism as an embattled, minority identity.

Atheist Identity Formation

In his study of 40 active members of atheist social groups, Smith identified a process that his participants underwent to become atheists. According to Smith, an atheist identity is one that consists of a 鈥渘ot-self鈥� identity (2010:14). This means that many atheists explain their identity based on beliefs that they do not have or activities in which they do not participate (2010:14). Smith likens atheists to vegans, nondrinkers, and virgins (2010:14) in that all of these particular identities imply behaviors in which the individual does not partake such as believing in God, eating animal products, drinking alcohol, or having sex. The process of 鈥渂ecoming鈥� an atheist that many atheists shared is outlined by Smith in four major steps: (1) the realization of the 鈥渦biquity of theism,鈥� (2)鈥渜uestioning theism,鈥� (3)鈥渞ejecting theism,鈥� and (4)鈥渃oming out atheist鈥� (2010:1).

According to Smith鈥檚 study, many of the atheists cited a prior belief in God, and many verified that even if they grew up in a secular household, they noticed the normative belief in their communities was the belief in God, and specifically in the Christian God. One participant said that even though she was raised in a secular household, she still went to church with her friends because she wanted to feel included (2010:7). This indicates how a belief in God can be a 鈥渄efault鈥� stance in the United States that one, especially as a child, may take in order to be accepted by peers and the dominant culture. Smith states, 鈥渆ven when there was low religious participation, the normative belief in God was still present鈥� (2010:7).

The step of questioning theism was experienced by many of Smith鈥檚 respondents upon entering college when they began to doubt the beliefs with which they had been raised. In college, respondents encountered new points of view, engaged in new relationships, and in the cases of those who had a religious background, attended church less and less. One respondent commented, 鈥淭hey [fellow students] didn鈥檛 go to church on Sunday and nothing bad was happening to them!鈥� (2010:8). Respondents in college also had interactions with atheist peers who would engage them in critical conversation about religion. This began a questioning process for the respondents who came to view religion as illogical and contradictory (2010:9). Smith explains the social process that respondents had to undergo in order to reject religion by stating, 鈥淏ecause religious beliefs and practices are socially learned, there must take place an unlearning process for those who eventually come to reject it鈥� (2010:9). People would undergo a 鈥渞ole-exit鈥� process where they would critique the beliefs with which they grew up and would realize that they were dissatisfied with the incongruence of their previous beliefs and their new worldviews informed by new experiences and interactions in college. They would then reinterpret the teachings of the Bible or other religious texts with a new, more skeptical understanding of their previous beliefs (2010:9) leading to a crossroads of deciding whether or not to reject their former religious beliefs.

The third step suggested by Smith in becoming an atheist was the rejection of theism where respondents explicitly rejected both the idea of God and religion and found a new way to legitimate a worldview without God using scientific explanations (2010:13). This 鈥渃oming out鈥� phase occurred when the new atheist acknowledged that atheism is a deviant view in a predominantly religious culture, and then found empowerment by actively stating the new beliefs. A respondent in Smith鈥檚 study reported upon 鈥渃oming out鈥�: 鈥淚t was very empowering鈥or me it was the idea of setting your own purpose, that it鈥檚 just you, [that] it comes from within; that鈥檚 pretty exciting!鈥� (2010:17).

For many, the new atheist identity became a salient part of self-identity even if the person experienced negative reactions from others after 鈥渃oming out鈥� (2010:15). Smith explains that even if these new atheists recognized their marginalized position in society, rather than feeling a sense of anomie, many respondents cited feelings of autonomy because it was through their own personal identity construction that they achieved this particular aspect of their identity (2010:17).

Another important piece of the literature in regard to atheist identity formation is a 2013 study by LeDrew. In his article, LeDrew explores the experiences of 鈥渁ctive atheists,鈥� who in LeDrew鈥檚 terms are actively engaged atheists who are involved in atheist groups and organizations. He explains their attitudes toward their own identity formation as well as attitudes toward the religious community. LeDrew explains identity development as a process of discovery including three steps: the discovery of ideas, self-discovery, and the discovery of the collective (2013:1).

LeDrew identified what he termed the 鈥渟tandard trajectory鈥� in atheist identity formation as: 鈥減eople experiencing a religious socialization who were believers, but went through a process of questioning and doubt that eventually resulted in abandoning their beliefs and adopting an atheist identity鈥� (2013:5). LeDrew references Smith鈥檚 work, mentioning the normative view and socialization including a belief in God (a 鈥渢heistic socialization and social environment鈥�) that the atheist comes to reject after progressing through questioning and doubt phases that result from education and exposure to new ideas (2013:5). LeDrew specifically identifies a 鈥減eriod of doubt鈥� in a person鈥檚 teenage years through their twenties (2013:5). One of LeDrew鈥檚 participants identified that between the ages of 13 to 16 he experienced doubt because 鈥渢hings weren鈥檛 making sense because questions weren鈥檛 being answered鈥� (2013:6). The participant educated himself more about science, realizing the inconsistencies of his religion and the perceived consistencies of science. This questioning continued into his college years, especially after his exposure to atheist friends and roommates (2013:6). The participant read The God Delusion by Richard Dawkins and was finally convinced of his full-fledged transformation to atheism, with Dawkin鈥檚 book providing him with the 鈥渓anguage and concepts to articulate the doubts and questions he had struggled with for years鈥� (2013:7).

An important distinction made by LeDrew relates to his participants鈥� backgrounds in terms of either secular socialization or religious socialization. Secular socialization is a background where religion was never important and the respondent鈥檚 atheism proceeded from socialization. Some participants actually explored religion before doubting and going back to atheism (2013:10). In respondents who described a religious socialization, some claimed to be skeptics from the beginning. Others were termed 鈥渟eekers鈥� who sought out religion, but 鈥渃ould not make themselves believe what they had been taught鈥� (2013:13). An interesting theme discovered was that many atheists adopted a narrative of a 鈥渄iscovery鈥� of atheism rather than a conversion or deconversion, meaning that they recognized something about themselves that had previously existed (2013:8).

听听听听听听听听听听听 Although the level of education is certainly a factor mentioned in the apostasy of his participants, LeDrew mentions that education is certainly not the only factor that influences religiosity or a lack thereof as there are contradictory findings in the area of the level of education in relation to the level of apostasy. LeDrew also cites socioeconomic status and namely financial stability as a factor in nonbelief. This is important to note as many celebrity atheists like Richard Dawkins link religiosity to a lack of education or just sheer ignorance. As LeDrew succinctly states: 鈥渞eligiosity is closely related to social conditions rather than simply a product of ignorance that is eradicated with science education鈥� (2013:7).

Emergence of Atheism in Adolescence and Early Adulthood

The emergence of atheism in adolescence and early adulthood in some individuals has been a rarely studied topic, however, some sociologists have attempted to understand the social conditions present that precipitate the decline of religiosity in young people. In his study of the emergence of atheism in adolescents, Hunsberger claims that many adolescents who become atheists are often raised in households where religion is not heavily emphasized or where parents are not necessarily good models of religion (2000:234). Therefore, 鈥渕ost apostasy does not represent a reversal in the socialisation process, but rather it might be characterized as a slight bend in the religious socialisation line鈥� (2000:234). Hunsberger explains that a weak level of religion in the home can cause adolescents to 鈥渄rift鈥� away from this religion during these years (2000:234). When compared to adolescents who come from fairly secular households and who actually convert to religion with very high levels of religiosity, apostates begin questioning their beliefs much earlier in life, mainly at the onset of adolescence. Hunsberger explains, 鈥渨hen young people start questioning a lot of things鈥heir edifice of beliefs then began to crumble bit by bit, until finally the whole structure came crashing down鈥� (2000:240). This process is described as taking a relatively large amount of time (usually a few years) because many apostates view leaving the family religion as a 鈥渄aunting鈥� process (2000:240). Hunsberger also states that low levels of religious influence in the home are not necessarily always indicative as causes of 鈥渄rift鈥� away from religion. This is simply a trend among adolescents, especially those who grow up with tendencies to question their beliefs and actively seek out truth and answers (2000:241).

听听听听听听听听听听听 In a study of German and U.S. adolescents, Schweitzer finds that age plays perhaps the largest role in an individual choosing to leave religion, and that young people within the ages of 18 and 30 are the group most likely to leave the church (2004:66). However, Schweitzer says that later in life, many of these people who previously leave the church may cycle back, and that this 鈥渕oratorium period鈥� will be abandoned as 鈥渆xperimental attitudes鈥� are left behind (2004:70). Regnerus and Uecker explain that people may simply abandon a high level of religious participation in college due to numerous factors such as a late-night orientation, deviance from religious norms (having sex, drinking, and so on), and the social norm of not appearing too outwardly religious (2007:4). Despite a decline in religious practice frequency, Regnerus and Uecker claim that the majority of college students maintain the same level and importance of religious faith that remain 鈥渓argely untouched for the duration of their education鈥� (2007:3-4). This faith 鈥渓ies dormant,鈥� during many students鈥� college years, and is often returned to later in life (Regnerus and Uecker 2007:3-4). Findings by Christianist sociologists such as Schweitzer (2004) suggest a similar cycle where apostates later return to religious practice.

Salience of Atheist Identities

Along with explanations of identity and identity formation of atheists, it is also important to discuss identity salience, as the salience of an atheist identity is particular to the fact that atheism is a minority status embroiled in a highly religious social climate in the United States. In her definition of identity salience, Morris refers to this concept as 鈥渢he likelihood that a given identity will be active across situations鈥� (2013:24). Identities reside in a hierarchy of importance related to identity commitment, where identities with the most commitment also are the most salient (Morris 2013:24). Identity salience is particular to an individual, and different identities can become more important or salient at different times or in different situations. For example, if a woman finds herself in a large room with many men and she is the only woman, her identity as a woman will become more salient than if she were in a room with all females.

Identity salience is important to discuss when considering atheism as an embattled, minority identity. Seifert exemplifies the formation of atheism as a salient part of some individuals鈥� identities in her article regarding 鈥淐hristian privilege鈥� (2007). Seifert introduces an idea that college students who identify as Christian in a heavily Christian-dominated society enjoy 鈥淐hristian privilege,鈥� defined as 鈥渢he conscious and subconscious advantages often afforded the Christian faith in America鈥檚 colleges and universities鈥� (2007:11). She explains that in some universities, Christian students are afforded privilege through having entire breaks planned around religious holidays like Christmas and through maintaining a predominant 鈥淐hristian ethos鈥� even though many universities no longer identify themselves in religious terms despite commonly religious foundations (Seifert 2007:11). In addition to the 鈥渇ormal privileges鈥� that Christian (and other religious) students enjoy such as having the Christmas holiday off from school,[5] Seifert explains 鈥渋nformal privileges鈥� that Christian students experience that can be applied to broader society. Many people, for example, have become accustomed to attributing behavior such as volunteering to being a 鈥済ood Christian鈥� or that these behaviors are 鈥淐hristian鈥� in nature (2007:12).

It may also be argued that Christian privilege has expanded to religious privilege in general. While college campuses seemed to once be seen as the bastions of science, reason, and secular thinking, Nash explains that today, religious pluralism on college campuses is booming while secularism and humanism are 鈥渋n retreat鈥� (2010:2). This is seen in the explosion of student-run and university-funded organizations for students of a variety of spiritual and religious affiliations (Nash 2010:2).[6] Nash explains, 鈥淧resently, it is the fashion on many campuses throughout America to celebrate religious difference and spiritual diversity, as a way of enlarging the meaning of multiculturalism鈥� (2010:2). Edgell, Gerteis, and Hartmann also explain that in general, the increasing tolerance of religious pluralism may actually lead to an increasing intolerance of people who reject religion (2006:214). They explain:

Religious acceptance may be driven largely by assumptions that religious people, of 听听听听听听听听 whatever faith, are 鈥榣ike me鈥� in two ways. In private life, they are understood to be moral 听听听听听听听 people, worthy of trust that is the basis for close personal relationships. In public life, the 听听听听听听听听听听 boundaries that separate religious identities (for example, evangelical versus mainline 听听听听听听听听听 Protestant versus Catholic or Jew) are understood to be encompassed by and to constitute 听听听听听听 a broader identity-being a good American. In such a setting, how do Americans view 听听听听听听听听 those who reject religion, and what does that tell us about how Americans view their 听 nation and themselves? (2006:214)

听

Nash echoes this idea that college campuses are actually becoming less tolerant of atheists. He introduces the idea of 鈥渁theophobia鈥� (the fear and hatred of atheists) that is 鈥渁live and well on college campuses鈥� (2010:3). Bowman and Small explain in their study of college students and their spiritual development that 鈥渟tudents from all faith backgrounds described a perceived three-tier structure of religious privilege in society with Christians at the top, atheists at the bottom, and other religions in the middle鈥� (Bowman and Small 2010:599). This may cause atheist college students to feel as if their needs are not being met by universities or to feel uncomfortable and defensive about their beliefs in a highly religious social climate on college campuses (Bowman and Small 2010:599).

These privileges extended to religious students on college campuses along with the 鈥渢heophile鈥�-dominated social culture[7] exclude individuals who do not identify with the dominant or normative campus culture and create situations for students where they feel ignored or invisible if they do not identify as religious or spiritual (Siner 2007:14). However, the experience of Christian privilege (or religious privilege) by an atheist college student also creates opportunities for this non-religious identity to become more salient.

Siner (2007) uses this same idea of Christian privilege to explain the experiences, identity development, and salience of college students who do not identify as Christian. According to Siner, 21 percent of college students do not believe in God and are often seen in a negative light by their peers, being described as 鈥渂itter, mean-spirited, Satanic, immoral, empty, or ignorant鈥� (2007:14). However, marginalization and judgment imposed on nonreligious students commonly leads to a greater importance in the irreligious aspect of the students鈥� identity. Siner explains that people have a multiplicity of identities, but some tend to be more salient or important than others at different times and places. He states that 鈥渙ppressed, marginalized, and minority identity statuses tend to be more salient to an individual鈥� while others fall to the periphery of one鈥檚 sense of self (2007:14). This can be likened to the Black Community, where many people experience the fact of being a person of color in a society that is predominantly white as a very salient part of their identity. Siner states: 鈥淛ust as any minority identity is likely to be more salient, faith identity may be particularly salient to those students from minority religions that are not valued in the Christian-dominated culture of the United States鈥� (2007:15).

听

III. College and Atheism

An interesting contradiction arises in the important works performed by both Smith and LeDrew in regard to the role of higher education in the making of an atheist identity. In Smith鈥檚 study, many of the participants did not begin to question their beliefs until college (Smith 2010). However, LeDrew makes an important reference to other research that shows that higher education might not cause apostasy (2013). Historical research by Caplovitz and Sherrow of 鈥渞eligious drop-outs鈥� points to college as 鈥渁 breeding ground for apostasy鈥� (1977:109), saying that in their study of college students, more students, regardless of religious background, left college as apostates than the number of apostates who entered college (1977:109). This study indicates factors such as being away from families who would otherwise sustain religious beliefs and contact with new peer groups as factors that push students away from religious beliefs in college (1977:108). In a more recent study, Regnerus and Uecker take on the issue of college and apostasy, explaining that college itself may not be the reason that people choose to disaffiliate themselves from religion (2007:1). They state that many people who became irreligious often are a product of 鈥減rocesses set in motion long before young people ever set foot on a college campus: Those students who 鈥榣ose their faith鈥� in college or drop out of organized religion after high school are primarily those already at considerable risk for doing so for other reasons that predate these actions鈥� (2007:3). The authors thus reject the common belief that college is a secularizing force that turns people into atheists. As the authors suggest, there is little evidence for this today as much of the research that made this claim came out of the 1960s and 1970s, when the counterculture was at its strongest in U.S. universities (Regnerus and Uecker 2007:2).

Regnerus and Uecker (2007) explain that the commonly believed explanations for a lapse in at least outward religiosity may be due to the 鈥渟ecularizing tendencies of higher education鈥� as well as 鈥渃ognitive dissonance鈥� that may result from activities such as drinking, having sex, or using drugs that deviate from religious teachings or parental guidance (Regnerus and Uecker 2007:1) Other reasons to temporarily abandon religion in college include logistical conflicts resulting from being faced with the responsibilities of young adulthood (2007:5) or adhering to norms that discourage a person from being too outwardly religious since 鈥渢o appear over-religious can be the social kiss of death鈥� (2007:4). Despite the lessening of outward religiosity, it is clear that many people remain privately religious, with four out of five students stating that they do not experience a diminishing religious salience during college years (2007:4). Regnerus and Uecker further questioned the assumption that college is the reason for deconversion or decreased salience of religion in young people, finding that while 64 percent of young people who went to church at least once per month decreased their attendance upon enrollment in a four-year institution, 76 percent of people the same age who never attended college curtailed their church attendance (2007:2). This suggests that religious disinterest probably occurs some time before college (or the typical college age). Other factors that lead to a decrease of salience in religiosity include cohabitation (which may imply sex before marriage), or other behaviors such as binge-drinking that are criticized by many religious traditions. However, these activities do not necessarily indicate a causation of secularity, but rather serve as a reflection of a secular orientation already present in the individual (2007:2).

IV. The Current Study

听听听听听听听听听听听 As demonstrated above, the existing literature has focused on atheist identity formation in a social context and on the institution of college as (historically) a direct cause of religious rejection or (more recently) as a force that causes religious practice to temporarily go underground. In order to gain a more nuanced understanding of the identity formation and experiences of atheists outside of the group context as well as the role that college plays in these processes, the next chapters of this thesis will describe my own qualitative analysis I used to study college atheists. A chapter describing the research methods will first be described, followed by the findings of the study. My findings will be discussed and placed into the context of the existing research in my final chapters.

听

CHAPTER 3: METHODS

听

听听听听听听听听听听听 Before embarking on this research project, I worked with two University of Colorado (CU) offices. First, I obtained CU Internal Review Board (IRB) approval for my recruitment materials, study design, interview, and data-storage protocol. Second, I applied for and was awarded CU Undergraduate Research Opportunity Program (UROP) funding to cover my material costs and costs for providing interview participants with a small compensation.

I. Setting

This study was conducted at the University of Colorado in Boulder, Colorado. The university is home to approximately 25,000 students and is predominantly white with a slightly greater number of male than female students (鈥淯niversity of Colorado-Boulder鈥� 2015). The city of Boulder is mainly composed of liberal, Democrat, upper-middle class, white, young, and physically active individuals. Notably, Boulder is one of the least religious towns in the United States with less than twenty percent of inhabitants claiming to be 鈥渧ery religious鈥� (Florida 2013). It is not surprising that this city claims less religious inhabitants as information previously stated in my literature review points to the least religious people tending to be predominantly white, male, liberal, Democrat, and so on. The status of Boulder as one of the least religious cities in the United States proved to be very important in my informants鈥� responses, as many of them felt comfortable affirming their atheist identity in this supportive normative context.

II. Recruitment Process

Recruitment methods included presenting information about the study to undergraduate classes, writing on classroom chalkboards, distributing flyers, advertising in the weekly 鈥淏uff Bulletin鈥� (an online publication that lists events, notices, services, and so on), presenting to an on-campus secular group, and through snowball sampling. In order to attract the most attention, recruitment materials included titles such as 鈥淒on鈥檛 Believe in God?鈥� or 鈥淎re you a Godless Heathen?鈥� The flyers (See Appendix A) included information about the study and my contact information. The flyers made explicit the parameters of the research, requesting participants who considered themselves to be atheist, were between the ages of 18 to 24, and who were current undergraduate students at the 彩民宝典. The flyers also stated that participants would be compensated with a $10 gift card. Interested individuals called me on my personal cell-phone to schedule an interview or e-mailed me on an account that was set up specifically for the study. After conducting an interview, I would give the participant an extra flyer and copy of the consent form along with a $10 gift card. I would then ask him or her to pass along the information about my study to any other friends who were atheists and who might be interested in participating. One woman referred her friend to me, but most participants reported that they found out about the study through seeing advertisements on chalkboards in their classrooms. One student responded after hearing my presentation in a student-run skeptics club meeting.[8] One participant was recruited through a presentation to an on-campus academic journal organization,[9] and several other participants were acquaintances who expressed interest in participating.

III. Sample

听听听听听听听听听听听 From September to December 2014, I conducted 14 semi-structured in-depth interviews. A sample size of 14 participants was large enough to yield data saturation and an ability to achieve a rich data set providing an understanding of atheist identity formation, atheist identity salience, and several other recurring themes among this college sample. Eight of the participants were women (six were men), with ages ranging from 18 to 23. The average age was 19.6 years, while the modal and median ages were 18 years and 19.5 years, respectively. Regarding class rank, the sample was bimodal, predominantly freshmen and seniors (6 freshmen, 1 sophomore, 2 juniors, and 5 seniors). There was a range of majors with an over-representation in the STEM (science, technology, engineering, and mathematics) fields, with ten participants having at least one major in one of those fields. Three of the participants had exclusively social science majors, and one was majoring in anthropology, which could encompass both natural and social sciences. Interview participants were 鈥渢raditional鈥� students at the university, meaning they followed a trajectory of entering their undergraduate education shortly after completion of high school and were on a four or five year track to finish their degree(s). Twelve of the participants were white, and two were bi-racial (Mexican-American and Japanese-American). The response rate was high, with 14 out of the 16 total respondents who contacted me participating in an interview. Most respondents contacted me over e-mail, and only two people who expressed interest over e-mail never responded. When I received an e-mail from an interested individual, I would reply within 24 hours and give them my phone number and more information about the study. If they did not respond within three days, I would send one follow-up e-mail.

IV. Interviews

During the interviews, my status as a college student helped me establish rapport with my participants, who were also students. Many of the participants also asked me about my own religious beliefs during the interviews. I answered them honestly, telling them that I, like them, am an atheist. As expected, this seemed to comfort many of the participants who may have otherwise not shared with me some of their reservations regarding organized religion and other personal beliefs. Since I was seen as part of the 鈥渋n-group,鈥� fellow stigma-bearer, or fellow sympathizer, it seemed that the participants鈥� knowledge of my own identity provided a more comfortable situation for them to share their experiences.

听听听听听听听听听听听 Most of the interviews were conducted in the CU campus library in private, reserved study rooms (one interview was conducted at the participant鈥檚 home). This comfortable and private setting allowed participants to freely talk about their experiences. Consent forms (see Appendix B) were signed before beginning the interviews. An additional consent form was provided to audio-record the interviews for later analysis. All participants agreed to participate in the study and to be audio-recorded. The interviews ranged from approximately fifteen minutes to slightly over one hour, depending on how much the participants chose to expand on their responses to the interview questions. On average, the interviews lasted about one-half of an hour. The interview questions were used as a guide, but flexibility was allowed depending on the flow of the conversation. Some participants did not need any follow-up questions or probing and would answer some interview schedule questions prior to me asking them. Conversely, other participants provided limited responses until I used probes to acquire more complete responses. 听

An interview schedule (see Appendix C) was used to structure the interview, but participants could expand on their experiences beyond what was in the interview guide. Further probing questions were asked in order to help clarify points or to provide a deeper understanding of the participants鈥� responses. The interview questions encouraged participants to take a narrative, conversational approach with an emphasis placed on their upbringing (religious or otherwise), family life, experiences with religion, experiences with atheism, and personal beliefs surrounding morality and what it means to be an atheist. Participants were asked to rate and explain the importance (salience) of their atheist identity in their lives and what role it played in family and other social settings and relationships. Questions were also asked regarding respondents鈥� perceptions of stigma,[10] with further probes added to see how the person coped with any perceived stigma or unacceptance. After the interviews were performed and recorded, they were transcribed verbatim in order to most accurately search for common themes. The participants鈥� actual names were substituted with pseudonyms in order to protect their anonymity. It is worth noting that although I told my participants that all names would be changed in my data transcriptions, analyses, and written reports, none of them were concerned about their identities being known and were fully willing to disclose their personal information.

听

CHAPTER 4: FINDINGS

听

听听听听听听听听听听听 The atheists in my study all proved to be individuals with nuanced views and varied ways of incorporating atheism into their daily lives and personal beliefs. After carefully combing through and coding my qualitative interviews with 14 CU student atheists, I identified three primary themes in my data: (1) precursors to atheism; (2) pathways to atheism; and (3) experiences with an atheist identity. All of these primary themes are further parceled into subthemes of the primary themes. Thus, this chapter of my thesis will elaborate on these three primary themes and their corresponding subthemes, providing examples through some of the participants鈥� quotes.

I. Precursors to Atheism: Backgrounds

听听听听听听听听听听听 The first theme in my findings, Precursors to Atheism, includes the participants鈥� religious backgrounds and upbringings. This discusses the level of religiosity in their homes as well as their own personal levels of religious belief and religious practice. This theme is further parceled into the five subthemes that are common background or upbringing experiences of the study participants. The first subtheme, Church Attendance, explains the backgrounds of participants in terms of religious participation as children and adolescents. The second subtheme, Christianity as the Norm, describes the normative belief that many of the atheists had growing up with a specifically Christian tradition. The third subtheme, Obligation by Friends and Family, describes how many participants adhered to their parents鈥� (or grandparents鈥�) religious beliefs and practices because they often had no other choice or because they felt obligated by their friend groups that were centered in the religious community. A discussion of participants鈥� experiences and opinions of their religious upbringings is included in the fourth subtheme: Experiences at Church. Finally, the fifth subtheme, A Background of Skepticism, describes participants who grew up in a household where religion was not emphasized, or where they had at least one parent who taught them to question or who gave them permission to seek out their own religious identities.

Church Attendance

听听听听听听听听听听听 Regardless of the presence of actual religious belief during childhood, there was a relatively high religious practice frequency among my sample early in the participants鈥� lives. Twelve of the fourteen respondents reported going to church during their childhood, but with varying degrees of frequency. Four of my respondents reported attending a religiously affiliated school. Two respondents (Sophie and Mark) went to a Catholic school. However, Sophie (who grew up in an 鈥渙pen-minded鈥� family with a very low rate of religious participation in the home) only attended Catholic school because she lived in an area with a poor public school system where the Catholic school was the only alternative. Although her family was not highly religious and only went to church on some holidays like Christmas, Danielle went to a Lutheran preschool. In addition to attending religiously affiliated schools, Aria and Mark both had very high religious practice frequencies. Aria went to a Quaker school, attended an Old Apostolic Lutheran Church, and participated in an intensive confirmation program one summer. Mark, who was raised a Catholic, went to mass every Sunday with his parents and frequently worked as a server in masses and at funerals.

听听听听听听听听听听听 Three participants (Julie, Matthew, and Bethany) went to Sunday school every Sunday when they were young (Matthew went on Wednesdays and Sundays until he was 8 years old). Five participants (Edward, Jennifer, Dillon, James, and Karen) went to church sporadically as children, either only on holidays, or when their parents wanted to go. Karen only reported going to church with her grandparents when she stayed at their house. Brian was somewhat of an anomaly in the sample because he grew up in a secular home but went to church with his grandmother often because she played the organ at the services. He also became interested in religion later in his childhood and began attending youth groups until his sophomore year of college. Only one respondent (Cora) had a completely secular upbringing where she never attended church, although she did go to a friend鈥檚 youth group several times.[11]

Christianity as the Norm

听听听听听听听听听听听 Of the 14 participants interviewed, 13 had at least some religious influence in their upbringing, whether it was in practice (going to religiously-affiliated events like church and youth groups) or in some level of a 鈥渘ormative鈥� belief in God (Smith 2010:7). All respondents with any religious influence had specifically Christian influences in their childhoods, whether with actual religious practice, or simply labeling oneself as a Christian because it was seen as the normal thing to do. James, who was raised in a reportedly secular family with a 鈥渟piritual鈥� mother and an atheist father (who lacked any spirituality or religious belief), said he considered himself to be a Christian during his childhood:

听听听听听听听听听听听 I kind of took myself to be a Christian because it was the norm. And so, I kind of viewed everyone else as being that way as well. But not necessarily because I believed in 听听听听听听听 religious things. Just because that鈥檚 just the way it was, and that鈥檚 what I was taught, and 听听听听听听听听听听听 that鈥檚 how everyone else seemed to be.

听

When asked where he learned that being religious was the norm, James replied that these messages came from school, people around him, and his mother taking him to Sunday school several times to give him a chance to experience it and decide what he wanted to believe. James certainly felt a social pressure to believe in something, and remembered thinking during his childhood:

听听听听听听听听听听听 There are so many people in the world that believe in a god that there must be one鈥t 听听听听 was just kind of taken-for-granted. And really, it was thought that if you didn鈥檛 believe in 听 a god, that was pretty stupid because clearly you鈥檙e not paying attention to what the rest 听听听听听听听听听听听 of the world is believing.

听听听听听听听听听听听 Karen grew up in a secular household but went to church sporadically when she stayed at her grandparents鈥� house. She said that before middle school, she adopted Christianity simply because she celebrated Christmas. However, she said, 鈥淚t wasn鈥檛 really anything about the actual practices of the Bible. It wasn鈥檛 like 鈥榯he Bible tells you to do this, and this, and this.鈥� It was 鈥榦h, we celebrate Christmas and it鈥檚 about Christ.鈥� 鈥� Many respondents also reported feeling a sense of comfort in the idea of religion growing up. Jennifer, who grew up going to church on Easter and Christmas said:

听听听听听听听听听听听 It鈥檚 just a nice thought when you鈥檙e a kid that your life means something. And when 听听听听听听听 you鈥檙e kind of learning about death as a kid, it鈥檚 a nice thought to have like 鈥榦h, your 听听听听听听听听听听 grandpa is in heaven, and oh, they went on to a better place.鈥� I think it鈥檚 a nice, 听听听听 comforting idea to have when you鈥檙e growing up and when you haven鈥檛 matured yet and 听 you don鈥檛 really have the capacity to understand what life is.

听

The idea of a heaven and an afterlife seemed compelling also to Julie, who was drawn to church and religion because she liked the idea of salvation and Heaven, saying: 鈥淲hen I did believe in salvation, or when I contemplated Heaven, I was like: 鈥榊eah, that鈥檇 be cool. Chill up on a cloud鈥︹€� 鈥� To many of the respondents, religion served as a comforting childhood idea that answered many often-difficult questions when they were younger about death and the meaning of life.

Obligation by Friends and Family

听听听听听听听听听听听 When asked why respondents adhered to their family鈥檚 religious beliefs when they were growing up (if applicable), most respondents with religious backgrounds answered that it was because they were obligated to attend religious services and show outward signs of religiosity as a part of being a member of their family or because many of their friends were centered in the church community. Mark, who attended his Roman Catholic church regularly said that he believed in and followed Roman Catholicism 鈥渂ecause they [my parents] did. One hundred percent. Because it was what was ingrained in me since I was zero years old.鈥� Jennifer commented: 鈥淵ou listen to your parents when you鈥檙e growing up. When they make you do something, you do it.鈥� Edward, Dillon, Matthew, Bethany, Danielle, and Mark all shared this sentiment, saying that they were given little to no choice in attending religious functions with their parents. Brian, Aria, and Julie cited both family and friend influences in going to church. Brian was an interesting case because although he had two parents who were not interested in religion at all, he was obliged to go to church with his grandmother, and later in his high school and early college years, he reported a high church attendance. Brian reported having a high religious practice frequency because of the social aspect of youth groups, saying:

听听听听听听听听听听听 My friend group was centered entirely in the church at one point鈥 enjoyed the culture 听 and the friendship aspect of it far more than I felt any sort of personal relationship with Jesus or anything. I guess I would say I sort of defaulted to that as a condition of considering myself a Christian.

听

For Aria, whose mother had her at a very young age, the Old Apostolic Lutheran Church was compulsory because she was raised in part by her grandmother who stipulated church attendance as part of her upbringing. Aria also attended a Quaker school and between these two religious groups, she had a circle of friends that served as a large reason for continuing her church attendance.

Experiences at Church

听听听听听听听听听听听 Some respondents reported enjoying many aspects of the church community growing up. Jennifer enjoyed getting dressed up and going out to eat as a family after a church service: 鈥淲e always did fun stuff afterwards, like we鈥檇 go out to eat or something. So while it was a little bit boring there when I was younger, it was still kind of a special event, so you felt like it was a sacred thing that you were doing.鈥� Brian expressed positive memories of attending church even though he did not have much of a choice when he was staying with his grandmother. He mentioned loving the music and the free cookies after the service. However, his enjoyment of church had nothing to do with the religious aspect of it and was mixed with irritation at his experiences at Sunday school. He remembered:

听听听听听听听听听听听 They had a little Sunday school in the place, and my sister and I at one point went there 听 for like one week. And it was terrible! We just didn鈥檛 have a good time [laughs]. It was 听听 annoying. They told us about this dumb Noah story, and it was stupid. And we had to 听听听 color. And I鈥檇 just rather sit and watch grandma play the organ because it was more 听 interesting.

听

听听听听听听听听听听听 Matthew, who was obligated by his mother to go to church on Wednesdays and Sundays, reported very negative memories of church as a child: 鈥淚 remember, I hated going to church because, I mean you have to wear the nice clothes all the time, you can鈥檛 get the clothes dirty. I remember my mama used to hit me a couple times if I spilled something on my clothes.鈥� Other participants reported other examples of negative experiences with the church they attended, but these will be explained in a later section, as many of these experiences actually proved to be instrumental in the slow progression of these participants out of their religious beliefs and toward atheistic beliefs. 听听听听听听听听

A Background of Skepticism

听听听听听听听听听听听 A final important trend in the backgrounds of participants was having at least one parent who was not religious or who taught the participant to question their beliefs (religious and otherwise) at an early age. Out of the fourteen participants, three had at least one parent who taught them to question their beliefs as children. Also, ten individuals had parents who were skeptical of religion themselves or who allowed their child the freedom to choose their own religious beliefs. Matthew, whose mother heavily stressed church attendance and religious belief, got a completely different message from his father, who himself was aggrieved at being forced to be a Christian by his parents while he was growing up. Matthew commented, 鈥淚 mean, my daddy came up with the idea: 鈥業 don鈥檛 want to have to force my children to be the religion that I am, that my parents forced me to be.鈥� 鈥� Matthew noted that his father would make fun of everything, including religion and emphasized to Matthew that he had a choice of what to believe: 鈥淗e stressed choice. He stressed that every man has a choice to make鈥nd that choice is never right nor wrong. It is his own.鈥� This led Matthew to give up going to church when he was 8 years old, and to take seriously the choice that his father gave him to believe or not to believe in Christianity. Edward had a similar experience at home where his mother was the one who brought the family to church on occasion, but whose father did not feel strongly about religion although he was raised a Christian. Edward commented:

听听听听听听听听听听听 He [my father] was raised a Christian, but he just neglected it. It wasn鈥檛 really an interest of his. So he didn鈥檛 pursue it. But he shares a lot of my views now too with it. A lot of 听听听听听听听听 organized religion doesn鈥檛 make sense to us鈥y dad went to church because my mom 听听 did.

听

For Edward, he and his father went to church because of his mother, but the indifference of Edward鈥檚 father clearly influenced Edward鈥檚 beliefs and vice-versa.

听听听听听听听听听听听 Both of Dillon鈥檚 parents went through a questioning of religion phase when he was still living at home. Dillon remembered:

听听听听听听听听听听听 They bought books like the Act of God and the Hand of God and How the Earth Came 听听听听 to Be. And things about the universe and whatnot. And I wouldn鈥檛 say they pushed it on 听听听听听听听 me, but I flipped through the books, and we had discussions鈥t was always 鈥榶ou have to 听听听听听听听听听听 make your own decisions and we鈥檙e not gonna make you think one way or the other.鈥� 听

听

Dillon and his family explored both religious belief and secularity, and he used these experiences to explore both religious and scientific explanations to questions he had. Karen and Sophie had similar relatively secular upbringings with unaffiliated parents. Sophie said that her parents never really spoke about religion, but told her that she could believe whatever she wanted. Karen mentioned that her father was forced as a child to go to church, and this influenced his decision to never make Karen go to church even though he expressed some interest in being a part of a community (religious or otherwise). Karen鈥檚 parents also raised her to be accepting of others and to question. They never answered her questions about religious matters however, asking her instead what she thought about it. Karen stated: 鈥淚 guess my parents were never really explicit about their own beliefs. Like they just told me about their own backgrounds, like what they grew up doing.鈥�

听听听听听听听听听听听 Cora had an interesting background that was different from the other 13 participants in that she grew up in a specifically atheist household. Although Cora grew up in an extremely conservative, predominantly Evangelical Christian community, she was raised in an upper-middle class family with two liberal, atheist university professors. For her parents, teaching Cora and her brother to question at an early age was of the utmost importance in a context where faith-based thinking and high levels of religiosity was more prevalent in the public schools and in the community. Cora remembered:

听听听听听听听听听听听 My parents raised my brother and myself with this mindset of always questioning. It was 听听听听听听听听听听 a really open household, so if we had any questions, they would really answer them鈥it 听听听听听听听 was] really important to my parents that my brother and I felt free to ask questions, but 听 they said you need to think critically and you should always ask for evidence and 听听听 shouldn鈥檛 take things based on faith value, which is good to keep in mind no matter what 听听听 you do, especially in college. If you鈥檙e going to say something, you should probably have 听听听听听听听听听听听 evidence to back it up.

听

It is important to note that Cora鈥檚 parents did not force her to be an atheist. If a friend invited Cora to a Christian youth group, her parents were excited for her, telling her that she should be open to everything.

II. Pathways to Atheism: 鈥淭he Spark鈥�

听听听听听听听听听听听 The second primary theme, Pathways to Atheism, refers to the different experiences and processes participants had that precipitated a questioning phase or the direct adoption of an atheist identity. This primary theme included five subthemes. The first subtheme, 鈥�I never really believed any of it,鈥� describes the trajectory of atheists who never viewed themselves as religious in their childhoods, regardless of the presence of a religious upbringing. The second subtheme, Negative Experiences with Religion, describes participants who strayed from religion after perceiving experiences with their church or religious acquaintances as wrong, contradictory, or negative in some way. The third subtheme, Traumatic Experience, is where participants described negative experiences that happened to them early in life that set them on a course of questioning their religious beliefs. The fourth subtheme, Friends, explains the impact some of the participants鈥� friends had in the raising of doubts regarding the participants鈥� religious beliefs. The final subtheme, Questioning Beliefs and Losing Religion, is included in order to explain the step after the 鈥渟park鈥� that many participants experienced where they questioned their beliefs. Two further sections are involved. The first, External Questioning, involves participants who went through a process of acquiring information about atheism and religion through seeking out information, and then making a conscious and rational decision to reject religious belief and adopt atheism. The second section, Internal Questioning, describes the experiences of people who still questioned their religious beliefs, but on an internal, personal level.

鈥淚 never really believed any of it鈥�

听听听听听听听听听听听 An interesting finding from my interviews is that many participants reported never truly believing in God. I asked respondents to rate their religious belief (a separate measure from religious practice frequency) on a scale of 1 to 10, with 10 denoting the most sincere and committed belief in God and in their religion. Only two respondents rated religious belief over a 5 at the most religious point in their lives. Jennifer, who only attended church occasionally and who grew up in a household where religion was rarely a topic of conversation, rated her most religious point as a 3 or 4 until she was around 13 years old. She remembered feeling indifferent about religion in general and not being sure of whether or not to believe in God. In fact, although Jennifer鈥檚 parents sometimes attended church, they expressed opposition to people being extremely religious, and thus steered Jennifer away from attending youth groups as a child. Edward rated himself as a 3 or 4 at his most religious point when he was age 4 or 5. He remembered believing Bible stories simply because he liked them. He followed with the comment: 鈥淚 believed in magic and stuff as a kid. Now I believe less in magic [laughs].鈥� When asked if he ever felt any closeness to God, James, who attended Sunday school a handful of times, responded: 听听听听听听听听听听

听听听听听听听听听听听 [At Sunday School] I always pretended that I was praying, but I could never figure out 听听听 what they were actually getting at. They always wanted me to connect with something, 听听听听听听听听听 and I was always waiting for some voice to answer me while I鈥檓 saying stuff in my head, 听听听听听听听听听听听 and I never heard anything鈥ut it wasn鈥檛 like I believed or knew anything about 听听听听听 scripture or took it very seriously.

听

Matthew commented that when he was a child, he would go through the motions of praying to God, but only when he got in trouble, saying, 鈥淕od, pull me a favor. I don鈥檛 want to get in trouble today.鈥� Matthew later stated that he never truly believed in God saying, 鈥淚 never wrapped my head around it, that there was something more than what was here.鈥� For Matthew, there existed a religious socialization in his family centered around religious practices like praying and attending church. Although Matthew participated in these practices, he could never make himself truly believe.

听听听听听听听听听听听 For all of the participants, even a minute level of religious belief did not last, almost as if religious belief was a childhood phase. Bethany, who grew up in a conservative Christian household, reaffirmed this idea of religiosity being likened to a phase that one eventually grows out of, saying:

听听听听听听听听听听听 I find it fascinating to see how these people can devote their lives to this religion. And 听听听听听 especially since I鈥elt like I almost grew out of Christianity. Like it was just like this 听听听听听听听听听听 phase when I was little. And to see these people that don鈥檛 grow out of it [laughs]. Like 听听 these large masses of people that still believe in it is just really interesting to find out why. 听听听听听听听听听 That鈥檚 what interests me. And to see what they get out of it. Because I felt like I didn鈥檛 get much out of it.听

听

Negative Experiences with Religion

听

听听听听听听听听听听听 Several respondents mentioned negative experiences with members of their religious community or religious family members as a reason for beginning a questioning process of their faith. Bethany grew up in a conservative Christian family that went to church every Sunday. Bethany explained that her family got kicked out of the Catholic Church when she was about six years old because her mother did not convert to Catholicism. At the next church she attended, the pastor got fired and disowned by the community due to an affair scandal. She also cited her parent鈥檚 intolerant beliefs as reasons for her questioning. She said, 鈥淎s a got I little bit older, I was always pretty accepting of people from other religions. My parents weren鈥檛 [laughs]. Yeah, like especially my dad. I mean, basically like 鈥楥hristian鈥� and 鈥榞ood person鈥� were synonymous to him. And so, that part鈥 never really understood that鈥︹€� She also said that both of her parents thought people of other religions were confused and did not know the truth. Bethany remembered being skeptical of this, saying, 鈥淗ow do I know that this [religion] is right, but then all these other people in the world are wrong?鈥�

听听听听听听听听听听听 Aria explained her experiences at the Old Apostolic Lutheran Church as a child. She remembered being appalled by the treatment of women in the church, saying: 鈥淲omen aren鈥檛 allowed to cut their hair, or you鈥檙e not allowed to watch TV. Women aren鈥檛 supposed to go to school.鈥� Women also had to cover their heads with a scarf after they were confirmed in the church. Aria said that as a small child, she did not mind going to church with her grandmother, but then as she got older, she would think, 鈥淭his is horrible. This is oppressive.鈥� Eventually, when she was around the ages of 10 to 14, in a small act of defiance during church services, she would pull off her headscarf if she disagreed with the pastors:

听听听听听听听听听听听 It was my own little form of protest. And so I would say there was some antagonism 听听听听听 towards organized religion in that form for me鈥nd I would say that questioning God 听听听听听听听 was a lot easier for me because of the Old Apostolic Lutheran Church. Because I was 听听听听听听 like, there鈥檚 no way this was right. And this belief that these people have鈥t can鈥檛 be 听听听听听听听听听 right. So they would say things like, 鈥榳e are the true Israelites鈥veryone else is going to 听听听听听听听 Hell. All of the Catholics are going to Hell; all of the other Lutherans are going to Hell. 听听听听听 Definitely the Jewish people are going to Hell.鈥� To me鈥aybe because the teachings 听听听听 were so intense鈥nd that鈥檚 where the questioning began because鈥ou鈥檙e not supposed 听听听听听听听听听听听 to have friends from the world. That鈥檚 what you鈥檙e called if you鈥檙e not in the church: 听听听鈥榝rom the world.鈥�

听

Interestingly, Aria鈥檚 very positive experiences with the Quakers from attending a Quaker school made her more skeptical of religion in general. Because she had the two influences: one very accepting and liberal, and one very oppressive and conservative, she opted for the critical thinking and questioning taught by the Quakers, eventually rejecting religious belief and a belief in God entirely.

听听听听听听听听听听听 Interestingly, Brian was the only participant who continued religious practice and some level (albeit very low) of religious belief into his college years. Brian attended Wednesday Bible studies in college until his sophomore year. He said that upon the first week of going to a Presbyterian church, 鈥渢wo things happened that sort of were the catalyst for the whole house of cards to come down.鈥� Apparently, the youth pastor at the church with whom Brian was very close was forced out of the church because of her unconventional style of teaching. Brian explained, 鈥淪he was very passionate, and she was not afraid to address difficult topics. She would talk about sex and cursing in a very real and tangible way. She was sort of more in touch with the world than the head pastor of the church.鈥� Brian explained that her contract was changed to the point that her pay was cut in half, essentially forcing her to resign. In the same week, Brian also heard that one of the staff members from the church he attended in high school had been fired because he was gay. This, according to Brian was the last straw. He said, 鈥淚 had come to realize that the church was fallible, and that was pretty much all it took for the entire house of cards to come down. And I just started asking questions and researching. And I read quite a bit on the topic over the course of two years from both sides.鈥�

听听听听听听听听听听听 Mark was also heavily influenced by negative experiences with religion. Mark described going to Roman Catholic mass every Sunday and attending a Catholic grade school. Mark developed reservations about the establishment of the church and power structure within the Catholic church when he got confirmed:

听听听听听听听听听听听 When we got confirmed, they were telling us that we had to bow and kiss the ring of the 听 Bishop. And it just infuriated me! No one is better than me, in terms of that I need to kiss 听听听听 their ring! And why should the person be the vassal for an omnipotent creature?

听

Mark also described noticing a spiritual difference between himself and the other members of his church saying, 鈥淚 couldn鈥檛 get all spiritual and be so devout鈥 have a problem with super conservative beliefs.鈥�

听听听听听听听听听听听 Cora, although never having a religious belief nor growing up in a religious family, had a phase of questioning sparked by a negative experience with religion when she was in middle school. In sixth grade, Cora remembered that a classmate told her that she was going to Hell because she was not religious and her parents were atheists. Some parents of her friends in grade school also would not let her come over to their houses for the same reason. Also, during her eighth grade year, Cora had a science class where the theory of continental drift was explained. Cora remembered that the teacher described the scientific theory, but then told the class that what really happened was that 鈥淕od鈥檚 hands came down and separated the continents. And that鈥檚 the theory just like the scientific theory.鈥� Cora explained how appalled she was about the lack of separation between church and state. She said that up until this point in her life, she was not 鈥渕ature enough or had enough background knowledge to really start thinking about [atheism].鈥� After these negative experiences with religious people however, Cora began to become very interested in the debate between religion and atheism, realizing: 鈥淚 need to start standing up for this.鈥�

听

Traumatic Experience

听听听听听听听听听听听 Some participants cited experiencing a life-changing event that made them question their religious beliefs, and especially their belief in God, reporting that if there was a god, it did not make sense for there to be evil and suffering in the world. Bethany cited both negative experiences in the church as well as her conservative parents鈥� intolerant views of others as reasons for her skepticism of religion. Moreover, the trauma of her parents鈥� divorce when she was 15 essentially shattered most of her childhood religious beliefs. Bethany remembered:

听听听听听听听听听听听 I mean, a lot of stuff was changing in my life then鈥nd since I was questioning 听听 everything and religion had been such a big part of my life, I just started questioning that 听听听听听听听听 too. And I mean, there was also the part where if there is a god and he鈥檚 supposed to be 听听 good, then why is all this happening to me?

听

After the divorce of her parents, which Bethany describes as a 鈥渢urmoil,鈥� Bethany鈥檚 entire worldview changed. She said:

听听听听听听听听听听听 I would go to church with her [my mother]. At that point I sort of started to see鈥ow, 听听 you know, judgmental they were. And the more negative aspects of it. And so, then I 听 started to question, why do I identify with these people? And, just thinking back to all the 听听听听听听听听听 negative experiences that I told you before about church and everything. And then I just 听听听听听听听听听 started on my own, just researching other religions, And just figuring out鈥nd I even 听听听听听听听听听听听 talked to some of my atheist friends too鈥nd so I guess after all of that, then I just kind 听听听听听听听听听听听 of realized that I actually don鈥檛 agree with this religion that I鈥檝e been practicing my whole life.

听

听听听听听听听听听听听 Matthew also experienced trauma at an early age when his friend died. According to Matthew:

听听听听听听听听听听听 He died when I was young. And it made me think if there is somebody who is all- powerful who can save anybody, who can kill any person, why would He let us die? And 听听听听听 it was that point I was like, there must not be anything or, I mean, if He鈥檚 better than us, He [God] wouldn鈥檛 let people die. So that was the big revelation.

听

For both Bethany and Matthew, life-changing experiences made them question the beliefs with which they had been raised. For Bethany, the divorce of her parents made her begin to question everything she had ever been taught, while for Matthew, the death of his friend seemed to be more of an affirmation of his suspicions that there was no God.

Friends

听听听听听听听听听听听 For some participants, friends in school were the most instrumental in setting a path of questioning for the participants. Julie grew up in a single-parent household and was very involved in the church community (going to Sunday school every Sunday), although she said her beliefs in God and Christianity were mainly out of fear of going to Hell. In fifth grade, she stopped going to Sunday school because her best friend also stopped going. Julie commented:

听听听听听听听听听听听 It was a motivation between us two. It was almost like a competition. Like who is more 听听 Christian than the other. So that was like some of the motivation I had to go. And as soon 听 as her interests turned, mine started to as well, and I decided not to keep up on it鈥here 听听听听听听听听听听听 for a while, my mom would make me go to church on Easter, with my grandma and 听听听听听听听听 everything, and I was like, 鈥榯his is stupid. I don鈥檛 want to hear anything about this.鈥� 听听听听听听听听听听听 Like yes, they鈥檙e good messages, but I just felt that they were鈥isinformed.

听

After she stopped going to church, Julie entered a 鈥渞ebellious phase鈥� when she was 12 and 13, listening to death metal and being 鈥済othic.鈥� She mentioned that being in the gothic crowd implied being a Satanist, which she did not want to identify with, saying:

听听听听听听听听听听听 Just because I don鈥檛 like what Christianity teaches doesn鈥檛 mean I have to go to the polar opposite鈥here could be a common ground. And that鈥檚 when I discovered atheism鈥nd 听听听听听听 then, yeah, science kind of stepped in...atheism just is more appealing for me. I guess the main appeal was I wasn鈥檛 gonna get judged for not believing in salvation鈥r not 听听听听听听听听 believing in the devil as well.

听

For Julie, atheism served as the 鈥渕iddle ground鈥� between her religious upbringing and her social group who were Satanists. For Julie, there did not seem to be a questioning phase that followed her rejection of Christianity and Satanism, but rather an adoption of atheism and the contentment that came with getting her questions answered with science.

听听听听听听听听听听听 Although Mark was skeptical about some of the more traditional and conservative practices in his Roman Catholic church, he said that up until high school, he was still a devout Catholic (though with a relatively weak belief in God) and 鈥渨ould have defended Catholicism.鈥� It was not until one of Mark鈥檚 friends began introducing him to new ideas around his sophomore year of high school that Mark began questioning his beliefs. Mark explained:

听听听听听听听听听听听 My friend, my good friend now, who is a year older than me, started talking a lot. And 听听听 he鈥檚 not atheist at all, but he has some crazy beliefs. Always changing his mind鈥 听听听听听听听听听听听 always loved reading. I knew all about these different things鈥o he would tell me 听听听听听听听听听听听 all these things, and I would look into them鈥� all this spiritual science, all this 听听听听 pseudoscience鈥� And then sort of day-by-day, I鈥檇 read different sources. So, it didn鈥檛听听听听 take me long to get rid of Christianity, to say I鈥檓 no longer Catholic, but it took me a 听听听 little bit longer to come to myself and be like, 鈥楢lright, yeah, I believe I鈥檓 an atheist. I 听听听听听听听听 am an atheist.鈥�

听

Mark went on to explain that although his friend was never an atheist, his strange views about conspiracy theories, eastern religion, Buddhism, and so on, made Mark open to beliefs other than the ones with which he had been raised. The friend being interested in far-fetched beliefs had a large impact on Mark allowing himself to begin the questioning process of his own religion, which eventually led him to become an atheist.

Questioning Beliefs and Losing Religion

听听听听听听听听听听听 Many of the participants reported numerous influences in their final adoption of an atheist identity. Never truly believing in God or religion, having negative experiences with religion or religious people, experiencing a traumatic event, and being influenced by friends all were common threads that were seen among my participants.[12] After experiencing certain events, some participants were 鈥渟parked鈥� into a questioning phase (although some went straight from the 鈥渟park鈥� to adopting atheism without a questioning phase) that was described just briefly in some of the quotations above. For many, this questioning period involved reading a variety of sources about God and religious belief as well as atheism. Not all participants went through an externally visible questioning phase of researching religion or irreligion, but it certainly was a common trend. Another trend among people who questioned themselves was a less visible and more internal personal questioning that led to the adoption of an atheist identity.

External Questioning

听听听听听听听听听听听 Of the 14 participants, Brian underwent perhaps the most dramatic questioning phase. After his youth leader was fired for encouraging questioning attitudes from her youth group and a staff member from Brian鈥檚 childhood church was fired for being gay, Brian repeated several times that the 鈥渨hole house of cards came down.鈥� He said that after these two events occurred during his sophomore year of college, he began reading literature from 鈥渂oth sides,鈥� and did so for two years. Brian explained:

听听听听听听听听听听听 I read a lot of Peter Kreeft, the Christian apologist. I read basically every book by Lee 听听听听听 Strobel, who is a converted- a reconverted atheist, I suppose you would say. And I read C.S. Lewis, but I also read Dawkins, Dennett, Harris, and Hitchens. And I just came to 听听听 realize that they [the atheists] had much more contact with reality. They had a lot more tangible arguments. I was watching debates on YouTube and listening to Podcasts on 听听听听听听听听听听听 both sides, so I just kind of mainlined information into my head for about two years. And 听听听听听听听听听听听 by the end of it, I was beyond convinced鈥�.after a while it was like reading the same 听听听听听 book over and over again, hearing the same arguments that had been put to bed over and 听听 over again鈥 guess that was sort of the deconversion story in a nutshell.

听

Brian鈥檚 questioning and research period was very long, and through a highly methodical process, he rejected religion and God and adopted an atheist identity by the time he was a junior in college.

听听听听听听听听听听听 Although Bethany described a period of questioning beliefs after her parents鈥� divorce when she was 15, she did not fully adopt an atheist identity until her freshman year of college.听 She explained that she began actively gathering information that she had previously ignored on a variety of topics including other religious beliefs and atheism. She describes beginning a process of researching other religions and talking to some of her friends who were atheists just to understand their points of view on religion and Christianity in general. She also remembered having conversations with her brother who adopted an atheist identity as well (interestingly, the two of them came to the same conclusions independently of one another). Upon entering college, Bethany took a philosophy class that presented various arguments for and against the existence of God. This sparked further questioning where Bethany began to read more articles for and against God and became convinced with 鈥渁ll the evidence against God鈥� that there was no God. 听

听听听听听听听听听听听 Aria began questioning her religious beliefs as young as fourth grade, sparked by her experiences in the Old Apostolic Lutheran Church. The church鈥檚 intolerant, rigid beliefs were in direct conflict with the very accepting views of the Quaker school that taught Aria to be a critical thinker and that the 鈥渢ruths that people tell you aren鈥檛 necessarily ultimately truth.鈥� Aria鈥檚 outward questioning and dissidence in the church was demonstrated through her open form of protest where she would pull off her headscarf in the middle of a church service. Aria later came to reject all of the beliefs with which she had been brought up at the Old Apostolic Lutheran Church and went a step further to reject a belief in God entirely.

听听听听听听听听听听听 During her questioning period when she was 12 or 13, Julie said that she began to realize a trend that led to her adoption of atheism. She remembered studying medieval times and noticing that whenever people could not explain something like a drought, they would often attribute the event to an act of God. Julie remembered realizing that religion and God were a 鈥渢heory for everything鈥� that people did not understand or that they could not explain. Julie came to reject religious explanations for events and began to adopt atheism and scientific reasoning as better explanations for natural events and her questions.

听

Internal Questioning

听听听听听听听听听听听 A common trend among people who were less personally religious growing up but who accepted a 鈥渘ormative鈥� belief in God, was undergoing a less outwardly visible questioning phase where they tended to simply question the normative belief they had and eventually reject it without doing heavy research, protesting against religion, or seeking out other atheists. James was a good example of this trajectory. He said that although his beliefs were in line with atheism since he was in middle school, he did not describe himself as an atheist until the very end of high school or the beginning of college. James said that he was in middle school when he began to think that the things he had learned about religion in Sunday school and from his religious friends were 鈥渒ind of silly.鈥� James said that he 鈥渞ationally鈥� came to the conclusion that there was no God because that seemed to be the most logical conclusion to him.听

听听听听听听听听听听听 Danielle also described a type of personal questioning. She said that she first became aware of atheism when she stayed with a friend whose entire family was atheist. That was the first time Danielle came across the idea that she did not have to believe in anything. Around eighth grade, Danielle solidified her identity as an atheist after her experiences of questioning whether or not there truly was a higher being. For Danielle, a failure to experience God served as a catalyst for her rejection of a religious belief. She remembered trying to pray, nothing happening, and subsequently concluding that God did not exist.

听听听听听听听听听听听 Jennifer experienced oscillation between religious and irreligious beliefs as a child, but described a point in high school when she concluded that there was no God. She admitted that she could not identify exactly what triggered this revelation. She went to a youth group with some of her Mormon and Christian friends and said: 鈥淐hurch just started giving me the heebie-jeebies. I just was really uncomfortable there, just because I felt like they were just kind of lying to me. People were just up at the pulpit, just lying.鈥� She said that she became very interested in science in high school, and that the more she learned about evolution, the more it made her feel like the world did not need a god. Jennifer stated: 鈥淭here鈥檚 so much science and so many things we know about the world, and so much beauty just in nature and biology and physics, and鈥ow this world works, that I didn鈥檛 feel like there needed to be a god.鈥� She also took a high school class called the Theory of Knowledge, which emphasized theories of how humans know what they know. In the class, Jennifer described feeling very negatively about religion when she learned about 鈥渁ll the oppressive things that happened out of religion鈥� in this specific class. All of these factors culminated, as Jennifer stated, 鈥淚 think all those things: the science, the growing up, some of the atheist friends, not liking church鈥t all kind of pushed [me]鈥t was a very gradual thing for me to go like, 鈥榤aybe [there is a god]鈥� to 鈥榥o, there is no God.鈥� 鈥�

III. Experiences with an Atheist Identity

听听听听听听听听听听听 The third primary theme, Experiences with an Atheist Identity, encompasses the lived experiences of college atheists after they adopted their atheist identity. This primary theme is further divided into the subthemes: Atheism Personally Defined, Atheist Identity Salience, and Perceptions of Stigma. In the first subtheme, Atheism Personally Defined, participants expand on their own personal definitions of what it means for them to be an atheist. The second subtheme, Atheist Identity Salience, constitutes the importance of atheism to the participants鈥� identities. The third subtheme, Perceptions of Stigma, covers the participants鈥� awareness of a stigma surrounding their atheist identity, if any, and how accepted participants felt as they navigate the social world as an atheist. The last two sections: Coping with Stigma: Redefining Friend Groups and Controlling Information and Coping with Stigma: Redefining Morality describe the two main ways that college atheists cope with a sometimes unaccepting society, how they are able to find accepting friend groups, and how they define (or redefine) themselves as moral people who are undeserving of a cultural stigma.

Atheism Personally Defined

听听听听听听听听听听听 When defining what atheism meant in terms of a personal definition, most participants mentioned that atheism at its most basic level is a lack of a belief in God. Most respondents also added that it is a rejection of any religion. James stated simply: 鈥淎t its most basic level, I think it [atheism] means that you don鈥檛 believe that there is a god. And I think that can mean different things to different people, but I think at its base level, that鈥檚 what it is.鈥� Bethany stated: 鈥淚 think to be an atheist is to not practice any religion. To not identify with any religion. And to not believe in any sort of god or other supernatural being.鈥� Danielle鈥檚 definition was: 鈥淚 don鈥檛 believe that there is a higher power or a man in the sky or whatever.鈥� She also added, 鈥淏ut, I don鈥檛 judge anyone who believes that.鈥� Aria also added this sentiment of a non-judgmental, personal belief to her definition of atheism, echoing that she does not believe in God, but adding that she does not feel compelled to tell people about this belief or to force it on other people. Several participants added this caveat of not wanting to push their atheism onto other people, mainly because they felt resentful of religious people who try to push their beliefs.

听听听听听听听听听听听 Some respondents had more involved and personal answers of what atheism meant to them based on science or other reasoning outside of religion. Jennifer, an evolutionary biology and secondary science education major responded:

听听听听听听听听听听听 I think being an atheist means not believing in any higher being. That, like what you see 听听听 on earth is what you get. There鈥檚 nothing after and there鈥檚 nothing before. You鈥檙e not 听听听听 here for a specific reason. There鈥檚 no one up there who has a plan for you and has a plan for the world. The world just鈥t is鈥 think because I鈥檓 a scientist, I very much believe 听听听听听听 in just the natural way of things. When you die, maggots eat 鈥榶a! You decompose [laughs]. And I think atheism has become this thing where you鈥檙e kind of combative 听听听听听听听听听听听 against people who are religious, and I don鈥檛 necessarily agree with that鈥 really try not to be combative.

听

To Edward, morality outside of religion informed his definition of atheism:

听听听听听听听听听听听 It鈥檚 to write your own moral code. Like the definition of atheism is to be, like if I were to look it up, it would be to abandon God. But to me, it鈥檚 to abandon the Bible and to write 听听听听听听听听听 your own moral codes and write your own beliefs and philosophies. A lot of it has to be 听 based on what you see and what you learn from the world鈥to] start from square one and 听听听听听 start writing everything over from the beginning.

听

For Edward, atheism and adopting an atheist identity was very much about reinventing himself and the way he viewed morality and the world. His analogy of 鈥渟tarting from square one鈥� implied that his atheism made him reevaluate everything he thought he knew about the world from a more critical, scientific standpoint than the religious, faith-based philosophies with which he had grown up.

听听听听听听听听听听听 Most participants also added something about the rejection of religion in general and that atheism implies critical thinking to their definitions of atheism. Cora stated that atheism is 鈥渘ot believing in a god or any gods鈥t鈥檚 looking at anything with a critical eye. So, for me, I don鈥檛 take things based on faith value; I want to know why something is this way. I love to ask questions鈥︹€� Cora also made an interesting point that atheists and religious people are not as different as they may seem, saying atheists are defined based on what they do not believe in, but that 鈥渁ny religious person could be considered an atheist because鈥he only god they do accept is their own, and so in any other religion, they鈥檇 be considered an atheist鈥 believe in one less god than they do, which doesn鈥檛 separate us by much, but they don鈥檛 like to hear that.鈥�

听听听听听听听听听听听 Notably, two participants did not even mention a lack of belief in a god as a defining feature of atheism. Sophie stated: 鈥淚 think for me, it means just keeping an open mind. Just not believing in anything that comes by, but also really deeply thinking about anyone鈥檚 opinions. Yeah, bottom line, keeping an open mind.鈥� Julie鈥檚 definition was very similar: 鈥淭o be an atheist to me is to definitely accept everyone鈥檚 religions鈥lso what atheism means for me is [to] educate yourself on all other religions and obviously science as well. So to me, atheism is just learning as much as you can about everything.鈥� These were two interesting responses because instead of citing a lack of a belief in God, the two women described atheism as encompassing characteristics such as acceptance, open-mindedness, knowledge, and knowledge of other religions as defining features of atheism. These specific features are undoubtedly characteristics of many religious people, but it was interesting that these two respondents saw them as stereotypical of atheists. Brian also underlined the stereotype of atheists as open-minded and scientific saying: 鈥淚 find that if somebody is an atheist, there鈥檚 a high likelihood that I will find them an interesting person because chances are, they are intellectually honest and intellectually curious.鈥� This was an interesting finding indicating that atheists themselves make generalizations and have stereotypes about other atheists. These included assuming that all atheists develop this identity through reason, curiosity, critical thinking, and so on, and that the antithesis to atheism is听 accepting any belief based on faith, which is a common stereotype of religious people held by some atheists.

Atheist Identity Salience

听听听听听听听听听听听 The importance of atheism as an identity or as part of participants鈥� identities varied greatly. In order to get an idea of how important an atheist identity was to participants, I asked each one to rate the importance of atheism to their identity on a scale of 1 to 10, with 1 being not at all important and 10 being the most important. If respondents did not feel that a scale captured their feelings about their atheist identity importance, I allowed them to explain it to me differently. A majority of the respondents (10 of the 14) rated atheism as being a very important part of their identity with ratings of 7 or higher (or saying that it was 鈥渋mportant鈥� or 鈥渧ery important鈥�). Four of the participants rated the salience of their atheism between 3 and 4 (or 鈥渘ot important鈥�). These responses will be expanded upon in the two following sections.

听Atheist Identity as 鈥淰ery Important鈥�

听听听听听听听听听听听 Jennifer rated atheism as a salient part of her identity as a 10, saying:

听听听听听听听听听听听 I think it鈥檚 very important to me in the sense that I think about it a lot, and it鈥檚 a big part of my life, just because I spend so much time doing science, and I like that I don鈥檛 have to 听听听听听听听听听听 reconcile the scientific part of myself with a religious side because I don鈥檛 have 听听听听 one鈥nd I think it鈥檚 made me more confident in some ways in how I move about the 听 world just because once you realize that you鈥檙e on your own, in a certain sense, I think 听听听 you can be very purposeful and very鈥onfident in how you make decisions.

听

Cora similarly rated atheism as very important to her identity (8-9), also saying that with an atheist belief, she does not have to reconcile science and religion. Cora stated: 鈥淚 feel like if I did have beliefs that were more on faith, it would be harder for me, especially in the sciences. So, I think that having that evidence-based kind of belief system or acceptance has really helped me鈥o I think that鈥檚 why it鈥檚 really important to me.鈥�

听听听听听听听听听听听 Edward rated his atheist identity as an 8, saying it 鈥済rounds a lot of the decisions I make. And I don鈥檛 go around making decisions like 鈥榳hat would an atheist do?鈥� I just have written my moral code along the lines of who I am, and who I am also happens to be along the lines of being an atheist.鈥�

听听听听听听听听听听听 James based his answer on of an idea of rationality, rating his atheism as a 7 (out of 10) in importance to his identity. He said that he did not feel strongly enough about his atheism to join an activist group (which is what he would consider to be a 10), but he stated:

听听听听听听听听听听听 I also think that a lot of who I am is based on my idea of rationality and thinking about 听听听 things in a certain rational way. And I think that鈥檚 a big part of it is how you see yourself 听听听听 in relation to the rest of the world and other people鈥o I鈥檇 say atheism is a topic that 听听听 really does a good job of representing how I think about the world.

听

Mark also mentioned rationality in his description of atheism as being 鈥渧ery important鈥� to his identity. He described wanting to gather knowledge and information, saying, 鈥渁theism is huge for me because I鈥檓 rational and, you know, this is just part of who I am. Huge. Huge part of who I am.鈥�

听听听听听听听听听听听 In contrast to his peers who answered that their atheism was very important to them, Brian rated his atheism as a 2 in importance to his identity, but rated his skepticism as a 10 saying:

听听听听听听听听听听听 That鈥檚 the main way I interact with the world. It just depends as what you mean by 听听听听听听听 atheism. So atheism in a vacuum is not extraordinarily important to me. I just happen to 听听听听 think it鈥檚 true. I also happen to think fairies and unicorns don鈥檛 exist. I鈥檓 exactly as 听听听听听听听听 passionate and as firm in my beliefs against fairies, leprechauns, and the tooth fairy as I 听听听听听听听 am against Yahweh. If you take atheism as a stand-in for skepticism and having a 听听听听听听听听听听听 realistic outlook, then it鈥檚 extraordinarily important.

听

This reveals an important distinction that not all atheists see their nonbelief as extremely important, but that their certain way of thinking about life (that one could define as skepticism, atheism, humanism, freethinking, and so on) simply differs from the norm of religious belief or faith. In this sense, semantics mattered for the participants, but the most important concept was establishing that their identity and worldview are informed by science, reasoning, and so on, rather than a non-belief in God.

Atheist Identity as 鈥淣ot Very Important鈥�

听听听听听听听听听听听 Of the participants that rated their atheism as not an important part of their identity, most stated that the lack of importance was due to an overall lack of importance of religion as a topic in their lives. For Karen, her atheism is not important because she said that she never really thinks about it, saying: 鈥淚t鈥檚 just not really a big part of my life. It鈥檚 like, why would I make a non-belief a big part of my life? It seems counterintuitive.鈥� For Aria, it is very important that people do not see her as an Old Apostolic Lutheran, but she stated that it is not at all important that people know she is an atheist because it is not a part of her 鈥渃ore identity.鈥� Aria explained, 听鈥淚t鈥檚 just, you know, just one of my beliefs. Like my political beliefs or anything like that.鈥� Dillon said that he does not define himself in relation to other people鈥檚 religious beliefs and that he is not influenced by others鈥� ideologies. He stated that atheism too is very low in importance to his identity.听

Perception of Stigma

听听听听听听听听听听听 Just as there were variations in how the atheist students defined atheism when referencing their own specific beliefs, there were variations of how accepted people felt in society as atheists. Several people experienced varying degrees of unacceptance from individuals and from society in general due to the general stigmas that surround an atheist identity. This section will discuss the experiences of individuals encountering various forms of acceptance and unacceptance from their family, friends, and society and will later explain the ways that individuals coped with perceived stigma.

听听听听听听听听听听听 In order to understand how participants navigate the world with a minority identity and if they perceive any stigma as attached to this identity, I asked a variety of questions in order to gain an understanding of how accepted these people felt as atheists by other individuals as well as society. The concept of acceptance was a way to operationalize perceived stigma with the idea being that levels of perceived unacceptance would translate to feelings of being stigmatized. I did not ask a direct question if participants felt that their identity as an atheist was stigmatized because I wanted to allow them to explain their experiences rather than run the risk of projecting an idea onto them that they might not experience. All participants that were interviewed had informed their parents of their identities and many of their friends. Many individuals cited Boulder or the university campus as overall accepting settings, but said that they might feel unaccepted if they lived in different areas.

听听听听听听听听听听听 When asked if she felt accepted by society as an atheist, Cora answered: 鈥淚 would say more and more. I think too though, as within college and then in graduate school, the people are more liberal鈥�.I think liberals generally are more open-minded, so especially being in Boulder, it has not been hard for me to say I鈥檓 an atheist鈥︹€� She explained that in the city in which she grew up, which was predominantly conservative and Evangelical Christian, she felt less accepted since she lost friends because she had different religious beliefs. Aria also mentioned academia as a very accepting place for atheists. Although she cited feeling like she had to defend her atheist identity around her boyfriend鈥檚 religious family and around her grandmother who would prefer that she went to church, Aria said that she generally feels accepted by the academic community and by society as a whole. 听

听听听听听听听听听听听 When asked about how accepted she felt by society and individuals, Karen stated that she felt as if the atheist part of her identity is often under attack by individuals and society because 鈥測ou see stuff that鈥檚 like, 鈥榶ou can鈥檛 have morals without a religious background.鈥� 鈥� Although Karen said she feels comfortable with her atheist identity and has told her parents, she still has not told her grandparents, saying that they would be 鈥渦pset鈥� and 鈥渘othing really good would come out of bringing it up.鈥� Otherwise, Karen had a positive overall feeling about the level of societal acceptance, saying that she feels accepted and especially among younger people. This may also be because Karen does not choose to talk about her nonbelief, saying that it is not a large part of her life. Dillon also felt accepted in general by his family and society as an atheist, but then added that he also does not talk about religion on a daily basis anyway. He also mentioned that he might feel differently outside the college environment, saying 鈥渋f you go out into society and you start taking with elders or your other cohort within your work environment, if your beliefs don鈥檛 match up with the majority of the beliefs, you might be discriminated against. I have not personally had that experience, but I鈥檇 imagine it鈥檚 out there.鈥�

听听听听听听听听听听听 Unfortunately, some participants did feel some level of fallout upon informing their family members of their rejection of religion and adoption of atheism. Bethany said that her mom was 鈥渁 little bit sad that I had adopted that [atheism]. Just because she had spent so much time trying to raise me as a Christian.鈥� She also does not have contact with her father, but said that he would likely be very unaccepting of her new identity. Apart from her parents, Bethany felt generally accepted by society, especially in Boulder. She added a disclaimer that she had not identified as an atheist for long, implying that she may feel differently as she spends more time in society with this particular identity.

听听听听听听听听听听听 Jennifer also said that her mom was 鈥渂ummed out鈥� upon finding out about Jennifer鈥檚 atheism although her father was fully supportive. She also mentioned that her religious grandmother does not know because Jennifer has simply avoided opening up 鈥渢hat can of worms.鈥� It seemed as if Jennifer felt more accepted as an atheist in areas and situations outside of her family where there remained some uncomfortable feelings surrounding her identity. She mentioned Boulder as a supportive context, saying, 鈥淏oulder is a very good place to be an atheist. I think it鈥檚 almost like a popular thing. It鈥檚 more common in some ways than being religious.鈥� Jennifer also mentioned the university setting as largely accepting and especially in her niche of an evolutionary biology lab which according to her, is a sort of atheist microcosm where most scientists in the lab are openly atheist.

听听听听听听听听听听听 James reported feeling accepted as an atheist by his father, who he suspects is also an atheist, but feels slightly different about his mother, saying: 鈥淢y mom, I don鈥檛 think really completely believes that I am [an atheist]. Although I think she鈥檚 fine with me believing whatever I want to. But I think she鈥檚 in denial of what I really am.鈥� James acknowledges his identity as stigmatized, but said that he is able to get by because religion does not come up in routine conversations. He commented: 鈥淚 don鈥檛 feel like it鈥檚 [the atheist identity] under attack on a daily basis because I don鈥檛 think it really comes up, but I think that鈥檚 because it鈥檚 a stigmatized topic鈥︹€� James mentioned a fear of discrimination by potential employers based on his atheist identity, saying that he keeps this information off of social media pages like Facebook.

听听听听听听听听听听听 Mark also had similar experiences to James, as his parents seemed to be in denial over his atheist identity. Mark seemed to have the most difficulty of all the participants interviewed informing his Roman Catholic family of his atheist identity. He said that he felt attacked by his parents and his whole family. When he first came out to his parents as an atheist, they forced him to continue attending church, saying: 鈥淵ou still live in our house.鈥� Mark responded by writing an essay to them about his beliefs. Eventually they stopped making him go to church, and his father eventually accepted him. However, his mother was apparently 鈥渨orse. Much, much worse.鈥� Mark stated, 鈥淢y mom still has this hope that I鈥檒l turn the corner, and one day I鈥檓 just going to show up at the door and be like 鈥業鈥檓 Catholic again!鈥� I don鈥檛 know. I think it鈥檚 a little far-fetched, but it鈥檚 alright. Whatever helps them sleep at night [laughs].鈥� Although Mark has told many of his cousins and other family members about his atheism, he has not told his grandparents saying, 鈥淚 love them too much.鈥� Mark鈥檚 aunt reacted poorly to his transition to atheism, questioning the possibility of morality without a belief in God. He stated, 鈥淭hey [atheists] are mistrusted. They are鈥pparently no one believes that anyone can have their own moral compass. It鈥檚 too bad.鈥� He explained that his aunt 鈥渄oesn鈥檛 understand how people can be moral without religion. Greatly upsetting. Because then it makes her [think] that humans must be so weak-minded that they can only act rightly out of fear鈥 think we should be able to act right out of what鈥檚 good inside of us.鈥� On a societal level, Mark feels that he could never come out publicly as an atheist if he wanted to run for public office, saying: 鈥淚 would feel entirely more accepted in Europe. I feel like if I ever did something major and came out as atheist, there would just be huge backlash鈥f I did that in America. One hundred percent.鈥�

听听听听听听听听听听听 Matthew said that his mother wished that he were still a Christian, although this fact is not a point of contention in their household. He also did not tell his grandparents about his identity 鈥渙ut of respect.鈥� Matthew said that he was sure his grandparents would still love him if they knew, but that they would undoubtedly 鈥減refer鈥� that he was a Christian. Matthew raised this idea of 鈥減reference鈥� by describing how he does not feel completely accepted in society as an atheist. He explained, 鈥淒oes society wish I wasn鈥檛 an atheist? Completely. I believe they [society] wished that I believed in some religion鈥�.One of the things is that鈥ertain [political] parties would never accept me for being an atheist. So while society will accept that I鈥檓 an atheist鈥hey will always try to change [me].鈥�

听听听听听听听听听听听 Brian mirrored this idea of his family 鈥減referring鈥� that he was not an atheist. His whole family is aware of his atheism and Brian said they still love him saying: 鈥淚t鈥檚 not like they don鈥檛 show me the same affection that they show the other cousins. It鈥檚 just sort of like passive aggressive. It鈥檚 sort of like 鈥業鈥檇 really prefer if you weren鈥檛 an atheist, but I still love you.鈥� 鈥� These examples show that although many participants do not experience dramatic consequences for their atheism such as being disowned as a family member, there is still a discomfort that accompanies an atheist identity that many respondents expressed. 听听听听听听听听听

Coping with Stigma: Redefining Friend Groups and Controlling Information

听听听听听听听听听听听 Many atheists interviewed coped with stigma by surrounding themselves with like-minded people, often other atheists. Cora mentioned that most of her friends are liberal, and her best friends are not religious. Aria too found herself in friend groups with less religious individuals. She said that she surrounds herself with other atheists as friends in order to avoid constant conflict: 鈥淚 choose to surround myself with people鈥攂ecause I can鈥檛 get into debates all the time about stuff. I can鈥檛 debate about feminism. I can鈥檛 debate about atheism. It would be so draining. So, my friends are all people who feel similarly, I鈥檇 say. None of them go to church.鈥�

听听听听听听听听听听听 Many participants coped with anticipated stigma by tightly controlling information about their identities to acquaintances to avoid judgment and discrimination. Cora mentioned that as far as romantic interests are concerned, she would avoid telling a person early in the relationship that she is an atheist. She also took any information off of Facebook concerning her atheist identity saying:

听听听听听听听听听听听 I did before have like 鈥楶astafarian鈥� and 鈥楲iberal Socialist,鈥� and now I鈥檓 making it more 听听听听听 of a professional鈥�.I don鈥檛 have those on there because I don鈥檛 want鈥hat to be the first 听听听听听听听 thing that people see and already start judging me. And when I鈥檓 applying for 听听听听听 things鈥ven in a liberal society, I do think I don鈥檛 want somebody to, you know, throw 听听听听听听听听 me out of a pool [of job candidates] or something because I am an atheist鈥nd I鈥檓听听听听听听听听听听听 fearful that if there is someone who is Evangelical and looks down upon atheists, that if 听听 they just see 鈥榓theist鈥� somehow on Facebook, that I鈥檇 been thrown out. You know, I 听听听听听听听听听听 wouldn鈥檛 want that, and I think that could happen鈥 wish it didn鈥檛 have to be that way. I鈥檓 proud to be a freethinker.

听

听听听听听听听听听听听 Jennifer said that the people she is very close to know she is an atheist, but she added that her atheist identity is 鈥渘ot something I advertise, just 鈥榗ause while I鈥檓 not ashamed of it in any way鈥t鈥檚 not something I advertise. I don鈥檛 know if it鈥檚 because I鈥檓 a private person or what鈥 don鈥檛 know.鈥�

听听听听听听听听听听听 James commented that he consciously filters himself around certain groups of people, saying, 鈥淚 guess there鈥檚 certain situations where I feel it鈥ould be under attack if I brought it up, and that鈥檚 why I鈥檇 be nervous to bring it up in certain situations like around certain鈥 guess family members or other people鈥hat I knew to be religious.鈥� James, like many of his other atheist peers interviewed, said that his closest friends are also atheists.

听听听听听听听听听听听 Dillon cited a specific experience where he had to control information for fear of being seen differently by his friends (he has a diverse friend group in terms of religious beliefs.):

听听听听听听听听听听听 I have a friend group [that] is religious, and I鈥檝e been told by other people not to bring it 听 up or talk about my identity鈥ecause 鈥榯hey鈥檒l think about you differently and might treat 听听听听听听听听听听听 you鈥� or hold grudges鈥� or something like that. So I鈥檝e held my tongue for a few people. 听听 But it hasn鈥檛 come up naturally. I haven鈥檛 held my tongue.

听

For Dillon, it seemed that while the threat of fallout from the discovery of his atheist identity by some of his friends finding out about his atheism has caused him to be cautious of what he says, but at the same time, it is possible that religion is not an important enough topic for him to feel truly fearful of being cast out from his friend group.

Coping with Stigma: Redefining Morality

听听听听听听听听听听听 Another method of coping with stigma associated with an atheist identity was defining or redefining morality outside of religion. A final important aspect to all interviews was the question of morality. All participants were very firm on the fact that they were moral humans, and they were moral without God. Many described a nuanced view of morality as something that had to be constantly negotiated and not black-and-white. Some participants also saw morality as being very different from a religious person鈥檚 morality, while others saw it as being similar due to a shared social construction of morality by society as a whole.

Actively Rejecting a Divine Morality

听听听听听听听听听听听 Brian actively rejected any idea of a morality based in religion, and specifically Christianity. He said:

听听听听听听听听听听听 I think the Bible is an awful source of morality. There鈥檚 things advocated in that book 听听听听听 that are cringe-worthy. I mean, stoning to death for minor offenses [like] picking up logs 听听听听听听听听听听 on a Sunday鈥here鈥檚 awful morality in the Bible. And it鈥檚 a very slippery slope when 听听 you have dogmatically imposed morality. Basically you have to believe the tenets of this 听 book.

听

Brian articulated his own ideas of morality, explaining the nuances of moral relativism, non-religious objective morality, and other theories, ultimately saying that he is still negotiating his own ideas of where morality originates for humans. He made an interesting point, saying:

听听听听听听听听听听听 I think morality is basically just a subset of human behavior and how we interact with 听听听听听 other people, and I think that because it鈥檚 been given this name and this special status, we鈥檝e sort of lost sight of the fact that it is just, you know, the way humans interact with 听听听听听听听听听听 one another and there鈥檚 nothing special or other-worldly about that. It鈥檚 just human 听听听听听听听听听听听 behavior. It falls into the same category as dietary habits, which no one would say there鈥檚听听听听听听听听听听听 a god-given diet.听

听

Brian, interestingly, also rejected a morality deriving from atheism saying: 鈥淢y moral system is not atheism and my intellectual system is not atheism. It鈥檚 [just] a consequence or conclusion that I鈥檝e reached.鈥�

听听听听听听听听听听听 Similarly, Jennifer thought that morality derived outside of religion is different from morality based on religious texts or tenets, saying:

听听听听听听听听听听听 I think in some ways, atheists have a very different morality. Or they come by their 听听听听听听听听 morality from a very different standpoint from religious people. I think sometimes 听听听听 religious people get stuck on this crutch of like, 鈥榳ell I did something bad, but if I go 听听听听听听听 pray about it later, it鈥檚 ok.鈥� Coming from an atheist perspective, I believe you get one 听听听 shot. Like there is no afterlife where you can have more time to live or you can make 听听听听听听听 better choices. And I think your morality comes from wanting to be a good person and 听听 wanting to be nice to the people around you rather than trying to please some higher 听听听 being or get to some higher place when you die.

听

听听听听听听听听听听听 Bethany also perceived morals as being possible without religion, and even implied that religion and morals should actually be separate, saying:

听听听听听听听听听听听 Your morals should not be defined by your [religious] beliefs. I think those are two 听听听听听听听听听 separate things. And that your morals are, you know, if you鈥檙e an honest person, if you 听听听听听听听 help others, if you are loyal and trustworthy, and everything like that. And that it can go along with your religion, but it does not have to.

听

Through this statement, Bethany posits herself as a moral human and also acknowledges that religion and morality can absolutely be compatible, but that they do not require each other.

Society as a Source of Morality

听听听听听听听听听听听 Several participants defined morality as coming from socialization and societal norms, and therefore concluded that morality exists as a socially constructed fact existing independently of religion. Aria emphasized this point saying: 鈥淩eligious morals are socially constructed as well as the morals that guide the rest of society, so we can all easily agree that like, you know, harming someone is wrong.鈥� In this way, Aria argued that because morals and norms are socially constructed regardless of whether they are constructed within a religion or not, all people in a society share these basic morals. James had a similar response, saying that morality is 鈥渂ased on both human nature and society鈥檚 norms鈥� and thus, 鈥減eople in general are going to share a lot of what they see to be moral just based on what they grow up with regardless of what their religion is, necessarily.鈥� James continued that he believed in a biological reason for morality, saying:

听听听听听听听听听听听 I think certain things like not killing people- that鈥檚 hardwired into your genes is to feel 听听听听 that way and to be repulsed by certain things. And I think that鈥檚 kind of where a lot of 听 religious people go wrong is they see that morals are derived from some book鈥ut I see 听 that as more of a human trait than anything.

听

Mark also thought of morality as an ingrained human trait saying:

听听听听听听听听听听听 There is morality. There鈥檚 something our consciousness inside of us knows what 听 decision [to make]. It knows the consequences. We鈥檙e very smart people. All of us鈥e 听听听听听听听听听听 have an ability to know. Do I think that there is an intrinsic鈥ight and wrong in the 听听听听听听听听 universe? No. I don鈥檛. I don鈥檛 think that there鈥檚 a black and white鈥�

听

听听听听听听听听听听听 Through articulating their own senses of morality and sources of morality, the participants coped with stigma attached to their identity as atheists and the stereotype that atheists do not have a sense of morality. Through negotiating personal theories of morality derived outside of religion, participants were able to conceptualize themselves as moral beings in order to contradict negative attitudes toward them held by family members, friends, and society. Some atheists even implied that morals that came from religious texts and beliefs were actually morally flawed and that it benefited individuals to look outside religion in order to be truly moral. In these ways, atheists were able to posit themselves as altruistic, moral beings while asserting their identity as atheists.

听

CHAPTER 5: DISCUSSION

听

听听听听听听听听听听听 Atheists have historically been an invisible 鈥渞eligious鈥� minority not only in society but in the field of sociology and within the subfield of the sociology of religion. However, the body of literature that documents the lived experiences of atheists and the different pathways to an atheist identity has expanded in recent years, including empirical studies on college atheists (see Bowman and Small 2010; Siner 2007; Nash 2010), the identity formation of atheists involved in atheist organizations (see Smith 2010; LeDrew 2013), and the collective identity work and stigma coping mechanisms atheists employ (Smith 2013). As existing studies have had a tendency to focus on group identities or to be comparative (comparing atheist identity formation to religious identity formation), this thesis research fills a gap in the extant research by placing the individual college atheist as the focal point in his or her own right. Due to their more private nature outside of a group context, individual processes of atheist identity formation and experiences with an atheist identity are rarely captured and are thus my main contribution to the extant literature. My findings both corroborate and contradict some of the existing literature in this field, indicating the need for further research on college atheists. In this chapter, I discuss the most prominent findings of my study and how they relate to or can be added to the existing literature.

Study Limitations

听听听听听听听听听听听 Although this research fills a gap in the missing literature on atheist identities by focusing specifically on a college population, there are several limitations of this study that should be considered. First, the setting in which the study was conducted seemed to be a supportive context in which most students felt comfortable with their atheist identities. As mentioned previously, the city in which the study took place is highly affluent, liberal, and educated. It also has a very low rate of religiosity. These contextual factors likely have a reciprocal relationship with the types of people who choose to study at the University of Colorado. A liberal university is likely to attract liberal students who in turn, influence the culture of the city. College students in Boulder also tend to be more privileged in terms of life chances and affluence, which could also bias results. Further, many of my participants cited feeling a high rate of acceptance in Boulder, but stated that they may feel otherwise if they lived in a different city, in the South for example. Further, the methods are limited by the fact that interviews have their own specific way of explaining the social world through a person鈥檚 perspective at one point in time. Responses to questions may vary based on the audience of the informant, the informant鈥檚 state of mind, and other external factors. A final limitation to this study is that respondents who were raised in a religious household were all raised in Christian households. Experiences with adopting an atheist identity after rejecting a religion other than Christianity may be an area that could be studied further. Future research should include Jewish, Muslim, and other religiously influenced college students. 听

Despite these limitations, my study provides an unprecedented exploration of the process by which college students identify as atheists and their experiences with atheism. The findings contribute to the extant research on atheism and on college students.

I. Backgrounds

听听听听听听听听听听听 As suggested by Hunsberger (2000), although many of the atheists that I interviewed came from religious homes, there was a high incidence of at least one parent showing a level of relative disinterest in the emphasis of religion in the lives of respondents. This tended to be the father figure who might acquiesce to the mother鈥檚 desire to attend church. Many of these 鈥渁cquiescing鈥� parental figures would eventually stop making the child go to church if resistance was perceived, or as seen in Edward鈥檚, Dillon鈥檚, and Matthew鈥檚 cases, family church attendance was simply discontinued. This reflects the findings of LeDrew who both suggests that children growing up in less religious homes may receive more of a secular socialization, and that some participants go through a 鈥渄iscovery鈥� process of their atheist identity rather than a deconversion from a high level of religious belief (2013:8).

听听听听听听听听听听听 This is not to suggest that even in more secular homes, respondents were not expected to believe in something, rather, the desire of parents for their children to choose to believe was very high. Some parents allowed their child to explore other religions and find their own way, but this freedom was allowed with the expectation that respondents would settle on a religious belief. Indeed, many participants felt this pressure believe in something, with most defaulting to Christianity (some in belief and practice, and some only in practice) because it was common in their communities and peer groups, because they celebrated Christmas, and so on. This suggests the extent to which Christianity saturates the culture of the United States. Adopting Christianity 鈥渂y default鈥� even among participants raised in relatively secular homes is consistent with the findings of Smith (2010:7). In terms of religious practice, levels were fairly high among my participants when they were children. This was due to the influence of parents (and often, grandparents) who required church attendance for practical reasons or as being a member of the family as well as the presence of friends and a social life centered in the church community.

II. Pathways to Atheism

听听听听听听听听听听听 A very important part of adopting an atheist identity was a questioning period as reported by many of my participants raised in a Christian household or who reported having some level of belief in God.[13] In contrast to previous findings that many atheists went through this phase upon entering college (Smith 2010; LeDrew 2013), almost all of my participants experienced this questioning phase sometime before college, most commonly in middle school and high school. All but three participants (Bethany, James, and Brian) declared themselves atheist before entering college. This is an important finding that contradicts the stereotype of college as an institutional moment for identity change, leading one to believe that as Regnerus and Uecker pointed out, many atheists are already on the track to adopting this identity before attending a university (2007:3). However, it is clear that in some cases, college does play a role in crystallizing an atheist identity. This could be through classes where students are encouraged to question their beliefs (such as the philosophy class that Bethany took her freshman year) or perhaps through participants being in a supportive normative context away from their homes where their atheist identity may not have been as accepted by their families. Many respondents cited Boulder and the academic community at the University of Colorado as being generally accepting places for atheists. Coming into contact with other atheists as roommates and as classmates may play a role in allowing a 鈥渄ormant鈥� atheist identity to emerge and become formally adopted while in college even if the person was already questioning their beliefs and had beliefs that were in line with atheism (such as James).

听听听听听听听听听听 The background effects of participants and the role of college as a force that crystallizes atheist identities rather than an institution that creates them, indicates that more research needs to be done on atheist identities throughout the life-course. As suggested by Schweitzer (2004), people may cycle back to religious practice after exiting the period in their lives associated with exploration and experimentation. However, Schweitzer鈥檚 findings seem to pertain more to religious practice that may be temporarily abandoned during adolescence and college years, while my results suggest an abandonment in religious beliefs. In order to gain a better understanding of atheism throughout the life-course, more longitudinal studies of atheists who reject religious belief and not just practice would add to the conversation being developed by scholars studying atheist identity formation in adolescents and youth (such as Schweitzer 2004; Regnerus and Uecker 2007; Hunsberger 2000).

III. Experiences with an Atheist Identity

听听听听听听听听听听听 As expected, most atheists interviewed had a conception of their own atheism as being defined as a lack of a belief in a higher power. However, there were some interesting themes that emerged where participants often added ideas to their definitions of atheism like stereotypes of atheists as knowledgeable about other religions, open-minded, and critical thinkers. This was an interesting theme that presented itself that raised the idea of atheists having stereotypes of other atheists. It is interesting to note that all of the stereotypes described were positive. This may be in response to commonly negative stereotyping of atheists by some religious people, and may in itself be a coping mechanism on the part of atheists. It may be a way to connect with people who share their 鈥渄is-belief鈥� and thereby creates an 鈥渦s鈥� versus 鈥渢hem鈥� situation where all atheists are seen as fellow stigma-bearers, as Goffman might say (1968). The experiences of stigma will be discussed in more detail later.

听听听听听听听听听听听 Another important finding was related to identity salience. Through my interviews, it seemed as if respondents distinguished between atheism as being part of their identity versus being a belief that had little bearing on their identity. Although a person鈥檚 self comprises multiple identities (Mead 1934), some respondents hesitated to say that atheism was very important to their identity. One of these respondents (Brian) said this was a matter of semantics. Although he said he would never deny being an atheist, he felt that in terms of an identity, skepticism was more central for him personally. For Brian, this term better informed his orientation toward the world, which was predominantly skeptical, questioning, and cerebral. To Brian, atheism was not a term he preferred, but this is not to say that skepticism (and many other beliefs that are typical of atheists) were not important to his identity. In contrast to Brian, other respondents who rated atheism as less important did so because it was just a topic that did not arise in daily life. These tended to be atheists who claimed atheist belief due to a 鈥渄efault鈥� status as a result of being raised in a secular household and never being able to make themselves believe, rather than actively adopting an atheist identity. One participant (Aria) saw atheism as being more similar to political beliefs rather than a part of her 鈥渃ore identity鈥� while another (Karen) also said that making a lack of a belief an important part of her identity seemed 鈥渃ounterintuitive.鈥� It is worth noting that several participants continued to be 鈥渙n the fence鈥� about their beliefs, with two preferring the term 鈥渁gnostic atheist鈥� instead of 鈥渁theist.鈥� If they were not fully ready to adopt this identity, it makes logical sense that this identity would not be extremely salient.

听听听听听听听听听听听 Of the participants who viewed atheism as very salient to their identities, many cited reasons for this being a scientific worldview. Some did not want to reconcile religion and science. Others felt that being an atheist and seeing the world in a more scientific light based on science and reason was simply more rational in their minds.

IV. Perceptions of Stigma

听听听听听听听听听听听 One of the larger implications of my findings is that atheists continue to feel some level of stigma based on their identities as atheists. My participants reported being told that they were immoral by religious people and feeling that they might be discriminated against in the job market if they were open about their identity. Although my findings indicate that atheists continue to feel unaccepted at some level, many participants said that they felt relatively accepted on the 彩民宝典 campus, perhaps contradicting the findings of previous scholars studying religious social climate on college campuses (see Bowman and Small 2010; Nash 2010). Despite citing the university campus as a supportive normative context, many atheists still acknowledged their identity as stigmatized and still felt unaccepted on a societal level. Interestingly, many atheists did not cite many examples of discrimination on an institutional or societal level, but there remained a fear of this type of discrimination and the recognition by many that their identity was still not widely accepted. Matthew and other participants used the concept of 鈥減reference鈥� to articulate that although society may be becoming more accepting, it would still 鈥減refer鈥� that individuals were not atheists.

听听听听听听听听听听听 These findings can be seen as part of a greater conversation about stigmatized identities on and off college campuses. Navigating the world with an embattled identity is not particular to atheists and is very prevalent among members of other minority groups, especially sexual minorities. Concepts of stigma, identity formation, and coping mechanisms have been applied frequently to the LGBT community who are often labeled as sexual deviants by U.S. society and who suffer similar marginalization, discrimination, and lack of access to the same rights that members of the heterosexual majority of U.S. society enjoy (See Kaufman and Johnson 2004; Bernstein 2002; Fassinger 1996). The LGBT community has been compared to the atheist community by several sociologists of religion as well, some even borrowing models of LGBT identity development (Fassinger 1996) and applying it to atheist identity development (Siner 2007; Smith 2010). On college campuses specifically, the experiences of atheists could also be related to other stigmatized religious minorities such as Muslim students. Muslim women on college campuses in particular are more visible if they choose to wear a hijab, and therefore navigate certain social stigmas on U.S. college campuses (See Mir 2009; Cole and Ahmadi 2003).

V. Individual Coping Mechanisms

听听听听听听听听听听听 Whether or not they felt accepted by society or individuals, many participants reported some sort of stigma coping mechanism whether it was surrounding themselves with other like-minded individuals, controlling the amount of information they shared with certain groups of people, and/or negotiating their own sense of morality. The previous research done on managing atheism as a stigmatized identity has focused on other stigmatized groups excluding atheists (Goffman 1968) or on the collective in terms of how atheists seek out groups of people with whom to connect and create community (Smith 2010). To the extent of my knowledge and research, little has been done on how atheists as individuals cope with stigma if they do not necessarily join atheist charities, activist groups, and so on. From my interviews, it was made clear that many respondents did have a way of managing their identities. Friend groups of respondents commonly comprised of mostly other atheists, people who were not very religious, or at the very least, liberally-minded individuals. Respondents explained not wanting to argue about core beliefs all the time and freely admitted to having like-minded people as their closest friends. This serves as evidence for Goffman鈥檚 theories of stigma that suggest that people create in-groups with 鈥渇ellow-sufferers鈥� of the stigma (1968:112). Many respondents cited great enjoyment out of being able to freely discuss any topic they wanted or to share information without feeling socially hindered or judged as a part of these in-groups.

听听听听听听听听听听听 Only sharing information about one鈥檚 atheist identity to a select group of people in one鈥檚 life corroborated Goffman鈥檚 idea of controlling information as a stigma-coping mechanism (1968:95). This strategy was exemplified by participants in this study who avoided openly 鈥渁dvertising鈥� their atheism, especially on social media sites where they feared discrimination from potential employers. Several respondents specifically said that they removed or avoided posting information on Facebook that stated their affiliation with atheist beliefs. Many also added that they do not talk about their atheism in groups of people who are not close friend or family members unless someone else brings it up first. None of the respondents said that they would deny their identities, but most simply chose to keep that particular topic off the table to avoid potential conflict. Thus, acquaintances, employers, and friends outside the participants鈥� inner circles make up the large group to which respondents do not share intimate information that could 鈥渙ther鈥� them in a social situation. The small group with which the respondents chose to openly share their atheism and believes included close friends, family, and intimate partners.

听听听听听听听听听听听 Morality was another key point that respondents continuously negotiated in coping with their identities as atheists. Since morality is the main point of contention between many religious and atheist people, with much of the stigma surrounding atheists arising from the common belief that atheists are rejecting common foundations of being a 鈥済ood鈥� person (and, in the United States, being a good 鈥淎merican鈥�), the idea for my respondents that they were moral beings was crucial. Every single respondent had their own thoughts about morality whether or not they believed that religious people and atheists share the exact same morals. Many thought that regardless of whether or not one is religious, morals come from socially constructed norms in society as well as biological drives to keep the human species alive. Declaring oneself a moral being was important for respondents, many of whom actively rejected, and even showed disdain for morality derived from religion. 听

听

CHAPTER 6: CONCLUSION

听

听听听听听听听听听听听 This study contributes to the understanding of college atheists and their adoption of an identity that remains stigmatized and embattled in the context of a society where being religious continues to be the norm. The participants in my study showed how important background factors are in the adoption of an atheist identity. I found that the role college played in relation to an atheist identity was to crystallize an identity-achieving process already set in motion in an individual鈥檚 childhood or adolescence. The specific campus of 彩民宝典 played a role as a supportive normative context where some atheists were able to finally adopt this identity and where most atheists felt comfortable being relatively open about their atheist identity. For some, atheism was not considered an identity, but rather a set of beliefs that were low in salience. However, most participants perceived atheism as a very important part of their identity. Regardless of identity salience, many atheists that I interviewed continued to perceive a social stigma, and whether or not direct discrimination based on this identity, it was still feared. It is unfortunate that some atheists feel that they have to censor themselves at any level for fear of experiencing discrimination or proselytizing. Although many religious people may mean well when they make comments about praying for an atheist or wanting them to come to church, for many atheists, this is offensive. It negates an atheist identity as legitimate and assumes that atheist people have an identity that is 鈥渓ess than鈥� a Christian or other religious identity. It also assumes that atheists are not moral beings, when the findings of my study show the exact opposite. Atheists can be just as moral as religious individuals. In fact, many of the atheists in my study stated that they greatly enjoy contemplating morality and its implications. Thus, it is time for U.S. society to dispel these negative stereotypes of atheists as rejecting morality and to see atheists as just as moral and just as 鈥淎merican鈥� as any other religious individual. This need extends to families of atheists, political parties, employers, and universities. Atheists should be able to live in a society alongside Christians and other religious groups without feeling uncomfortable or stigmatized. It is time to hear the atheist point of view without clouding the social climate with fear, stereotypes, and judgments.

听

听听听听听听听听听听听听听听听听听听听听听听

听

REFERENCES

Arkansas Constitution, Article 19, Section 1.

Ayres, Ian and William Eskridge. 2014 鈥淯.S. hypocrisy over Russia鈥檚 anti-gay laws.鈥� The 听听听听听听听听听听 Washington Post. Jan. 31. Retrieved March 10, 2015 ( 听听 opinions/us-hypocrisy-over-russias-anti-gay-laws/2014/01/31/3df0baf0-8548-11e3-9dd4-听听听听听听听听听听听 e7278db80d86_story.html).

Bainbridge, William S. 2009. "Atheism."听Pp. 319-335 in The Oxford Handbook of the Sociology 听 of Religion, edited by P.M. Clarke. Oxford: Oxford University Press.

Beit-Hallahmi, Benjamin. 2011. 鈥淎theists: A Psychological Profile.鈥� Pp. 300-313 in The 听 Cambridge Companion to Atheism, edited by M. Martin. New York: Cambridge 听听听听听听听 University Press.

Bernstein, Mary. 2002. 鈥淚dentities and Politics: Toward a Historical Understanding of the 听听听听听听听听听听 Lesbian and Gay Movement.鈥� Social Science History 26(3): 531-581.

Bowman, Nicholas A. and Jenny L. Small. 2010. 鈥淒o College Students Who Identify with a 听听听听听听听 Privileged Religion Experience Greater Spiritual Development? Exploring Individual and 听听听听听听听 Institutional Dynamics.鈥� Research in Higher Education 51(7): 595-614.

听听听听听听听听听听听 doi:10.1007/s11162-010-9175-2.

Bulger, Matthew. 2013. 鈥淯nelectable Atheists: U.S. States that Prohibit Godless Americans from 听听听听听听听听听听 Holding Public Office.鈥� American Humanist Association. Retrieved Feb. 8, 2015

听听听听听听听听听听听 (听听听听 prohibit-godless#comment-1108615841).听

Campbell, Colin. 1972. Toward a Sociology of Irreligion. New York: Herder and Herder.

Caplovitz, David and Fred Sherrow. 1977. The Religious Drop-Outs: Apostasy Among College 听听听 Graduates. Beverly Hills, CA: Sage.

Carriere-Kretschmer, Erin and Kathleen Holzwart. 2008. 鈥淐hapter 2. Religiosity.鈥� Pp. 18-24 in 听鈥淯nfavorable Views of Jews and Muslims on the Increase in Europe.鈥� The Pew Global 听 Attitudes Project. Pew Research Center. Washington D.C. Retrieved Oct. 30, 2014 听听听听听听听听听听听 ().

Cline, Austin. 2015. 鈥淒efinition of Irreligion, Irreligious.鈥� About Religion. Retrieved Feb. 9,

听听听听听听听听听听听 2015 (听听听听听听听 Definition.htm).

Cole, Darnell and Shafiqa Ahmadi. 2003. 鈥淧erspectives and Experiences of Muslim Women 听听听听听听听听 Who Veil on College Campuses.鈥� Journal of College Student Development 44(1): 47-66.

Cooperman, Alan, Greg Smith, Besheer Mohamed, Jessica Martinez, Elizabeth Sciupac, 听 Katherine Ritchey. 2014. 鈥淗ow Americans Feel About Religious Groups.鈥� Religion and 听听听听听听听听听听 Public Life Project. Pew Research Center. Washington D.C.: Pew Research Center. 听听听听听听听听听听 Retrieved Oct. 28, 2014 (听听听听听听听听听听听 about-religious-groups/).听

Dawkins, Richard. 2013. "The Out Campaign."听Retrieved Nov. 5 2013. ().

Edgell, Penny, Joseph Gerteis, and Douglas Hartmann. 2006. 鈥淎theists As "Other": Moral 听听听听听听听听 Boundaries and Cultural Membership in American Society.鈥� American Sociological

听听听听听听听听听听听 Review 71(2): 211-234.

Fassinger, Ruth E. 1998. 鈥淟esbian, Gay, and Bisexual Identity and Student Development Theory.鈥� Pp 13-22 in Working with Lesbian, Gay, Bisexual, and Transgender College Students, edited by R.L. Sanlo. Westpor, CT: Greenwood Press.

Florida, Richard. 2013. 鈥淎merica鈥檚 Most (and Least) Religious Metro Areas.鈥� The Atlantic.

听听听听听听听听听听听 Retrieved Jan. 3, 2015 (

听听听听听听听听听听听 least-religious-metro-areas/5180/).

Frank, Adam. 2014 鈥淒oes Being 鈥楽piritual But Not Religious鈥� Really Mean Anything?鈥� 听 National Public Radio. Oct. 21. Retrieved Feb. 23, 2015 听听听听听听听听听听听 (听听听听听听听 religious-really-mean-anything).

Gervais, Will M., and Ara Norenzayan. 2012. "Reminders of Secular Authority Reduce 听听 Believers' Distrust of Atheists."听Psychological Science听23 (5): 583-91. Retrieved Oct. 听听听听听听听听听 17, 2013. doi: 10.1177/0956797611429711.

Goffman, Erving. 1968. Stigma: Notes on the Management of Spoiled Identity. New York, NY: 听听听听 Simon & Schuster, Inc.

Greenberg, Neil. 2008. Can Spirituality be Defined? University of Tennessee. University Studies Interdisciplinary Colloquy on Spirituality and Critical Inquiry. Retrieved Feb. 23, 2015 听听听听听听 ( 58d14a852573b00072525d).

Hamblin, James. 2013. 鈥淏ullied for Not Believing in God.鈥� The Atlantic. Sept. 13. Retrieved on 听听 March 13, 2015 (听听听听听听听听听听听 believing-in-god/279095/).

Hill, Chris and Peter Sprigg. 2013. 鈥淪lippery slope to accepting atheist Boy Scouts.鈥� USA 听听听听听听听听听听 Today. May 25. Retrieved March 13, 2015 (听听听听听听听 /05/25/boy-scouts-gay-god-column/2357959/).

Hunsberger, Bruce. 2000. 鈥淪wimming against the current: exceptional cases of apostates and 听听听听听听听 converts.鈥� Pp. 233-247 in Joining and Leaving Religion: Research Perspectives, edited 听听听听听听听听听听听 by L.J. Francis and Y.J. Katz. Herefordshire, England: Redwood Books.

Kaufman, Joanne M. and Cathryn Johnson. 2004. 鈥淪tigmatized Individuals and the Process of 听听听 Identity.鈥� The Sociological Quarterly 45(4): 807-833.

LeDrew, Stephen. 2013. 鈥淒iscovering Atheism: Heterogeneity in Trajectories to Atheist Identity and Activism.鈥� Sociology of Religion: 1-23.

Lipka, Michael. 2013. 鈥�5 facts about atheists.鈥� Pew Research Center. Retrieved Nov. 13, 2014 听听 ().

Mead, George Herbert.听1934. Mind, Self, and Society: From the Standpoint of a Social 听听听听听 Behaviorist. Edited by C.W. Morris. Chicago, IL: University of Chicago Press.

Mir, Shabana. 2009. 鈥� 鈥楴ot Too College-Like,鈥� Not Too Normal: American Muslim 听听听听听听听听 Undergraduate Women鈥檚 Gendered Discourses.鈥� Anthropology and Education Quarterly 听听听听听听 40(3): 237-256.

Moreno, Carlos S. 2014. 鈥淎n atheist for Congress?鈥� CNN. Sept. 1. Retrieved Feb. 8, 2014

听听听听听听听听听听听 ().

Morris, R.C. 2013. 鈥淚dentity Salience and Identity Importance in Identity Theory.鈥� Current 听听听听听听 Research in Social Psychology: 23-36. Retrieved Nov. 7, 2014听听 ().

Nash, Robert J. 2010. 鈥淚nviting Atheists to the Table: A Modest Proposal for Higher Education.鈥� 听听听听听听听听听听 Religion and Education 30(1): 1-23. doi:10.1080/15507394.2003.10012315.

Regnerus, Mark E. and Jeremy E. Uecker. 2007. "How Corrosive is College to Religious 听 Faith 听听 and Practice?鈥� Social Science Research Council: 1鈥�6.

Sahgal, Neha. 2009. 鈥淭he Religious Affiliations of U.S. Presidents.鈥� Pew Research Center: 听听听听听听听听听 Religion and Public Life. Jan 15. Retrieved March 5, 2015. 听听听听听听听听听听听 ().

Schweitzer, Friedrich L. 2004. The Postmodern Life Cycle: Challenges for Church and 听听听听 Theology. Danvers, MA: Chalice Press.

Seifert, Tricia. 2007. 鈥淯nderstanding Christian Privilege: Managing the Tensions of Spiritual 听听听听听听 Plurality.鈥� About Campus. 10-17.

Siner, Sam. "A Theory of Atheist Student Identity Development."听 2007.听 Journal of the 听 Indiana University Student Personnel Association: 14-21.

Smith, Jesse M. 2010. "Becoming an Atheist in America: Constructing Identity and Meaning 听听听听听听 From a Rejection of Theism."听Sociology of Religion. 72(2): 1-23.

Smith, Jesse M. 2013. 鈥淐reating a Godless Community: The Collective Identity Work of 听听听听听听听听听听听 Contemporary American Atheists.鈥� Journal for the Scientific Study of Religion. (52): 80-听 99.

彩民宝典. 2015. 鈥淩eligious Campus Organizations.鈥� Retrieved March 11, 2015 ().

鈥淯niversity of Colorado-Boulder. 2015. U.S. News and World Report. Retrieved March 19, 2015

听听听听听听听听听听听 ().

U.S. Constitution, Article 6, Section 3.

听

听听听听听听听听听听听

听

APPENDIX A: Recruitment Material

Email Advertisement text (Body of the text will also be used for class presentations)

听

Subject: Opportunity to be involved in a research study

听

Body: University of Colorado senior is performing a study regarding atheist identity formation. If you identify as an atheist and are a CU student between the ages of 18-24, please contact to learn more about the study and consider if you would be interested in being interviewed. Interviews will likely last between a half hour to an hour and will cover your personal experiences and views regarding atheism and religion. Participants will be compensated with a $10 gift card.

听

Contact information: email: atheist.study@gmail.com

Phone: 970-333-4618

听

Flyer 1

1. Are you a Godless Heathen?

2. Do you want to be part of a research study?

3.听 Are you a CU student between the ages of 18 and 24?

4. Could you use $10?

听

If you answered 鈥測es鈥� to the above questions, I want to talk to you! I鈥檓 a CU senior conducting a sociological study involving 30-minute one-on-one interviews regarding Atheist identity formation for my honors research project. If you are interested in learning more about my study and possibly participating, please call 970-333-4618 or email atheist.study@gmail.com.

听

Participants will be compensated with a $10 gift card 听

听

Flyer 2

Don鈥檛 Believe in God?

Get paid to share your personal experiences as an Atheist.

A senior at CU is conducting a sociological study involving 1/2 -1 hour one-on-one interviews regarding Atheist identity formation. If you are a CU student between the ages of 18 and 24 and identify as an Atheist, please call 970-333-4618 or email atheist.study@gmail.com to learn more about the study and see if you鈥檇 like to participate.

Participants will be compensated with a $10 gift card to your choice of Starbucks or Amazon.com.

听

APPENDIX B: Consent Form (HRP-502)

Title of research study: The Atheist Experience: A Sociological Approach to College Students鈥� Identity and Identity Salience

Investigator:

Why am I being invited to take part in a research study?

I invite you to take part in a research study because I would like to understand how college students who identify as atheist come to adopt this identity and how they experience it, for example, do they feel there is stigma associated with being an atheist? There is little research conducted on this particular topic as pertaining to college students and as a result, they will be the focus of my study.

What should I know about a research study?

I will explain this research study to you.
Whether or not you take part is up to you.
You can choose not to take part.
You can agree to take part and later change your mind.
Your decision will not be held against you.
You can ask all the questions you want before you decide.
You can skip any of the questions or decide to quit your participation at any time.

Who can I talk to?

If you have questions, concerns, or complaints, or think the research has hurt you, talk to the research team at 970-333-4618. If you would rather speak to someone other than the primary researcher, you can contact Dr. Joanne Belknap at joanne.belknap@colorado.edu.

This research has been reviewed and approved by an Institutional Review Board (鈥淚RB鈥�). You may talk to them at (303) 735-3702 or irbadmin@colorado.edu if:

Your questions, concerns, or complaints are not being answered by the research team.
You cannot reach the research team.
You want to talk to someone besides the research team.
You have questions about your rights as a research subject.
You want to get information or provide input about this research.

Why is this research being done?

This research is being done to better understand why college students identify as atheist and to identify the pathways there are to adopting or maintaining this identity. This research, ideally, will better explain concepts of identity and especially stigmatized identity as many groups and individuals in the United States view atheists as immoral and 鈥渦n-American.鈥� The goal of this study is to advance the understanding of an atheist identity formation, identity salience, and whether stigmatizing is something atheists experience and if so, how.

How long will the research last?

I expect that the interview will take approximately one-half to one hour, but it may be longer if you want to answer the questions in more depth.

How many people will be studied?

I hope to interview between 10 and 30 people in this study.

What happens if I say yes, I want to be in this research?

You will be interviewed by a senior studying Sociology at 彩民宝典. The interview will last from a half-hour to one hour, depending on the depth of your responses and will be scheduled taking into account both your and my schedules. A series of questions will be asked about your experiences of being or becoming an atheist. Questions will include background information about religious experiences, deconversion experiences, and reasons for becoming (or remaining) an atheist. If you give your consent, a recording device may be used to make data collection more accurate. If you would like to participate but not have your interview recorded, I will not record it. The interview will be performed wherever you feel the most comfortable (within reason) with recommended sites being library study rooms or a coffee shop. All personal information that participants provide will be kept confidential. Names and other identifying information will be changed. The study findings will not be written in a manner that could identify any of the participants.

What happens if I do not want to be in this research?

You can leave the research at any time with no repercussions. If you decide to leave before the interview is complete (with the exception of skipping a few questions), no compensation will be given.

What happens if I say yes, but I change my mind later?

You can leave the research at any time and it will not be held against you.

Is there any way being in this study could be bad for me?

If it is difficult for you to talk about your atheist identity, your pathway to atheism, and/or response to your atheist identity by family, friends, and so on, it is possible that you will feel uncomfortable during the interview. For example, you will be asked about reasons for leaving a previous religion (if applicable) and whether as an atheist you have experienced any discrimination.

Will being in this study help me any way?

Other than the $10 gift card, I cannot promise any benefits to you or others from your taking part in this research. However, possible benefits include being able to talk openly about a commonly stigmatized identity and understanding your own identity. Ideally, I will publish my findings in order to broaden the knowledge base and improve the understanding of atheists, particularly college student atheists.

What happens to the information collected for the research?

I will make every effort to ensure that your participation is confidential and anonymous. You can tell people you participated, but I will not. The only individuals that could possibly have access to the data are members of the University of Colorado-Boulder Institutional Review Board.

Information collected will be used to discover trends and better explain atheism as a sociological phenomenon. Your information will be used to write a formal research paper that will be presented as an honors thesis to a committee of faculty. During the study, your information will be kept in a secure location. Your real name will be replaced with a number and later (for publishing purposes or for the final research paper) you will be given a fictitious name. Any recordings and other documents with personal information will be destroyed after the data have been collected.

What else do I need to know?

If you agree to take part in this study, you will be compensated with a $10 gift card for your time and effort. No amount will be paid if the interview is abandoned without being completed (unless only a few questions are skipped).

Signature Block for Capable Adult

Your signature documents your permission to take part in this research.
听	听	听
Signature of subject	听	Date
听	听
Printed name of subject	听
听	听	听
Signature of person obtaining consent	听	Date
听	听	听
Printed name of person obtaining consent	听	IRB Approval Date

听

Permission to be Audio-Recorded

听

**not required**

听

Your signature documents your permission to be audio-recorded as part of this research.
听	听	听
Signature of subject	听	Date
听	听
Printed name of subject	听
听	听	听
Signature of person obtaining consent	听	Date
听	听	听
Printed name of person obtaining consent	听	IRB Approval Date

听

APPENDIX C: Interview Schedule (HRP-503: Protocol)

The Atheist Experience:

A Sociological Approach to College Students鈥� Identity and Identity Salience

听

Interview Schedule

听

Section I: Demographics

听

1. Participant Name:

Recruitment Method:

听

2. Age:

听

3. Year in School:

听

4. Field of Study/Major:

听

Section II: Background Information

听

5. Tell me a little bit about your background (family, siblings, etc.) Where did you grow up? (neighborhood, type of house, etc.)

听

6.听 What was it that your parents/guardians do/did for a living? (Note if parents are still together or divorced)

听

7. Did you grow up religious? If yes, what religion? If no, see below (#8. A-E)

听

8. What was your religious practice frequency; how often did you attend religious functions, worship, and so on?

听

9.听 On a scale of 1 to 10 with 1 being not at all religious and 10 being extremely religious, what number would be the highest you would say describing you over your lifetime when you were the most religious?听 Tell me more about this鈥攈ow old were you, what do you think made you the most religious then?

听

10. Growing up, what was your general attitude towards your religion? (positive, negative, indifferent) What were your experiences as a member of your church, synagogue, etc.?

听

11. If you adhered to your family鈥檚 religion/religious practices, why would you say that you did so?

听

12. What was your attitude towards people of other religions when you were growing up?

听

13. What was your attitude towards atheists when you were growing up?

听

Section III: Experiences with Atheism

听

14. What does it mean to you to be an atheist? (Prompt: Can you give your definition?)

听

15. What were your first experiences with atheism? How old were you?

听

16. Why would you say that you identify as an atheist now?

听

17. What has been your general attitude towards atheists/atheism now that you identify as an atheist?

听

18. Do you feel like as an atheist your identity is under attack?

听

19. Does your family know you鈥檙e an atheist?听 If not, why? If so, how did/does your family react to your atheism?

听

20. Do most of your friends, co-workers, and other people in your life know you鈥檙e an atheist? If not, why?听 If so, how do they react (to your atheism)?

听

21. Do you feel accepted by society as an atheist? If yes, in what ways, and if not, in what ways don鈥檛 you feel accepted?

听

22. What are your views about other religions or religious people?

听

23. On a scale of 1 to 10, with 1 being not at all important and 10 being extremely important, how important is your atheism to you?听 Can you explain why?

听

24. Does your atheism affect your personal relationships? If so, how?

听

25. Does your atheism affect the way you see the world? If so, how?

听

26. What about morality? What is your view of morality from an atheistic standpoint?

听

27. Are you a part of any atheist groups/organizations?

听

Part IV: Closing

听

28. Thank you for taking part in my interview.听 Are there any questions you wish I鈥檇 asked? Do you have any more to add that wasn鈥檛 captured in the questions I asked?听

听

29. Finally, little research has been conducted on atheists.听 Can you tell me why you agreed to participate in this research study?

听

30. Do you have any friends who you think may be interested in this study? If so, would you be willing to give them my contact information?

**If person grew up atheist, refer to questions below:

听

8. A. What was your attitude towards other religions/religious people?

听

B. Did the religion of others affect you in any way? If so, how?

听

C. Did your parents stress their nonbelief/atheism? If so, how?

听

E. Did your atheism ever become a very important part of your life while you were growing up? If so, how?

听

**Return to question #15

听

APPENDIX D: Religious Information of Participants

Name	Age	Major	Most Religious Point (1-10)	Childhood Religious Denomination	Age of Adopting Atheist Identity	Atheist Identity Salience
Edward	20	Aerospace Engineering and Engineering Physics	3-4 (age 4 or 5)	Lutheran	12^th grade	8
Cora	20	Math and E-Bio	___________	_________________	8^th grade	8-9
Jennifer	21	E-Bio and Secondary Science Education	3-4 (up to age 13 or 14)	United Church of Christ (Methodist)	High School	10
Dillon	21	Environmental Studies and Geology	___________	Christian	High School	(didn鈥檛 provide number) Low Salience
James	21	MCDB and Political Science	___________ (Described normative belief at an early age)	______________	Senior Year of High School-Freshman Year College	7
Karen	18	Linguistics and German	(Described normative belief at an early age)	________________	8^th or 9^th grade (end off middle school/beginning of high school)	3
Aria	23	Women鈥檚 Studies and Anthropology	5 (age 8)	Old Apostolic Lutheran and Quakerism	4^th or 5^th grade	3
Danielle	18	Environmental Design	鈥淚n between鈥� (very young age)	Lutheran	10^th grade	7
Mark	18	History	7 (before 7^th grade)	Roman Catholic	10^th grade	Very Important (No number provided)
Bethany	18	Anthropology	8 (through middle school)	Non-denominational Christian (Conservative)	Freshman Year of College	4
Matthew	18	Biological and Chemical Engineering	________	Christian	Early Middle School	Important (No number provided)
Brian	22	Calculus, Physics, Astronomy	5 (up to sophomore year of college)	(Presbyterian)	Junior Year of College	Atheism:2 Skepticism:10
Julie	19	Computer Sciences	Low belief, but high practice (up to 5^th or 6^th grade)	Christian	Age 12-13	8-9
Sophie	18	IPHY and Neuroscience	__________	Catholic school, but no religion in home	鈥渁lways an atheist鈥�	7

听

[1] In my own interviews, it was clear that some atheists were still in a 鈥渟eeking鈥� phase, or they preferred other labels since their own views were nuanced and evolving. Some participants even referred to themselves as 鈥渁gnostic atheists.鈥�

[2] A note on spirituality: It is a widely held misconception that atheists cannot be spiritual. Spirituality does not necessarily require a belief in a higher being. Therefore, spiritual people can indeed be atheists and vice versa. According to Greenberg (2008), spirituality can be defined as involving 鈥渇eelings of significance, unity, awe, joy, acceptance, and consolation. Such feelings are intrinsically rewarding and so are sought in their own right, but they also help us in dealing with different situations involving death, loss, and disappointment.鈥� Sam Harris, a famous neuroscientist, philosopher, and author explained: 鈥渟pirituality begins with a reverence for the ordinary that can lead us to insights and experiences that are anything but ordinary鈥� (Harris as cited by Frank 2014).

[3] Although Goffman never applied his work on stigma to atheism and atheists, it is certainly relevant for this stigmatized group. Smith (2010:15) also uses Goffman鈥檚 concept of 鈥渟poiled identities鈥� (stigmatized identities) in reference to atheists as a group, but to my knowledge, this thesis along with Smith鈥檚 work are the first works to apply Goffman鈥檚 concepts to atheism.

[4] It is worth noting that the United States has never seen an atheist president, nor a president with a religious faith that falls outside of Christianity. Over half of the presidents have been either Episcopalian or Presbyterian, with other Protestant groups represented (Sahgal 2009). To date, there have been no Jewish, Muslim, Buddhist, or other religious tradition represented in the White House. This is perhaps telling of how deeply ingrained Christianity is in U.S. society.

[5] All students have the Christmas holiday off from school, but the Christian ethos appears in that this holiday is formally recognized in the academic calendar.

[6] Interestingly, the campus culture at the 彩民宝典 (the context of this study) supports Nash鈥檚 idea. 彩民宝典 is home to over 30 religious or spiritual clubs and groups available to students. The vast majority of these clubs are Christian (彩民宝典 2015). There is only one secular student group on the 彩民宝典 campus.听

[7] 鈥淭heophile鈥� is a term used by Nash to refer to a 鈥淕od-loving鈥� culture that permeates U.S. society in general as well as university campuses (Nash 2010:3).

[8] This is a club for students who are atheists, agnostics, skeptics, and so on. Members of the club meet to converse about recent scientific findings and to provide an accepting place for people who are not religious (though not necessarily atheist).

[9] This student-run organization publishes University of Colorado undergraduate work in all fields including the natural sciences, social sciences, humanities, art, poetry, fiction, and creative nonfiction. The names of the student-run clubs are omitted to protect the privacy of participants.

[10] When analyzing interview responses, stigma was operationalized by the perceptions of an atheist identity being under attack and the perceptions of acceptance of this identity by friends, family, and other people in the participants鈥� lives. The exact questions were: 18. 鈥淒o you feel like as an atheist your identity is under attack?鈥�; 19. 鈥淒oes your family know you鈥檙e an atheist? If not, why? If so, how did/does your family react (to your atheism)?鈥�; 20. 鈥淒o most of your

friends, co-workers, and other people in your life know you鈥檙e an atheist? If not, why?听 If so, how do they react (to your atheism)?鈥�; 21. Do you feel accepted by society as an atheist? If yes, in what ways, and if not, in what ways don鈥檛 you feel accepted?鈥�

听

[11] See Appendix C for a table with childhood religious information for each participant.

[12] It is worth noting that several of the participants (Cora, Sophie, and Karen) adopted an atheist identity less through a process, but more as a 鈥渄efault鈥� status. Cora, who never had any level of religious belief nor a religious upbringing, actively adopted and reinforced her atheist background and identity, while Sophie and Karen described never being able to make themselves believe. Karen said: 鈥渏ust learning about other religious traditions, none of them really seem super compelling to me. And maybe I鈥檒l find on that seems really legitimate to me, but it鈥檚 just like, none of them seem compelling for the questions that I have.鈥� Sophie had a very similar comment saying: 鈥淪o far, since I haven鈥檛 run into any religions that I do believe in, for the most part, I鈥檓 an atheist.鈥�

[13]For a complete list of the participants and their childhood religious beliefs, denominations, and so on, see Appendix D.

彩民宝典

Search

Other ways to search:

Breadcrumb