Living in Sodom's Shadow:Essays on Attitudes Towards Gay Men And

Living in Sodom’s Shadow: Essays on attitudes towards gay men and lesbians in the Commonwealth Caribbean

A thesis submitted to the University of Manchester for the degree of PhD in Applied Social Research in the Faculty of Humanities

2016

MAHALIA JACKMAN Cathie Marsh Institute for Social Research School of Social Sciences

~This page was intentionally left blank

Table of Contents

List of Figures ...... 8

Figures for Appendix B ...... 8

List of Tables ...... 9

Tables for Appendix B ...... 10

Tables for Appendix C ...... 10

Abbreviations ...... 11

Abstract ...... 12

Declaration ...... 13

Dedication ...... 15

Acknowledgements...... 16

Published Work ...... 17

Chapter 1 : Introductory remarks ...... 18

1.1 Significance of the study ...... 25

1.2 Objectives and research questions ...... 26

1.2.1 Public opinions about gay rights ...... 26

1.2.2 Analysis of anti-gay rhetoric ...... 29

1.3 Thesis structure ...... 31

Chapter 2 : Review of the related literature ...... 32

2.1 Negative attitudes towards lesbians and gay men ...... 33

2.1.1 Functionalist approach...... 33

2.1.2 The perceived threat of homosexuality ...... 35

2.2 Individual-level correlates of anti-gay prejudice ...... 36

2.2.1 Demographic variables ...... 37

2.2.2 Religion ...... 40

2.2.3 Interpersonal contact ...... 42

2.2.3 Beliefs about the origins of homosexuality ...... 44

2.2.4 Attitudes towards other groups ...... 45

2.3 Beyond the individual – Contextual effects ...... 46

2.3.1 Socioeconomic development ...... 47

2.3.2 Religious culture ...... 48

2.3.3 Laws ...... 50

2.4 Attitudes towards homosexuality in the Commonwealth Caribbean ...... 53

2.5 Summary ...... 59

Chapter 3 : The socio-legal situation of sexual minorities living in the Commonwealth Caribbean: Some stylised facts ...... 61

3.1 State-sponsored discrimination ...... 61

3.2 Levels of socioeconomic development ...... 68

3.3 The religious climate ...... 70

3.4 The policing of sexual minorities in the media and popular culture ...... 77

3.5 Support services for sexual minorities ...... 82

3.6 Summary ...... 83

Chapter 4 : Views on the ‘abominable’ crime: Heterosexual support for and perceived usefulness of the anti-gay laws in Barbados, Guyana and Trinidad and Tobago ...... 85

4.1 Heterosexual attitudes towards the criminalisation of same-sex intimacy in Barbados, Guyana and Trinidad and Tobago ...... 86

4.1.1 General support for bans on same-sex intimacy...... 88

4.1.2 Perceived usefulness of the anti-gay laws...... 93

4.2 Who is likely to support the anti-gay laws? ...... 95

4.2.1 Description of dependent variables ...... 95

4.2.2 Description of independent variables ...... 96

4.2.2.1 Demographics (gender, age, education and country of residence) ..... 98

4.2.2.2 Religion ...... 98

4.2.2.3 Beliefs about the origins of homosexuality ...... 99

4.2.2.4 Interpersonal contact ...... 99

4.2.2.5 Discrimination against other groups ...... 100

4.2.3 Statistical method ...... 101

4.3 Results ...... 107

4.3.1 Factors influencing heterosexual support for the criminalisation of same- sex intimacy ...... 107

4.3.1.1 Consistent and mixed support for the criminalisation of same-sex intimacy ...... 113

4.3.1.2 The unsure/prefer not to say responses to questions about banning same-sex intimacy ...... 116

4.3.2 Subscriptions to the notions that the current laws protect the fabric of society ...... 121

4.3.2.1 Laws achieve all objectives ...... 125

4.3.2.2 The ‘unsure/won’t say’ response when asked about the usefulness of the laws ...... 127

4.4 Discussion ...... 131

Chapter 5 : The Commonwealth Caribbean in context: Comparing attitudes towards same-sex marriage across the Americas ...... 138

5.1 Attitudes towards same-sex marriage in the Americas: Is the Commonwealth Caribbean less tolerant? ...... 138

5.2 Exploring the differences in attitudes towards same-sex marriage in the Americas ...... 143

5.2.1 Socioeconomic development ...... 143

5.2.2 Religious Climate ...... 146

5.2.3 Laws ...... 148

5.3 A multi-level analysis of attitudes ...... 149

5.3.1 Statistical method ...... 149

5.3.2 Description of the variables ...... 150

5.4 Results ...... 152

5.4.1 The impact of the age and religion: Does context matter? ...... 160

5.5 Discussion ...... 166

Chapter 6 : ‘Man to man, gyal to gyal…dats wrong’: A qualitative analysis of anti-gay rhetoric in Caribbean [Jamaican] popular culture ...... 170

6.1 Music Trends ...... 171

6.2 Thematic analysis of the music ...... 174

6.2.1 Homosexuality is a violation of gender norms ...... 176

6.2.2 Homosexuality is a sin ...... 178

6.2.3 Homosexuality is unnatural ...... 181

6.2.4 Homosexuality is a threat to the fabric of society ...... 183

6.2.5 Homosexuality as a foreign concept and intolerance is part of Jamaican identity ...... 184

6.2.6 Combinations of themes ...... 185

6.3 Discussion ...... 186

Chapter 7 : Summary and concluding remarks...... 190

7.1 Key findings ...... 191

7.2 Policy implications ...... 194

7.3 Limitations and areas for future research ...... 197

7.4 Conclusion ...... 199

References ...... 201

Appendix A – Appendix to Chapter 3 ...... 220

Antigua and Barbuda ...... 220

Barbados ...... 221

Belize ...... 222

Dominica ...... 222

Grenada ...... 224

Guyana ...... 224

Jamaica ...... 225

St Kitts and Nevis...... 225

St Vincent and the Grenadines ...... 227

Trinidad and Tobago ...... 228

Appendix B – Appendix to Chapter 4 ...... 230

Sexual orientation of CADRES survey respondents by country...... 230

Missing values ...... 230

‘Discrimination against minorities’ variable ...... 231

Appendix C – Appendix to Chapter 5 ...... 232

HDI and cohort ...... 232

Religious climate and religious identity ...... 234

Legal context and cohort ...... 236

Legal context and religious identity ...... 238

Final word count (not including Appendices): 64,795

List of Figures Figure 1.1: Gay friendliness of some tourist destinations ...... 20 Figure 2.1: Acceptance of homosexuality around the world ...... 46 Figure 3.1: Percentage of countries banning same-sex intimacy by level of human development ...... 70 Figure 3.2: Ethnic makeup of the Commonwealth Caribbean ...... 73 Figure 4.1: Heterosexual support for the anti-gay laws ...... 89 Figure 4.2: Heterosexual support for the penalisation of sex between persons of the same- sex among those who want the current laws maintained and enforced ...... 91 Figure 4.3: Heterosexual support for the criminalisation of same-sex intimacy by country ...... 92 Figure 4.4: Beliefs about the usefulness of the current laws ...... 93 Figure 4.5: Beliefs about the usefulness of the current laws (by country) ...... 94 Figure 5.1: Average approval of same-sex marriage across the Americas – Regional Differences ...... 141 Figure 5.2: Average approval of same-sex marriage across the Americas – Country differences ...... 142 Figure 5.3: Approval of same-sex marriage by level of socioeconomic development . 145 Figure 5.4: Socioeconomic development and approval of same-sex marriage ...... 145 Figure 5.5: Share of persons who identify as Evangelicals (%) ...... 147 Figure 5.6: Evangelicalism and approval of same-sex marriage ...... 147 Figure 5.7: Legality of same-sex relations and approval of same-sex marriage ...... 149 Figure 5.8: Plot of country effects (Random intercepts) ...... 159 Figure 5.9: Approval of same-sex marriage by birth cohorts and level of development ...... 161 Figure 5.10: Approval of same-sex marriage religious identities and Evangelical presence ...... 162 Figure 5.11: Approval of same-sex marriage by birth cohort effects and legal context163 Figure 5.12: Approval of same-sex marriage by religious identity and legal context .... 165

Figures for Appendix B Figure B1: Responses to question about sexual orientation by country of residence .. 230

List of Tables Table 3.1: Penalties for private, consensual same-sex intimacy between adults in the Commonwealth Caribbean ...... 66 Table 3.2: HDI values, categories and rankings in the Commonwealth Caribbean in 2014 ...... 69 Table 3.3: Religious identities in the independent Commonwealth Caribbean (population share in %) ...... 71 Table 3.4: Popular Christian denominations in the independent Commonwealth Caribbean (population share in %)...... 74 Table 3.5: Popular derogatory terms used to refer to homosexuals in dancehall and reggae ...... 79 Table 4.1: Descriptive statistics for predictors ...... 97 Table 4.2: Responses to questions about discrimination against minorities ...... 100 Table 4.3: Matrix showing tetrachoric correlation between the items measuring support for the criminalisation of same-sex behaviours ...... 103 Table 4.4: Matrix showing tetrachoric correlation between the items measuring the perceived usefulness of the anti-gay laws ...... 103 Table 4.5: Nested models – Support for the criminalisation of same-sex intimacy ...... 105 Table 4.6: Nested models – Perceived usefulness of the anti-gay laws ...... 105 Table 4.7: Support for criminalisation of same-sex intimacy ...... 110 Table 4.8: Joint probability models – Consistent and mixed support for the criminalisation of same-sex intimacy ...... 114 Table 4.9: Modelling ‘unsure/prefer not to say’ responses for items relating to support for the criminalisation of same-sex intimacy ...... 118 Table 4.10 Perceived usefulness of the current anti-gay laws ...... 122 Table 4.11: Perceived usefulness of the laws - Joint probability model...... 126 Table 4.12: Modelling ‘Unsure/Won’t say’ responses for items relating to the usefulness of the laws ...... 128 Table 5.1: Countries included in the 2014 AmericasBarometer survey ...... 139 Table 5.2: Pairwise comparisons – The Commonwealth Caribbean vs. other country groupings ...... 141 Table 5.3: Socioeconomic development across the Americas ...... 143 Table 5.4: Regulation of same-sex behaviours as at 2014 ...... 148 Table 5.5: Descriptive statistics for individual level predictor variables ...... 151

Table 5.6: Multi-level analysis of approval towards same-sex marriage in the Americas ...... 153 Table 5.7: Changes in (average) approval of same-sex marriage and legal environment by country 2010-2014 ...... 157 Table 5.8: Differences in approvals of same-sex marriage ...... 158

Tables for Appendix B Table B1: Missing values for dependent variables ...... 230 Table B2: Tetrachoric correlations ...... 231 Table B3: Item discrimination parameter and difficulty estimates ...... 231 Table B4: Goodness of fit statistics ...... 231

Tables for Appendix C Table C 1: Interacting HDI with birth cohort ...... 232 Table C2: Significance of cohort effects by level of human development ...... 233 Table C3: Interacting religious climate with religious identity ...... 234 Table C4: Significance of religious identity variable by religious climate ...... 235 Table C5: Interacting legal context with birth cohort ...... 236 Table C6: Significance of cohort effects by legal context ...... 237 Table C7: Interacting legal context with religious identity ...... 238 Table C8: Significance of religious identity by legal context ...... 239

Abbreviations

Commonly used abbreviations:

AMEs Average Marginal Effects CHRI Commonwealth Human Rights Initiative CADRES Caribbean Development Research Services GDP Gross Domestic Product HDI Human Development Index IGLHRC International Gay and Lesbian Human Rights Commission ICC Intraclass correlation ILGA International Lesbian, Gay, Bisexual, Trans and Intersex Association LAPOP Latin American Public Opinion Project MP Member of Parliament PM Prime Minister St. Saint SMMC Stop Murder Music Campaign UN United Nations UNAIDS United Nations Program on HIV/AIDS UNDP United Nations Development Program WTTC World Travel and Tourism Council

Abstract

Over the last few decades, there has been a significant increase in the political and public acceptance of gay men and lesbians. However, this trend of acceptance is not a global phenomenon. Currently over 70 countries still criminalise private consensual same-sex intimacy, among which are 11 of the 12 independent Commonwealth Caribbean states. It should be noted that the anti-gay laws of the Caribbean are rarely used to police consensual private sexual activities. Thus, if private same-sex conduct is rarely penalised, why keep the laws in place, especially in the age where such bans are considered a violation of basic human rights? Many policy makers in the region have cited public opinions about homosexuality as a significant barrier to law reform. However, while a common view is that these laws are anchored by public support, very few studies have emerged to test whether the attitudes and behaviours of the general population are in line with this view. Against this backdrop, this thesis analyses attitudes towards lesbians and gay men and their legal rights in the Commonwealth Caribbean. The thesis begins with an analysis of support for the anti-gay laws in Barbados, Guyana and Trinidad and Tobago. The analysis revealed that a majority of the sample supported the maintenance and enforcement of the laws, but did not want same-sex couples to be penalised for having sex in private. This suggests that attitudes may not be as stark as policy makers suggest. The descriptive statistics also show that a significant share of individuals think that the laws (1) reflect moral standards; (2) stop the spread of homosexuality; (3) are important from a public health perspective, and (4) protect young people from abuse. Support for the laws are thus related to beliefs that homosexuality is a ‘threat’ to the fabric of society. The empirical analysis of support for the laws revealed that religiousness, interpersonal contact and beliefs about the origin of homosexuality were the most reliable predictors of public support. However, age and education were only statistically significant in a few models, and there was no evidence that attitudes varied across religious denominations. This is a contrast to the findings of studies in the West. It was hypothesised that macro-level factors – such as the large share of Evangelicals, anti-gay laws and level of socioeconomic development – could be exerting an influence on attitudes that is stronger than that of these personal characteristics. As such, the study conducted a cross-national analysis of attitudes towards same-sex marriage in 28 countries in the Americas, 6 of which were members of the Commonwealth Caribbean. In general, countries with higher levels of development, smaller shares of Evangelicals and more liberal laws on homosexuality were more approving of same-sex marriage. The results also suggest that the impact of age and/or religion is less prominent in countries with restrictions on same-sex intimacy, lower levels of development and a strong Evangelical presence, confirming the hypothesis that contextual factors could mitigate the impact of some of the individual-level variables. Finally, to get a nuanced view of anti-gay prejudice in the region, a thematic analysis of anti-gay speech in dancehall and reggae – music originating from Jamaica but popular in the region – was presented. The thematic analysis revealed that homosexuality is presented as ‘sinful’, a ‘violation of gendered norms’, ‘unnatural’, a ‘threat to society’ and a ‘foreign lifestyle’. The presentation of homosexuality as a ‘foreign’ lifestyle suggests that anti-gay prejudice could be related to fears of neo-imperialism and could be a means of rejecting ideological intrusions from the West. This is not surprising, as currently, the fight for the advancement of gay rights is being headed by activists in the West. Based on the thematic analysis, efforts to remove the anti-gay laws should be (or at least appear to be) home-grown to limit public backlash.

Declaration

No portion of the work referred to in the thesis has been submitted in support of an application for another degree or qualification at this or any other university or other institute of learning;

The author of this thesis (including any appendices and/or schedules to this thesis) owns certain copyright or related rights in it (the “Copyright”) and she has given The University of Manchester certain rights to use such Copyright, including for administrative purposes.

Copies of this thesis, either in full or in extracts and whether in hard or electronic copy, may be made only in accordance with the Copyright, Designs and Patents Act 1988 (as amended) and regulations issued under it or, where appropriate, in accordance with licensing agreements which the University has from time to time. This page must form part of any such copies made.

The ownership of certain Copyright, patents, designs, trademarks and other intellectual property (the “Intellectual Property”) and any reproductions of copyright works in the thesis, for example graphs and tables (“Reproductions”), which may be described in this thesis, may not be owned by the author and may be owned by third parties. Such Intellectual Property and Reproductions cannot and must not be made available for use without the prior written permission of the owner(s) of the relevant Intellectual Property and/or Reproductions.

Further information on the conditions under which disclosure, publication and commercialisation of this thesis, the Copyright and any Intellectual Property and/or Reproductions described in it may take place is available in the University IP Policy (see http://documents.manchester.ac.uk/DocuInfo.aspx?DocID=487), in any relevant Thesis restriction declarations deposited in the University Library, The University Library’s regulations (see http://www.manchester.ac.uk/library/aboutus/regulations) and in The University’s policy on Presentation of Theses

Dedication

I am a product of the good coaching and mentoring that I received over the years. That being said, it only seems fit to dedicate my greatest achievement to date to those who took the time to teach me, advise me, encourage me and motivate me. So:

To my teachers and mentors with love. I hope I made you all proud!

Acknowledgements

I would first like to thank my supervisors, Rob Ford, and Ingrid Storm, for guiding me throughout this project. My work has truly benefited from their constructive criticisms, suggested improvements and support. A very special thank you goes to Siobhan McAndrew, who was a member of my supervisory team during my first year of the Ph.D., and, even after transferring to the University of Bristol, continued to give me advice and encouragement.

I am also grateful to Gindo Tampubolon, Angelia Wilson, and Rachel Gibson, who were independent reviewers during my annual reviews, and provided detailed commentaries on the work. A big thank you to Xiomara Archibald, Lisa Drakes, Stacia Howard, Andrew Flores and Ryan Bynoe, who willingly read and commented on various chapters of the thesis.

I am indebted to Pratima Srivastava for providing me with (and also explaining) the GUASS codes she used in her book Recreational Drug Consumption – An Economic Perspective. Her assistance enabled me to run the joint probability models featured in Chapter 4 of this thesis. A special thank you goes to the Caribbean Development and Research Services for allowing me to use their datasets, as well as the Latin American Public Opinion Project and their major supporters for making their data publicly available.

I would also like to express my gratitude for the generous funding I received via an Economics and Social Research Council’s ‘Advanced Quantitative Methods’ studentship and the University of Manchester’s ‘President’s Doctoral Scholar’ award.

Last, but certainly not least, I would like to thank my family and friends, who provided me with the much needed emotional support throughout this PhD. You lot kept me sane throughout this process, and for that, I will be eternally grateful!

Published Work

Some of the work done in Years 1 and 2 of the PhD program was accepted for publication. The papers are listed below and are accessible via the accompanying links:

They called it the ‘abominable crime’: An analysis of heterosexual support for anti-gay laws in Barbados, Guyana and Trinidad and Tobago in Sexuality Research and Social Policy, 13(2): 130-141. http://dx.doi.org/10.1007/s13178-015-0209-6

Protecting the fabric of society? Heterosexual views on the usefulness of the anti- gay laws in Barbados, Guyana and Trinidad and Tobago, forthcoming in Culture, Health and Sexuality. http://dx.doi.org/10.1080/13691058.2016.1207806

Chapter 1 : Introductory remarks

In the last few decades, countries in the Global North have made significant strides in their efforts to reduce state discrimination against lesbians and gay men. For instance, many Western countries prohibit discrimination in employment based on sexual orientation and have equal ages of consent for same and opposite sex intimacy. Quite a few states also recognise same-sex unions/marriage and allow joint adoption by same- sex couples. While the socio-legal acceptance of gays and lesbians appears to be on the rise in the Global North, in several countries, gays and lesbians have very few rights. According to the 2016 International Lesbian, Gay, Bisexual, Trans and Intersex Association’s (ILGA) State-Sponsored Homophobia report, over 70 countries still penalise private consensual same-sex intimate acts (Carroll 2016). Among these are 11 of the 12 Commonwealth Caribbean states1—that is, Antigua and Barbuda, Barbados, Belize, Dominica, Guyana, Grenada, Jamaica, Saint (St.) Kitts and Nevis St. Lucia, St. Vincent and the Grenadines, and Trinidad and Tobago. The Bahamas currently stands as the only Commonwealth Caribbean state to have removed the ban on same-sex intimacy.2 However, there is still some bias in Bahamian laws, evidenced by different ages of consent for same and opposite sex sexual acts. According to the Human Dignity Trust (2015), the anti-gay laws in the Commonwealth Caribbean are not currently used to prosecute persons engaging in consensual same-sex sexual activity in private. However, it has been reported that anti- gay laws are sometimes used in conjunction with more serious offences (such as rape) in Trinidad and Tobago (Human Dignity Trust 2015). Meanwhile, in Jamaica, there have been reports of the laws have being enforced in instances where same-sex couples were caught engaging in sexual activities in public (Human Dignity Trust 2015). Even though the laws are not currently used to police private consensual sexual activity, this does not mean that anti-gay laws are harmless. Laws are part of a larger socio-political process, and so they help shape norms by prefiguring values, prejudices and interests (Gordon 1984; van den Akker et al. 2013). The anti-gay laws in the Caribbean thus create the presumption that lesbians and gay men are criminals, and legitimise and strengthen prejudice against them.

1 The Commonwealth Caribbean refers to the independent English-speaking Caribbean—that is, former colonies of Great Britain. 2 The Bahamas repealed its bans on private same-sex intimacy in 1991. 18

In the last decade, the prevalence of state-sponsored discrimination against sexual minorities in the Commonwealth Caribbean has received considerable international attention, and many Caribbean governments have been criticised for their retention of anti-gay laws. For instance, the International Gay and Lesbian Human Rights Commission (IGLHRC) and the Commonwealth Human Rights Initiative (CHRI) have both accused Commonwealth Caribbean governments of failing in their obligation to protect fundamental human rights (CHRI 2011; IGLHRC 2011, 2012). Most of these states are signatories to the various human rights documents whose treaty bodies stipulate the right to equality, non-discrimination and privacy for all persons. The Joint United Nations Program on HIV/AIDS (UNAIDS) has also called for the overturning of the Caribbean’s anti-gay laws, arguing that in addition to violating human rights, they also have a negative impact on HIV/AIDS prevention and access to healthcare (UNAIDS 2010). Notably, gay-hostile regions often link HIV/AIDS with homosexuality (Reding 2003), and so persons may avoid getting tested or obtaining treatment for fear of being stigmatised (Corrêa et al. 2008; Haas and Cowell 2011). This should be a fundamental concern of Caribbean governments, given that the region has the second highest rate of HIV/AIDS infections after sub-Saharan Africa (UNAIDS 2010). It is also possible that retaining anti-gay laws could have economic repercussions. For instance, at the 2011 Commonwealth Heads of Government Meeting in Perth, Australia, the UK Prime Minister (henceforth PM) David Cameron hinted that his Government may cut or withhold foreign aid to countries still criminalising homosexuality. He noted that recipients of UK aid should ‘adhere to proper human rights’ (BBC 2011). There have also been several calls for tourists to avoid the ‘homophobic’ Caribbean (Caribbean360 2006; Cowell and Saunders 2011). According to the 2016 Spartacus World Gay Travel Index (Spartacus World 2016), which ranks destinations regarding their ‘gay friendliness’, Commonwealth Caribbean countries lie at the negative end of the tolerance continuum (see Figure 1.1), which ranged from a high score of +9 to a low of -14. The perceived lack of gay friendliness could deter potential tourists, particularly members of, and advocates for, the lesbian and gay community. This should be a concern for Caribbean governments, given that the region’s economic fortune is intimately intertwined with the performance of its tourism industry (Jackman and Naitram 2015; Jackman et al. 2011)

Figure 1.1: Gay friendliness of some tourist destinations (Members of the Commonwealth Caribbean in red)

9 8

-4 -4 -4 -4 -5 -5 -6 -7 -9 -9 -9

-13

Iran

Belize

Russia

Canada

Guyana

Jamaica

Grenada

Bahamas

Barbados

Dominica

StKitts & Nevis

UnitedKingdom Trinidad & Tobago Antigua & Barbuda Notes: Data sourced from Spartacus World (2016). Scores for St. Lucia and St Vincent and the Grenadines were not available, and so, these states could not be included in the above figure.

As is well documented, Caribbean countries are characterised by their limited natural resource endowments, limitations on import substitution possibilities, small domestic markets and weak inter-industry linkages (Briguglio 1995). This translates to a high import content relative to its Gross Domestic Product (GDP), making the economies very dependent on foreign earnings to pay their large import bills and to spur economic growth (Worrell 2012; Worrell et al. 2013). Also, recent erosions in trade preferences and increases in the global market share of their competitors has made trade- in-goods very challenging for these states, forcing policy makers to focus more on non- traditional exports, such as tourism, as a means of attaining the much needed foreign exchange (Jackman 2014). According to the World Travel and Tourism Council’s (WTTC) 2016 economic impact report (WTTC 2016), the Caribbean currently stands as the most tourism-dependent region in the world with respect to the sector’s contribution to the national economy. The WTTC estimates that in 2015, tourism’s total contribution (direct + indirect + induced) to the Caribbean’s GDP was approximately 14.8 per cent. Tourism also directly employs about 4.2 per cent of the labour force and stands as a significant earner of foreign exchange, accounting for 18.8 per cent of the region’s foreign

exchange earnings. Thus threats to tourism can be viewed as a threat to economic well- being. At face value, one would assume that Commonwealth Caribbean governments would find it desirable, or even necessary, to adhere to international demands. However, even in the face of intense international pressure to end human rights abuses, and threats related to international support and markets for their goods and services, policy makers have opted to keep the laws on the books. The choice to put international relations at risk would suggest that strong domestic politics must be at play. Indeed, it would seem plausible to assume that policy makers in the region are merely carrying out the will of the electorate. Though the consent of the governed is not always necessary or sufficient to bring about changes in laws discriminating against sexual minorities, it is considered an important policy benchmark in democratic societies (Lax and Phillips 2009; Evans and Kelley 2004a). A perceived lack of public appetite for advancing gay rights may encourage policy makers to prioritise the preferences of the electorate above the demands of foreign interest groups, to gain or remain in public favour and influence voting behaviours. Recent statements made by some public officials in the region suggest that public opinion of homosexuality serves as a barrier to law reform. For example, in response to recommendations from the United Nations (UN) Human Rights Committee to remove the anti-gay laws in 2009, the Government of Barbados noted that:

“Barbados cannot accept at this time, the recommendation to decriminalize such sexual acts between consenting adults of the same sex. Decriminalization of sexual acts between adults of the same sex has not received the consensus of religious denominations or the public of Barbados as a whole. In fact, significant sections of the community are opposed to such decriminalization…It must be noted that Barbados is a heavily religious society, and there is a significant lobby by the church on such issues.” (UN Human Rights Committee 2009:4)

Similar sentiments were expressed to the UN Human Rights Council by the policy makers in St. Lucia in 2010:

“The Government acknowledges the contribution and concerns expressed by United and Strong, representing gay and lesbian interests. However, deeply rooted religious, cultural and moral values and practices on the island create a formidable challenge towards mobilization and general acceptance of gay rights by society” (UN Human Rights Council 2010:23) and by government officials in St Vincent and the Grenadines in 2011:

“The Government of Saint Vincent and the Grenadines cannot accept at this time this recommendation to repeal provisions against lesbian and gays, as the 21

public sentiment favours the retention of provisions which criminalize buggery and sexual relations between adults of the same sex” (UN Human Rights Council 2011:7)

In fact, fear of electoral sanction may have caused political elites in Jamaica and Trinidad and Tobago to recant on promises to tackle state-sponsored discrimination against gay and lesbians in their countries. Specifically, in the lead up to the 2011 general elections in Jamaica, the President of the People’s National Party (PNP) and then Leader of Opposition, Portia Simpson Miller, stated that:

“no one should be discriminated against because of their sexual orientation” (Boseley 2012).

She then promised to review the anti-gay laws should her party come to power (Hamilton 2011). The PNP won the 2011 election, and in June 2013, Senator Sandrea Falconer informed the Jamaican press that the PM would make good on her promise and would soon ask parliament to take a conscience vote on the anti-gay laws (Reynolds 2013). However, after many delays to the vote, in April 2014, then PM Portia Simpson Miller told reporters

“…we have to go to our constituents, consult our constituents and then we go with the decisions of those consultations” (Roberts 2014).

A similar line of reasoning was expressed by the then Opposition Leader Andrew Holness. In an interview with Cliff Hughes on the Nationwide radio programme Impact in May 2014, Mr Holness stated:

“I don’t discriminate against people. Whatever my personal views may be, there is a public space and the public space must guarantee everyone’s protection, and the ability to fulfil their aspirations…as [a] political leader, we would have to get an indication – which is what the referendum would do – as to where Government should go” (Nationwide 2014)

With respect to policy makers in Trinidad and Tobago, in 2012, then PM Kamla Persad- Bissessar wrote a letter to the Kaleidoscope Trust3 in which she promised to address the discrimination in the country based on gender and sexual orientation (Baboolal 2012).

3 The Kaleidoscope Trust is a non-governmental organisation based in the UK. It campaigns globally for gay/LGBT rights and diversity. 22

However, in 2014, the PM backpedalled, stating that the decriminalisation of homosexuality was not an issue for her Government to decide at this time. She said:

“It is an issue that the people must decide and therefore one that may require a referendum to get the views of the people… personal views are not good enough” (Day 2014).

At the time of writing this thesis (July 2016), a referendum had yet to be called in either country. A key observation is that government officials in both Barbados and St Lucia cite religious actors (that is, faith-based groups and religiously inclined voters) as strong veto players in the politics of lesbian and gay rights. In their guidelines for moral living, many religious institutions stress the virtue of sex and procreation within the bounds of heterosexual marriage (Herek 1990). Many religious groups condemn homosexuality and often deem it ‘sinful’ or ‘unnatural’ (Hough 2004; Wilson 1971 ). The fact that religious groups condemn homosexuality is rather important. Religious institutions have long been recognised as intermediaries between individual and public life (Gilbert 1993; Jelen 1992; Roozen et al. 1984). They effectively function as political communities, shaping the viewpoints of their members (Wald et al. 1988). It is quite common for political messages to be intertwined into informal conversations among parishioners and via messages from the pulpit (Beatty and Walter 1988; Djupe and Gilbert 2003; Greenberg 2000; Woolfalk 2012). In this way, religious institutions can significantly influence the political views of their members, among which are matters regarding gay rights (Olson et al. 2006; Sherkat et al. 2010). As will be discussed later in the thesis, countries in the Commonwealth Caribbean can be characterised as religious states with a bias towards conservative Christian values—at least on matters of sexual orientation. Opposition from religious interest groups, coupled with public subscriptions to conservative Christian views about sexual orientation, make the advancement of gay rights in the region difficult. The main takeaway is that policy makers in the region justify the retention of the anti-gay laws under the premise of catering to public preferences. This has led them to adopt a ‘wait until society evolves’ approach to law reform. Interestingly, while citizens in the Commonwealth Caribbean are often portrayed as largely intolerant of sexual minorities (Atluri 2001; Douglas 2007; Gutzmore 2004), anecdotal evidence suggests that current attitudes may not be as stark as presented. For example, in Barbados, a well- known self-identified transgender woman (DiDi Winston) has won the coveted title of

Best Flag Person at the Grand Kadooment (often referred to as Kadooment Day)4 on multiple occasions (Murray 2009). Meanwhile, in 2015, Jamaica hosted its first public gay pride celebration (Davis 2015). The hosting of pride in Jamaica is particularly significant, as Jamaica is presented as the most homophobic country in the Caribbean (Schleifer 2004). Currently, there are no official statistics for hate crimes against sexual minorities in the region. However, newspapers articles describing physical violence towards gays and lesbians appear more frequently in Jamaica than any other country in the region (West and Cowell 2015). While hosting various events celebrating sexual diversity does not necessarily mean that these societies are tolerant, it does suggest that tolerance may be on the rise and also calls depictions of a uniform anti-gay stance into question. Since policy makers in the region are adamant that social tolerance must precede law reform, the main question for activists then becomes ‘how do we speed up the process?’ There is thus a need for research on public opinion on homosexuality and gay rights in the Commonwealth Caribbean. Firstly, an analysis of public opinion would help policy makers and activists understand how the general public would react to policy intervention. Secondly, knowing the social and psychological determinants of anti-gay attitudes can aid in developing public policies to reduce prejudice and engender an environment that is receptive to law reform. To date, very few studies have quantitatively evaluated public attitudes towards homosexuality in the Commonwealth Caribbean, and the few quantitative studies that do exist mostly sampled from university student bodies (Chadee et al. 2013; Gromer et al. 2013; West and Hewstone 2012). This research seeks to fill the previously mentioned gap in the literature and uses mixed methods to shed some light on public attitudes towards gays and lesbians in a sample of Commonwealth Caribbean countries. Nationally representative survey data will be employed to explore the ways in which—and the degree to which—Caribbeans make use of their existing toolkits of moral reasoning to make judgements about the rights of gay men and lesbians. The use of representative national samples is important, as it will allow us to make generalisations to the whole population. Attitudes in the Commonwealth Caribbean will then be compared with other countries in the Americas. Cross-national comparisons will help us to determine if there is something distinctive about attitudes in the Commonwealth Caribbean or if their patterns

4 Kadooment is the grand finale of Barbados’ largest annual national festival ‘Crop Over’. It is a carnival parade and features several bands whose members, ‘the revellers’, make their way down the streets dressed in elaborate costumes and dancing to soca and/or calypso music. The Crop Over season normally begins in May/June and ends on Kadooment Day, which is the first Monday in August. 24

are similar to many other countries in the Americas. Finally, the research will move from the general picture of attitudes towards homosexuality to the specific case of negative attitudes in popular culture. A thematic analysis of anti-gay speech in Caribbean (Jamaican) music will be presented to provide additional insights into the roots and posture of anti-gay prejudice.

1.1 Significance of the study

The findings of this thesis are particularly useful to the Commonwealth Caribbean, as they could aid in developing effective policies to reduce sexual prejudice in these states. The study also has relevance outside the Caribbean context. Bans on private sexual acts are currently an important human rights issue, one that has captured the attention of key international bodies, economic stakeholders and governments around the world. Theoretically, the fight to eradicate the ban on same-sex intimacy is at the global level. This international humanitarian effort requires improving attitudes towards gay and lesbians in every country (Lin et al. 2016). Improving attitudes, in turn, requires an understanding of the drivers of anti-gay prejudice in each country—or, at the very least, an understanding of attitudes towards homosexuality across a variety of cultural contexts. As is well documented, attitudes towards homosexuality vary significantly across nations (Collier et al. 2015; Jäckle and Wenzelburger 2015), and these differences have been shown to reflect specific national circumstances.(Adamczyk and Cheng 2015; Adamczyk and Pitt 2009; van den Akker et al. 2013). In fact, it has been argued that cultural factors have a bigger impact on attitudes towards homosexuality than individual- level characteristics such as education, age, gender and personal religiosity (Adamczyk and Cheng 2015). To date, most national-level studies on opinions on gay rights tend to focus on a subset of Western countries (mainly the US and Western Europe). Given the importance of context in determining attitudes, the extent to which studies based on the developed world can be used to inform policies aimed at achieving positive social and legal change for lesbians and gays in countries where same-sex intimacy is criminalised is not clear. Firstly, most of the countries with anti-gay laws are classified as developing nations (see Chapter 3). Secondly, in most Western countries, the fight for the legality of same-sex sexual acts has already been won. Societal differences between Western nations and countries with anti-gay laws may not only lead to differences in the level of anti-gay

prejudice expressed by the public but could also result in differences in how various factors affect attitudes. The lack of studies on attitudes towards homosexuality outside North America and Western Europe limits our understanding of anti-gay attitudes, which in turn is an impediment to collective action to reduced state-sponsored discrimination against gay men and lesbians at the global level. By focusing on the Commonwealth Caribbean, this study adds to the sparse body of work examining anti-gay prejudice in countries that are not classified as developed and/or still criminalise same-sex intimacy. While the findings of this study cannot be assumed to be universal for all developing countries with anti-gay laws, it does seem plausible that better parallels can be drawn from the Commonwealth Caribbean than from countries in the developed world. A primary benefit of the study is that it allows us to investigate the extent to which cultural and sociological factors, often cited as key determinants of anti-gay prejudice in the developed world, also explain anti-gay prejudice in the Commonwealth Caribbean.

1.2 Objectives and research questions

This study builds on democratic theory and so begins with the premise that there is some match between government choice and citizen preference. The basic idea is that if we are to achieve positive social change for lesbians and gays in the Commonwealth Caribbean, then a better understanding of public attitudes towards homosexuality is necessary. The broad objective is to provide a thorough empirical investigation of attitudes towards gay men and lesbians and their respective rights in the Commonwealth Caribbean and contribute to the current debates about the anti-gay laws in these states. The primary objectives of the research are: (1) to determine the level of public support for lesbian and gay rights; (2) to determine the individual and contextual factors influencing said support; and (3) to present and analyse the anti-gay rhetoric often heard in the Commonwealth Caribbean.

1.2.1 Public opinions about gay rights To address objectives (1) and (2), I draw on two databases. First, I employ data attained from the Caribbean Development Research Services (CADRES). CADRES is a popular political research organisation based in Barbados that specialises in collecting extensive and detailed data on nationally representative samples of Caribbeans. In 2013, CADRES carried out a three-country survey project on attitudes towards homosexuals. CADRES

surveyed a total of 2,871 adults living in Barbados, Guyana, and Trinidad and Tobago.5 In this thesis, I focus on eight survey items which capture two dimensions of support for the criminalisation of same-sex intimacy. The first four survey items can be classified as measuring general support for the current laws. These items measured support for the maintenance and enforcement of the anti-gay laws currently in place, as well as support for penalising private same-sex intimacy. The last four survey items capture the extent to which citizens view the laws as protectionist measures. Specifically, they capture subscriptions to notions that the anti-gay laws: (1) reflect moral standards; (2) stop the spread of homosexuality; (3) are important from a public health perspective; and (4) protect young people from abuse. Descriptive statistics are used to answer the following questions:

1. What is the level of support for the criminalisation of same-sex intimacy in Barbados, Guyana and Trinidad and Tobago? 2. To what extent do citizens in the countries view the laws as protectionist measures?

Then, a series of multivariate probit models are used to evaluate the main drivers of support and perceived usefulness of the anti-gay laws. In building these models, I rely heavily on previous analyses done in Western countries. While the use of secondary data constrained my choice of independent variables, I was able to control for some of the key variables that previous studies on the West cited as key determinants of anti-gay prejudice: these are gender, age, education, religion, interpersonal contact, beliefs about the origins of homosexuality and attitudes towards other minorities (Mason and Barr 2006; Haider-Markel and Joslyn 2008; Smith et al. 2009; van den Akker et al. 2013). As alluded to earlier, it is possible that the individual-level correlates of anti-gay attitudes may differ across national contexts. This raises a third research question:

3. To what extent do the social and psychological factors identified in the literature on the West also explain views on the criminalisation of same-sex intimacy in Barbados, Guyana and Trinidad and Tobago?

5 Further details about this survey are provided in Chapter 4 of the thesis. 27

After tracing the contours of attitudes towards anti-gay laws in Barbados, Guyana, and Trinidad and Tobago, I then conduct a cross-national comparison of attitudes. Earlier, I noted that various studies have shown that context is an important determinant of public opinion on homosexuality. To determine how context may be affecting attitudes in the Commonwealth Caribbean, I utilise data from the 2014 wave of the AmericasBarometer surveys by the Latin American Public Opinion Project (LAPOP). The AmericasBarometer survey covered 28 countries in the Americas, six of which were members of the Commonwealth Caribbean.6 I focus on the survey item capturing approval of same-sex marriage, as marriage is considered the ultimate civil right (Gwartney and Schwartz 2016). State recognition of same-sex marriage sends a strong message that same-sex couples and opposite couples are equal, and allows same-sex couples to enjoy a number of financial benefits currently available to married persons, such as tax allowances, inheritance rights and state bereavement benefits. In addition to financial security, marital rights would also grant same-sex couples access to the social support structures and well-being associated with marriage, such as the ability to publicly express and celebrate their love or ensure visitation rights when a partner is sick (Gwartney and Schwartz 2016). Using descriptive statistics and multilevel models, I try to answer the following questions:

4. How do attitudes towards same-sex marriage in the Commonwealth Caribbean compare with attitudes in the rest of the Americas? 5. If comparatively different, to what extent can societal context —specifically, the level of socioeconomic development, religious climate and policies affecting lesbians and gay men — explain the differences observed in attitudes?

My choice of contextual variables was largely inspired by the existing literature on how context shapes public attitudes. Specifically, theories of modernisation suggest that socioeconomic development has a liberalising effect on attitudes (Andersen and Fetner 2008b). One would thus expect that countries with higher levels of development should be more supportive of gay rights. However, research has found that even as countries develop, religious beliefs and behaviours remain strong (Norris and Inglehart 2011). It follows that attitudes towards homosexualiy may be more negative in societies that place

6 Commonwealth Caribbean countries included in the 2014 wave are the Bahamas, Barbados, Belize, Guyana, Jamaica, and Trinidad and Tobago. 28

a high value on relgious norms, even in the face of modernisation. Finally, while it is expected that the legal system will adapt to changing social needs (the core premise of this study), we cannot ignore the fact that laws also impact society. One of the oldest debates about the relationship between laws and public opinion concerns whether or not the laws should be considered the independent variable or the dependent variable (Gordon 1984). In this thesis, I do not choose a side of the debate. Similar to other researchers (Flores and Barclay 2016; Slenders et al. 2014), I adopt the view that there is a bi-directional relationship between laws and public opinions: the legal regulation of homosexuality changes with public opinion, but at the same time opinions are affected by policy. Hence, it is hypothesised that individuals living in countries with more progressive laws on homosexuality will also show greater support for gay rights. A brief overview of countries in the Americas suggests that there are significant differences in the levels of socioeconomic development, religious traditions and the civil and legal rights of gay and lesbians. In line with previous literature, it is expected that these differences in societal context will result in substantial differences in approval of same-sex marriage across countries. On the region of interest, I posit that due to the large Evangelical presence and the stricter regulations of homosexuality (at least in comparison to other countries in the Americas), citizens of the Commonwealth Caribbean will report lower levels of approval than their neighbours in the Americas.

1.2.2 Analysis of anti-gay rhetoric To meet the third objective, I conduct a thematic analysis of reggae and dancehall—music indigenous to Jamaica, but popular throughout the Caribbean. Music and public opinion are two separate but highly related social categories. Similar to other forms of media, the link between music and society is typically viewed through two lenses. The first is related to the role of the media in shaping public attitude towards various social and political issues (Bryant and Zillman 2009; Potter 2013). The impact of the media on attitude formation and change is often presented within the ambient of priming, agenda setting and framing models (Scheufele and Tewksbury 2007). Agenda setting theory tell us that the media is capable of shaping the public’s beliefs about what is important (Potter 2013). The basic idea is that the more emphasis the media places on an issue, the more likely the public is to see it as important (Scheufele and Tewksbury 2007). Priming suggests that exposure to certain symbols used in the media can trigger users to activate related thoughts and ideas when they come across these symbols in real life (Jang

and Lee 2014; Potter 2013). Finally, the framing impact is based on the assumption that how the media selects information and structures its message influences how an issue is characterised and understood by audiences, and this in turn influences attitudes (Brewer et al. 2003; Potter 2013; Scheufele and Tewksbury 2007). On the other hand, researchers also view media content as a manifestation of society’s values (Adamczyk et al. 2015). As pointed out by Adamczyk et al. (2015), the professionals who create the local media content are members of the same society and are likely to be influenced by the context in which they live. The authors also note that media professionals seek to maximise profits, and so are often mindful of their audiences’ preferences. It is entirely reasonable to assume that there would be some relationship between how the public views an issue and how it is presented in the media. In this thesis, I opt to focus on the latter view of media and use the anti-gay rhetoric found in dancehall and reggae as a means of providing additional insight into attitudes towards homosexuality. Of the various forms of media, I opt to focus on reggae and dancehall since the most fervent opposition of homosexuality within Jamaican society comes from these two musical genres (Smith and Kosobucki 2011). Since the music is indigenous to Jamaica, it can be argued that it is more indicative of Jamaican attitudes than the rest of the Caribbean. As mentioned earlier, Jamaica is often classed as the most homophobic country in the region. However, songs with anti-gay rhetoric are popular throughout the Commonwealth Caribbean—the music is not restricted to Jamaican airwaves. So while I recognise that the music is more representative of societal values in Jamaica, the popularity of songs with strong anti-gay messages in the Commonwealth Caribbean suggests that many consumers either agree with the sentiments from the music or are indifferent to anti-gay rhetoric. An analysis of the music could provide us with a nuanced view of beliefs about homosexuality and how anti-gay attitudes are justified in Jamaica and, to some extent, other states in the Commonwealth Caribbean. The thematic analysis seeks to answer the following questions:

6. What are the main arguments used to condemn homosexuality in reggae and dancehall? 7. Are these arguments unique, or are they similar to arguments used to justify anti- gay prejudice elsewhere?

1.3 Thesis structure

The rest of this thesis proceeds as follows. The next chapter (Chapter 2) reviews the literature on anti-gay prejudice. Chapter 2 begins with an overview of the main theories used to explain anti-gay bias; it then reviews empirical studies on attitudes towards lesbians and gay men. Chapter 2 ends with a brief overview of the state of research (or lack thereof) on anti-gay attitudes in the Commonwealth Caribbean. Chapter 3 provides some stylised facts on the socio-legal environment of the Commonwealth Caribbean. In this chapter, I present some of the features of these societies that both influence and reflect attitudes towards homosexuality—specifically, the anti-gay laws, the level of socioeconomic development, the strong Evangelical presence and the media. However, even though the current environment arguably leads to intolerance, as alluded to in the introduction, attitudes towards homosexuality are not uniform and there are still inclusive spaces for sexual minorities. The chapter ends with a brief overview of support services for gay men and lesbians and local activism. The empirical analyses are covered in Chapters 4–6. Chapter 4 presents the analysis based on the CADRES data—that is, the analysis of heterosexual views of the anti-gay laws in Barbados, Guyana, and Trinidad and Tobago. In Chapter 5, I present the analyses of attitudes towards same-sex marriage in the Americas, and in Chapter 6, the thematic analysis of the policing of homosexuality in reggae and dance is presented. Finally, Chapter 7 summarises the main findings of this thesis and the implications of these findings. The thesis ends by outlining the limitations of the research and highlighting some areas for future work.

Chapter 2 : Review of the related literature

The equal treatment of people, irrespective of their sexual orientation, is a fundamental human right principle, one that is recognised and widely supported by international bodies and governments in all regions of the world (Badgett et al. 2014). Despite the widespread support for equality in principle, the 2016 IGLA State-Sponsored Homophobia report highlights that in almost every country in the world, lesbians and gay men are not afforded the same legal, social or civil privileges as heterosexuals (Carroll 2016). A consensus in the literature is that discrimination against lesbians and gay men is rooted in negative attitudes towards homosexuality. First, people holding anti-gay attitudes are more likely to engage in anti-gay behaviours (Alden and Parker 2005; Franklin 2000; Herek 2004). Lesbians and gay men are thus more likely to encounter negative reactions and discrimination if public attitudes in their society are very negative. Second, in democratic societies, there is a relationship between public attitudes towards homosexuality and public policies affecting gays and lesbians (Hooghe and Meeusen 2013; Slenders et al. 2014). Politicians have a strong desire to remain in office (Oldendick 2002), and the threat of electoral sanction forces politicians to take the views of the electorate into consideration (Hobolt and Klemmemsen 2005). For activists, social acceptance of homosexuality is an important policy benchmark, especially since lesbians and gay men often make up a small percentage of the electorate, and this can sometimes limit their political influence. Although there is clear evidence that proactive pro-gay social movements were crucial to the advancement of gay rights (Frank and McEneaney 1999; Gwartney and Schwartz 2016), the success of these movements was not solely due to strong partnerships with larger movements or political actors. Success was also highly correlated with the ability of activists to frame the issue of gay rights in ways that resonated with the sentiments of the public (Corrales 2015). As noted by Hough (2004), state prohibitions on moral-laden issues are often discarded when the societal majority no longer disapproves of the activity. Indeed, there is empirical evidence to suggest that homosexual law reforms were precipitated by an increase in social acceptance of homosexuality in the US (Lax and Phillips 2009; Lewis and Oh 2008) and Europe (Hooghe and Meeusen 2013; Slenders et al. 2014).

However, causality may also run the other way. Laws are essentially socialising agents: many people accept the results of legislation and adopt the legal view of what is right or wrong (Carmines and Woods 2002; van den Akker et al. 2013). Extending legal and civil rights to gay men and lesbians can stimulate more positive and/or inclusive attitudes (Kreitzer et al. 2014). Currently, the common view is that there is a bi-directional relationship between public attitudes and laws related to gay rights, wherein public policy responds to attitudes, but attitudes also change with policy (Flores and Barclay 2016; Slenders et al. 2014). The links between public attitudes towards homosexuality and anti-gay behaviours and policies relating to gay men and lesbians means that public attitudes towards homosexuality should not be downplayed as mere opinions (Kuyper et al. 2013). Rather, they are important determinants of equality and non-discrimination. Against this backdrop, a burgeoning body of literature has emerged examining public attitudes towards gays and lesbians and their rights. However, a majority of quantitative studies have focused on the US (Copolov and Knowles 2016; Mason and Barr 2006), and to some extent, Europe. Although the extent to which these findings can be generalised to the states of the Commonwealth Caribbean is unclear, they do provide a starting point for studying anti-gay prejudice in the region. This chapter reviews the literature that is relevant to the research questions and empirical models. The discussion presents an overview of popular theories on attitudes towards lesbians and gay men and also summarises some key findings from the literature.

2.1 Negative attitudes towards lesbians and gay men

In recent decades several theories have been used to explain why some heterosexuals tend to be prejudiced against gays and lesbians. Among the most popular theories are: (1) the functionalist approach to sexual prejudice, and (2) the perceived threat of homosexuality.

2.1.1 Functionalist approach In his classic functionalist analysis of prejudice against gay men and lesbians, Gregory Herek (Herek 1990, 1991) suggests that stigma, prejudice and violence against lesbians and gay men serve at least three functions. First, sexual prejudice may serve a value- expressive function, allowing people to express values that are important to them. For

example, some individuals may express sexual prejudice as a means of affirming their religious beliefs, teachings and values (Kuhar 2013). The sacred texts of most mainstream religions contain scriptural passages that are frequently interpreted as condemnations of homosexuality (Comstock 1991; Duran 1993; Siker 2007). These texts are often seen as having a divine origin — that is, God’s inspired and inerrant word — and so, are thought to be universally and perennially valid (Piazza 2012). Some religious texts also contain tales of punishments for people who transgress the divine law (for example, the biblical story of Noah and the flood, or the story of Lut in the Qur'an). The texts also contain after-life rewards (such as heaven) for those who adhere to the rules, thus adding a dimension of divine accountability. Thus, for some individuals, opposition to homosexuality both reflects their religious commitment and is seen as necessary to gain after-life rewards or to avoid God’s wrath. It is therefore not surprising that a principal justification for anti-gay hostility is religious morality (Herek 1991). Another popular value expressed through anti-gay prejudice is traditional gender roles. This narrative is based on the premise that there are specific roles, behaviours and physical characteristics that society deems appropriate for men and women (Whitley Jr 2001; Whitley Jr and Ægisdóttir 2000). Under a traditional gender belief system, heterosexuality is often purported as a quality of ‘normal’ males and ‘normal’ females, and homosexuality is seen as a violation of gendered norms (Herek 1990). Within this gendered belief system are perceptions and implicit theories about the personality traits, interests and behaviours of the individuals who violate these norms (Gato and Fontaine 2016). Homosexuality is often conflated with gender inversions — where gay men are believed to be more akin to heterosexual women, and lesbians are expected to be similar to heterosexual men (Gato and Fontaine 2016; Kite and Deaux 1987). Thus, people who rigidly adhere to traditional gender role ideologies are also likely to be hostile to individuals whom they perceive as violating the ‘norm’, chief among which are gay men and lesbians (Whitley Jr 2001; Whitley Jr and Ægisdóttir 2000). Anti-gay prejudice can also serve a social-expressive function by increasing individuals’ social acceptability. Under the social expressive function, lesbians and gay men are the epitome of outsiders. Some people may be hostile toward this group to ensure that they themselves are viewed as insiders, simultaneously winning approval from others. Finally, anti-gay prejudice may serve a defensive function, whereby individuals act in ways that vilify homosexuals to reduce the anxiety associated with some psychological

conflict, for example, their sexual impulses or gender conformity. In general, people who express hostility towards gay men and lesbians do so as a means of confirming their heterosexuality and to avoid being stigmatised. Herek (1990; 1991) maintains that the need to be defensive tends to be particularly strong for those who are insecure about conforming to gendered norms and for those who have either experienced same-sex attraction or engaged in same-sex intimacy. The defensive function is thus somewhat related to the popular notion that some people who are hostile towards lesbians and gay men are revealing their own (suppressed) homoerotic feelings (Kuhar 2013).

2.1.2 The perceived threat of homosexuality The link between the threat of homosexuality and anti-gay prejudice is best described within the context of intergroup threat theory. Stephan and Stephan’s (2000) integrated theory of intergroup threat suggests that symbolic and/or realistic threats may act as causal antecedents of intergroup prejudice. Realistic threats pertain to perceptions that an out-group endangers the welfare of the in-group (including the in-group’s existence, political and/or economic power or physical well-being) (Brambilla and Butz 2013; Stephan and Stephan 2000). Symbolic threats refer to the perception that the out-group threatens the in-group’s way of life due to perceived differences in values, standards, beliefs, and morals (Brambilla and Butz 2013; Stephan and Stephan 2000). The anxiety caused by these feelings of threat can lead members of the in-group to dislike members of the out-group. Specifically, the more the in-group views the out-group as a threat (symbolic or realistic), the more negative the attitudes are likely to be (Vincent et al. 2009). Gay men and lesbians are often perceived as sources of both symbolic and realistic threats to society. This perception is often grounded in the acceptance of negative stereotypes — mostly exaggerated and ill-founded beliefs — about gay men and lesbians (Herek 1991). As noted in the previous section, homosexuality is perceived as violating traditional gendered norms, as well as sacred religious values and, as such, it could be considered a source of symbolic threat. Another common belief about homosexuals is that they are promiscuous, and by extension more likely to have venereal diseases (Bhugra 1987). Homosexuals are thus stereotyped as a threat to public health. This stereotype became more prominent during the 1980s when HIV/AIDS first emerged on the scene. The virus had a disproportionate impact on men who had sex with men (Herek and Capitanio 1999a) so much so that medical practitioners informally labelled the virus as ‘gay-related immune deficiency’ (Herek and Capitanio 1999a;

Gwartney and Schwartz 2016), reflecting the assumption that the disease only affected gay men. The HIV/AIDS epidemic led to an increase in anti-gay prejudice during the 1980s (Clements and Field 2014; Crockett and Voas 2003). Individuals blamed the gay community for starting the epidemic and believed that gay men were not only dangers to themselves, but also to heterosexuals (Herek and Capitanio 1999a). Whilst the spread of HIV/AIDS among heterosexual populations has significantly reduced the prevalence of the belief that HIV/AIDS is a gay disease, a link still exists between HIV/AIDS stigma and anti-gay prejudice (Taylor 2014). Another particularly powerful stereotype that affects people’s attitudes towards gay men is that they prey on children (Bhugra 1987; Niedwiecki 2013; Wiley and Bottoms 2013). Kort (2012) and Herek (2012) suggest that because of prejudice against gay men, and attempts by anti-gay groups to present gay men as a danger to justify discrimination against them, male-male molestation was often mischaracterised as homosexuality. It became common practice to refer to a man that molests a boy as a homosexual, but to refer to a man that molests a girl as a paedophile (Salter 2003). Although most of the empirical evidence to date suggests that there is no link between homosexuality and child molestation (Herek 2012), a significant number of individuals continue to accept this stereotype (Wiley and Bottoms 2013). Niedwiecki (2013) notes that the scandal in the Catholic Church — that is the molestation of underage males by priests and the attempted cover-up by the church — has, in recent years, served to reboot and strengthen this narrative. Taken together, people who accept these negative stereotypes about lesbians and gay men would be less inclined to extend basic human rights to people that they believe threaten their way of life, their health, and their children.

2.2 Individual-level correlates of anti-gay prejudice

Currently, there is a well-established body of research examining individual differences in anti-gay attitudes in the West, and the principal findings of this research have been summarised several times (see Andersen and Fetner 2008b; Bhugra 1987; Herek 1991, 2000b; Lee and Hicks 2011; Mason and Barr 2006; van den Akker et al. 2013). These studies suggest that attitudes towards lesbians and gay men and their rights can be predicted by demographic variables, religion, interpersonal contact with gay men and

lesbians, beliefs about the origin of homosexuality, and attitudes towards other minorities.

2.2.1 Demographic variables There is no shortage of scholarly work on the impact of demographics on anti-gay bias. Most empirical studies include measures of gender, age, and education as predictors. Gender is arguably one of the most researched demographic predictors. Numerous studies have shown that, on average, heterosexual men tend to display higher levels of sexual prejudice than heterosexual women (Gato and Fontaine 2016; Herek 1990, 2000a, 2002; Herek and Capitanio 1999b; Kite 1984; Kite and Whitley Jr 1996; LaMar and Kite 1998; Petersen and Hyde 2010, 2011). The literature also suggests that there is a second gender gap in attitudes towards homosexuality, which relates to the sex of the target (that is, whether it is a gay man or a lesbian). In a meta-analysis of 112 studies on anti-gay attitudes, Kite and Whitley Jr. (1996) found that heterosexual men displayed more negative attitudes towards homosexual people and behaviours than heterosexual women. However, men reported more negative attitudes towards gay men than they did lesbians, whereas women reported the same level of acceptance or prejudice towards the two sexual orientations Various theories have been used to explain why men tend to be more prejudiced than women. Among the most popular is that of gender role beliefs and the defensive function of anti-gay prejudice, both of which were discussed earlier in this chapter. To recap, homosexuality is often conflated with gender inversion and viewed as a violation of gendered norms. People who hold on to traditional gender role ideologies are thus more likely to dislike lesbians and gay men. The defensive function suggests that some people may be hostile towards homosexuals to solidify their status as a ‘real’ man or a ‘real’ woman (Schwartz 2010). The literature suggests that men tend to be more negative to homosexuals, gay men in particular, since men adhere to gender roles more rigidly than women (Guittar and Pals 2014; Whitley Jr 2001), and they are also more likely to feel the pressure to conform to gendered norms than women (Herek 2002), making them more defensive. This line of reasoning bears much similarity to early theories about gendered identities. To summarise, in many cultures, masculinity is often defined as anti- feminine (Kimmel 1994), and throughout their lives, men feel pressured to assert their masculinity to demonstrate that they are not women (Connell 1987, 2005). Feminity is also associated with certain behaviours, but these are not as restricted as masculinity

(Badinter and Davis 1997). Under this construct of manhood, masculinity is exclusively heterosexual, which in turn generates negative attitudes towards men whose behaviours contradict this definition and also creates a need for men to ascertain their heterosexuality. Another popular correlate of attitudes towards homosexuality is education. The traditional and dominant view in the literature is that there is a positive relationship between education and liberal values, such as tolerance of diversity (Fosse et al. 2014; Gross and Fosse, 2012; Phelan et al. 1995). The relationship between education and social liberalism is largely explained through theories of cognitive upgrading and a socialisation mechanism. The cognitive model argues that education increases an individual’s general knowledge, stimulates critical thinking, and expands people’s frame of reference (van den Akker et al. 2013). Educated people are better able to contend with new ideas, and to assess information in an unbiased manner (Schwartz 2010), which in turn might induce tolerance for those who do not conform to what is considered the ‘norm’. Higher education is also accompanied by increased awareness of human diversity, and this increased awareness legitimzes variations in values, behaviours and beliefs – particularly those that are different from the individuals’ (Stubager, 2008). The socialisation model, on the other hand, proposes that individuals who enrol in educational institutions are exposed to liberal values, which they internalise. The literature suggests two mechanisms through which higher education expose students to liberal values. Some authors place emphasis on the interactions between peers. Specifically, attending university expands one’s social circle and informational sources (Ohlander et al. 2005). University graduates are thus more likely than less educated people to interact with different types of individuals and, as a result, they learn more about diverse cultures and types of people. In addition to meeting different kinds of people, university students are also more likely to interact with individuals with different socio- political views. Thus, changes in attitudes can come about via direct peer pressure (Campbell and Horowitz 2016), where sustained exposure to diverse peers and diverse views may cause some students to question their views of the social and political world (Fosse et al. 2014). Other researchers propose that attitudes can be transmitted to students through the actual teaching of liberal values such as tolerance (Stubager, 2008; Phelan et al. 1995). This often comes about from the implementation of diversity courses, as well as courses that promote the importance of tolerance (Radloff, 2007).

These lines of reasoning suggest a straight-forward hypothesis about the relationship between education and attitudes towards homosexuality. Specifically, it suggests that educated people, largely university graduates, should hold more positive attitudes towards gay men and lesbians. The literature to date confirms this, with several studies showing that college graduates tend to report lower levels of anti-gay prejudice and are more supportive of gay and lesbian rights (Baunach 2012; Loftus 2001; Ohlander et al. 2005; Perry 2013; Schwartz 2010; van den Akker et al. 2013). Recent studies have also produced evidence that age is an important determinant of attitudes towards gay men and lesbians. This is not surprising, as age is a widely-used predictor of social/political liberalism, with research suggesting that younger persons tend to more liberal than older persons (Tilley and Evans, 2014). When studying the differences in social and political attitudes across age groups, special attention should be devoted to two processes: the effects of ageing and cohort effects. The ageing effects model suggests that a person’s values and preferences change as they get older. As noted by Tilley and Evans (2014), ageing is often associated with increased authoritarianism, closed-mindedness and resistance to change, and this has traditionally been used to explain the positive relationship between age and conservative values. Age effects can also be a consequence of social changes that occur over the life cycle, where events over person’s lives (such as getting married, having children and retiring) are thought to be directly connected with opinions (Keleher and Smith 2012). Taken together, under the ageing effects model, the observed differences in attitudes between younger and older persons are largely explained by differences in their positions in the life cycle. The alternative to the ageing effects model is that of cohort effects. Here, differences in attitudes across age groups are often presented as a consequence of the unique social and economic circumstances that members of a birth cohort experience at the same time, particularly, at a point in time when they are in the process of forming opinions (Pew Research Center, 2015). Under the cohort effects model, persons belonging to younger cohorts (for instance millennials) tend to be more liberal since they encountered more liberal social structures when they were ‘coming to age’ that is, the point in their life cycle when personal identities and value systems were being shaped (Hart-Brinson, 2014). In line with the literature on the impact of age and social liberalism, recent studies have shown that the elderly tend to hold more negative attitudes towards gay men and

lesbians than younger people (Andersen and Fetner 2008a, 2008b; Crockett and Voas 2003; Keleher and Smith 2012; Schwartz 2010; van den Akker et al. 2013). Although the impact of age is now a strong predictor of anti-gay prejudice in the US, studies suggest that this is a recent development. Using data from the General Social Survey, Baunach (2012) and Sherkat et al. (2011) analysed how a range of factors structured support for same-sex marriage in the US between 1988 and 2010. Both studies report that in the late 1980s, Americans from nearly all birth cohorts/ages had similar, largely negative, views on same-sex marriage. Support for same-sex marriage has increased in recent years (2000s) and became more localised to specific groups, among which are younger people/birth cohorts. Sherkat et al. explained that there was a shift in attitudes in the post-1980s that came on the heels of increased awareness, positive media attention, and the ethos of sexual tolerance in higher education. They also suggested that this social change had a greater impact on the formation of values in younger people than it did in the older generations. Thus, the impact of age on attitudes towards homosexuality has largely been explained as evidence of cohort effects (largely because younger people were raised further along in the period of the gay rights social movement) rather than an ageing effect. Some studies even report that attitudes towards homosexuality have liberalised across cohorts (Andersen and Fetner 2008a; Duffy n.d.; Keleher and Smith 2012), which is inconsistent with the hypothesis that an individual’s attitude towards homosexuality becomes more conservative as they age. Rather, the liberalisation of attitudes across cohorts is indicative of a period effect – where events and social forces impact everyone in the society, regardless of their age (Pew Research Center, 2015).

2.2.2 Religion Religion is irrevocably bound up with the concept of right and wrong and, so, is often correlated with public values (Fairbanks 1977; Kellstedt and Green 1993; Layman 2001; Pettys 2010). It has a strong impact on the believer’s world views, what they classify as moral issues, and, by extension, their definition of which areas of life should be governed by law and which areas should be determined by individual choice (Davis and Robinson 1999; Evans and Kelley 2004b; Kelley et al. 1993; Pettys 2010). Most religions promote tolerance through brotherly love and care: many have texts that contain Golden Rule-like passages — that is, that we should treat others as we ourselves would want to be treated (Vilaythong et al. 2010). Taken at face value, one would expect that religious groups would be more willing to extend basic rights to same-

sex couples. However, in addition to tolerance, religions also emphasise the importance of other values and behaviours that are considered central to maintaining social cohesion and favourable relations with a higher power. As such, many people justify their anti-gay stance under the premise of religious morality. One of the most consistent findings in the academic literature is that religious individuals tend to be less accepting of lesbians and gay men than the religiously unaffiliated and, by extension, less willing to grant them civil liberties (Walls 2010; Whitley Jr 2009). Whilst religious people tend to be more disapproving of homosexuality, several studies have also reported drastic differences both across and within religious traditions. This most likely reflects the fact that some denominations tend to be more liberal than others on a range of issues. It is generally accepted that people who belong to more conservative denominations tend to display more negative attitudes than those belonging to more liberal denominations (Adamczyk and Pitt 2009; Cadge et al. 2008; Hooghe et al. 2010; Sherkat et al. 2011). For example, research on the US suggests that within the Christian faith, mainline/liberal Protestants have been cited as the most accepting of homosexuals, whereas Evangelicals tend to show the most opposition to same-sex intimate behaviours (Loftus 2001; Olson et al. 2006; Schulte and Battle 2004; Sherkat et al. 2010; Whitehead 2013). Arguably, differences in theological liberalism may shape how various religious sects interpret the Bible, as well as their stances on gender and sexuality. For example, Evangelicals tend to adhere to literal interpretations of the Bible (Gordon 2006), and this has been used to explain the lack of tolerance found among this group. A case in point is that Evangelicals often cite the Old Testament account of the destruction of Sodom and Gomorrah as evidence that those who engage in same-sex sexual acts will face God’s wrath (Burdette et al. 2005). Thus, sexual morality is not presented as a matter of personal choice but as something that can affect the whole society. Conservative sects also emphasise gendered norms, that is, the idea that there are distinct roles for men (leadership) and women (nurturance) (Bean and Martinez 2014), and this gendered belief system contributes to their intolerance of same-sex unions. Along with religious belonging, researchers have also linked religious involvement (most notably, the frequency of religious service attendance) to attitudes towards sexuality. As noted by Scheepers et al. (2002), this line of thinking was largely derived from Durkheimian theory on social integration, which suggests that the more one is integrated into a social group, the more likely one is to conform to its teachings and norms. From this, one would expect that those with greater religious exposure would

be more likely to see homosexuality as wrong and to oppose gay rights. To date, research on the US and Europe supports this view (Adamczyk and Pitt 2009; Gerhards 2010; Hayes and Dowds 2015; Herek and Capitanio 1995; Hicks and Lee 2006; Jaspers et al. 2007; Olson et al. 2006; Perry 2013; Rowatt et al. 2009; Schwartz 2010; van den Akker et al. 2013; Whitehead and Perry 2016). However, a shortcoming of most studies on the impact of religion on anti-gay attitudes is that they assume an exogenous relationship between religiosity and anti-gay attitudes. Whilst a person’s views of homosexuality alone are unlikely to cause them to change their religious affiliation, it may influence where they worship or how frequently or fervently they practice. The observed relationship between religion and anti-gay attitudes could be due to selection effects. To address this issue Yen and Zampelli (2016) used a selection effects framework to examine the relationship between religiosity and support for gay rights in the continental US. Specifically, using an ordered probit model7 and incorporating an ‘importance of religion’ variable as an ordinal endogenous treatment, the authors evaluated the effects of religious affiliation and attendance at services of worship on support for same-sex marriage. The instrumental variables included8 a measure of risk aversion, an indicator of respondent values, and measures of the respondent’s religious background (that is, the religion they grew up in and the frequency of religious service attendance while growing up). Yen and Zampelli (2016) found that even after controlling for selection effects, religion was still a significant predictor of attitudes towards gay rights, suggesting that religion has independent and direct effects on attitudes. They found that high religious service attendance (at least once a week) increased the likelihood of an individual strongly opposing same-sex marriage, regardless of the current importance of religion in their lives. They also found that across religious traditions, Evangelicals had the greatest probability of opposing same-sex marriage and that the impact of Evangelicalism was roughly the same across all levels of religious importance.

2.2.3 Interpersonal contact Intergroup contact theory is one of the most prevalent theories of prejudice. Though originally developed by Allport (1954) to study racial attitudes and tolerance, this theory has proved useful in explaining prejudice against a wide variety of marginalised groups

7 The model included controls for education, age, gender, marital status, contact with gays and lesbians, and political ideology. 8 In treatment models, instrumental variables are needed to allow for model identification. 42

(Pettigrew and Tropp 2006). The basic idea is that interpersonal contact with members of an out-group is one of the most effective ways of reducing in-group bias towards that out-group. By interacting with members of the out-group, the in-group learns about the out-group and acquires new information that may challenge the stereotypes and over- generalisations that shape their attitudes (Skipworth et al. 2010). With less stereotypical information, direct attitudes towards the out-group may become more tolerant. The contact hypothesis is argued to be more powerful in reducing prejudice against gay men and lesbians than it is in reducing prejudice against other minority groups (Herek and Capitanio 1996; Overby and Barth 2002). This is because homosexuality is a ‘concealable’ stigma, meaning that persons can get to know gay men and lesbians before knowing their sexual orientation (Herek and Capitanio 1996). Recently, a significant amount of cross-sectional research has been dedicated to studying the impact of interpersonal contact on attitudes towards sexual minorities (Smith et al. 2009), and most of these studies have found evidence that is in line with the predictions of the contact hypothesis. Specifically, heterosexuals who have interpersonal contact with gay men and lesbians tend to express more positive attitudes towards homosexuals and their rights than do individuals with no contact (Barth and Parry 2009; Baunach et al. 2009; Becker 2012; Becker and Scheufele 2011; Herek and Glunt 1993; Lewis 2011). Moreover, the number of contacts also matters: contact with two or more gay people has been linked to more favourable attitudes than contact with only one person. As noted by Herek and Capitanio (1996), having multiple contacts means that heterosexuals have increased opportunities for learning about what it means to be gay, and for observing behaviours or characteristics that are inconsistent with stereotypes of gay men and lesbians. Such individualisation is believed to reduce intergroup prejudice. There is also evidence of an intimacy effect — the closer the connection, the greater the impact. It is often theorised that having a gay friend or family member is more likely to be associated with less anti-gay bias than it is with contact with gay acquaintances (Becker and Scheufele 2009; Becker 2012). By having close relationships with homosexuals, heterosexuals gain intimate information, and this personalisation is believed to assist in reducing prejudice (Herek and Capitanio 1996). Cross-sectional studies with findings favouring the interpersonal contact hypothesis always raise the question of reverse causality. It is entirely possible that heterosexuals have gay friends or chose to have gay friends because they had pre-existing favourable views of gay men and lesbians. There could also be a selection effect as gays

and lesbians are more likely to befriend and reveal their sexual orientation to heterosexuals from whom they expect a positive response (Herek and Glunt 1993). The only sure way to identify the direction of causality is via panel data. The work of Anderssen (2002) addressed this issue. Using two-year longitudinal data, Anderssen assessed attitudes towards and contact with lesbians and gay men among Norwegians. The author found evidence that causality runs in both directions: an increase in contact was associated with a positive attitude change and attitude change also predicted contact.

2.2.3 Beliefs about the origins of homosexuality Some authors focus on how beliefs about the origins of homosexuality influence anti-gay bias. Indeed, the crux of the argument for gay rights in the US often concerns whether homosexuality is a choice or is genetic (Whitehead 2010). The link between beliefs about the origins of homosexuality and anti-gay bias is usually explained within the context of attribution theory. Attribution theory supposes that the perceived cause or controllability of behaviour influences how individuals view a stigmatised group or behaviour (Weiner 1979; Weiner et al. 1988). The theory predicts that attitudes towards a stigmatised group or behaviour tend to be more negative when individuals view the cause as a personal choice, rather than an uncontrollable condition. In line with attribution theory, several researchers have found that causal attributions for homosexuality are key determinants of attitudes towards homosexuals and their respective rights. Specifically, the cross- sectional studies to date strongly suggests that people who think that same-sex attraction is a choice (controllable) tend to show more prejudice towards gay men and lesbians than those who believe it is innate (Haider-Markel and Joslyn 2008; Hans et al. 2012; Lewis 2009; Sakalli 2002a; Tygart 2000; Whitley Jr 1990; Wood and Bartkowski 2004; Whitehead 2010). Yet, it is possible that attribution may not be the engine that drives attitudes. Instead, the observed correlation in cross-sectional research could be that of reciprocal causation: pre-existing attitudes towards homosexuality may be driving whether or not a person will accept biological explanations of homosexuality (Lewis 2009). Frias-Navarro et al. (2015) proposed an alternative strategy to test the impact of attribution. Using a sample of Spanish university students they investigated whether or not individuals were more likely to change their attribution positions if they were exposed to certain types of evidence about the origins of homosexuality. Their experimental manipulation showed that beliefs about the causes of homosexuality could be changed with exposure to

information that was labelled as ‘scientific’. Participants who were given a text that described homosexuality as genetically inherited increased their support for this etiology. Those who were exposed to an environment etiology text lowered their support for the genetic causation of homosexuality. In the post-test, people who were exposed to the environmental argument also reported lower support for same-sex marriage and adoption. This does lend some credence to the view that attribution could affect tolerance.

2.2.4 Attitudes towards other groups Some of the academic literature suggests that people holding negative attitudes towards other minorities are also more likely to be prejudiced against lesbians and gay men. Indices of anti-gay attitudes have been shown to be correlated with indices of racism (Henley and Pincus 1978; Ficarrotto 1990; Kite and Whitley Jr 1996; Laythe et al. 2001), sexism (Davies 2004; Sakalli 2002b), and ageism (Aosved and Long 2006). This could be because many of the variables that predict anti-gay prejudice are also correlated with other forms of prejudice. For example, Altemeyer (2002) claims that for some individuals, anti- gay attitudes have little to do with homosexuality per se, but are more related to personality traits that lead to a general intolerance of minorities and stigmatised groups. These personality traits are often subsumed under the labels of right-wing authoritarianism (RWA) and social dominance orientation (SDO) (Altemeyer and Hunsberger 1992; Altemeyer 2002; Whitley Jr and Ægisdóttir 2000). Individuals with high levels of RWA are often characterised as submissive to legitimate authorities, aggressive towards out-groups when authorities permit such aggression and adhering to values and beliefs endorsed by the established authorities (Altemeyer 1981; Whitley Jr and Lee 2000). On the other hand, SDO refers to the extent to which an individual prefers hierarchy within a social system, specifically, the domination of his/her in-group over the out- groups (Pratto et al. 1994; Smith 2012). Both measures have been shown to be highly correlated with various forms of prejudice, including anti-gay bias (Herek 1991; Ekehammar et al. 2004; Whitley Jr 1999; Whitley Jr and Ægisdóttir 2000). Other individual variables that simultaneously predict sexual prejudice and other forms of bigotry include age, religiosity, education, and political orientation (left or right) (Borooah and Mangan 2007). Given the propensity for some people to be intolerant of various out- groups, it is not surprising that anti-gay bias is often correlated with other forms of prejudice.

2.3 Beyond the individual – Contextual effects

Not only does anti-gay prejudice vary across individuals, but there are also significant differences across countries. Specifically, the Pew Research Center (2014b) surveyed populations in 39 countries and found large cross-national variances in responses to the question of whether homosexuality should be accepted or rejected by society. Figure 2.1 maps the percentage of persons that state that homosexuality should be accepted by society in each country surveyed. The map shows that there is country-level variation, but also some degree of regional clustering. Based on this map, there appears to be a higher level of acceptance in North American and Western European countries. Rejection seems to be more prominent in Sub-Saharan Africa, Asia, and the Middle East. Differences in population compositions are unlikely to be the sole cause of these large variations in cross-national attitudes. Rather, the differences across countries may also reflect specific national circumstances, that is, contextual effects (van den Akker et al. 2013).

Figure 2.1: Acceptance of homosexuality around the world

Note: Image obtained from the Pew Research Center (2013).

Contextual effect theory would suggest that an individual’s behaviour and attitudes are affected by the social, economic, and legal properties of a population (Huckfeldt 1986). There is currently work to suggest that national differences in anti-gay attitudes are related to the country’s level of socioeconomic development, religiosity, and laws concerning homosexuality.

2.3.1 Socioeconomic development The theoretical support for a link between living conditions in a country and attitudes towards homosexuals is largely derived from Inglehart’s post-materialist thesis (Inglehart 1990, 1997). The central claim is that once sufficient levels of economic security have been reached, individuals shift their value priorities from an emphasis on survival to a concern for the values of personal freedom and self-expression. Those who experience scarcity maintain materialistic values, whereas those whose basic needs are fulfilled are more likely to develop post-materialistic values, which leads to greater tolerance of diversity (Inglehart and Baker 2000). Inglehart’s thesis implies that the worldview of people living in developed and developing countries differs profoundly (Inglehart 2006). In particular, the economic security resulting from socioeconomic development frees individuals to reconsider social norms and exclusion. As noted by Inglehart (2006):

“Rising resources mean that there’s enough to go around. Newcomers can be accommodated. Foreigners seem much less threatening than when survival is a zero-sum game, and societies become increasingly tolerant toward foreigners and other out-groups. As they attain high levels of existential security, people come to take survival for granted… instead of being threatening, different cultures come to be seen as interesting and stimulating” (Inglehart 2006: 26)

In a nutshell, socioeconomic development reduces the need to fight for survival and, so, makes out-groups less threatening. This, in turn, results in tolerance for various out- groups, such as gay men and lesbians. Recent empirical studies lend credibility to the notion that people living in more modern economies will express less anti-gay prejudice than those living in less developed countries. In their studies of homonegativity across Europe, Štulhofer and Rimac (2009), Slenders, Sieben and Verbakel (2014) and Gerhards (2010) all report that economic development significantly increases social tolerance and/or acceptance of homosexuality. Similarly, in their multilevel analysis of attitudes towards homosexuality in 79 countries,

Jäckle and Wenzelburger (2015) finds that approval of homosexuality tends to be higher in states with higher levels of human development. In addition to explaining differences in attitudes towards homosexuality across nations, Inglehart’s work on modernisation provides additional explanations for the significant cohort differences found in attitudes towards homosexuals in the industrialised countries. For example, in his cohort analysis of Britain, France, West Germany, Italy, Belgium and the Netheralnds, Inglehart (1971) reported that ‘materialist’ values were very high in the older cohorts and that ‘post-materialist’ values tend to be more prominent in younger cohorts. He noted that the process of modernisation altered the value priorities of different generations. The younger cohorts were socialised in an environment where survival was secure and therefore took survival for granted and placed more emphasis on self-expression than did older generations. If this holds true, then a logical assumption would be that the link between age and attitudes towards homosexuality may be more significant in developed countries than developing countries. Specifically, due to differences in value priorities brought about by development, the younger generations in the West may be more receptive to pro-gay messages. However, in less developed countries, materialist values may still be prominent among younger cohorts. Hence, the generation gap could be smaller in developing countries.

2.3.2 Religious culture Whilst sustained economic development increases the chances of citizens adopting post- materialistic values, a shift in attitudes is not guaranteed — traditional values may persist post-development. In their analysis of modernisation and the maintenance of traditional values, Inglehart and Baker (2000) note that situation-specific factors, such as religious values, can leave an imprint on values that endure despite economic security. A common view is that as a country undergoes economic development, religiosity declines (Barro and McCleary 2003; Inglehart and Welzel 2005). However, although religion may become less important after modernisation, it does not become irrelevant (Henshaw 2014). As pointed out be Inglehart and Welzel (2005):

“The fact that a society was historically shaped by a Protestant or Confucian or Islamic cultural heritage leaves an enduring impact, setting that society on a trajectory that continues to influence subsequent development – even if the direct influence of religious institutions is modest today” (Inglehart and Welzel 2005: 22).

The persistence of religious norms, even in the face of modernisation, could allow traditional values to persist. In fact, the seminal works of Max Weber (Weber 1958) and Samuel Huntington (Huntington 1993) imply that in some instances the religious tradition in a country provides more insight about citizen values than its level of economic development. This means that differences in the religious climate of a country could further explain cross-national differences in attitudes. As noted earlier in this chapter, many religious institutions promote traditional values and condemn homosexuality. Hence, people living in religious countries9 are more likely to be exposed to anti-gay rhetoric than those living in more secular countries — regardless of whether or not they are personally religious (Adamczyk and Pitt 2009). This occurs for two reasons. First, in religious countries religious values are not transmitted by the religious institutions alone: the educational systems, mass media, and legal codes are also infused with religious norms (Norris and Inglehart 2011), among which could be opposition to homosexuality. In secular countries, religion has a less important role in culture and in legal codes and therefore tends to have less influence on traditional values and norms (Slenders et al. 2014; van den Akker et al. 2013). Second, in religious countries, there is a larger pool of religious people and therefore both religious and non-religious people are more likely to socialise with religious people. Such interactions could reinforce anti-gay sentiment among the religious, and for the nonreligious, they could increase their likelihood of conforming to the anti-gay ‘norm’ (Slenders et al. 2014; van den Akker et al. 2013). For this reason, it is often hypothesised that individuals living in more secular countries will be more tolerant towards homosexuals than people living in religious countries. Slenders et al. (2014), van den Akker et al. (2013) and Kuntz et al. (2015) all found evidence in support of this hypothesis. Specifically, they all report that Europeans living in countries with lower levels of religiosity (aggregate church attendance) tend to report greater levels of tolerance towards homosexuality than individuals living in countries with higher levels of religiosity. Religious tradition is also expected to be important. As I alluded to earlier, religious denominations often differ in their levels of conservatism and, by extension, their restrictions on sexuality. Hence, it is expected that citizens of countries where a majority subscribe to one of the more conservative religious denominations would be less accepting of homosexuality than those in countries where liberal religious traditions

9 In this text, the term religious countries refer to countries with high levels of religiosity in the population, as well as those countries where there are strong ties between religious groups and the state. 49

dominate. Štulhofer and Rimac (2009) report that people living in European countries with an Orthodox Christian tradition were more likely to object to having a homosexual neighbour than those with a Protestant or Catholic religious heritage. Further, in their analysis of 33 countries, Adamczyk and Pitt (2009) found that, relative to individuals living in countries with a Muslim majority, people living in Catholic countries consistently showed less anti-gay prejudice.

2.3.3 Laws A common theory is that differences in attitudes across countries could stem from differences in the regulation of homosexuality. As alluded to earlier, a consensus in the literature is that there is a bi-directional relationship between public attitudes towards homosexuality and policies affecting the rights of gays and lesbians: policy responds to opinion, but at the same time opinions are also affected by policy. Work on the impact of laws on attitudes is largely framed within the context of the expressive function of laws. Here, laws are presented as a key mechanism through which policy-makers signal what are ‘good’ values or what is the acceptable norm (McAdams 2000a, 2000b). Once a law is in place, it creates feedback effects, which are either normative or backlash. In the normative feedback model the values signalled by policy-makers are internalised by citizens and thus the laws shape public attitudes and behaviours. Yeh (2012) contends that this happens because most people prefer not to deviate too much from existing social norms, so they adjust their attitudes and behaviours to be in line with the newly established norm. This is similar to McAdam’s attitudinal theory (McAdams 2000b). According to McAdams

“People care about attitudes of approval and disapproval, but make mistakes about such matters; legislation is correlated with public attitudes so that the enactment of legislation provides a signal of public attitudes; and those who observe the signal will update their prior beliefs about public attitudes in the direction of expecting more disapproval for behavior the law condemns. Expecting disapproval for the behavior provides an incentive, independent of legal sanctions, to comply with the law.” (McAdams 2000b:372)

Thus, under the normative model, laws that prohibit anti-gay discrimination and promote equality across the sexual orientations would lead to increased intolerance towards homosexuals.

To test this theory, Kreitzer et al. (2014) utilised a panel survey of Iowa registered voters and evaluated opinions on same-sex marriage immediately before and after the Iowa Supreme Court decision to legalise same-sex marriage on April 3, 2009. The first wave of the survey was conducted between March 23 and April 1, 2009, and the second wave was conducted between April 7 and April 14, 2009. Due to the short time-span, any change in attitudes would likely reflect the signalling power of the law, where attitudes change to avoid cognitive dissonance. Kreitzer et al. (2014) found that, on average, support for same-sex marriage increased after the ruling. However, not all respondents were receptive to the signals of the new law. The impact of legalising same-sex marriage was strongest on those whose characteristics would generally predict support for gay rights. Specifically, Democrats10, people with close gay contacts, non-religious people, non-Evangelical people, educated, and younger respondents were more likely to accept the court ruling and change their opinions on same-sex marriage. However, Kreitzer et al. (2014) also report that the increase in support for same- sex marriage found in their study was not short-lived. Subsequent data polls (Hawkeye Polls of Iowa) showed that post-April 2009, support for public same-sex marriage in Iowa continued to trend upward. Although in the short term attitudes may ‘change’ to avoid cognitive dissonance (Kuntz et al. 2015), the long run it is expected that the changed environment would lead people to reconsider their preferences. One of the principal mechanisms for normative policy feedback is exposure and, as pointed out by Gusmano et al. (2002):

“…the longer citizens interact with a policy or institution, the more comfortable they become” (Gusmano et al. 2002: 735).

It is also possible that the enactment of such laws could result in public backlash. A change in the law is said to have a backlash effect when attitudes shift against the law’s values (Chen and Yeh 2014). Backlash is most likely to occur in the subset of people whose characteristics would predict that they would strongly oppose the values being signalled by the law (Chen and Yeh 2014). In this way, it is possible that if a majority of a population holds very conservative views about sex and sexuality, there could be significant backlash from homosexual law reform. Significant backlash could also occur if the legislative changes are not imposed by democratic institutions that are thought to be in line with the public will (Flores and Barclay 2016), but, rather, are imposed by

10 In the US, identification with the Democratic Party is normally associated with more liberal values. 51

institutions such as the Supreme Court or the European Court of Human Rights. However, if public trust in these institutions is high, then the possibility of public backlash is unlikely. Indeed, the literature suggests that trust in policy-makers is the main determinant of whether or not the values expressed by laws are internalised (Ellickson 2001; McAdams 2000b). A few studies have recently emerged to test the relationship between laws and differences in attitudes across Europe (Hooghe and Meeusen 2013; Kuntz et al. 2015; Slenders et al. 2014; Takács and Szalma 2011; van den Akker et al. 2013). These studies all find evidence in favour of normative feedback effects: citizens of European countries with liberal gay and lesbian policies tend to be more tolerant of homosexuals than those with more restrictive policies. However, from these models it is hard to determine whether or not the observed relationship is indicative that pro-gay laws have a legitimising impact on attitudes, or if the models are just picking up policy consensuses — that is, if the laws are simply the enactment of the majority opinion. In an attempt to differentiate between these two effects, two of the above studies (Slenders et al. 2014; Takács and Szalma 2011) complemented their multi-level analyses with impact monitoring, focusing on attitudinal changes following law reform Takács and Szalma (2011) compared the temporal changes in attitudes between 2002 and 2008 in European countries where: (1) same-sex marriages or registered partnerships had been legally recognised before 2002; (2) same-sex or registered partnerships were recognised between 2002 and 2008, and (3) countries without same- sex partnership legislations during the review period. They found that there was a significant increase in the social acceptance of lesbians and gay men in the first two groups of countries. However, there was no significant change in attitudes for countries without same-sex partnership legislation. Similarly, Slenders et al. (2014) compared developments in the average tolerance towards homosexuality in three types of countries: (1) countries that introduced laws on same-sex marriages or recognised partnerships before 1999; (2) countries that implemented these laws between 1999 and 2008; and (3) countries without this legislation during the review period. The results of Slenders et al. (2014) hint at a bi-directional relationship between laws and attitudes. They found that countries that implemented same-sex union laws between 1999 and 2008 (group 2) showed significantly higher levels of tolerance towards homosexuality in 1999 than countries without these laws (group 3). This implies that the tolerance levels of group 2 were higher to start with and that public

opinion could have driven the change in legislation. Slenders et al. (2014) also reported that for all three groups of countries, attitudes liberalised over the period, but the liberalisation of attitudes in countries that implemented same-sex union laws before 1999 (group 1) was statistically greater than that for the other two groups. They concluded that laws could indeed influence attitudes, though after some time, which in line with the exposure mechanism.

2.4 Attitudes towards homosexuality in the Commonwealth Caribbean

The vast majority of the reviewed empirical research was sampled from Western countries. The extent to which these findings can be applied to countries within the Commonwealth Caribbean has not been well established. Only a small body of work has been published on anti-gay bias in the Caribbean, most of which tends to be qualitative/ethnographic in nature (Gromer et al. 2013). However, these studies suggest that some of the theories about anti-gay prejudice in the Global North also hold in the Caribbean. Specifically, sexual prejudice in the Commonwealth Caribbean has been argued to serve a value-expressive function and is also assumed to be related to the perceived threat of homosexuality. A general theme in the published work on the Caribbean is that anti-gay prejudice is largely driven by conservative religious beliefs (Douglas 2007; Gaskins 2013; Gutzmore 2004)11. Public discourse on homosexuality often echoes the sentiments of the church (Cowell 2011). Homosexuality is cited as immoral, unnatural, and a sin so grievous that God destroyed a nation (Sodom and Gomorrah) because of it (Gutzmore 2004). Yet although Caribbeans hold conservative views on homosexuality, they are more lenient on non-homosexual sexual ‘sins’, such as pre-marital sex, adultery, and sexual promiscuity by heterosexuals. This is surprising, as attitudes towards homosexuality has been presented as part of an underlying dimension of sexual attitudes (see for instance, Finke and Adamczyk, 2008). Under such models, negative attitudes towards homosexuality are argued to be the consequence of an underlying dimension measuring sexual conservatism. This does not seem to be the case in the Caribbean. Rather, sexual

11 The religious environment will be discussed in detail in the following chapter 53

conservatism mainly showcases itself in Caribbean attitudes towards homosexuality. In her description of sexuality in the Caribbean, Kempadoo (2009) noted that

“Caribbean sexuality is both hypervisble and obscured…it is celebrated in popular culture as an ingredient in Caribbean social life and flaunted to attract tourists to the region, yet it is shrouded in double entre, secrecy and shame” Kempadoo (2009:1).

As will be shown in Chapter 6, sexual prowess and promiscuity are celebrated in Caribbean popular culture, but, homosexuality is strictly condemned. This suggests that the Caribbean does not show consistently permissive or consevative sexual attitudes, which would be expected in these conservative religious contexts. Some Caribbean researchers have argued that the mismatch between religious values and attitudes towards sex could be the result of the Africanisation of Christianity. Christianity largely emerged in the Caribbean during European colonialism (Bisnauth 1996). Although it is possible that the African slaves and freemen who turned to Christianity were completely orthodox in their understanding of the religion, a common belief among Caribbean academics is that many of them only had a partial grasp of the faith (Bisnauth 1996). It is often argued that many converted African slaves and freemen understood Christianity in terms of religious ideas based on their cultural and spiritual beliefs in Africa. In her analysis of the sexual mores of Jamaica, Lafont (2001, 2009) argues that many of the slaves did not fully embrace all European Christian sexual ideologies. Interpretations of sexual morality were not focused on fornication or adultery (probably due to the fact that slaves were initially not allowed to marry). Rather, they were confined to forms of unnatural sex, such as same-sex sexual relations. Today, although the Church condemns extra-marital affairs and promiscuity, it is possible that the Africanisation of Christianity served to create a set of sexual mores where homosexuality is ranked as a more grievous sin than adultery or promiscuity by people of the opposite sex. Researchers have also speculated that anti-gay prejudice (particularly prejudices against gay men) in the Caribbean is rooted in social perceptions of gendered identities (Atluri 2001; Hope 2010; Lewis 2003; Maiorana et al. 2013). In the Caribbean, promiscuity with women is argued to be a key feature of Caribbean masculinity (Kempadoo 2009) and therefore men who have sex with men challenge the cultural beliefs about what constitutes socially appropriate male behaviour (Maiorana et al. 2013). It is quite possible that these ideas about what it means to ‘be a man’ are also related to

why people are strict about male homosexuality but show leniency towards other sexual sins. A subset of the Caribbean literature suggests that negative attitudes towards gay men and lesbians are related to the perceived threat of homosexuality. Similar to Western countries, many people in the Commonwealth Caribbean often conflate homosexuality with paedophilia, and causally link homosexuality with HIV/AIDS (Genrich and Brathwaite 2005; Gutzmore 2004; Rutledge and Abell 2005; White and Carr 2005). Since these are considered serious threats to society, some individuals in the Caribbean may believe that the harsh condemnation of gay men and lesbians is somewhat justified. Some public officials in the region have even justified the anti-gay laws on the grounds of public morality and the need to protect the fabric of society. For example, after UK Prime Minister, David Cameron, stated in 2011 that the UK Government could cut or withhold foreign aid to countries with anti-gay laws (BBC 2011), the then Minister of Youth, Family and Sport in Barbados Stephen Lashley noted:

“The question is: are we going to surrender our tried and tested values for all these things that are now becoming the flavour of the international community? I say no.” (Marshall 2011)

Meanwhile, in 2007, religious leaders in Trinidad and Tobago unsuccessfully tried to convince the House of Assembly to bar Elton John — an openly gay singer and musician — from performing in Tobago’s national jazz festival for fear that he would encourage others to adopt his lifestyle (Daily Mail Online 2007). There are also documented cases of gay cruises being barred from docking in St Kitts and Nevis (Kiesnoski 2005). To date, very few studies have quantitatively investigated whether public attitudes and behaviours are in line with the qualitative/ethnographic research on the region or if they match the findings of the West. The few quantitative studies that do exist have predominantly sampled university students. Gromer et al. (2013), for example, studied attitudes towards gay men using a convenience sample of university students in Barbados. They did not find a relationship between contact with gay men or lesbians and sexual prejudice. However, they did find that men had more negative attitudes towards gay men, but there was no evidence of a gender gap with respect to attitudes towards lesbians. They also found that higher levels of religiosity were associated with greater prejudice toward both gay men and lesbians. Chadee et al. (2013) studied the attitudes of a sample of university students in Trinidad and Tobago and found that religious people were the most likely to be intolerant of homosexuality. Finally, West and Hewston (2012)

investigated the relationship between contact with, and attitudes towards gay men using a sample of Jamaican students and they compared this to the corresponding relationship for British students. They found that the Jamaican students reported more negative attitudes than their British peers, but, interestingly, contact was more strongly associated with a reduction in anti-gay prejudice for the Jamaican sample than for the Britons. This could be evidence that the impact of contact is stronger when attitudes are initially less favourable. However, the generalisability of studies using university students is questionable. University students tend to be younger and better educated than the national population, making it difficult to use these studies to draw inferences about the attitudes of the general public. West and Cowell (2015) have recently attempted to remedy this problem. Using nationally representative survey data on Jamaica they focused on the link between demographics (age, gender, education, and income), religiosity (measured by church attendance), music choice (specifically, whether or not the participant had a preference for dancehall12), and four measures of anti-gay prejudice: negative attitudes, social distance, opposition to rights, and negative behaviours. Each of these variables was created using summative scores of responses to various questions based on a five-point Likert scale (1 = strongly disagree, 5 = strongly agree). The negative attitudes index was based on three items about the emotional reactions towards gay people, and two questions about the immorality and sinfulness of homosexuality. Social distance was measured as respondents’ willingness to accept homosexuals in their lives in different social roles. Opposition to gay rights was based on respondents’ views about same-sex marriage, bans on homosexual behaviour, attitudes towards organisations promoting gay rights, and whether or not they felt that society should recognise homosexuality as ‘normal’. Finally, the negative behaviours index was based on items that asked participants if they have ever damaged the property of gay people, if they make derogatory remarks about or to gay people, if they tease gay individuals, and if they avoid them. Only gender, religiosity, and music choice were significantly related to all four measures of anti-gay prejudice. Men and those with a preference for dancehall music were more likely to express anti-gay prejudice than women or people preferring other genres of music. On the other hand, religiosity was associated with more negative attitudes,

12 As noted in the introduction, some of this music contains harsh condemnations of homosexuality. 56

greater social distance, and greater opposition to gay rights, but was also associated with less negative behaviours towards homosexuals. As suggested by the authors, the mixed impact of religion (that is associated with more negative attitudes but less negative behaviours) could be related to the ‘love the sinner, hate the sin’ mantra that is often repeated by public church leaders in Jamaica. However, the impact of age and education on sexual prejudice in Jamaica was a little different to that frequently reported by recent research on the US and Europe. West and Cowell (2015) found no evidence of a generational gap when it comes to attitudes towards, and comfort with, homosexuals. Even though feelings about, and being around, homosexuals were on par for younger and older generations, younger people were less likely to oppose gay rights but were also more likely to engage in anti-gay behaviour. However, the opposite holds for educated people. Participants with higher levels of education and income generally reported less negative attitudes, social distance, and negative behaviour towards homosexuals, but these feelings did not extend to educated people being willing to grant them civil rights. Taken together, this marks a deviation from the literature on the Global North where we often find that younger people and people who are highly educated general report less anti-gay prejudice in all forms. Although West and Cowell’s (2015) study on Jamaica has added to the paucity of empirical research using nationally representative data, they were unable to control for three popular predictors of anti-gay prejudice often used in studies based on the US and Europe. These are interpersonal contact, attitudes towards other minorities, and beliefs about the origin of homosexuality. The authors were also unable to evaluate the extent to which disapproval could stem from subscriptions to negative stereotypes that portray gays and lesbians as a threat to society. As noted earlier, the qualitative research on the Caribbean suggests that anti-gay prejudice is related to the perceived threat of homosexuality. In this thesis, I will try to fill this gap in the Caribbean literature by quantitatively assessing support for the laws in general, and as protectionist measures. I will evaluate the extent to which demographic variables, religion, interpersonal contact, beliefs about the origins of homosexuality, and attitudes towards other minority groups affect support for anti-gay laws in the region. Another observation from my review of the literature on the Caribbean is that, to date, there have been no quantitative studies explicitly focusing on how features of Caribbean society could be influencing attitudes. In the review of the related literature (Section 2.3), the country context — particularly the level of modernisation, religious culture and the

legal environment — has been cited as an important determinant of the level of prejudice in various countries. These variables not only impact the level of prejudice but could also impact the predictors of prejudice in a country. For example, modernisation could widen the attitudinal gap between older and younger generations. Thus, in more developed countries we may observe larger cohort effects. Activists, researchers, and social commentators in the Caribbean often cite widespread subscription to conservative religions and the presence of anti-gay laws as compounding anti-gay prejudice (CHRI 2011; Phillip and Williams 2013). It is therefore surprising that no studies have emerged to test this hypothesis. Of course, this could be due to the limited number of cross- national datasets that include a subset of Commonwealth Caribbean states in the sample. The recently released 2014 AmericasBarometer survey included respondents from six Commonwealth Caribbean countries in their sample and so allows us to explicitly test whether certain features of Commonwealth Caribbean society breed intolerance. This is important because developing effective policies to reduce sexual prejudice in these states requires a thorough understanding of the determinants of this prejudice at both the individual and country level. The quantitative methods, however, would only allow us to scratch the surface of Caribbean anti-gay attitudes. For example, the data and methods will not allow us to confirm whether or not condemnation of homosexuality is due to sexual conservatism. Furthermore, it is possible that there could be other motivators of anti-gay prejudice in the Caribbean that had not been captured in the quantitative enquiries. In an attempt to better understand the attitudes, I complement the analysis with an evaluation of the anti- gay rhetoric found in reggae and dancehall music. It should be noted that this music is indigenous to Jamaica and can be argued to be more reflective of Jamaican attitudes than that of other Caribbean states. However, songs with vicious condemnations of homosexuality are widely played throughout the region, suggesting that although the level of anti-gay prejudice in the other states may not be as intense as in Jamaica, anti-gay rhetoric is largely tolerated. For some individuals though, it is likely that the music is reflective of their views on sexuality. Thus, the analysis of the music is used to provide additional insights into how some individuals justify their prejudice.

2.5 Summary

This chapter has summarised the state of research on attitudes towards lesbians and gay men. The review has revealed that anti-gay prejudice tends to be higher among men, those who belong to a conservative religion, those who are older, the less educated, those with little or no interpersonal contact with gay men or lesbians, people who do not believe that homosexuality is innate, and those who are generally also prejudiced towards minorities. The review has also highlighted how contextual variables may shape these attitudes. It suggests that people in more developed and secular countries tend to report lower hostility towards homosexuals than those living in less developed and/or more religious states. More than this, public policies towards gay men and lesbians in the country are also important. A limitation of the literature is that the vast majority of the studies, particularly those using national level data, are sampled from Western societies. This raises the question: to what extent can these findings be applied to the countries within the Commonwealth Caribbean? The research by West and Cowell (2015) suggest that there could be a difference in how certain variables predict various forms of anti-gay prejudice in Jamaica, but studies on other states are needed to complement this research. However, based on the review of the literature it is likely that the context in which Caribbeans’ live may not only influence their attitudes but also affect the individual level predictors. Specifically, the modernisation hypothesis speculates that although development leads to more liberal attitudes, it could also widen the gap between the attitudes of the younger and older generations. We have also learned that in very religious countries, people might be influenced by the dominant religion in their country even if they do not subscribe to that faith. Hence, the impact of religious identity on anti-gay attitudes may be smaller in countries with a monopolistic conservative religious environment. Further, although less regulation of homosexual behaviour tends, on average, to increase tolerance, it is also likely that the message enshrined in the laws will not be fully internalised by everyone, leading to further gaps in attitudes among some groups. Taken together, it would appear that living in a developed country that is more secular and has more liberal policies on homosexuality could lead to more tolerance in general, but could also exaggerate differences between religious identities and socio- demographics. As will be shown in the next chapter, Commonwealth Caribbean countries are classified as developing nations, have a large Evangelical presence, and still carry bans on intimate same-sex acts. Thus, although the variables that often predict anti-

gay prejudice in Western contexts could be useful in predicting attitudes towards homosexuality within the Caribbean, it is also possible that they may not. By focusing on the Commonwealth Caribbean, this study can evaluate the extent to which findings on the Global North also work in other contexts and the ways in which they do not.

Chapter 3 : The socio-legal situation of sexual minorities living in the Commonwealth Caribbean: Some stylised facts

Any discussion of attitudes towards lesbians and gay men in the Commonwealth Caribbean must account for the environment in which these sexual minorities live. As a preliminary step to the empirical analyses, this chapter highlights the legal and socio- cultural conditions that strengthen, reduce and/or reflect anti-gay prejudice in these states. Specifically, I provide an overview of the legal status of same-sex intimacy, level of socio-economic development, religious culture, media and activism in the region.

3.1 State-sponsored discrimination

The anti-gay laws of the Commonwealth Caribbean can generally be classified as part of the dark legacy of British imperialism. As is well documented, Britain had a long tradition of criminalising ‘unnatural’ sex (West and Woelke 2002). Buggery (that is, anal sex) was first criminalised under the 1553 Buggery Act13 and was punishable by death. Buggery remained a capital offence until the enactment of the Offences against the Person Act 1861 (West and Woelke 2002), which lowered the punishment for the ‘abominable’ crime of buggery to life imprisonment, or a minimum of ten years. While the early British laws criminalised penetrative sex between men, male homosexuality was not the sole target of the law. It was also an offence to commit an act of ‘buggery’ with a female or with any animal. The first laws to explicitly target male homosexuality came in 1885. Section 11 of the 1885 Criminal Law Amendment Act said:

“Any male person who, in public or private, commits or is a party to the commission of, or procures or attempts to procure the commission by any male person of, any act of gross indecency with another male person, shall be guilty of a misdemeanour, and being convicted thereof shall be liable at the discretion of the Court to be imprisoned for any term not exceeding two years, with or without hard labour” (Criminal Law Amendment Act, 1885)

Though the 1885 laws gave no formal definition of what exactly constituted gross indecency, in practice, acts of gross indecency have often been interpreted as all intimate

13 The Buggery Act was repealed that same year (in 1553) on the accession of Queen Mary, but re-enacted by Queen Elizabeth I in 1563. 61

acts between men other than anal sex (Waites 1998). The decriminalisation of intimate acts between men came decades later, under the 1967 Sexual Offences Act (Han and O'Mahoney 2014). Needless to say, the Commonwealth Caribbean territories did not escape their colonial masters’ condemnation of homosexuality. In an attempt to bring European morality to these ‘uncivilised’ colonies (Gupta and Long 2008), Britain exported its views on sex to its colonies (LaFont 2001) and subjected its colonies to various anti-sodomy provisions. Today, these laws are often seen as representative of Caribbean culture. Since 1986, several Caribbean nations have revised their laws governing sex crimes (Robinson 2009), and these revisions have been largely influenced by independence movements and contemporary constitutional writings. While the laws are largely unreformed in some territories, as will be shown in the following paragraphs, some countries have sharpened their definition of buggery to explicitly target men who have sex with men and revised acts of indecency to target all same-sex couples. Thus, several decades after independence,14 Commonwealth Caribbean states have chosen to carry laws that not only define what acceptable sex between consenting adults is, but also label heterosexuality as the ‘norm’ (Alexander 1994), the only exception being the Bahamas. As mentioned in Chapter 1, the Bahamas is the only state that has decriminalised consensual (private) homosexual practices. The Bahamas is an interesting case study. In 1989, the Bahamian Government amended its Sexual Offences and Domestic Violence Act, replacing the colonial laws from 1861 (Gaskins 2013). At the time, the then National Security Minister Paul Adderley said that the changes to the bill sought to limit choices on sexual preferences (The Nassau Guardian 2014). The new laws were stricter and more explicit than the previous colonial laws. Specifically, under Section 16 of the 1989 Sexual Offences and Domestic Violence Act, ‘buggery between any two persons’ and ‘unnatural connection with any animal’ now carried a prison term of 20 years (Alexander 1994; Gaskins 2013). The amendment also introduced bans on sex between women, and women found guilty of the ‘offence of lesbianism’ were liable to a prison term of 20 years (Alexander 1994; Gaskins 2013). However, members of parliament (MPs) were divided on whether or not the government should be concerned with private sexual acts. For instance, in contrast to the sentiments expressed by MP Paul Adderley, MP Tennyson

14 Jamaica and Trinidad and Tobago gained independence from England in 1962, Barbados and Guyana in 1966, the Bahamas in 1973, Grenada in 1974, St Lucia and St Vincent and the Grenadines in 1979, Dominica in 1978, Antigua and Barbuda and Belize in 1981, and St Kitts and Nevis in 1983. 62

Wells believed that it was the Government’s place to police the private lives of Bahamians (The Nassau Guardian 2014). Two years later, the government amended the laws. The 1991 Bahamas Sexual Offences Act no longer criminalised anal sex or lesbianism between consenting adults in private, though Section 5b and Section 16 of the bill carried harsher punishments for same-sex sexual acts performed in public or with a minor than for heterosexual sex (Gaskins 2013). The bill also introduced different ages of consent for same-sex and opposite-sex sexual acts (Gaskins 2013). As noted by Gaskins (2013), the exact reason for the Bahamian government strengthening its opposition to homosexuality in 1989 and then changing its stance in 1991 is not clear. Gaskins (2013) also notes that at the time of law reform, the general public did not appear concerned enough to voice their concerns about law reform. However, this does not mean that the Bahamian public is more accepting of sexual minorities than the other Commonwealth Caribbean countries. As will be shown in Chapter 5, Bahamians are less supportive of same-sex marriage than Trinidadians or Barbadians. Also, years after law reform, religious leaders have organised protests against gay cruise ships docking in the Bahamas and petitions to reinstate sodomy laws and ban openly gays and lesbian people from holding positions in government (Gaskins 2013; The Nassau Guardian 2014). In 2005, Miss Teen Bahamas, Gari McDonald, was stripped of her crown after publicly coming out as a lesbian (The Nassau Guardian 2014), and in 2014, a gay pride event was cut short due to violent threats (McCartney 2014). Taken at face value, it does appear that the Bahamian public has become more concerned about homosexuality in recent years. On one hand, this could be evidence of a backlash effect, which was discussed in Chapter 2. Others attribute this rise in anti-gay sentiments to the rise in Evangelicalism. Specifically, Bahamian activist, Erin Green was quoted as saying

“Homophobia in these colonial communities is complex, but the starting point could be when American Southern Baptist churches started coming in, you started seeing the homophobia” (Gaskins 2013:439).

It is also possible that Bahamians are not really concerned about same-sex intimacy in private, but about gay men and lesbians openly expressing their sexuality. MPs who were against the laws in 1989 and 1991 mostly framed their arguments within the realms of the government ‘policing the bedroom business of Bahamians’ (The Nassau Guardian 2014). Thus, concerns about privacy outweighed concerns about ‘morality’, at least at that time. The recent backlash seems to be correlated with incidents that make gay men and lesbians

visible. Based on the anecdotal evidence, it may be inferred that gays and lesbians are tolerated in the Bahamas as long as they do not publicly assert their sexual identity. Regardless of the recent public backlash against homosexuality, the Bahamas continues to have the most progressive laws on homosexuality in the Commonwealth Caribbean. Today, Bahamian laws no longer carry differences in punishment between same-sex or heterosexual sexual acts in public or with minors.15 However, the current Bahamian legislation (that is, the Sexual Offences Act 2010) still carries different ages of consent for same and opposite-sex sexual acts. Under Sections 10 and 11 of the 2010 Act, an adult engaging in sexual activities with a person under the age of 16 is liable to life imprisonment. The wording of the laws is gender neutral, leading one to believe that the age of consent is 16 years. However, Section 16 of the Act specifically criminalises ‘sex with a minor of the same sex’. In the Act, a minor is defined as a person under the age of 18. In addition to this, the Bahamas does not carry any legal protections for sexual minorities: there are no laws prohibiting discrimination on the basis of sexual orientation in employment, housing, education or healthcare (Carroll 2016). In the remaining 11 Commonwealth Caribbean states, it is irrelevant whether the participants are adults or minors, whether the acts take place in public or private, or whether consent is involved—same-sex sexual acts are criminalised. However, apart from the illegality of same-sex sexual acts, anti-gay legislation in these 11 jurisdictions differs considerably (see Appendix A)16. As summarised in Table 3.1, Antigua and Barbuda, Barbados, Dominica, St Vincent and the Grenadines and Trinidad and Tobago have the crimes of buggery and acts of gross/serious indecency as part of their laws. Guyana, Jamaica, St Lucia, and St Kitts and Nevis criminalise buggery, attempted buggery and acts of gross/serious indecencies. Belize punishes acts ‘against the order of nature’; and Grenada has the crime of ‘unnatural connexion’. Furthermore, in addition to banning homosexual acts, Belize and Trinidad and Tobago also ban homosexuals from entering their countries in their Immigration Acts.17 Even in instances where states have similar anti-gay laws, the applicability of these laws varies. As noted earlier, Antigua and Barbuda, Barbados, Dominica, Guyana, St Lucia, St Vincent and the Grenadines and Trinidad and Tobago all criminalise acts of gross/serious indecency. The laws of Guyana and Jamaica largely resemble the 1885

15 This change came in 2008 under the Sexual Offences and Domestic Violence (Amendment) Act 2008. 16 The precise language of the laws criminalising same-sex intimacy is presented in Appendix A 17 The bans on homosexuals entering Belize and Trinidad & Tobago are presented in Appendix A. 64

Criminal Law Amendment Act and only prohibit acts of gross indecency between men. However, the legislation of Antigua and Barbuda, Dominica, St Lucia, St Vincent and the Grenadines and Trinidad and Tobago go a step further and prohibit ‘indecent’ acts between persons of the same-sex, thus also criminalising female-to-female sexual relations, a feature that was not in the original colonial laws. There is an interesting difference in the language used to ban ‘indecent’ acts between same-sex couples. Specifically, the laws in St. Vincent and the Grenadines explicitly state that acts of gross indecency between same-sex persons are punishable. In the remaining five states, the bans are subtle, as subsections of their laws point out that the bans on acts of serious/gross indecency do not apply to adult opposite-sex couples engaging in said acts in private. Currently, Barbados is the only state where acts of indecency are not limited to individuals of a specific gender or sexual orientation. A similar story emerges if one looks at the laws criminalising buggery. For instance, Antigua and Barbuda, Dominica, and Trinidad and Tobago describe buggery as anal sex between two males or by a male person with a female person, suggesting that two women engaging in anal sex is permissible. In St. Lucia, the crime of buggery is defined as sexual intercourse per anus by a male person with another male person. In Barbados and Guyana, buggery is left undefined and assumed to be applicable to all persons regardless of gender or sexual orientation or orifice(s) used. Finally, in Jamaica and St. Vincent and the Grenadines, buggery is not defined per se, but there is a resemblance to the 1861 laws which prohibit persons from engaging in buggery with another person or with an animal. Even in instances where the laws are neutral to gender or sexual orientation, they are often misinterpreted as applying to individuals of a specific sexual orientation and gender identity, rather than sexual acts (AIDS-Free World 2010; Gaskins 2013). For instance, Barbadians tend to use the terms ‘decriminalisation of buggery’ and ‘decriminalisation of homosexuality’ interchangeably (Abramschmitt 2008), even though the laws are not limited to anal sex between men.

Table 3.1: Penalties for private, consensual same-sex intimacy between adults in the Commonwealth Caribbean Country Law Offence (Applicability) Maximum penalty Antigua and Barbuda Sexual Offences Act of 1995 (Act No. 9) Articles 12 (1b) and 12 (2) Buggery (men; opposite sex persons) 15 years Articles 15 (1b), 15 (2) and 15(3) Serious indecency (same-sex persons) 5 years

Barbados Sexual Offences Act 1992 Chapter 154 Section 9 Buggery (undefined, assumed all persons) Life Sections 12 (1) and 12(3) Serious indecency (undefined) 10 years

Belize Chapter 101 of Belize Criminal Code (revised in 2000) Section 53 Carnal intercourse against the order of nature 10 years (undefined)

Dominica Sexual Offences Act 1998 Section 14 Gross indecency (same-sex persons) 5 years Section 16 (1b) and 16 (3) Buggery (men; opposite sex persons) 10 years

Grenada Criminal Code of 1987 (amended in 1993) Article 431 Unnatural connexion (undefined) 10 years

Guyana Chapter 8 of Criminal Law (Offences) Act Section 352 Gross indecency (men) 2 years Section 353 Attempt to commit unnatural offences (men) 10 years Section 354 Buggery (undefined) Life

Table 3.1 cont’d Country Law Offence (Applicability) Maximum penalty Jamaica Offences Against the Person Act Article 76 Unnatural crime [buggery] (humans or animals) 10 years Article 77 Attempted buggery (men) 7 years Article 79 Outrages on indecency [gross indecency] (men) 2 years

St Kitts and Nevis Offences Against the Amendment Act 1986 (revised in 2002) Section 56 Buggery (undefined) 10 years Section 57 Attempt to commit an infamous crime – that is 4 years attempted buggery (undefined) and indecent assault (men)

St Lucia Criminal Code No.9 of 2004 Sections 132 (1) and 132 (2) Gross indecency (same-sex persons) 10 years Sections 133 (1b) and 133 (3) Buggery (men) 10 years Sections 133 (2) and 133 (3) Attempted buggery (men) 5 years

St Vincent and the Criminal Code 1990 Grenadines Section 146 Buggery; (humans or animals) 10 years Section 148 Gross indecency (same sex persons) 5 years

Trinidad and Tobago Sexual Offences Act 1986, (amended in 2000) Sections 13 (1b) and 13 (2) Buggery (men; opposite sex persons) 25 years Sections 16 (1b) and 16 (2) Serious indecency (same-sex persons) 5 years Note: Information obtained from Human Dignity Trust (2015) and Carroll (2016)

The final difference in the laws comes in terms of the penalties assigned to these offences. Punishments for these crimes vary substantially. For example, penalties for penetrative sex (buggery) range from a minimum of ten years in Jamaica, Belize, Grenada and St Lucia to life imprisonment in Barbados and Guyana. At this juncture, it is important to point out that in most states, the laws are rarely used to prosecute persons on the basis of their sexuality. Legal, psychological and economic theory has traditionally focused on the deterrent or direct incentive effects of laws. A simplified version of the theory is that laws influence behaviours by imposing costs or payoffs that make particular behaviours more attractive than others (Bilz and Nadler 2009; Yeh 2012). Hence, if the laws are unenforced and agents are aware that this is the case, this can reduce the deterrent power of the law and can lead to the misguided perception that the laws are harmless. This line of reasoning often causes the public and legislators to believe that law reform is unnecessary (Leslie 2000). However, as noted in Chapter 2, laws are not just for deterrence: they are also key mechanisms through which policy makers signal community values, or what the social norm is or should be. Though unenforced, the anti-gay laws of the Caribbean signal to residents that homosexual behaviours are offensive and less accepted in society. They also help shape attitudes to homosexuality by prefiguring anti-gay values, prejudices and interests—particularly by referring to intimate same-sex acts as gross/serious indecency, unnatural or abominable. The laws prevent the legal protection of sexual minorities from discrimination and hate crimes and essentially relegate a subset of Caribbeans to a legal state of second-class citizenship.

3.2 Levels of socioeconomic development

As discussed in Chapter 2, the level of existential security and socioeconomic development in a country can influence anti-gay prejudice. Hence in this subsection, I briefly discuss the level of socioeconomic development in these states. Following Norris and Inglehart (2011) and Fernández and Lutter (2013), the level of existential security and socioeconomic development in a country is operationalised by the United Nations Development Programme’s (UNDP) Human Development Index (HDI). The HDI is a composite statistic of life expectancy, education and income per capita variables. This thesis uses the 2014 index, which ranged from 0.30 to 0.95 (UNDP 2015)18. The UNDP

18 At the time of writing, this was the latest data available. 68

classified countries into four levels of human development. Countries with an HDI value greater than or equal to 0.8 were classified as having ‘very high human development’ (49 countries attained this rank). Those with an HDI between 0.7 and 0.8 (56 countries) were categorised as countries with ‘high human development’. Countries with values in the range of 0.555 and 0.7 were placed in the ‘medium human development’ category (39 countries) and finally, those countries with an HDI value below 0.555 were classed as having ‘low human development’ (44 countries). Table 3.2 provides information on the levels of human development in Commonwealth Caribbean states, specifically their HDI values, their HDI category and their rank in the world. The table suggests that while none of these countries qualifies as having a ‘very high level of human development’, most states (11 of the 12) are in the ‘high level of human development’ category.

Table 3.2: HDI values, categories and rankings in the Commonwealth Caribbean in 2014 HDI value HDI rank (out of 188 countries) High level of human development Bahamas 0.790 55 Barbados 0.785 57 Antigua and Barbuda 0.783 58 Trinidad and Tobago 0.772 64 St Kitts and Nevis 0.752 77 Grenada 0.750 79 St Lucia 0.729 89 Dominica 0.724 94 St Vincent and Grenadines 0.720 97 Jamaica 0.719 99 Belize 0.715 101 Medium human development Guyana 0.636 124 Data obtained from 2015 UNDP HDI report (UNDP 2015).

As shown in Figure 3.1, the chance of a country having a ban on same-sex relations is lower for countries with higher scores on the human development index. However, modernisation does not guarantee that a country will not criminalise homosexuality. As noted in the literature review (Chapter 2), sometimes traditional values persist even in the presence of modernisation, partly due to religious tradition.

Figure 3.1: Percentage of countries banning same-sex intimacy by level of human development

70%

60%

50%

40%

30%

20%

10%

0% Low human Medium human High human Very high human development development development development

Data sources: The 2016 IGLA State Sponsored homophobia report (Carroll 2016) and the 2015 UNDP HDI report (UNDP 2015).

3.3 The religious climate

A common narrative in the published work on the Commonwealth Caribbean is that anti-gay prejudice is related to the region’s highly Christian climate (Douglas 2007; Gaskins 2013; Gutzmore 2004). In fact, the anti-gay laws and hostility towards homosexuality in the Caribbean are often justified under the guise of adhering to basic Christian principles (Chadee et al. 2013; Reding 2003; Smith and Kosobucki 2011). For instance, when asked why homosexual acts are illegal in Jamaica in 2010, former PM Bruce Goulding was quoted as saying:

“…we are predominately a Christian country and a fervently Christian country. It may not be reflected entirely in terms of how we live sometimes, but we are passionately committed to certain basic Christian principles” (AIDS-Free World 2010:20)

Christianity’s status as the dominant system of values can be characterised as a relic of European colonialism. European explorers were permeated by religious thought and brought Christianity to the Caribbean (Campbell 2011). The first doctrinal form of Christianity in the Caribbean was Catholicism, which was introduced after the ‘discovery of the new world’ by Spanish explorers in the late 1400s (Bisnauth 1996). Protestantism came later with the British colonisation of the Caribbean in the sixteenth century

(Bisnauth 1996). By the 1800s, Christianity had flourished throughout the Caribbean (Campbell 2011), and today it continues to serve as the dominant system of values. As shown in Table 3.3, a majority of Commonwealth Caribbean citizens claim a religious identity and Christianity has the largest number of adherents, though its distribution varies across countries.

Table 3.3: Religious identities in the independent Commonwealth Caribbean (population share in %) Christian Hindu Muslim Other Unaffiliated Not Stated Antigua and ≥ 82.8 0.4 0.2 ≤ 5.2 5.9 5.5 Barbuda Bahamas 95.0 0.1 0.1 0.3 1.9 2.6 Barbados 75.6 0.5 0.7 1.4 20.6 1.2 Belize ≥ 74.0 0.2 0.2 ≤ 9.5 15.5 0.6 Dominica ≥ 90.1 0.1 0.2 ≤ 2.4 6.1 1.1 Grenada ≥ 89.1 0.2 0.3 ≤ 9.9 3.6 0.0 Guyana 63.9 24.8 6.8 1.4 3.1 0.0 Jamaica ≥ 68.9 0.1 0.1 ≤ 7.3 21.3 2.2 St Kitts and Nevis ≥ 82.4 0.8 0.3 ≤ 8.1 5.2 3.2 St Lucia ≥ 92.7 0.2 0.1 ≤ 1.1 4.6 1.6 St Vincent and ≥ 81.5 0.1 0.1 ≤ 8.0 8.8 1.5 Grenadines Trinidad and ≥ 55.2 18.2 5.0 ≤ 8.5 2.2 11.1 Tobago Notes: (1) For some countries, the census information did not provide an ‘Other Christian’ category and it is unclear whether or not the ‘other’ category included some of the smaller Christian denominations. Hence, the use of ≥ and ≤ in the Christian and Other categories. (2) The data in this figure are based on the most recent publicly available census reports, specifically, the Antigua and Barbuda 2011 Population and Housing Census Book of Statistical Tables; the Bahamas 2010 Census of Population and Housing; Barbados 2010 Population and Housing Census Volume 1; Belize 2010 Population and Housing Census Country Report; Dominica 2001 National Census Report; Grenada 2001 National Census Report; Guyana 2012 Census Compendium; Jamaica 2011 Population and Housing Census Report; St Lucia 2000 National Census Report; St Kitts and Nevis National Census Report; St Vincent and the Grenadines 2001 National Census Report; and, Trinidad and Tobago 2011 Population and Housing Census Demographic Report.

Guyana and Trinidad and Tobago show the most religious diversification. Though Christianity is the largest religion in these two states, they also have a large population of Hindus: 24.8 per cent and 18.2 per cent of the population in Guyana and Trinidad and Tobago, respectively. There is also a significant number of persons who adhere to the Muslim faith, that is, 6.8 per cent of Guyana’s population and 5.0 per cent of persons in Trinidad and Tobago. Even with a significant non-Christian religious presence in Guyana and Trinidad and Tobago, Christianity is still considered the

dominant system of values. This can be traced back to how Hinduism and Islam first emerged in these states. After the abolition of slavery, colonial sugar planters looked for alternative sources of labour (van der Veer and Vertovec 1991) and brought labourers from India to the Commonwealth Caribbean. As is well documented, (Roberts and Byrne 1966; Singaravelou 1990; Tinker 1993), several other Caribbean territories also received Indian workers, but their shares were not as large as those of Guyana or Trinidad and Tobago. According to the latest censuses, Indo-Caribbeans represent the largest ethnic group in both Trinidad and Tobago (35.4 per cent) and Guyana (39.8 per cent), followed closely by Afro-Caribbeans (34.2 per cent in Trinidad and Tobago and 30.2 per cent in Guyana). This is very different from the ethnic makeup of the other Commonwealth Caribbean states (Figure 3.2). While Indian immigrants held on to their religious beliefs and practices, they were largely socialised under colonial oppression and a well-established Christian society (Naidu 2007). So, even though these religions survived in Guyana and Trinidad and Tobago, it was largely through cultural adaptation and to some extent religious syncretism (Naidu 2007). A recent report by the Hindu American Foundation (Hindu American Foundation 2013) noted that in Trinidad and Tobago, Hindus face restrictions on religious freedoms and the State has a preference for Christianity – evidenced by the inequitable funding of religious activities. So, while Guyana and Trinidad and Tobago are religiously diverse —at least in comparison to other Caribbean states— there is still some bias towards Christian values and principles.

Figure 3.2: Ethnic makeup of the Commonwealth Caribbean

Black Indo - Caribbean White Other ANTIGUA & BARBUDA

BAHAMAS

BARBADOS

BELIZE

DOMINICA

GRENADA

GUYANA

JAMAICA

ST KITTS AND NEVIS

ST LUCIA

ST VINCENT & GRENADINES

TRINIDAD & TOBAGO

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Note: (1) The ‘Other’ in Belize mainly consists of Mestizos/Spanish/Latinos and Creoles. (2) The data in this figure was are based on the most recent publicly available census reports. S, specifically, the Antigua and Barbuda 2011 Population and Housing Census Book of Statistical Tables; the Bahamas 2010 Census of Population and Housing; Barbados 2010 Population and Housing Census Volume 1; Belize 2010 Population and Housing Census Country Report; Dominica 2001 National Census Report; Grenada 2001 National Census Report; Guyana 2012 Census Compendium; Jamaica 2011 Population and Housing Census Report; St Lucia 2000 National Census Report; St Kitts & and Nevis National Census Report; St Vincent and& the Grenadines 2001 National Census Report; and, Trinidad & and Tobago 2011 Population and Housing.

Table 3.4: Popular Christian denominations in the independent Commonwealth Caribbean (population share in %)

Evangelicals Mainline Protestants Seventh Church Day Other Catholics Baptists of God Pentecostals Adventist Evangelicals Anglicans Methodists Antigua and Barbuda 8.2 3.6 4.1 12.2 12.4 2.3 17.6 5.6 Bahamas 12.0 34.9 1.9 8.9 4.4 n.a. 13.7 3.6 Barbados 3.8 1.8 2.4 19.5 5.9 n.a. 23.9 4.2 Belize 40.1 3.6 n.a. 8.4 5.4 n.a. 4.7 2.9 Dominica 61.4 4.1 1.2 2.5 6.1 6.7 0.6 3.7 Grenada 44.6 2.9 2.6 11.3 3.6 1.6 11.5 n.a. Guyana 7.1 n.a. n.a. 22.8 5.4 n.a. 5.2 1.4 Jamaica 2.2 6.7 25.7 11 12.0 1.4 2.8 1.6 St Kitts and Nevis 6.7 4.8 6.8 8.2 4.7 2.6 20.6 19.1 St Lucia 67.2 1.9 1.5 5.7 8.6 1.9 2.0 0.5 St Vincent and Grenadines 7.5 10.0 2.5 17.6 10.2 2.8 17.8 10.9 Trinidad and Tobago 21.6 6.9 n.a. 12.0 4.1 n.a. 5.7 0.7 Note: n.a. means not available. The data in this table are sourced from the most recent publicly available census reports, specifically, the Antigua and Barbuda 2011 Population and Housing Census Book of Statistical Tables; the Bahamas 2010 Census of Population and Housing; Barbados 2010 Population and Housing Census Volume 1; Belize 2010 Population and Housing Census Country Report; Dominica 2001 National Census Report; Grenada 2001 National Census Report; Guyana 2012 Census Compendium; Jamaica 2011 Population and Housing Census Report; St Lucia 2000 National Census Report; St Kitts and Nevis National Census Report; St Vincent and the Grenadines 2001 National Census Report; and Trinidad and Tobago 2011 Population and Housing Census Demographic Report.

Altogether, the Commonwealth Caribbean can be classified as religious states with a bias towards Christianity. This could serve as a barrier to the advancement of gay rights in these states. As discussed in the previous chapter, religious morality is often used to justify denying basic rights to sexual minorities. For instance, in Jamaica, recent efforts to expand gay rights have been met with strong criticisms from religious groups and conservative religious voters. On June 23, 2013, several Church leaders and church goers came together to protest against a challenge to the anti-gay laws, which was scheduled to be heard in the Supreme Court the next day. The challenge was initiated by Jamaican Javid Jaghai with the aid of the US gay rights advocacy group AIDS-Free World and questioned whether the bans on intimacy between men breached rights guaranteed under Jamaica’s Charter of Fundamental Rights and Freedoms, which was passed in 2011 (Jamaica Gleaner 2013). Then, roughly one year later, 25,000 Jamaicans came out in support of Jamaica’s anti-gay laws in a demonstration organised by the Jamaica Churches Action Uniting Society for Emancipation (CAUSE)19 mass rally. As reported by the Jamaican Observer:

“The rally called on Jamaicans to stand up for strong and healthy families, righteousness and justice as well as to resist the homosexual agenda and the repealing of the buggery act” (Skyers 2014)

Eventually, Mr. Jaghai dropped his lawsuit, allegedly due to threats made to him and his family (Stewart 2014). However, the aforementioned challenge to the laws is currently being revisited by Jamaican-born activist Maurice Tomilson, and this time around, the challenge is being supported by both the Canadian HIV/AIDS Legal Network and AIDS-Free World (Canadian HIV/AIDS Legal Network and AIDS-Free World 2015). According to recent reports, the renewed challenge is also being met with strong resistance from religious groups. In fact, on July 8, 2016, the Jamaican Supreme Court accepted applications from religious groups to stand as interested parties in the challenge. As noted by Maurice Tomilson:

“…the court found that the opinion of the majority of Jamaicans – as reflected by the religious groups – was important in deciding on what two consenting adults do in the privacy of their bedrooms”(Canadian HIV/AIDS Legal Network and AIDS-Free World 2016).

19 On their Facebook page (which was accessed on June 30, 2016) the Jamaica CAUSE describes themselves as a “grouping of Jamaican churches concerned about the advancing attack on our sovereignty and conscience, expression and religious liberty”. 75

There is also evidence to suggest that faith-based groups successfully thwarted policy initiatives aimed at reducing discrimination against gay men and lesbians in Guyana and Trinidad and Tobago. In 2001, the parliament of Guyana voted to include prohibitions that outlaw discrimination on the grounds of sexual orientation in an amendment to its constitution. However, many religious people opposed the bill. According to local newspaper Stabroek News, Guyana’s (IRO)20 called for the sexual orientation category to be deleted from the Constitutional Amendment bill (Stabroek News 2003). The article in the newspaper also noted that while the organisation noted that it opposed discrimination against sexual minorities, it believed that

“…it is God’s will in his creation of humanity, that sexual intercourse, an image of God’s own power of creation, should be practised only within the context of marriage between members of the opposite sex” (Stabroek News 2003).

Eventually, Parliament succumbed to religious pressures (Maitland 2013), and the amendment was passed in 2003 without any prohibitions on discrimination on the basis of sexual orientation (Naipaul 2003). Turning to the case of Trinidad and Tobago, at the end of 2012, various media outlets carried stories that the country’s new policy on gender and development would address discrimination against sexual minorities (Baboolal 2012; Kaleidoscope Trust 2012). However, as reported by the Trinidad and Tobago Guardian, prior to submitting a draft of the policy to Cabinet, the Government was warned by the IRO21 that it would not get support for the legislation of homosexuality (Baboolal 2013). Then, on May 18, 2013, the then Minister of Gender, Youth and Child Development Marlene Coudray, announced that the policy would not address any gay rights, and was quoted as saying:

“It really is not up to me…I convened a meeting with faith-based organisations and the IRO and its members and we had a very extensive, open and frank discussion with them. We had to extend the discourse to two days and we came to terms with what they would like to see (in the gender policy). They reminded us the Constitution is based on the supremacy of God. So those views are documented to be put forward” (Baboolal 2013).

The main takeaways from these examples are that (1) religious institutions constitute a strong anti-gay rights lobby (as suggested by some government officials—

20 The IRO represents a large number of diverse faith-based organisations in Guyana. It members include, but are not limited to, representatives from the Hindu, Muslim, Christian and Baha’i faiths. 21 Similar to the IRO in Guyana, the IRO in Trinidad represents a large number of diverse faith-based organisations. 76

see Chapter 1); and (2) religious bodies appear united on issues relating to homosexuality, which certainly strengthens their influence on the regulation of homosexuality. It is possible that the strong resistance to homosexuality in these states could also be related to the type of religion that dominates. As shown in Table 3.4, Evangelicalism is currently the largest Christian denomination in 8 of the 12 states. From the literature, two things were clear. Firstly, within the Christian faith, conservative sects, such as Evangelicals, tend to show most opposition to intimate same-sex relations. Secondly, the religious climate of a country can influence residents, regardless of whether they subscribe to the dominant religion. Based on the previous analyses, it is possible that there is some type of religious syncretism on the issue of gay rights. Actors—who may not even be religious—may be unconsciously taking cues from the dominant religion on the issue of homosexuality. It could thus be argued that the strong Evangelical presence is a major contributor to the maintenance of the laws and by extension, negative attitudes towards lesbians and gay men in these states.

3.4 The policing of sexual minorities in the media and popular culture

In the introduction, I noted that while the media can influence attitudes, it is also a reflection of societal values (Adamczyk et al. 2015). First media elites are not isolated from the wider society, and their views are likely to be influenced by the dominant culture in the country. Second, media elites are mindful of their core audiences and in their attempt to maximise profits, some media outlets may either opt to frame messages in ways that echo public sentiment or opt not to cover stories that challenge the status quo. Given the presence of anti-gay laws and large Evangelical sects in Commonwealth Caribbean states, it is somewhat expected that media coverage in the region will reflect the dominant narratives in these countries. In July 2014, the IGLHRC and United and Strong (a Lesbian, Gay, Bisexual, Transgender and Intersex (LGBTI) organisation in St Lucia) collaborated with various LGBTI organisations in Belize, Grenada, Guyana and Jamaica to conduct a LGBTI-focused media monitoring project in the five aforementioned Commonwealth Caribbean countries (IGLHRC and United and Strong 2015). The results of the monitoring exercise suggest that in these states, news coverage of LGBTI-related issues lacked nuance and coverage on LGBTI persons and issues were often sensationalised, negatively framed, and served to reinforce stereotypes. For example, the report noted that news stories often referred to an undefined ‘gay

agenda’ or concerns about an increase in sexual minorities. The media analysis also suggested that media outlets were largely silent on issues related to the LGBTI community – particularly policy issues or concerns about social exclusion – and failed to provide coverage that presented LGBTI persons in a positive light. In addition to the unbalanced and unflattering coverage of LGBTI issues by various news agencies, sexual minorities are also discriminated against in Caribbean popular music. Some of the most vehement condemnations of homosexuality come from the region’s iconic musical genres: dancehall and reggae. Dancehall and reggae are two styles of music indigenous to Jamaica, but widely popular throughout the Commonwealth Caribbean. On a visit to any of the 12 states, one is likely to be exposed to them: dancehall and reggae are played on the radio, in nightclubs, blasting in public transportation, the marketplace, in sports bars and local rum shops, etc. Though anti-gay rhetoric is present in both dancehall and reggae, the condemnations in dancehall tend to be more aggressive. This points to some of the differences between dancehall and reggae22. It is sometimes argued that dancehall is distinct from reggae due to its use of digital instrumentation, faster lyrics and less politically conscious rhetoric. Stanley (2005), for instance, maintains that ‘slackness’ (references to women’s body parts and men’s sexual prowess) and violence have emerged as a master narrative in dancehall music, a contrast to the cultural and conscious values that permeate reggae. In both genres, homosexuals are widely referred to using disparaging slurs (see Table 3.5). However, the use of homosexual slurs tends to be more prominent in dancehall, particularly in the ‘clashing’ aspect of dancehall, where artists use their music as a means of insulting and publicly shaming a rival. However, the main criticism of the music (mainly dancehall) is its (perceived) glorification of violence against homosexuals. Somewhat of a poster child for songs with violent anti-homosexual content in dancehall is Buju Banton’s hit Boom Bye Bye (Buju Banton 1993). The entire song is dedicated to (violently) condemning homosexuality, but the chorus is the real clincher:

22 It should be noted that not everyone is agreed on whether dancehall should be considered as separate from reggae (the older of the two genres), or just a reincarnation of reggae (Farquharson, 2005), as they are both used as forms of social commentary and criticism of the system (local and foreign). In fact, many persons use the terms dancehall and reggae interchangeably. 78

Lyrics Translation “(It’s like) Boom bye bye innaa batty bwoy head Shoot the gay man in his head Rude bwoy no promote no nasty man Rude boys don’t accept gay men Dem haffi dead” They have to die Note: Lyrics taken from http://www.lyricsmode.com. Translation by author.

Table 3.5: Popular derogatory terms used to refer to homosexuals in dancehall and reggae Terms Meaning Batty bwoy/man Batty is slang for the buttocks; bwoy means boy. So the term ‘batty man’ refers to a man who prefers the buttocks or who engages in anal sex with another man

Bottomist Similar to batty bwoy

Bugga man A man who engages in the act of buggery

Chi-Chi bwoy/man Gay man

Cocks man A man who prefers the penis

Fish A weak/inferior man

Freaki bwoy Gay man

Funni Guy A man who acts strange. He is not conforming to society’s norms. The term is used to describe homosexuals

Mauma/mama man An effeminate or weak man

Pungga-man Similar to mauma/mama man

Sadomite Often used to refer to a lesbian, but can also be used to describe a person who identifies as homosexual

Uman-man Similar to mauma/mama man

While Boom Bye Bye was not the first song with anti-gay lyrics, it was the song that ‘outed’ the overtly vicious condemnations of the lifestyle, at least in the international arena (Farquharson 2005). Soon after Boom Bye Bye’s 1992 release,23 dancehall quickly came under scrutiny by lesbian and gay lobbyists in the UK and the US. Buju Banton was not the only artist labelled as giving voice to anti-gay violence with murderous anti-

23 Boom Bye Bye was originally penned by Buju Banton back in 1988 when he was 15 years old. The song was re-released in the 1990s after he achieved international fame. 79

homosexual lyrics: other internationally renowned artists—for instance, Beenie Man, Bounty Killer, Capleton, Elephant Man, Shabba Ranks, Sizzla and Spragga Benz, just to name a few—have songs that appear to advocate the beating, hanging, shooting, drowning or burning of homosexuals. In fact, the music was labelled ‘murder music’ by LGBT societies and advocates, which led to the now infamous Stop Murder Music Campaign (henceforth SMMC)24. It should be noted that lyrical criticisms of homosexuality are by no means unique to dancehall/reggae. At the international level, traces of anti-gay speech have been found in other genres of music such as rap/hip-hop, country and pop (Friedrichs n.d.), and music indigenous to other Caribbean countries (like calypso and soca). However, the exceptionality of dancehall/reggae (to date) seems to lie in the ferocity of homosexual condemnations and the sheer quantity of songs featuring some type of anti-gay rhetoric, particularly during a period where global acceptance of homosexuality has been perceived to be on the rise. For instance, it has been estimated that during the early 2000s, in any given week, at least one of the top twenty hits in dancehall contained some form of anti- gay sentiment (LaFont 2009). The anti-gay lyrics caused a stir internationally, and this contributed to a significant decline in music condemning homosexuality (discussed later in Chapter 6). Though many artists currently refrain from releasing songs with anti-gay rhetoric (at least through mainstream media outlets), sometimes during an ‘old school’ music session, anti- gay songs from the 1990s may play on some local radio stations across the region. This hints that, even now, some Caribbeans are either tolerant of or indifferent to anti-gay rhetoric. In fact, the controversy concerning dancehall and reggae has been highly criticised. There have been theories of a mythical ‘international gay rights lobby’ who singled out Jamaica because of its convenience, economic vulnerability, national pride and/or hidden agenda of bringing same-sex tourism to Jamaica (Larcher and Robinson 2009). Adding to existing beliefs about Jamaica being scapegoated was a 2009 article by independent media outlet Alternet (Scherr 2009), which claimed that while iTunes blocked anti-gay songs such as Buju Banton’s Boom Bye Bye, it still featured music spewing messages of white supremacy. Other arguments have focused on the fairness of labelling dancehall as homophobic. Whether or not it was the intention of the SMMC, the entire genre was

24 To be discussed in Chapter 6. 80

perceived as an advocate of violence against gays (Moore 2014). It has been argued that though the number of songs with homophobic slurs or lyrical assaults was very high (as will be shown in Chapter 6), there were many more songs produced during the era that did not focus on homosexuality. More than this, homosexuals are not the only ones to be on the receiving end of violent condemnations: informers, gossips, ‘bad-minded’ persons, corrupt politicians, persons who engaged in oral/anal sex with their (heterosexual) partners, persons who practise Obeah25 and any perceived rival also faced overtly vicious condemnations. This brings us to the third major criticism, which was framed within the context of cross-cultural conflict. Individuals critiqued these ‘outsiders’ for imposing their interpretations and values on Jamaican music without fully understanding the context or tradition out of which the music emerged (Saunders 2003). For example, in a piece written for VIBE magazine, Joan Morgan (Morgan 1993) chastised dancehall’s critics for their “ignorance of Jamaican street culture” and their inability to “grasp the metaphoric richness of Jamaican patois”. Similarly, Cooper (1994) argued that the violence in dancehall lyrics cannot be taken literally, but that it largely falls into the realm of metaphor and role-play. With respect to the ‘murder music’ debate, Cooper (2004) argued that the music does not represent the artists’ intentions to end the lives of gays and lesbians or to incite violence, as the SMMC suggested. She noted that:

“…in its cultural context, this battle cry, which is appropriated by Buju Banton in ‘Boom Bye-Bye,’ primarily articulates an indictment of the abstraction, homosexuality, which is rendered in typically Jamaican terms as an indictment of the actual homosexual: The person (the homosexual) and the project (homosexuality) are not identical.” (Cooper, 2004: 160)

While one can agree that the music should not be universally labelled as homophobic and that the violent lyrics probably should not be taken literally, one cannot dismiss the argument that the music does spread messages of hate and intolerance. The fact that the music continues to be played on mainstream radio and in public spaces in many Caribbean states says a lot about attitudes towards homosexuality. While dancehall and reggae are indigenous to Jamaica and so are more representative of anti-gay bias in Jamaica than other Caribbean states, the continued popularity of this music in some

25 In the Caribbean, the term ‘Obeah’ refers to black magic, sorcery and a religious practice developed among West African slaves, and in this vein is somewhat similar to voodoo. 81

public spaces suggests one of two things: either the messages of the music resonate with members of the Caribbean public or they are largely indifferent to anti-gay rhetoric.

3.5 Support services for sexual minorities

Thus far, the thesis has provided an overview of the various features in Caribbean society that contribute to and/or reflect anti-gay prejudice. The charged social and legal atmosphere is indicative of the obstacles that lie ahead for advocates seeking to reduce anti-gay prejudice in these states. However, the situation is not monolithic. As suggested by Murray (2012), even in places where anti-gay prejudice is high, there are still safe/inclusive spaces for sexual minorities. Unfortunately, I was unable to find detailed information about the support services available for sexual minorities in the region. However, according to the Equality Network, most Commonwealth Caribbean countries have at least one local Lesbian, Gay, Bisexual, Transgender and Intersex (LGBTI) organisation that works to provide a safe and secure environment for sexual minorities and advocate for their rights (Equality Network n.d.). For instance, United and Strong, St Lucia’s sole LBTI organisation conducted human rights training for a subset of police, customs and correctional officers on both general and LGBTI-specific content, and the Jamaica Forum of Lesbians, All-Sexuals and Gays partnered with the Ministry of Health to train over 200 health workers and sensitise them to LGBTI patients in 2014 and 2015 (Stewart 2016). However, activists in the region report that limited funding and societal attitudes impede operations (Kaleidoscope Trust 2014; Kaleidoscope Trust and Kaleidoscope Australia Human Rights Foundation 2015). In recent years, LGBTI activists have begun partnering with the international LGBTI rights lobbyists, such as the Kaleidoscope Trust, UNAIDS, IGHLRC, AIDS-Free World and the Canadian HIV/AIDS Legal Network. There is evidence to suggest that these alliances have been helpful in advancing the fight for gay rights. For example, the previously mentioned challenges to Jamaican laws that criminalise buggery and acts of serious indecency (past and present) came about through the help of international advocacy groups. It should be noted that there were other challenges to the anti-gay laws that were also supported by international lobbyists. Specifically, in 2013, Maurice Tomlinson26, with the help of AIDS-Free World, launched a challenge to laws in Belize and Trinidad and Tobago that explicitly ban homosexuals

26 As noted earlier, Mr Tomlinson is leading the current challenge to the anti-gay laws in Jamaica. 82

from entering the country (see Appendix A) in the Caribbean Court of Justice (Tomlinson 2016). Mr Tomlinson argued that the two countries were in breach of the Caribbean Community (CARICOM) treaty that guarantees the free movement of CARICOM persons across the regional bloc (Tomlinson 2016). On June 10, 2016, the Caribbean Court of Justice denied Mr Tomlinson’s application for a declaration that the countries violated his right of free movement within Belize and Trinidad and Tobago, as he was never denied entry at either border (Tomlinson 2016). Even though the challenge was unsuccessful, this case can be considered an important milestone in the LGBTI liberation movement in the Commonwealth Caribbean.

3.6 Summary

This chapter has provided a description of the atmosphere in which lesbians and gay men in the Commonwealth Caribbean exist, paying attention to the anti-gay laws, the level of socioeconomic development, the religious climate, media and support services for sexual minorities. An important observation is that religious groups seem united in their opposition to gay rights. This differs from the current religious climate in the West, where religious denominations tend to be divided on the issue of gay rights, particularly same-sex marriage (Gerhards 2010). This religious alliance means that one cannot ignore the role of religion. It is possible that many could be conforming to the view of the dominant religion (Evangelicalism) on this issue. The large proportion of Evangelicals, coupled with the anti-gay laws, paints a picture of a charged social environment where anti-gay attitudes are prevalent. In fact, these two factors may serve to mitigate the impact of development in these states. However, as discussed in this chapter, the situation is not monolithic. There are a few organised groups in the region working to protect sexual minorities and engender change. While the existence of these groups is crucial to law reform, it is not sufficient. As noted by Corrales (2015), to obtain change, these organisations would need to (1) operate in countries where courts are progressive, (2) have strong political connections, and (3) frame the issue in a way that resonates with majoritarian sentiments. Based on the strong ties between religion and the state, the chances of LGBTI organisations establishing strong political ties or successfully arguing their cases in various courts look slim. Hence, I opt to focus on the latter requirement presented by Corrales (2015). My analysis in this chapter and the introduction suggests that on the issue of gay rights, policy makers relegate the issue to the public or faith-based

groups. Thus, in order to increase their chances of success, advocates must understand public attitudes towards homosexuality and gay rights. By understanding public attitudes, advocates would not only be aware of the challenges that lie ahead, but may also gain some insight on how to frame the issue of gay rights in a way that appeals to public sentiment. In what follows, I empirically evaluate the prevalence of, and contributors to, anti-gay prejudice in the three Caribbean states: Barbados, Guyana and Trinidad and Tobago.

Chapter 4 : Views on the ‘abominable’ crime: Heterosexual support for and perceived usefulness of the anti-gay laws in Barbados, Guyana and Trinidad and Tobago

The previous chapter provided an overview of the current socio-legal situation of lesbians and gay men in the Commonwealth Caribbean. Based on the analysis, it seems as though the environment is conducive to anti-gay prejudice. This finding prompts the following question: do public attitudes echo the sentiments of the laws, religion and music? To determine the prevalence of anti-gay prejudice in these states, national level survey data should be analysed. But, as in noted in Chapter 2 (section 2.4), most studies on attitudes towards lesbians and gay men in the Commonwealth Caribbean predominantly sampled from university student bodies This likely reflects the fact that in the Caribbean, nationally representative public opinion surveys are rare, and when they do exist, are often not publicly available. However, in 2013, CADRES conducted a national survey of attitudes towards homosexuals in three Caribbean states – Barbados, Guyana and Trinidad and Tobago. As such, this chapter uses this rare dataset to determine heterosexual views on banning same-sex sexual behaviours in these three states. At this juncture, I should point out that an analysis of only 3 of 11 of the Commonwealth Caribbean states with anti-gay laws cannot be seen as representative of the region. It would have been ideal to work with a dataset that included a greater number of Commonwealth countries, however, I was unable to attain additional data. Even with a limited databse, the analysis is still important. Particularly, it can be seen as a step towards filling the gap in the literature in terms of attitudes towards gay rights within the region. Against this backdrop, this chapter seeks to answer the following questions: 1) Do a majority of heterosexuals in Barbados, Guyana and Trinidad and Tobago support the bans on same-sex intimacy? 2) To what extent do the social and psychological factors identified in Chapter 2 (specifically gender, age, education, religion, interpersonal contact, beliefs about the origins of homosexuality and attitudes towards

other minorities) explain the views on the criminalisation of same-sex intimacy?

4.1 Heterosexual attitudes towards the criminalisation of same-sex intimacy in Barbados, Guyana and Trinidad and Tobago

To gain a better understanding of heterosexual views on the anti-gay laws in Barbados, Guyana and Trinidad and Tobago, I use secondary data obtained from the 2013 Caribbean Development Research Services (CADRES) ‘Attitudes towards homosexuals’ survey. According to the CADRES reports, the surveys employed stratified random sampling, with the primary strata being age and gender in each country (Caribbean Development Research Services 2013a, 2013b, 2013c)27. Interviewers were randomly assigned to areas associated with the polling districts each country. The polling districts are almost equal subdivisions of constituencies, which are normally used for electoral purposes. The polling districts were numbered, allowing for a random selection of specific divisions, which CADRES then translated into geographical districts. In each instance, interviewers selected 12, 24, 36 or 48 households in a random manner (1 in 3), and conducted interviews at each of the 12, 24, 36 or 48 households. One respondent in each of these households became the interviewee. The survey was largely administered via face-to-face interviews, though interviewers complied with requests from respondents to complete the questionnaire on their own. The section of the survey that asked about sexual orientation and interpersonal contact with lesbians and gay men was set aside for self-administration, as CADRES viewed these as confidential issues. Interviewers were advised to only complete this section of the form if requested by the respondent. A total of 2,871 adults (age 18 and over) were surveyed: 830 Barbadians, 1,034 Guyanese and 1,007 persons from Trinidad and Tobago. I focus on heterosexual support for the laws: I removed individuals who identified as gay/lesbian (71 persons) and bisexual (88 persons). A large segment of persons (287) chose the option ‘I prefer not to say’ when asked about their sexual orientation, and 37 persons did not provide a response. As the orientation of these individuals was unclear, I opted to remove them from the analysis (see Appendix B for details by country). This initially reduced the

27 I was unable to attain information about response rates. 86

sample size to 2,388 individuals, which included 689 Barbadians, 790 Guyanese and 909 persons from Trinidad and Tobago. Five questions from the CADRES surveys allow for the operationalisation of heterosexual views on the criminalisation of same-sex intimacy. These questions were formulated as follows:

 “Presently, the laws of (insert country name here) outlaw the act of Buggery/Sodomy, whether between two men or a man and a woman regardless of whether the act is in public or private, consensual or forced. Do you generally support the maintenance of this law?” o Yes, I think the buggery/sodomy laws should be maintained o No, I think the buggery/sodomy laws should be changed (removed or modified) o I am unsure/prefer not to say how I feel about this law

“Currently, the laws of (insert country name here) with respect to Buggery/Sodomy are NOT being enforced (except in instances of forced sex/rape). Do you think that the state should enforce these laws by investigating and prosecuting persons who engage in these acts (by consent)?”

o Yes, start enforcing the laws o No, continue with non-enforcement o Unsure/prefer not to say

 “Do you think that the laws of (insert country name here) should penalise two men for having ‘sexual intercourse’ in private?” o Yes o No o Unsure/prefer not to say

 “Do you think that the laws of (insert country name here) should penalise two women for having ‘sexual intercourse’ in private?” o Yes o No o Unsure/prefer not to say

 “Based on your understanding of the (Insert country name here) laws with respect to Buggery/Sodomy, which of these objectives do you think that the laws achieve in their present form?”

a. “The laws are a fair and reasonable expression of our moral standards.” o Yes o No o Unsure/won’t say

b. “The laws help to stop the ‘spread’ of homosexuality.”

o Yes o No o Unsure/Won’t say c. “The laws are important from a public health perspective.” o Yes o No o Unsure/Won’t say

d. “The laws protect young people from abuse.” o Yes o No o Unsure/Won’t say

The first four questions capture general support for the criminalisation of same-sex behaviours – that is the maintenance and enforcement of the anti-gay laws. The last question captures the perceived usefulness of the laws, specifically, law support that emerges from the perceived threat of homosexuality. As alluded to in Chapter 2, if persons genuinely believe that homosexuals or homosexual acts harm society, they would be more inclined to support the anti-gay laws. Quite a few survey participants did not respond to these questions, resulting in an analytical sample that stood between 2,210 and 2,350 observations (see Appendix B for more details).

4.1.1 General support for bans on same-sex intimacy Figure 4.1 shows the distribution of the responses to the first four questions. For the pooled data, population size weighting was used to ensure that each county is represented in proportion to its population size. The descriptive statistics suggest that 59 per cent of heterosexuals support the maintenance of the anti-gay laws as is, and a slightly higher share of persons (65 per cent) believes that authorities should begin enforcing the laws. Also, a significant proportion of participants (19-21 per cent) said they were unsure or declined to state their opinions about the maintenance and enforcement of laws. At first glance, the slightly higher support for the enforcement of said laws, relative to support for their retention, may seem surprising. It could be a measurement error or perhaps representative of a subset of persons who either did not understand one of the questions or were unsure about their opinions concerning the laws. However, it is plausible that a small share of respondents who do not support the maintenance of the laws in the present form may be supportive of changes to the laws (e.g. changes to the penalties or removing penalties for heterosexual couples who engage in anal sex) but do not think the laws should be abolished. It could also be capturing a subset of persons

who believe that laws should be enforced after being established, even if they do not agree with them. Interestingly, residents are much less supportive of the state penalising private sexual acts between persons of the same sex: 52 per cent of heterosexuals in this sample do not support the penalisation of two men for having sex in private, and 57 per cent oppose punishing two women for having sex in private. Also, persons seem more willing to state their opinions on these measures, as the share of individuals who selected ‘unsure/prefer not to say’ for these items is much lower (10-11 per cent). A further evaluation of the data reveals that of the persons who supported both the maintenance and the enforcement of the laws, 45 per cent of them oppose penalising men for having sex in private, and 51 per cent oppose the penalisation of sexual acts between women in private (Figure 4.2). Breaking down the sample even further, I discovered that 19 per cent of the sample consistently supported the laws, while only 7 per cent consistently rejected the laws and 9 per cent consistently chose the ‘unsure/prefer not to say’ option, confirming that many persons hold mixed views about these laws.

Figure 4.1: Heterosexual support for the anti-gay laws

10.0% 10.5% 21.6% 19.1%

16.0% 19.1% 51.9% 56.8%

64.9% 59.3% 38.1% 31.8%

Maintain the laws Enforce the laws Penalise two men in Penalise two women private in private Yes No Unsure/prefer not to say

Note: Based on weighted data

A similar story of mixed attitudes emerges when one examines the support for the laws across countries. As shown in Figure 4.3, across each state, most persons support the maintenance and enforcement of the anti-gay laws. However, at the same time, they do not support the penalisation of same-sex couples having consensual sex in private. The plots also suggest that across all states, persons seem more willing to penalise male homosexual acts than sex between women. They also hint that there may be some cross- country differences in support, as heterosexuals in Trinidad and Tobago consistently show the highest support for the criminalisation of same-sex sexual acts. Taken together, the analysis suggests that a significant subset of citizens living in Barbados, Guyana and Trinidad and Tobago hold mixed attitudes about penalising persons that engage in same-sex sexual relations. It is my view that though all four survey items arguably capture attitudes towards the criminalisation of consensual same-sex acts, the items may be tapping into two different value dimensions and, as a result, elicit different responses. For instance, when asked ‘general’ questions about the retention and enforcement of the anti-gay laws, persons may give responses that reflect their beliefs about homosexuality. However, by reformulating the question and asking about homosexual conduct in private, the latter two survey items lead respondents to think about the issue within the context of a government meddling in someone’s sex life, which would represent a gross intrusion into the personal and private life of an individual. So, while some persons may believe that homosexuality is wrong and believe that the laws should reflect this, they may simultaneously believe that individuals – regardless of sexual orientation – have a right to privacy. This view could make it difficult for them to hold a firm position on the bans on same-sex intimacy. An alternative explanation could be that individuals view the laws as a means of limiting public displays of affection between same-sex individuals or dissuading them from being open about their sexual orientation. Similar to the case of the Bahamas (see Chapter 3), it is possible that many individuals in Barbados in these states do not care if a person has sex with someone of the same sex in private, as long as he or she does not publicly assert their sexual identity. Hence, the observed mixed attitudes could be evidence of a typical ‘don’t ask, don’t tell’ type of tolerance.

Figure 4.2: Heterosexual support for the penalisation of sex between persons of the same- sex among those who want the current laws maintained and enforced

7.5% 7.4%

44.7% 51.3%

47.8% 41.3%

Penalise two men in private Penalise two women in private

Yes No Unsure/prefer not to say

Note: Based on weighted data

Figure 4.3: Heterosexual support for the criminalisation of same-sex intimacy by country

Barbados Guyana Trinidad and Tobago

11.9% 11.7% 11.2% 13.3% 9.1% 8.9% 21.2% 19.7% 18.3% 22.5% 17.9% 27.4% 12.9% 19.8% 19.3% 16.4% 25.8% 50.4% 53.4% 54.4% 55.1% 24.0% 60.9% 58.6%

69.2% 61.9% 59.5% 54.5% 61.2% 48.6% 34.8% 34.4% 40.4% 36.0% 27.4% 28.1%

Maintain the Enforce the Penalise two Penalise two Maintain the Enforce the Penalise two Penalise two Maintain the Enforce the Penalise two Penalise two laws (as is) laws men in women in laws (as is) laws men in women in laws (as is) laws men in women in private private private private private private

Yes No Unsure/prefer not to say

4.1.2 Perceived usefulness of the anti-gay laws In this subsection, I analyse the prevalence of the view that the current anti-gay laws protect the fabric of Caribbean society. Figure 4.4 shows the distribution of responses. Based on the analytical sample, 46 percent heterosexual respondents believe the laws are in line with moral standards, 46 per cent of the sample thinks the laws are important from a public health perspective and 40 per cent believe that the laws protect young people from abuse. The argument ‘Laws help to prevent the spread of homosexuality’ has the lowest number adherents, with only 23 per cent of respondents agreeing with this statement. A large share of respondents (20 to 27 per cent) was either unsure or did not want to state opinions on the usefulness of the laws. Figure 4.5 takes the analysis a step further and looks at support across the three countries.

Figure 4.4: Beliefs about the usefulness of the current laws

20.9% 20.2% 26.9% 27.2%

26.8% 27.0% 40.0% 56.5%

46.3% 45.9% 39.9% 22.6%

Reflect morals Stop spread of Important for public Protect children from homosexuality health abuse

Yes No Unsure/prefer not to say

Note: N = 2,210 to 2,226. Based on weighted data.

Figure 4.5: Beliefs about the usefulness of the current laws (by country)

Barbados Guyana Trinidad and Tobago

17.2% 22.5% 22.4% 17.6% 21.1% 21.6% 20.1% 28.6% 28.1% 24.3% 27.0% 27.7%

20.2% 24.1% 43.4% 20.3% 22.6% 36.0% 30.7% 29.2% 42.4% 52.1% 66.2% 56.3%

57.3% 53.5% 51.1% 49.3% 39.0% 43.1% 39.7% 42.4% 37.5% 26.8% 16.6% 22.1%

Reflect morals Stop spread of Important for Protect young Reflect morals Stop spread of Important for Protect young Reflect morals Stop spread of Important for Protect young homosexuality public health people from homosexuality public health people from abuse homosexuality public health people from abuse abuse Yes No Unsure/prefer not to say

In line with the pooled data, the ‘Laws are a fair and reasonable expression of our morals’ statement has the largest number of supporters in each country, followed by the statement that the laws protect public health. The statement that the ‘laws stop the spread of homosexuality’ has the least support. The plot also suggests that the intensity of anti- gay sentiment differs across the countries. For instance, Barbados had the largest shares of respondents agreeing that the laws are an expression of the moral standards and that the laws are important from a public health perspective, while respondents from Guyana were more likely to believe that the laws curtail the spread of homosexuality and protect children from abuse.

4.2 Who is likely to support the anti-gay laws?

The analysis thus far suggests that, in general, there are mixed attitudes towards penalising persons who engage in same-sex intimacy, which I speculated could be reflective of contradictory impulses. Also, a significant share of people perceives homosexuality as a threat to society and, as such, believes that laws are useful. In what follows, I outline the empirical approach that will be used to analyse the data further. As noted in earlier chapters of this thesis, the literature to date on attitudes towards gay rights in the Commonwealth Caribbean is very scant. As such, this thesis strongly relies on the previous literature (based on North America and Europe) to model the general support for the anti-gay laws and the perceived usefulness of the anti-gay laws in Barbados, Guyana and Trinidad and Tobago. Public support for the laws is thus modelled as a function of age, gender, education, religion, beliefs about the origins of homosexuality, interpersonal contact and attitudes towards other minorities. The following subsections describe how the variables are coded for the statistical analysis.

4.2.1 Description of dependent variables The dependent variables are responses to the five survey items described in the previous sections. For the empirical investigation, the dependent variables are dichotomised to form eight binary variables defined as follows: 1. ‘Maintain the laws’, which takes on a variable of 1 if the respondent supports the maintenance of the laws as is; 0 otherwise. 2. ‘Enforce the laws’, which takes on a value of 1 if the respondent believes that the laws should be enforced; 0 otherwise.

3. ‘Penalise men in private’, which takes on a value of 1 if the respondent thinks that the state should penalise two men for having sex in private; 0 otherwise. 4. ‘Penalise women in private’, which takes on a value of 1 if the respondent thinks that the state should penalise two women for having sex in private; 0 otherwise. 5. ‘Laws express moral standards’, which takes on a value of 1 if the respondent thinks that the laws are a fair and reasonable expression of moral standards; 0 otherwise. 6. ‘Laws stop the spread of homosexuality’, which takes on a value of 1 if the respondent thinks that the laws help to stop the spread of homosexuality; 0 otherwise. 7. ‘Laws important from a health perspective’ which takes on a value of 1 if the respondent thinks that the laws are important from a public health perspective; 0 otherwise. 8. ‘Laws protect young people from abuse’ which takes on a value of 1 if the respondent thinks that the laws protect young people from abuse; 0 otherwise.

4.2.2 Description of independent variables The dependent variables are modelled as a function of demographics (age, gender, education and country of residence), religion, beliefs about the origin of homosexuality, interpersonal contact, and discrimination against other minorities Table 4.1 provides some descriptive statistics. Most of the independent variables were directly taken from the survey questions (very few variables were derived) hence the form of the variables largely reflected the structure of the CADRES survey items. Specifically, respondents were presented with a few response options (derived by CADRES), with some variables including the response option ‘unsure/prefer not to say’. It should be noted that none of the items used in this study were multi-response. In the description that follows, I indicate and describe the variables that I derived or altered (for example: discrimination against other minorities). A general concern with surveys that give respondents a few response options is whether or not the response options are adequate. It could be the case that some persons may opt to not answer the question or even choose ‘unsure/prefer not to say’ as the response options given did not capture their views or situation. While in my opinion, the CADRES surveys cover popular responses, as shown in Table 4.1, the share of persons choosing ‘unsure/prefer not to say’ for some items was as high as 12 percent. It is quite possible that some persons chose this category as a default as their beliefs or feelings were not captured by the available response options. Perhaps a better approach would

have been to give respondents the option to state their views if different than what was being provided. This could be considered in future surveys.

Table 4.1: Descriptive statistics for predictors

Categorical Variables Share in (%) Gender Men 48.6 Women 51.4 Age 18-30 35.6 31-50 34.1 51 + 30.4 Education Primary/Secondary 55.9 Post-secondary 18.0 Tertiary 26.2 Religious Identity Evangelical Christian 36.5 Non-Evangelical Christian 25.6 Muslim 5.5 Hindu 12.7 Other religion 13.2 Not religious 2.7 No response 3.9 Religious participation Active 49.2 Passive 41.4 Unsure/Won’t say 9.4 Source of views on sexuality Religion 48.5 Socialisation 29.5 Popular Culture 10.7 Unsure/Prefer not to say 11.3 Contact No gay friends/family 47.1 Gay friend OR family 30.9 Gay friend AND family 10.6 Prefer not to say 11.5 Origin of homosexuality Born that way 16.9 Birth defect 5.4 Psychological trauma/abuse 14.8 Poor religious/moral groundings 16.1 Choice 34.3 Unsure/prefer not to say 12.4

Table 4.1 cont’d Continuous Variable Mean (standard deviation) Discrimination against minorities 0.2 (0.5) Based on weighted data

4.2.2.1 Demographics (gender, age, education and country of residence) The gender variable is in binary form, with men serving as the reference category. The CADRES survey captured age using three distinct age groups: (1) 18-30 (contrast), (2) 31-50 and (3) 51 and over. Education is divided into three categories: (1) compulsory schooling (that is, primary/secondary education), (2) post-secondary/vocational education and (3) tertiary education, where the base category is compulsory schooling. It should be noted that the education item on the survey initially had 5-categories (Primary, secondary, post -secondary, vocational and tertiary), but was simplified by the author. The recoding of the variable had no impact on the results. Finally, I added a three- category variable to account for any cross-country differences. As highlighted in Chapter 3, even though Caribbean states are highly similar, there are slight contextual differences. The three states evaluated here differ slightly in terms of their socio-economic development, the intensity of punishments for same-sex behaviours and their religious compositions – all of which can lead to differences in attitudes. All of these factors could lead to some difference in attitudes. Trinidad and Tobago is arbitrarily chosen as the reference category.

4.2.2.2 Religion Unfortunately, the CADRES dataset does not contain widely used measures of religiosity, such as the importance of religion to one’s life or the frequency of attendance at religious services. Nevertheless, three of the survey’s items can be used to test the role of religion in shaping anti-gay attitudes. First, respondents were asked about their religious identities, which in this study is presented as a 7-category variable: (1) Evangelical Christians (the reference category); (2) non-Evangelical Christians; (3) Muslim; (4) Hindu; (5) other religion; (6) religiously unaffiliated and (7) no response28. As shown in Chapter 3 (see Table 3.3), the religious categories presented here adequately cover the major religions in these three Caribbean countries. To recap, according to the latest publicly available census (presented in Chapter 3) in Barbados, over 90 percent of persons who identity

28 A significant portion of persons did not answer the question about religious affiliation. As such, I added a non-response category so the sample size could be maximised. 98

with a religion adhere to Christianity. In the other two countries, Christianity is also the largest religion, but they also have a large Hindu and Muslim sects. As noted in Chapter 3, about 18 percent of the Trinidad and Tobago population identify as Hindus and roughly 5 percent are Muslims. In Guyana, about 25 percent of the population subscribe to Hinduism and 7 percent are followers of Islam. The other religions in these states tend to be smaller, and together account for by less than 1 percent of the population in Barbados and Guyana, and less than 8.5 percent of persons living in Trinidad and Tobago. Second, participants were asked to identify their religious participation and were given three choices: (1) ‘active’ (which I use as the reference category), (2) ‘passive’ and (3) ‘unsure/won’t say’. Finally, respondents were asked about the main source of their views on human sexuality. Respondents were asked to choose one of the following (1) Religion (Holy Bible/Bhagavad Gita/Holy Koran); (2) Socialisation (family and friends); (3) Popular culture (music, literature, television and the social scene) and (4) I am unsure/prefer not to say. Religion served as the reference category.

4.2.2.3 Beliefs about the origins of homosexuality Following the measures used in the survey, this variable is a 6-item categorical variable based on the CADRES survey question, ‘In your opinion/to the best of your knowledge, what is at the origin of homosexuality?’ The six response options were given by CADRES were (1) Born that way (reference category); (2) Birth defect; (3) Chose to be that way; (4) Psychological trauma/Sexual abuse; (5) Lack of/Poor moral or religious groundings/Bad parenting and (6) I am unsure/Prefer not to say.

4.2.2.4 Interpersonal contact The interpersonal contact variable is constructed using two survey items. The CADRES survey asked if the respondents had any gay friends as well as any gay family members. For both questions, respondents were given the response choices ‘Yes’, ‘No’ and ‘Prefer not to say’. Responses were recoded to create a four-category variable, such that if a person did not have friends or family members who identified as gay, they were denoted as having no close gay associates. If they chose ‘Yes’ to only one category (‘Yes’ to having either gay friends or a gay family member), they were denoted as having a gay friend or family member. If they chose ‘Yes’ for both items, they were denoted as having both gay friends and gay family members. Finally, if individuals chose ‘Unsure/Prefer not to say’

for both categories or a combination of ‘No’ and ‘Unsure/Prefer not to say’, they were denoted as having an undetermined number of associations.

4.2.2.5 Discrimination against other groups To capture general intolerance of other groups, a variable was constructed using four survey measures. Specifically, the CADRES survey asked respondents whether or not it was acceptable to treat persons differently on the basis of (1) Race; (2) Gender; (3) Religion and (4) Age. To each of these queries, respondents were given the choices of (1) Yes, acceptable, (2) No, not acceptable and (3) Unsure/Won’t say. Some descriptive statistics for responses to these items are presented in Table 4.2.

Table 4.2: Responses to questions about discrimination against minorities

Yes No Unsure/Won’t Say Race 6.8 87.9 5.3 Gender 9.3 84.4 6.3 Age 16.0 78.0 6.0 Religion 6.8 85.6 7.6

Arguably, persons with a predisposition to discriminate against minorities are more likely to choose ‘Yes’ to these questions. Responses were thus dichotomised such that a ‘Yes’ response was given a value of 1, and all other responses were given a value of 0. A look at the tetrachoric correlations confirms that all the items are highly correlated with one another (correlation coefficients range from 0.70 to 0.86; see Appendix B), suggesting that the items may be tapping into a single dimension. Measurement models are used to test whether or not a single dimension is being tapped. Since the responses were dichotomised, item response theory is used to generate and test the latent variable driving these responses. In this study, a two-parameter model is used to fit the latent model, as it allows items to vary in terms of their difficulty (the location of the item on the latent) and discrimination (how the items discriminate between persons on low and high levels of the latent trait). The goodness of fit statistics test confirms that a single dimension model is adequate (see Appendix B for more details). The values of the latent variable ranged from -0.5 to 1.5, where higher values indicated greater support for discrimination against minorities.

100

4.2.3 Statistical method To recap, the dependent variables were recoded so that they were binary. One could estimate a series of probit or logit models for each of the dependent variables. However, though there are some idiosyncrasies present, Tables 4.3 shows that the items capturing support for the criminalisation of same-sex intimacy are significantly correlated. As would be expected (based on Figure 4.1), the strongest pairwise correlation occurs between the ‘Penalise men in private’ and ‘Penalise women in private’ variables (coefficient is 0.95). The ‘Maintain the laws’ and ‘Enforce the laws’ variables are also strongly correlated (coefficient is 0.75). A much weaker, albeit statistically significant, correlation occurs among the other items. For instance, the correlation between the ‘Maintain’ and ‘Penalise Women in private’. The items capturing the perceived usefulness of the laws are also significantly correlated, and the tetrachoric correlations for these items range between 0.55 and 0.79 (Table 4.4). While the items in each group are correlated, the varying strengths of these correlations hint that the eight items may not necessarily be capturing two distinct underlying dimensions of attitudes. In fact, attempts to use latent variable modelling to measure two underlying dimensions (that is general support for the criminalisation of same-sex intimacy and support for laws as protectionist measures) resulted in two poorly fitted models. So, while the four variables in each group are correlated, it does not appear that these societies show consistently permissive or conservative attitudes on either subset of items, suggesting that responses to these questions may not represent a simple conservative-liberal dichotomy. The model needed is one that can account for the correlation between items, but would still allow us to investigate the variation in each item capturing attitudes towards the anti-gay laws in these countries. Hence, multivariate probit models are used to investigate anti-gay attitudes. Multivariate probit models are typically used when there are two or more binary dependent variables that are correlated, and this correlation is believed to persist even after regressing the binary variables on a set of predictors29 (Greene 2012). If true, and if one opts to use a univariate approach (like probit or logit models), vital information is lost. Failure to control for the relationship between the dependent variables would result in correlated error terms, which in turn would lead to inefficient estimates. To overcome this problem, the multivariate probit regression estimates the model within a system (that

29 The ‘multivariate’ term thus refers to the fact that there are multiple binary dependent variables and not the number of independent variables. 101

is, estimate all regressions jointly) and thus permits cross-equation error correlation. The model can be specified as:

∗ ′ ∗ 푌푚 = x푚훽푚 + 휀푚, 푌푚 = 1 푖푓 푌푚 > 0, 0 표푡ℎ푒푟푤푖푠푒, 푚 = 1, … , 푀

∗ where 푌푚 is a vector outcome variables, x is a matrix of predictor variables and 푀 represents the number of equations The error terms (휀푚) are assumed to jointly follow a multivariate normal distribution such that (휀1, … . , 휀푀)~푁푀[0, 푅], where 푅 is the variance-covariance matrix, with values of 1 on the diagonal (that is,

푉푎푟[휀푚|푥1, … . , 푥푀] = 1) and the correlations, that is, 퐶표푣[휀푗, 휀푚|푥1, … . , 푥푀] = 휌푗푚, as the off-diagonals. While the multivariate probit model is commonly used to efficiently evaluate the factors influencing the unconditional marginal probabilities (that is, how a set of predictors affects each dependent variable), it can also be extended to evaluate joint probabilities. Specifically, the system estimation can also be used to assess the joint probability that the dependent variables are all one, all zero or some combination of ones and zeros (Srivastava 2013). For instance, the multivariate probit model can be used to evaluate the probability that a person agrees with all the ‘laws protect the fabric of society’ arguments or that an individual wants the anti-gay laws maintained and enforced, but does not want same-sex couples penalised for having sex in private. Since the analysis is conducted within the system, it is much more efficient than creating and analysing a combined variable from the data. The multivariate joint probabilities are given by:

푃(푌1 = 푑1, 푌2 = 푑2, … , 푌푚 = 푑푚; x1, x2, … , x푚)

′ ′ ′ = ϕ푚 ((2푑1 − 1)x1훽1, (2푑2 − 1)x2훽1, … , (2푑2 − 1)x푚훽푚 ; 푅̃푚) where

̃푅푚 1 (2푑1 − 1)(2푑2 − 1)휌21 ⋯ (2푑1 − 1)(2푑푚 − 1)휌21 (2푑 − 1)(2푑 − 1)휌 1 … (2푑 − 1)(2푑 − 1)휌 = ( 2 1 12 푚 푚 21) ⋮ ⋮ ⋱ ⋮ (2푑푚 − 1)(2푑1 − 1)휌1푚 (2푑푚 − 1)(2푑2 − 1)휌2푚 ⋯ 1

where 푑1, 푑2, … , 푑푚 = 0 표푟 1, and ϕ푚is the cumulative distribution function of standard multivariate normal distribution with variance-covariance ̃푅푚.

102

Table 4.3: Matrix showing tetrachoric correlation between the items measuring support for the criminalisation of same-sex behaviours Maintain the laws Enforce the laws Penalise men in private Penalise women in private Maintain the laws 1.00 Enforce the laws 0.75*** 1.00 Penalise men in private 0.26*** 0.40*** 1.00 Penalise women in private 0.22*** 0.39*** 0.95*** 1.00 Note: ***, **, * and + indicate significance at the 0.1, 1, 5 and 10 per cent levels, respectively.

Table 4.4: Matrix showing tetrachoric correlation between the items measuring the perceived usefulness of the anti-gay laws Laws express moral Laws stop spread of Laws important from Laws protect young

standards homosexuality health perspective people from abuse Laws express moral standards 1.00 Laws stop spread of homosexuality 0.55*** 1.00 Laws important from health perspective 0.60*** 0.71*** 1.00 Laws protect young people from abuse 0.66*** 0.79*** 0.64*** 1.00 Note: ***, **, * and + indicate significance at the 0.1, 1, 5 and 10 per cent levels, respectively.

103

One key concern about the empirical models is that of mediation, as it is believed that some of the variables are correlated with each other. For instance, there is literature to suggest that age, education and gender are key predictors of religiosity (McCleary and Barro 2006; Voas and Crockett 2005; Voas and McAndrew 2012; Storm 2009). Religion and the demographic variables, in turn, are assumed to affect the socialisation patterns of individuals (McPherson et al. 2001), which could include whether or not persons have gay friends. The religious and demographic variables may also be related to views about whether or not homosexuality is innate (Whitehead 2010; Whitehead and Baker 2012) and influence attitudes towards other minorities (Borooah and Mangan 2007). Unfortunately, the multivariate probit models do not allow us to adequately test for mediation or break down the impact of the variables into direct and indirect effects. Hence, as a preliminary step to the empirical analysis, block regressions are utilised as an informal means of identifying potential interrelatedness among the variables. Three multivariate probit models are estimated: Model 1 only contains the demographic variables; Model 2 adds the religious variables; Model 3 adds the remaining regressors (interpersonal contact, beliefs about the origins of homosexuality and attitudes towards other minorities). A log likelihood ratio test is used to determine whether the increase in model fit from the additional variables is statistically significant, and Z and Chi-square tests are used to track if any of the predictors lose statistical significance as new predictors are added. The results for the model of support for keeping the anti-gay laws are shown in Table 4.5, while the results for the perceived usefulness of the current anti-gay laws are shown in Table 4.6. In each instance, the introduction of additional variables improved model fit. There is also minimal evidence of mediation. Only the age variable in the ‘Penalise men in private’ equation loses significance in Model 2 when the religious variables are added. With little evidence of mediation, the models are run with all the predictors so that one can assess the impact of a variable, ceteris paribus. Another concern is the coefficient values of the multivariate probit model. A consensus in the literature is that the interpretation of the coefficients in probit models is not straightforward (Greene 2012; Cameron and Trivedi 2010). While the coefficients provided are a good indication of the sign and statistical significance of the predictors, their interpretation is not appealing intuitively. As such, I opted to calculate the average marginal effects (AMEs), which are easier to interpret (Greene 2012; Cameron and Trivedi 2010).

104

Table 4.5: Nested models – Support for the criminalisation of same-sex intimacy χ2 tests for categorical variables, Z tests for continuous variables

Maintain Enforce Penalise men Penalise women the laws the laws in private in private

Model 1 – Log Likelihood = -4688.83

Gender (χ2 test) 0.018 0.019 36.298*** 5.931* Age (χ2 test) 7.861* 3.013 6.79* 11.001** Education (χ2 test) 2.130 3.927 4.051 2.351 Country (χ2 test) 7.351* 64.275*** 4.032 12.899**

Model 2 – Log Likelihood = -4518.55 Change in Log Likelihood = 170.28***

Gender (χ2 test) 1.181 0.229 45.068*** 10.810** Age (χ2 test) 7.614* 1.801 3.377 6.860* Education (χ2 test) 0.404 1.176 4.321 2.530 Country (χ2 test) 16.214*** 67.323*** 4.023 14.289*** Religious identity (χ2 test) 13.331* 12.199** 14.657* 16.331* Religious participation (χ2 test) 5.127+ 5.277+ 10.778** 21.910*** Source of views on Sexuality (χ2 39.009*** 52.812*** 19.257*** 27.275*** test)

Model 3 – Log Likelihood = -4285.30 Change in Log Likelihood = 233.25***

Gender (χ2 test) 0.441 0.111 31.138*** 6.561* Age (χ2 test) 5.310+ 1.027 3.914 7.605* Education (χ2 test) 2.380 3.105 1.202 1.009 Country (χ2 test) 8.027* 47.369*** 0.185 6.066* Religious identity (χ2 test) 11.878+ 15.851** 14.181** 17.601** Religious participation (χ2 test) 4.710+ 4.830+ 7.510* 19.207*** Source of views on sexuality (χ2 23.277*** 36.167*** 7.859* 14.802** test) Contact (χ2 test) 20.898*** 7.721+ 68.750*** 14.820*** Beliefs about the origins of 30.531*** 42.350*** 111.014*** 77.125*** homosexuality (χ2 test) Discrimination against other 0.021 0.875 4.051*** 2.910** minorities (Z-test) Note: ***, **, * and + indicate significance at the 0.1, 1, 5 and 10 per cent levels, respectively.

Table 4.6: Nested models – Perceived usefulness of the anti-gay laws χ2 tests for categorical variables, Z test for continuous variables

105

Moral Spread Health Abuse

Model 1 - Log Likelihood = -4719.53

Gender (χ2 test) 1.460 1.550 0.507 0.018 Age (χ2 test) 1.000 6.587* 3.421 3.054 Education (χ2 test) 1.021 8.357* 1.811 0.775 Country (χ2 test) 30.364*** 20.221*** 18.643*** 16.276***

Model 2 – Log Likelihood =- 4593.00 Change in Log Likelihood = 126.53***

Gender (χ2 test) 0.740 3.212+ 0.001 0.161 Age (χ2 test) 0.600 5.764+ 0.771 0.931 Education (χ2 test) 2.590 6.123* 1.374 0.854 Country (χ2 test) 31.415*** 12.421** 20.020*** 9.353** Religious identity (χ2 test) 10.511 9.460 5.561 8.870 Religious participation (χ2 test) 7.121* 3.831 13.742** 10.055** Source of views on sexuality (χ2 29.955*** 2.780 3.831 8.276* test)

Model 3 – Log- Likelihood = -4475.13 Change in Log Likelihood = 117.84***

Gender (χ2 test) 0.198 2.904+ 0.040 0.168 Age (χ2 test) 0.549 6.031* 1.401 1.159 Education (χ2 test) 1.907 5.021+ 1.112 0.864 Country (χ2 test) 34.185*** 15.953*** 23.781*** 7.761* Religious Identity (χ2 test) 10.301 8.047 6.083 6.201 Religious Participation (χ2 test) 6.291* 3.541 12.311** 10.399** Source of Views on Sexuality 21.612** 2.370 1.400 7.161+ (χ2 test) Contact (χ2 test) 7.260+ 12.392** 18.613*** 2.634 Beliefs about the origins of 19.888*** 21.601*** 27.087*** 13.721* homosexuality (χ2 test) Discrimination against other 2.030* 3.681*** 1.391 3.467*** minorities (Z-test)

Note: ***, **, * and + indicate significance at the 0.1, 1, 5 and 10 per cent levels, respectively.

The AMEs are quite similar to the coefficients estimated in simple regression models. For categorical variables, the AMEs indicate the average percentage point differences in probability between the reference category of a variable and the other categories of that variable. For continuous variables, they represent the average

106

percentage point differences in probability when the independent variables change by 1 unit.

4.3 Results

4.3.1 Factors influencing heterosexual support for the criminalisation of same-sex intimacy I began by looking at the items capturing the retention and enforcement of the current laws, as well as support for penalising persons engaging in same-sex sexual behaviours (Table 4.7). This 4-equation model is characterised by six cross-equation correlation terms, which represent the correlation between each of the dependent variables that exist even after controlling for the various predictors. As shown in Table 4.7, each cross-correlation remains statistically significant, confirming that the system estimation is more appropriate than individual probit regressions. Based on the literature on anti-gay sentiment in the Western world (see Chapter 2), one would expect to find that demographics account for a significant share of the differences in the attitudes towards gay rights in Barbados, Guyana and Trinidad and Tobago. In contrast, the findings suggest that demographics play a limited role in explaining heterosexual support for the bans on same-sex intimacy in these three countries. Education is statistically insignificant across the board, and gender only matters when the sex of the target is the focus, such as in the ‘Penalise men in private’ and ‘Penalise women in private’ equations. In line with the literature, there is some evidence that the gender gap is more prominent for attitudes towards the rights of gay men than that of lesbians. To further understand these differences, I calculated the predictive margins. The margins suggest that there is a small difference between attitudes towards female homosexuality and male homosexuality among women; however, men reported more negative attitudes towards male homosexuality than towards lesbianism. The probability of a man supporting the penalisation of two men having sex in private is 0.42, and the probability that he will support the penalisation of two women having sex in private is 0.34. For women, the probability of them opposing sex between men is 0.31, while the probability of them opposing sex between women is only slightly lower (0.29). There is also weak evidence of generational differences. While younger respondents are less likely than older ones to support the maintenance of the laws and to support the penalisation of two women having sex in private, there is no evidence of a

107

generational gap in the ‘Enforce the laws’ and ‘Penalise men in private’ equations. It is not exactly clear why there would be an age difference for some measures of law support and not others. A look at cross-tabulations of the ‘Maintain’ and ‘Enforce’ items by age group suggests that younger persons were more likely to support the enforcement of the laws but not their maintenance, which could explain why we found significant variation by age in the Maintain equation, but not in the Enforcement model. As discussed in the previous Section 4.1, it is possible that this ‘switching’ could either be measurement error, or represent persons who may want slight modifications to the laws, but not abolishment. On the lack of age effects in the ‘Penalise Men in private’ equation, it should be noted that in the preliminary analysis (the block regressions in Table 4.5), age was significant prior to the addition of the religious variables. It is possible that the religion variables could be mediating the impact of age on this particular item. There is also evidence that one’s place of residence matters. For instance, Guyanese respondents seem least likely to state that the laws should be maintained, and Barbadians are the least likely to support the penalisation of two women who have sex in private. Meanwhile, respondents from Trinidad and Tobago appear to offer the most support for enforcement. However, the country dummies are statistically insignificant in the ‘Penalise men in private’ equation. Turning to the religious variables, for the least strict measure of law support (that in the ‘Maintain’ equation), there is no evidence to suggest that religious identity matters. However, in the last three equations, there is a statistically significant difference between those who are religiously affiliated and those who do not assume a religious identity. The results suggest that the probability that the religiously unaffiliated will support law enforcement or the penalisation of persons engaging in same-sex intimacy is 13-22 percentage points lower than those individuals who identified as Evangelicals. Since there is no evidence of significant differences between religious denominations for the last three measures of law support, the results suggest that the religiously unaffiliated are less likely to support the prosecution of gay men and lesbians than those who identify with a religion. These results suggest that even persons who are religiously unaffiliated may be either supportive or indifferent to having the laws on the books. However, religiously unaffiliated persons are more likely than religious individuals to oppose intrusive actions, such as prosecution. The results also imply that religious participation and whether or not one’s views on sexuality are theologically based influence the support for gay rights, and they are largely in line with what the literature would suggest. Across all equations, the probability

108

that a respondent who states that he/she is actively involved in a religion will support the criminalisation of same-sex intimacy is significantly greater than persons who identify as passively involved. Also, in each equation, an individual whose views on sexuality are theologically based is more likely to support bans on same-sex behaviours than a person whose views on sexuality are not religiously inspired. The gap between individuals who are actively involved in their religion and those passively involved is much larger in the ‘Penalise women’ equation than it is in the ‘Penalise men’ equation. To further examine this discrepancy, I once again calculated predictive margins. These suggest that while both groups display more negative attitudes towards gay men, the sex of the person engaging in same-sex intimacy mattered less for persons actively involved in their religion. The probability of active persons supporting penalties for men was 0.39 compared with 0.35 for penalising women. The probability of persons passively involved in their religion supporting penalties for men was 0.34, and their support for penalties for women was 0.28. For persons passively involved in their religion, it is likely that the results are picking up attitudes towards gay men based on gendered norms. In Chapter 2, I noted that in the Caribbean, male homosexuality is more condemned than female homosexuality – possibly due to the definitions of masculinities. Persons who are actively involved in their religion may be more exposed to scripture verses that are interpreted as condemnations of both male and female homosexuality. Thus, their attitudes are less likely to vary across the two sexual orientations. In fact, a similar story emerges when we compare attitudes for those whose views on sexuality were formed by religion and those whose views were formed through socialisation. For persons with views on sexuality that are theologically based, the probability of them supporting the penalisation of sex between men is 0.39; for sex between women, it is 0.35. For those whose views were formed by socialisation, support for the penalisation of sex between men was 0.35, and for women it was 0.28. This suggests that religious persons show less variation in their views about the criminalisation of homosexual acts by sexual orientation than persons whose views about homosexuality were mainly influenced by social norms.

109

Table 4.7: Support for criminalisation of same-sex intimacy Maintain the laws Enforce the laws Penalise men in private Penalise women in private

Gender Men (base) Women -0.014 -0.007 -0.110*** -0.019*** Age 18-30 (base) 31-50 0.045+ -0.020 0.038 0.059** 51 + 0.056* 0.007 0.041 0.045+ Education Primary/Secondary (base) Post-secondary 0.020 -0.042 -0.017 -0.012 Tertiary 0.038 -0.031 -0.025 -0.023 Country Trinidad and Tobago (base) Barbados -0.027 -0.179*** -0.005 -0.059* Guyana -0.091* -0.097*** -0.009 -0.035 Religious Identity Evangelical Christian (base) Non-Evangelical Christian 0.005 -0.025 -0.024 -0.011 Muslim 0.016 0.010 -0.005 -0.002 Hindu -0.009 0.048 0.012 0.007 Other religion -0.034 0.021 0.012 0.015 Not religious -0.088 -0.216** -0.143** -0.133** No response -0.168** -0.010 0.035 0.107+

110

Table 4.7 cont’d Maintain the laws Enforce the laws Penalise men in private Penalise women in private

Religious participation Active (base) Passive -0.045+ -0.036+ -0.058** -0.091*** Unsure/Won’t say -0.036 -0.001 -0.034 -0.053 Source of views on sexuality Religion (base) Socialisation -0.082** -0.136*** -0.040+ -0.061** Popular Culture -0.051+ -0.049+ -0.060* -0.059+ Unsure/Prefer not to say -0.156*** -0.115** -0.058+ -0.103*** Contact No gay friends/family (base) Gay friend OR family -0.066** -0.053+ -0.161*** -0.105*** Gay friend AND family -0.069* -0.088** -0.184*** -0.127*** Prefer not to say -0.178*** -0.025*** -0.131** -0.064** Origin of homosexuality Born that way (base) Birth defect 0.121* 0.095+ 0.251*** 0.158*** Psychological trauma/abuse 0.182*** 0.130*** 0.122*** 0.080*** Poor religious/moral groundings 0.119** 0.192*** 0.321*** 0.254*** Choice 0.142*** 0.135*** 0.193*** 0.174*** Unsure/prefer not to say 0.070+ 0.009 0.067* 0.071* Discrimination against minorities 0.007 0.019 0.080*** 0.056***

111

Table 4.7 cont’d Cross-equation correlations: 휌(푚푎푖푛푡푎푖푛, 푒푛푓표푟푐푒) 0.775*** 휌(푚푎푖푛푡푎푖푛, 푚푒푛) 0.218*** 휌(푚푎푖푛푡푎푖푛, 푤표푚푒푛) 0.170*** 휌(푒푛푓표푟푐푒, 푤표푚푒푛) 0.320*** *** 휌(푒푛푓표푟푐푒, 푚푒푛) 0.316 휌(푚푒푛, 푤표푚푒푛) 0.957*** Notes: (1) n = 2,265; (2) ρij indicates the errors in equation i and equation j; (3) ***, **, * and + indicate significance at the 0.1, 1, 5 and 10 per cent levels, respectively.

112

Support for the bans on same-sex intimacy is also susceptible to interpersonal contact. In line with the contact hypothesis, respondents with affective ties to gay men and/or lesbians are less supportive of banning intimate same-sex acts than those without. The association is slightly stronger for those with contacts in both their social circle and family. In line with attribution theory, the results also suggest that individuals who believe that homosexuality is innate are less supportive of the laws than those who think otherwise. Finally, persons who think it is okay to discriminate against persons based on gender, race, age and/or religion are more likely to think that persons of the same-sex should be penalised for having sex in private. However, the discrimination variable had no impact on views about the retention and enforcement of the current anti-gay laws and could indicate that these two items are capturing more than just intolerance to out- groups.

4.3.1.1 Consistent and mixed support for the criminalisation of same-sex intimacy The results presented thus far have identified the factors influencing the marginal probabilities for the various items. Or, who is likely to consistently support the criminalisation of same-sex intimacy? And, who is likely to have be ambivalent about the bans? As mentioned in the Section 4.2.3, one of the key advantages of the multivariate probit model is that one can move beyond investigating the marginal probabilities and evaluate other types of probabilities via the system. Hence, I use the model to evaluate the factors influencing the following joint probabilities:

1. The probability that an individual consistently supports the anti-gay laws, algebraically written as Pr (Maintain=Yes, Enforce=Yes, Penalise Men= Yes, Penalise Women=Yes) 2. The probability that an individual supports the maintenance and enforcement of the law but opposes penalising private same-sex intimate acts, that is, Pr (Maintain=Yes, Enforce=Yes, Penalise Men= No, Penalise Women=No)30

The results for the joint probabilities are detailed in Table 4.8.

30 For this model, I generate two additional dummy variables: (1) ‘Not men in private’, which takes on a value of 1 if the respondent thinks that men should not be penalised for having sex in private, and (2) ‘Not women in private’, which takes on a value of 1 if the respondent disagrees that women should be penalised for having sex in private. 113

Table 4.8: Joint probability models – Consistent and mixed support for the criminalisation of same-sex intimacy Consistent support for Mixed support for the

the laws laws

Gender Men (base) Women -0.032* 0.042** Age 18-30 (base) 31-50 0.049** -0.000 51 + 0.043** -0.001 Education Primary/Secondary (base) Post-secondary -0.011 0.004 Tertiary -0.009 0.014 Country Trinidad and Tobago (base) Barbados -0.080*** -0.092*** Guyana -0.049** -0.060*** Religious Identity Evangelical Christian (base) Non-Evangelical Christian -0.008 0.007 Muslim 0.013 0.049 Hindu -0.014 -0.019 Other religion 0.038 -0.016 Not religious -0.109*** -0.004 No response 0.011 -0.078* Religious participation Active (base) Passive -0.067*** 0.026+ Unsure/Won’t say -0.039+ 0.001 Source of views on sexuality Religion (base) Socialisation -0.066*** -0.029+ Popular Culture -0.047* -0.011 Unsure/Prefer not to say -0.089*** 0.076*** Contact No gay friends/family (base) Gay friend OR family -0.084*** 0.056** Gay friend AND family -0.098*** 0.054* Prefer not to say -0.105*** -0.074**

114

Table 4.8 cont’d Consistent support for Mixed support for the

the laws laws Origin of homosexuality Born that way (base) Birth defect 0.105** -0.067* Psychological trauma/abuse 0.063** 0.015 Poor religious/moral groundings 0.174*** -0.089*** Choice 0.120*** -0.022 Unsure/prefer not to say 0.046* -0.041+ Discrimination against minorities 0.035** -0.029* Notes: (1) n =2,265; (2) ***, **, * and + indicate significance at the 0.1, 1, 5 and 10 per cent levels, respectively.

The model suggests that the individual most likely to consistently support having anti-gay laws (1) is a man; (2) is older than 30; (3) resides in Trinidad and Tobago; (4) is religious – i.e. is religiously affiliated, is active in his religion and his views on sexuality are religiously inspired; (5) has no close contacts that identify as a gay man or lesbian; (6) does not believe that homosexuality is innate; (7) thinks it is acceptable to treat persons differently based on their race, gender, age and/or religion. The education variable is the only predictor that is not statistically significant in this model. With the exception of education level, the characteristics of the individual least supportive of law reform match the profile of persons most likely to hold negative attitudes towards homosexuality in the West (see Chapter 2). The model also suggests that neither age nor education impacts the likelihood that a person will be ambivalent about the laws. Rather, those most likely to be inconsistent in their law support are (1) women; (2) respondents from Trinidad and Tobago; (3) respondents who are passively religious; (4) persons who cite religion or popular culture as their main source of views on human sexuality; (5) persons with close gay contacts; (6) those who believe homosexuality is innate, due to trauma or a choice and (7) persons less likely to approve of discrimination against minorities. With the exception of persons who cite religion as the source of views about the laws, and those who reported that homosexuality is a choice or trauma, the factors predicting the probability that a person is ambivalent about criminalising same-sex intimacy is quite similar to the factors assumed to increase tolerance in the West (see Chapter 2).

115

4.3.1.2 The unsure/prefer not to say responses to questions about banning same- sex intimacy An area hitherto unexplored concerns the subset of the persons who selected ‘Unsure/Prefer not to say’ when asked about their appetite for law reform (see Figure 4.2). The current modelling technique lumps this subset of respondents with those who opposed the anti-gay laws. Given the number of individuals who selected this category – particularly for the ‘Maintain the laws as is’ and ‘Enforce the laws’ survey items – one cannot help but wonder about the characteristics of this group. Specifically, are these individuals really unsure about their stance? Or, is this a hidden liberal or conservative class of individuals who prefer not to state their true opinions for fear of ridicule by the interviewers? As an attempt to find answers to these questions, this sub-section supplements the previous analysis by investigating who is likely to choose ‘Unsure/Prefer not to say’. Four additional binary variables are created, each taking on a value of 1 if the respondent chooses ‘Unsure/Prefer not to say’ and 0 otherwise. The recoded variables are re-run as a function of the previously identified independent variables. This analysis allows us to determine if respondents who opted for ‘Unsure/Prefer not to say’ have characteristics that would normally predict support for gay rights (and so could be classed as a hidden liberal group) or opposition to gay rights (hidden conservative group). The results are presented in Table 4.9. The gender, age, education and religious participation variables are insignificant across the board. There is some evidence of a country effect, as Barbadians seem the most likely to choose ‘Unsure/prefer not say’ when asked about whether or not the laws should be enforced, and Guyanese seem most likely to choose this category when asked about the penalisation of two men or two women for having sex in private. Religious identity also appears to matter. In the ‘Maintain the laws’, ‘Enforce the laws’ and ‘Penalise women in private’ equations, persons who are not religiously affiliated are more likely to state their opinions (i.e. not choose ‘Unsure/Prefer not to say’) than those who are religiously affiliated. Respondents who identify as Muslim are the least likely to choose this category when asked about their views on penalising men who have sex with other men. In addition, Muslims and persons from other religions are less likely than Evangelicals to choose this category when asked about penalising two women for having sex in private. When it comes to the source of views on sexuality, contact and attribution variables, it seems that the type of respondents most likely to choose ‘Unsure/Prefer not

116

to say’ when asked about having anti-gay laws also opted not to state their preferences on these items of the survey. In the ‘Enforce the laws’ and the ‘Penalise women in private’ equations, persons with close contacts were also less likely to choose this category. Finally, there is no relationship between the discrimination against minorities variables and choosing the option ‘Unsure/prefer not to say’. Altogether, there is no clear pattern to suggest that the share of persons choosing not to state their opinions are part of a hidden liberal group or a conservative group. They are more likely a collection of individuals who either do not have a strong opinion on these issues or prefer to keep their views to themselves for whatever reasons. However, this does not preclude that there could be people who do not state their opinions for fear of being judged by the interviewer.

117

Table 4.9: Modelling ‘unsure/prefer not to say’ responses for items relating to support for the criminalisation of same-sex intimacy Maintain the laws Enforce the laws Penalise men in private Penalise women in private

Gender Men (base) Women 0.002 -0.013 0.008 0.005 Age 18-30 (base) 31-50 -0.036 -0.029 -0.016 0.002 51 + -0.019 -0.006 0.013 0.011 Education Primary/Secondary (base) Post-secondary -0.031 0.015 0.010 -0.008 Tertiary -0.024 -0.008 0.019 0.002 Country Trinidad and Tobago (base) Barbados 0.007 0.094*** 0.020 0.024 Guyana 0.002 0.028 0.039* 0.070*** Religious Identity Evangelical Christian (base) Non-Evangelical Christian 0.007 -0.005 0.014 -0.018 Muslim -0.014 -0.015 -0.056* -0.052+ Hindu 0.003 -0.014 -0.017 -0.029 Other religion 0.003 0.003 -0.033 -0.079*** Not religious -0.075+ -0.130*** -0.011 -0.070** No response 0.156** 0.017 -0.007 0.027

118

Table 4.9 cont’d Maintain the laws Enforce the laws Penalise men in private Penalise women in private Religious participation Active (base) Passive -0.006 -0.003 -0.010 -0.013 Unsure/Won’t say 0.053 0.004 0.045+ -0.012 Source of views on sexuality Religion (base) Socialisation 0.007 0.051* -0.019 0.014 Popular Culture -0.040 -0.022 0.039+ 0.053* Unsure/Prefer not to say 0.114*** 0.061* 0.053* 0.061* Contact No gay friends/family (base) Gay friend OR family 0.033 -0.041* -0.013 0.001 Gay friend AND family -0.004 0.016 -0.012 -0.036* Prefer not to say 0.143*** 0.092* 0.102** 0.092** Origin of homosexuality Born that way (base) Birth defect 0.027 0.058 0.004 0.023 Psychological trauma/abuse -0.026 0.035 0.025 0.019 Poor religious/moral groundings 0.040 -0.021 0.025 0.038 Choice 0.036 0.017 0.020 0.027 Unsure/prefer not to say 0.124*** 0.156*** 0.122*** 0.102*** Discrimination against minorities -0.014 -0.020 -0.012 -0.021

119

Table 4.9 cont’d Correlated errors: 휌(푚푎푖푛푡푎푖푛, 푒푛푓표푟푐푒) 0.860*** 휌(푚푎푖푛푡푎푖푛, 푚푒푛) 0.140*** 휌(푚푎푖푛푡푎푖푛, 푤표푚푒푛) 0.164*** 휌(푒푛푓표푟푐푒, 푤표푚푒푛) 0.203*** *** 휌(푒푛푓표푟푐푒, 푚푒푛) 0.243 휌(푚푒푛, 푤표푚푒푛) 0.938*** *** ** * + Notes: (1) n = 2,265; (2) 휌ij indicates the errors in equation 푖 and equation 푗; (3) , , and indicate significance at the 0.1, 1, 5 and 10 per cent levels, respectively.

120

4.3.2 Subscriptions to the notions that the current laws protect the fabric of society Figure 4.4 showed that a significant portion of persons believe that the anti-gay laws reflect the morals of the society, stop the spread of homosexuality, are important from a public health perspective and protect young people from abuse. Multivariate probit models were again employed to determine who is most likely to subscribe to each of these arguments (Table 4.10). The cross-equation correlations are statistically significant, confirming again that the system estimation is more appropriate than individual probit/logit regressions. Looking first at the demographics, the results suggest that gender, age and education do not consistently predict heterosexual beliefs about the usefulness of the current laws. In fact, these variables are only significant in the ‘Laws stop the spread of homosexuality’ equation. The most recognisable supporters of this argument are men, persons between the ages of 31 and 50 and persons who are not educated at the tertiary level. The only demographic variable that is significant in each model is the country of residence. Residents of Trinidad and Tobago are the least likely to agree that the laws are an expression of the moral standards and that the laws are important from a public health perspective, while respondents from Guyana had the highest probability of stating that the laws stop the spread of homosexuality and protect children from abuse. The observed limited impact of demographic variables is very similar to what was observed in the previous subsection. Turning now to the impact of religion, the results provide no evidence that opinions vary substantially across religious identities. Rather, religious participation seems to be more important: the probability that persons who identify as active in their religion want the antigay laws in place to protect moral standards, public health and children from abuse is greater than those who rate themselves as passive and/or persons who did not state their religious status. Meanwhile, individuals whose views are shaped by religion have the greatest likelihood of supporting the anti-gay laws from a moral perspective or to limit child abuse. As a side note, none of the religious variables has a significant impact on the argument relating to the prevalence of homosexuality.

121

Table 4.10 Perceived usefulness of the current anti-gay laws Laws express moral Laws stop spread of Laws important from Laws protect young

standards homosexuality health perspective people from abuse

Gender Men (base) Women 0.021 -0.030+ 0.004 -0.008 Age 18-30 (base) 31-50 -0.014 0.051* 0.019 0.011 51 + 0.002 0.029 0.031 0.028 Education Primary/Secondary (base) Post-secondary 0.032 -0.015 -0.029 0.017 Tertiary -0.009 -0.043* -0.016 0.022 Country Trinidad and Tobago (base) Barbados 0.158*** -0.028 0.130*** 0.009 Guyana 0.062*** 0.066** 0.078** 0.072** Religious Identity Evangelical Christian (base) Non-Evangelical Christian -0.013 0.008 0.008 -0.041 Muslim 0.025 0.068 -0.026 0.008 Hindu 0.057 0.039 0.036 0.054 Other religion 0.006 0.010 0.017 -0.060 Not religious -0.103 -0.030 -0.095 -0.104 No response -0.015 0.070 0.020 0.024

122

Table 4.10 cont’d Laws express moral Laws stop spread of Laws important from Laws protect young

standards homosexuality health perspective people from abuse Religious participation Active (base) Passive -0.014 -0.028 -0.062* -0.021 Unsure/Won’t say -0.101* -0.052 -0.123** -0.122** Source of views on sexuality Religion (base) Socialisation -0.067* -0.012 -0.006 -0.051+ Popular Culture -0.122** -0.042 0.034 -0.052+ Unsure/Prefer not to say -0.135*** 0.001 -0.015 -0.074* Contact No gay friends/family (base) Gay friend OR family -0.016 -0.055* -0.035 -0.014 Gay friend AND family -0.043+ -0.075** -0.068* -0.037 Prefer not to say -0.118** -0.081* -0.179*** -0.059 Origin of homosexuality Born that way (base) Birth defect 0.006 -0.038 -0.004 -0.051 Psychological trauma/abuse 0.098* 0.081* 0.167*** 0.092* Poor religious/moral groundings 0.026 0.066* 0.076+ -0.029 Choice 0.046 0.035 0.065* 0.022 Unsure/prefer not to say -0.069* -0.006 -0.006 -0.015 Discrimination against minorities 0.052* 0.067*** 0.032 0.072**

123

Table 4.10 cont’d Correlated errors: 휌(푚표푟푎푙, 푠푝푟푒푎푑) 0.549*** 휌(푚표푟푎푙, ℎ푒푎푙푡ℎ) 0.592*** 휌(푚표푟푎푙, 푎푏푢푠푒) 0.654*** 휌(푠푝푟푒푎푑, ℎ푒푎푙푡ℎ) 0.718*** *** 휌(푠푝푟푒푎푑, 푎푏푢푠푒) 0.778 휌(ℎ푒푎푙푡ℎ, 푎푏푢푠푒) 0.647***

*** ** * + Notes: (1) n = 2,147; (2) 휌ij indicates the errors in equation 푖 and equation 푗; and (3) , , and indicate significance at the 0.1, 1, 5 and 10 per cent levels, respectively

124

Views that the laws are a means of upholding the moral standards, stopping the spread of homosexuality and protecting the public health varied with the amount of contact with gay men/lesbians. In each of these three equations, those who stated they have no close contact with homosexuals consistently had the highest predicted probabilities of agreeing with the anti-gay statements. The models also imply that those with more avenues of contact consistently had the smallest probability of viewing homosexuals as a threat – suggesting that the number of interactions matters. Beliefs about the origin of homosexuality were significant in each equation: in absolute terms, persons who believe homosexuality is caused by some traumatic event are consistently the most likely to view the anti-gay laws as necessities31. Relative to persons who believe homosexuality is innate, respondents who believe that homosexuality is a consequence of poor religious/moral groundings are more likely to see the laws as a way to stop the spread of homosexuality and to regard them as important from a public health perspective. Finally, views on discrimination against other minorities mattered for three of the four arguments: those with more biased views are more likely to argue that the laws are necessary from a moral perspective, that the laws help curtail the spread of homosexuality and that the laws protect children from abuse. Thus, those who are inherently keener to discriminate against other minorities find all but one of the arguments for discriminatory legislation persuasive.

4.3.2.1 Laws achieve all objectives In this subsection, I focus on modelling the joint probability that an individual agrees with each of the ‘laws protect the fabric of society’ arguments, written algebraically as Pr (Laws express moral standards =Yes, Laws stop spread of homosexuality =Yes, Laws important from health perspective =Yes, Laws protect young people from abuse=Yes). As shown in Table 4.11, the respondents most likely to think the laws are useful in all dimensions (1) are over 30; (2) are not tertiary educated; (3) reside in Guyana; (4) chose either socialisation or religion as their main source of views on sexuality; (5) have no close gay contacts; (6) believe that the cause of same-sex attraction is either psychological trauma/abuse or poor religious/moral background; (7) are more likely to be prejudiced against other minorities.

31 Based on the value of the AME 125

Most of these characteristics are in line with Western definitions of persons most likely to be prejudiced against homosexuals (see Chapter 2).

Table 4.11: Perceived usefulness of the laws - Joint probability model Believes laws achieve all objectives

Gender Men (base) Women -0.019 Age 18-30 (base) 31-50 0.054** 51 + 0.050** Education Primary/Secondary (base) Post-secondary -0.028 Tertiary -0.033+ Country Trinidad and Tobago (base) Barbados -0.015 Guyana 0.055*** Religious Identity Evangelical Christian (base) Non-Evangelical Christian -0.004 Muslim 0.022 Hindu 0.033 Other religion -0.010 Not religious -0.030 No response -0.012 Religious participation Active (base) Passive -0.020 Unsure/Won’t say -0.047+ Source of views on sexuality Religion (base) Socialisation -0.032 Popular Culture -0.048* Unsure/Prefer not to say -0.039 Contact No gay friends/family (base) Gay friend OR family -0.046* Gay friend AND family -0.056** Prefer not to say -0.088**

126

Table 4.11 cont’d Believes laws achieve all objectives Origin of homosexuality Born that way (base) Birth defect 0.007 Psychological trauma/abuse 0.094** Poor religious/moral groundings 0.095** Choice 0.032 Unsure/prefer not to say 0.017 Discrimination against minorities 0.061*** Notes: (1) n = 2,147; (2) ***, **, * and + indicate significance at the 0.1, 1, 5 and 10 per cent levels, respectively

4.3.2.2 The ‘unsure/won’t say’ response when asked about the usefulness of the laws To recap, between 20 to 27 per cent of respondents were either unsure or refused to state their opinions about what they believed the laws in current form achieved (Figure 4.4). As with the general support model, an additional multivariate probit model is run for these responses (Table 4.12). Generally, respondents most likely to choose ‘Unsure/Won’t say’ are either in the 18-30 age category or the 51+ age category. They also opted for the ‘Unsure/Prefer not to say’ response for other survey questions: namely, questions relating to religious participation, the source of views on sexuality, personal contact with gay men and lesbians and beliefs about the origins of homosexuality. There is some evidence that country residence, religious identity and educational level matter for some items. Specifically, Muslims, Barbadians and persons with post-secondary school qualifications are less likely to choose ‘Unsure/Won’t say’ when asked about the moral objectives of the laws. Muslims, persons in the ‘other’ religion category and the religiously unaffiliated are less likely to choose this response for the ‘Laws stop the spread of homosexuality’ item. Barbadians and persons from other religions were less willing to give a response to the public health perspective item. Finally, Guyanese people were more likely to choose ‘Unsure/Won’t say’ when asked about the ability of the laws to protect young people from abuse, while Muslims and Hindus were less likely to choose this response option for this survey item.

127

Table 4.12: Modelling ‘Unsure/Won’t say’ responses for items relating to the usefulness of the laws Laws express moral Laws stop spread of Laws important from Laws protect young

standards homosexuality health perspective people from abuse

Gender Men (base) Women 0.002 0.013 0.013 0.036+ Age 18-30 (base) 31-50 -0.048* -0.050* -0.040+ -0.046* 51 + -0.027 -0.017 -0.020 -0.021 Education Primary/Secondary (base) Post-secondary -0.045+ 0.013 -0.011 -0.009 Tertiary 0.008 0.025 -0.002 -0.016 Country Trinidad and Tobago (base) Barbados -0.047* -0.032 -0.064** -0.003 Guyana 0.036 0.022 0.016 0.057* Religious Identity Evangelical Christian (base) Non-Evangelical Christian 0.006 -0.007 0.029 0.002 Muslim -0.139*** -0.100** -0.056 -0.093** Hindu -0.046 -0.029 -0.043 -0.063* Other religion -0.048 -0.064* -0.074* -0.039 Not religious -0.075 -0.076+ 0.045 -0.054 No response 0.129* 0.028 -0.020 -0.002

128

Table 4.12 cont’d Laws express moral Laws stop spread of Laws important from Laws protect young

standards homosexuality health perspective people from abuse Religious participation Active (base) Passive -0.008 -0.002 0.013 -0.003 Unsure/Won’t say 0.149*** 0.110** 0.138*** 0.057* Source of views on sexuality Religion (base) Socialisation 0.010 -0.010 -0.001 0.026 Popular Culture 0.004 -0.003 0.011 0.006 Unsure/Prefer not to say 0.132*** 0.121*** 0.073* 0.121*** Contact No gay friends/family (base) Gay friend OR family -0.017 -0.020 -0.022 -0.018 Gay friend AND family 0.005 -0.027 -0.007 0.001 Prefer not to say 0.190*** 0.201*** 0.202*** 0.211*** Origin of homosexuality Born that way (base) Birth defect 0.013 0.007 0.059 -0.035 Psychological trauma/abuse -0.005 -0.001 0.010 0.002 Poor religious/moral groundings 0.048 0.016 0.051 0.006 Choice 0.002 0.008 0.037 0.015 Unsure/prefer not to say 0.151*** 0.076* 0.151*** 0.072* Discrimination against minorities -0.025 -0.009 -0.016 -0.024

129

Table 4.12 cont’d Correlated errors: 휌(푚표푟푎푙, 푠푝푟푒푎푑) 0.745*** 휌(푚표푟푎푙, ℎ푒푎푙푡ℎ) 0.725*** 휌(푚표푟푎푙, 푎푏푢푠푒) 0.799*** 휌(푠푝푟푒푎푑, ℎ푒푎푙푡ℎ) 0.830*** *** 휌(푠푝푟푒푎푑, 푎푏푢푠푒) 0.884 휌(ℎ푒푎푙푡ℎ, 푎푏푢푠푒) 0.802***

*** ** * + Notes: (1) n = 2,147; (2) 휌ij indicates the errors in equation 푖 and equation 푗; (3) and , , and indicates significance at the 0.1, 1, 5 and 10 per cent levels, respectively

130

4.4 Discussion

This chapter investigated the prevalence of support for anti-gay laws in three Commonwealth Caribbean states. First, the study looked at the extent to which heterosexuals wanted the current laws maintained/enforced and supported the penalisation of persons engaging in same-sex sexual acts in private. There was mixed support for the preservation of anti-gay laws: while the majority stated they wanted the current laws maintained and enforced, only a minority endorsed state regulation of private sex between individuals of the same-sex. So, while persons disagree with the government regulating private sexual acts, this belief is currently not strong enough for them to support the removal of the laws. In these states, law support appears symbolic – in other words, persons view homosexuality as wrong and believe that laws should be in line with this perspective. Some Caribbeans may see the laws as a means of stopping sexual minorities from asserting their sexuality in public. The mixed support could be a manifestation of a typical ‘don’t ask, don’t tell’ attitude towards homosexuality. The observed ambivalent attitudes could also be related to beliefs associated with the perceived usefulness of the laws — specifically, the ability of the laws to ‘protect the fabric of society’. While Caribbeans may not view sex between two adults as harmful, ill- founded beliefs about what lesbians and gay men do outside this scenario could be driving law support. In this chapter, I also evaluated the extent to which persons subscribed to notions that the anti-gay laws (1) reflect moral standards; (2) stop the spread of homosexuality; (3) are important from a public health perspective and (4) protect young people from abuse. The survey responses suggest that to some extent, public support for these laws is grounded in the perceived threat of homosexuality, as 23-46 per cent of respondents believe that the laws achieve at least one of the above objectives. In the second stage of the study, I evaluated the factors correlated with support for and perceived usefulness of the anti-gay laws. The aim here was to determine the extent to which popular correlates of support for gay rights (at least in the West) also explain heterosexual views of the anti-gay laws in Barbados, Guyana and Trinidad and Tobago. The results were mixed: some of the findings aligned with research on the West, and others did not. Looking first at the similarities, the results suggest that interpersonal contact and beliefs about homosexuality were significantly correlated with support for gay rights. In line with Allport’s contact hypothesis, individuals who reported having personal contact

131

with lesbians or gay men were less likely to support the maintenance and/or enforcement of the current laws and were also less likely to want same-sex couples penalised for having sex in private. The evidence also suggests that the degree of interaction mattered, as suggested by Herek and Capitanio (1996). For instance, individuals who reported having both gay friends and family members were less likely to believe that the anti-gay laws were useful than those with no close gay contacts or one form of contact (i.e. either friends or family). With respect to the beliefs about the origin of homosexuality, the empirical model suggested that persons who believed that homosexuality was innate also reported more favourable views of the anti-gay laws than those who believed otherwise. This finding is in line with the predictions of attribution theory (see Chapter 2), which suggests that attitudes towards a stigmatised group or behaviour tend to be more positive when individuals view the cause as an uncontrollable condition. The extent to which one integrates religion into one’s life is also a reliable predictor of anti-gay prejudice. The results suggested that individuals who were actively involved in a religion and those whose views on human sexuality were shaped by religion were more likely to support the laws and view them as useful. These findings are in line with Durkheimian theories on social integrations, which suggest that the more one is integrated into a social group, the more likely one is to conform to its norms (Scheepers et al. 2002). Given that several religious doctrines characterise homosexual acts as ‘sinful’, it seems logical that those with greater religious exposure and those whose views are shaped by religion are more likely to see homosexuality in a negative light and, by extension, support regulations that limit its ‘harm to society’. There was also evidence that persons who discriminate against other minorities are also prejudiced against lesbians and gay men, even after controlling for variables such as education, age and religiosity, which are sometimes hypothesised to be the reason for the relationships with other types of prejudice. This could indicate that for some members of the population, prejudice towards homosexuals may just be an extension of their general intolerance of out-groups. However, prejudice against other minorities did not always co-extend to a disapproval of homosexuality. This variable was insignificant in the ‘Maintain the laws’, ‘Enforce the laws’ and ‘Laws are important from a health perspective’ equations. It is unclear as to why this item was related to some measures and not others. But, it is possible that for many individuals, their responses to this question were not stimulated by their general intolerances. This is somewhat plausible. Many of the arguments used to justify anti-gay prejudice may not apply to other prejudices

132

(Whitley Jr and Ægisdóttir 2000). For instance, many persons (currently) do not justify anti-Muslim sentiment by saying ‘persons were not born Muslim’. With respect to the public health item, it seems plausible to assume that among Caribbeans, homosexuals are viewed as a threat to public health in a way that other minority groups do not. Hence, in answering these particular questions, many persons may have been stimulated by something other than their general intolerance of minority groups. In contrast to studies on the West, the results suggest that some demographic variables (such as education) were not reliable predictors of support for the anti-gay laws, even in the absence of other control variables. In the marginal probability models, education was only significant in the ‘Laws stop the spread of homosexuality’ equation. At first glance, the limited impact education is surprising. One would assume that highly educated persons would have greater cognitive abilities than the less educated and so would be more likely to reject prejudice. However, one must consider the other hypothesised reasons for why education matters for prejudice, and it is likely the reasons given may be more relevant in Western contexts. For example, one way higher education influences attitudes is through social interaction with persons whose views differ from one's own. However, this would only work in societies where a diverse set of persons (in terms of views) attends higher education. Most persons attending a tertiary institution in Barbados, Guyana or Trinidad and Tobago are likely to come from within the Commonwealth Caribbean. Thus, attending a local university may not be coextensive to a greater exposure to diverse views, at least on the issue of anti-gay laws. Unfortunately, the CADRES data do not allow us to differentiate between those who attended a local tertiary institution and those who attended university in a more liberal country – e.g. the UK, US or Canada. However, statistics from the World Bank (World Bank 2016) suggest that many students opt to study locally. In 2013, the gross outbound enrolment ratio32 for Barbados was 6.0 per cent; it was 2.1 per cent for Guyana and 4.8 per cent for Trinidad and Tobago. It is likely that most persons who participated in the survey were educated in the region. Also, there is a bigger push to provide alternative views about sexuality in Western colleges and universities than in the Caribbean. As mentioned in Chapter 2, Western universities often offer free spaces to hold rallies, protests or discussions about the issues

32 The World Bank defines the gross outbound enrolment ratio as the total number of mobile tertiary students coming from a country/region as a percentage of the population of tertiary student age in their home country. 133

faced by minority groups, thus widening the views of attendees. Moreover, in recent decades, administrators in Western colleges and universities have increased their efforts to reduce sexual prejudice. Specifically, in Western universities and schools, there are currently several rules and regulations that explicitly prohibit discrimination on the basis of sexuality, and curriculums have been updated to include lessons on sexual diversity. We do not see such intense efforts in tertiary institutions in Barbados, Guyana and Trinidad and Tobago. It should be noted that West and Cowell (2015) also found that education was not a significant factor in attitudes towards gay rights in Jamaica, suggesting that this finding may not be limited to these three states. Many theories about the features of tertiary education mentioned in Barbados, Guyana and Trinidad and Tobago also hold for Jamaica. This points to how differences in context could affect the predictors of anti-gay prejudice. Age, unlike education, was able to predict some aspects of heterosexual support for the criminalisation of same-sex intimacy. But, age did not have a statistically significant impact on the marginal probability that respondents viewed the laws as a reflection of society’s morals or as a means of protecting public health or young persons from abuse. One would expect that younger persons would be more likely than older persons to reject many of the ill-founded beliefs about gays and lesbians – particularly, that they are immoral, child abusers or that the spread diseases. However, as noted in the review of the literature, while recent work often shows evidence of generational differences in attitudes in the US, there are studies that suggest that the ability of age to predict attitudes towards gay rights only came about in recent years (for instance, Baunach 2012; Sherkat et al. 2011). This shift came about due to increased awareness, positive media attention and the ethos of sexual tolerance in higher education brought about by LGBT social movements. As highlighted in Chapter 3, in the Commonwealth Caribbean, the LGBT rights movement is still very much in its infancy stages. The inability of the age variable to reliably predict whether or not a person subscribes to the above three ill- founded stereotypes of homosexual’s in these countries could then be indicative of a lack of competitive frames and public awareness. Hence, negative stereotypes of homosexuals are still being passed on to succeeding generations. Perhaps, over time, as the gay rights movement evolves in the Caribbean, we will see age and educational differences that parallel those in the US and Western Europe. The level of modernisation and the fact that anti-gay laws still exist could also be mitigating the impact of age and education in these states. According to Inglehart (1971),

134

the increase in post-materialist values brought about by socio-economic development is more likely to be adopted by younger generations (see Chapter 2). Moreover, the work by Kreitzer et al. (2014), suggests that the change in attitudes brought about by homosexual law reform is more substantial for younger and more educated individuals (see Chapter 2). Hence, it is possible that age and education could be more reliable predictors of attitudes towards gay rights in developed and in countries without anti-gay laws. Another deviation from the existing literature concerns the impact of religious subcultures on antigay sentiment. Research on the US and Europe tends to report drastic differences in the extent to which persons of different religious affiliations condemn homosexual behaviours. Within the context of Barbados, Guyana and Trinidad and Tobago, there is no evidence that views on the bans on same-sex intimacy differ across religious denominations. The results suggest that there may be religious syncretism regarding support for, and the usefulness of, the bans on same-sex intimacy. Indeed, in Chapter 3, we saw that religious organisations in the Commonwealth Caribbean show a united front on the issue of homosexuality. It is not surprising that we see this translated into public attitudes. A particularly interesting observation is that while we see deviations from the previous literature regarding the factors driving the individualised measure of law support when we conduct the joint probability models, a different story emerges. Specifically, the profile of the person likely to consistently support the laws (i.e. want the laws maintained and enforced and want persons penalised for engaging in homosexual behaviours in private) generally fits the typical anti-gay profile found in the West. The person is female, young, religious, has no contact with gays, does not believe that homosexuality is innate and is prejudiced against other minority groups. We see a similar narrative for persons who believe that the laws meet all the objectives given in the survey — the laws protect the morals, health and children of the society, and they also stop the spread of homosexuality. This person has a low education level, is religious, has no contact with gays, believes trauma or poor morals/upbringing is the cause of homosexuality and is prejudiced towards other groups. The joint probability models were also used to determine the characteristics of persons most likely to support the maintenance and enforcement of the laws but do not support penalising men or women who are intimate with persons of the same sex in private. Many of the characteristics held by persons who were ambivalent towards the

135

criminalisation of homosexual behaviours fit the profile of a pro-gay person in Western contexts. Mixed attitudes were more likely to be found among women as well as those who were passively involved in their religion, had close gay contacts, believed that homosexuality was innate and were less prejudiced against other minority groups. Based on these results, it can be argued that within the context of Barbados, Guyana and Trinidad and Tobago, tolerance manifests itself in an ‘I do not mind what you do in private’ attitude, but it does not stretch to a willingness to repeal the laws. A consistent finding (in both marginal and joint probability models) was that there were no differences in attitudes across religious denominations. This finding lends more credit to the view that on the issue of the sodomy laws, religious persons are united. It could also be related to the fact that non-Christian religious individuals in the Caribbean have been socialised under well-established (conservative) Christian social norms (see Chapter 3). As suggested by Norris and Inglehart (2011), the values of the dominant religion in society are often transmitted via the main institutions. Non- Christian religious persons may (unconsciously) be taking cues from the dominant religion, resulting in little to no denominational differences in views about gay rights. A finding hitherto unexplored in this chapter is the observed country differences in attitudes. Attitudes vary across countries, but the country effects also varied across items. For example, the joint probability models suggest that Barbadians were the least likely to consistently support the criminalisation same-sex intimate acts. Trinidadians were the most likely to consistent support the laws but were also most likely to be conflicted about the criminalisation of same-sex acts. At the same time, Trinidadians were consistently the least likely to subscribe to any one of the ‘laws protect the fabric of society’ arguments. Barbadians were the most likely to support the view that the laws protect the morals and health of the society, while Guyanese were more likely to view the laws as a means of stopping the spread of homosexuality and protecting young persons from abuse. These findings suggest that while Caribbean states are similar, Caribbean people are not uniform in their views on homosexuality – which is somewhat expected. In Chapter 3, it was highlighted that even though Caribbean states are highly similar, there are contextual differences. To be specific, in terms of socio-economic development, Barbados and Trinidad and Tobago fall into the high human development category, while Guyana falls in the medium development category. With respect to the laws, while all countries carry harsh penalties for engaging in acts of ‘buggery’ and acts of serious/gross indecency. Barbados and Guyana carry much harsher penalties for engaging in anal sex

136

(maximum penalty life imprisonment) and acts of indecency (10 years) than Trinidad and Tobago (maximum penalty 25 years for anal sex, and 5 years for acts of serious indecency). At the same time, in Barbados, bans on indecent acts are neutral in terms of gender and sexual orientation. In Guyana, these prohibitions mainly target men who have sex with other men. However, Trinidad and Tobago is the only one of the three to target both lesbianism and sex between men are considered in their laws, and to ban homosexuals from entering the country in their immigration laws. It is possible that the differences in attitudes observed across these states could be due to differences in context. In what follows, I will empirically investigate the role of context in shaping attitudes towards gay rights in the Commonwealth Caribbean and the rest of the Americas.

137

Chapter 5 : The Commonwealth Caribbean in context: Comparing attitudes towards same-sex marriage across the Americas

In Chapter 3 of this thesis, I highlighted some features of the Commonwealth Caribbean that possibly contributes to anti-gay prejudice in the region. However, one can’t help but wonder about the extent to which these characteristics differentiate the independent Commonwealth Caribbean from its neighbours in the Americas. A cursory glance at the socio-legal situation of homosexuals across the Americas suggests that persons living in the states outside the Commonwealth Caribbean would generally display greater tolerance towards homosexuals. These states tend to have a smaller Evangelical presence and no laws banning same-sex intimacy. In fact, many states in the Americas have opened the door to same-sex marriage/unions or offer some legal recognition of same-sex relationships. This chapter attempts to shed some light on the role of societal context in shaping attitudes in the Commonwealth Caribbean. Specifically, it attempts to answer two questions:

1) Are citizens of the independent Commonwealth Caribbean less supportive of gay rights than those in the rest of the Americas? 2) If so, to what extent can societal context – specifically, the level of socioeconomic development, a large Evangelical presence, and bans on same- sex sexual acts – explain the lower levels of tolerance?

5.1 Attitudes towards same-sex marriage in the Americas: Is the Commonwealth Caribbean less tolerant?

To answer the first question, I analyse attitudes towards homosexual rights through a secondary analysis of the 2014 wave of the AmericasBarometer surveys conducted by LAPOP. The AmericasBarometer is a large-scale nationally representative survey, designed to enable a cross-national comparison of political viewpoints across the Americas. As noted in the introduction, the AmericasBarometer is one of the few cross- national surveys to include a subset of Commonwealth Caribbean states. For instance,

138

the current World Values Surveys (up to Wave 6) only includes one Commonwealth Caribbean country – Trinidad and Tobago33. The 2014 Americas Barometer captures the opinions of 53,456 persons of voting age from 28 countries in the Americas, 6 of which are Commonwealth Caribbean countries. (see Table 5.1)34.

Table 5.1: Countries included in the 2014 AmericasBarometer survey Country Group Country Sample Size US and Canada Canada 1,517 US 1,500

Mexico and Central America Mexico 1,535 Costa Rica 1,537 Guatemala 1,506 El Salvador 1,512 Honduras 1,561 Nicaragua 1,546 Panama 1,508

Southern Cone Argentina 1,512 Brazil 1,500 Chile 1,571 Paraguay 1,503 Uruguay 1,512

Andean States Bolivia 3,066 Colombia 1,496 Ecuador 1,489 Peru 1,500 Venezuela 1,500

Commonwealth Caribbean Belize 1,512 Bahamas 3,828 Barbados 3,429 Guyana 1,557 Jamaica 1,503 Trinidad and Tobago 4,203

Other Caribbean Dominican Republic 1,520 Haiti 1,512 Suriname 4,000

33 It should be noted that one cannot merge the AmericasBarometer data with that from the World Values survey or even the Eurobarometer as the sample designs and survey items are not comparable 34 Unlike the CADRES survey, the AmericasBarometer does not capture the sexual orientation of the respondents. Hence, the full sample is used. 139

Samples in each country were developed using a multi-stage probabilistic design and were stratified by major regions of the country, size of the municipality, and by urban and rural areas within municipalities (Latin American Public Opinion Project 2014) 35. Face-to-face interviews were conducted in Latin America and the Caribbean, while web surveys were used for the US and Canada. The use of different data collection method does raise some concern. However, the data are weighted using the AmericasBarometer weighting scheme (Latin American Public Opinion Project n.d.), which thus takes into account the survey design. Using the weighted data is likely to reduce some of the biases and errors (like representation) that could arise due to differences in modes of data collection. The survey question “How strongly do you approve or disapprove of same-sex couples having the right to marry?” was used to operationalise attitudes towards homosexual rights. For this survey item. respondents were given a card with a 10-point ladder, which ranged from 1 to 10, where 1 indicated ‘strongly disapprove’ and 10 indicated ‘strongly approve’. For the comparative analysis, responses were recalibrated to yield a 0–100 scale, where the higher the value, the greater the approval of homosexual rights.36 For this question, there were 1,782 missing values, which initially reduced the sample size to 51,674. Figure 5.1 shows the mean values for all countries and the various regional clusters. The average index lies below the midpoint of the scale (that is, 50.0), hinting that in general, countries of the Americas do not approve of extending marital rights to gays and lesbians. In fact, a look at the raw data revealed that 57.7 percent of respondents chose “1” (strongly disapprove) when asked about their opinion of same sex marriage. But, as would be expected, attitudes towards same-sex marriage differed across the regional groupings. Focusing first on the differences across regions, the data suggests that the US and Canada are the most tolerant cluster, and they are the only set of countries with an average index lying above the mid-point of the scale. The Commonwealth Caribbean cluster displays the greatest amount of anti-gay prejudice. Moreover, according to an analysis of variance test and corresponding pairwise comparisons, the observed differences between the Commonwealth Caribbean and the other country groupings are statistically significant (see Table 5.2).

35 At the time of writing I was unable to attain information about response rates. 36 Studies in the AmericasBarometer Insights Series often recalibrate the main variable of interest. I follow suit in this thesis. Scale conversion is accomplished by reducing each score by 1 and dividing by 9 and then multiplied by 100. A benefit of this approach is the ease of interpretation of the scale. 140

Figure 5.1: Average approval of same-sex marriage across the Americas – Regional Differences

Americas 23.9 All

US and Canada 64.8

Southern Cone 48.6

Andean States 25.3

Mexico and Central America 21.3 Country Groupings Other Caribbean 16.3

Commonwealth Caribbean 10.7

0 10 20 30 40 50 60 70

Data sourced from the 2014 Americas Barometer by LAPOP (www.LapopSurveys.org). Figure based on weighted sample.

Table 5.2: Pairwise comparisons – The Commonwealth Caribbean vs. other country groupings Pairwise comparisons Difference t-statistic p-value in means US and Canada vs. the Commonwealth Caribbean 54.12 75.40 0.00 Southern cone vs. the Commonwealth Caribbean 37.91 80.02 0.00 Andean states vs. the Commonwealth Caribbean 14.59 31.38 0.00 Mexico and Central America vs. the Commonwealth 10.60 23.53 0.00 Caribbean Other Caribbean vs. the Commonwealth Caribbean 5.65 10.11 0.00

Looking now at the differences within regions (Figure 5.2), the greatest amount of variation within the country groupings comes from the Southern cone. Whereas Uruguay shows clear signs of support for gay rights (even higher than that of the US), the mean values for Argentina, Brazil and Chile lie close to the midpoint of the scale, and Paraguay shows little support for extending rights to gays and lesbians. The Commonwealth Caribbean shows the least variation across countries, with their scores lying in the range of 5 to 12. We also see some country differences within the Commonwealth Caribbean, with Trinidad and Tobago showing the greatest support for

141

same-sex marriage and Jamaica showing the least support. In fact, Jamaica is the least supportive state in the full sample.

Figure 5.2: Average approval of same-sex marriage across the Americas – Country differences

Canada 71.7

US US & US 51.1 Canada Uruguay 70.6 Argentina 57.5 Brazil 45.5

Chile 47.9 Southern Cone Paraguay 20.8 Colombia 34.8 Venezuela 29.6 Peru 26.4

Bolivia 22.2 AndeanSates Ecuador 16.5 Mexico 43.3 Costa Rica 28.2 Panama 25.0 Nicaragua 14.5 Honduras 14.1

CentralAmerica El Salvador 13.9 Guatemala 10.8 Dom. Rep. 22.0

Suriname 17.8 Other

Caribbean Haiti 6.7 Trinidad & Tobago 14.4 Barbados 11.3 Bahamas 10.4

Belize 8.4 Caribbean

Guyana 7.6 Commonwealth Jamaica 5.1

0 10 20 30 40 50 60 70 80

Data sourced from the 2014 Americas Barometer by LAPOP (www.LapopSurveys.org). Figure based on weighted sample.

142

Thus far, the descriptive analysis suggests that discrimination against homosexuals varies considerably across the Americas, and suggests that, in this sample, members of the Commonwealth Caribbean not only display the highest levels of homonegativity at a regional level, but it also shows the least within-group variation. With such differences across groups, it seems logical to conclude that the attitudes may be affected by the societal context. In what follows, the extent to which societal context explains the higher levels of anti-gay prejudice observed in the Commonwealth Caribbean relative to other parts of the Americas is examined – looking, chiefly, at differences in socioeconomic development, religious climate, and laws.

5.2 Exploring the differences in attitudes towards same-sex marriage in the Americas

5.2.1 Socioeconomic development Similar to Chapter 3, the level of socioeconomic development is represented by the HDI. As shown in Table 5.3 the level of human development varies significantly across the Americas.

Table 5.3: Socioeconomic development across the Americas

HDI value Very high level of human development US 0.915 Canada 0.913 Argentina 0.836 Chile 0.832 High level of human development Uruguay 0.793 Bahamas 0.790 Barbados 0.785 Panama 0.780 Trinidad and Tobago 0.772 Costa Rica 0.766 Venezuela 0.762 Mexico 0.756 Brazil 0.755 Peru 0.734 Ecuador 0.732 Colombia 0.720 Jamaica 0.719

143

Table 5.3 cont’d HDI value High level of human development Dominican Republic 0.715 Belize 0.715 Suriname 0.714 Medium human development Paraguay 0.679 El Salvador 0.666 Bolivia 0.662 Guyana 0.636 Nicaragua 0.631 Guatemala 0.627 Honduras 0.606 Low human development Haiti 0.438

The post-materialist thesis suggests that persons living in countries with higher levels of existential security and socioeconomic development would be more approving of same-sex marriage than those living in less developed societies (see Chapter 2). Does this hold true in the Americas? Figure 5.3 plots attitudes towards same-sex marriage by the level of development. On average, countries with higher levels of human development appear to be more approving of same-sex marriage. However, as shown in Figure 5.4, members of the Commonwealth Caribbean are considerably less tolerant than their level of development would suggest. For instance, Barbados, the Bahamas, Belize, Jamaica and Trinidad and Tobago, qualify as having high levels of human development. Yet, the average approval for homosexual rights in these states was 11.1, significantly below the mean value of 30.8 recorded for other countries in the ‘high level of human development’ category (that is, Brazil, Costa Rica, Colombia, Dominican Republic, Ecuador, Mexico, Panama, Peru, Suriname, Uruguay and Venezuela). In fact, the relationship between public attitudes towards same-sex marriage and HDI strengthens when members of the Commonwealth Caribbean are removed from the bivariate analysis: the correlation coefficient increases from 0.69 to 0.83. However, differences in HDI do seem to explain some the variation within the Commonwealth Caribbean. As shown in Figure 5.4, the Commonwealth Caribbean states on the upper end of the development spectrum (that is, Barbados, Bahamas, and Trinidad and Tobago) lend more support to same-sex marriage than states on the lower end/

144

Figure 5.3: Approval of same-sex marriage by level of socioeconomic development

60 57.9

sex mariage sex 40 -

22.3 20 Approval of same of Approval 15.7 10 6.7

0 Very High Human High Human Medium Human Low Human Development Development Development Development

Data sourced from the 2015 UNDP HDI report (UNDP 2015) and the 2014 Americas Barometer by LAPOP (www.LapopSurveys.org). LAPOP data based on weighted sample.

Figure 5.4: Socioeconomic development and approval of same-sex marriage

70 Uruguay Canada

60 Argentina

50 United States Brazil Chile Suriname

sex sex marriage Colombia Mexico - 40 Paraguay Venezuela 30 Bolivia Peru Costa Rica Panama El Salvador Dom. Rep. 20 Nicaragua Ecuador Honduras Trinidad & Tobago Guatemala Barbados 10 Belize Haiti Bahamas Guyana Jamaica

0 Average Average approval same of 0.4 0.5 0.6 0.7 0.8 0.9 1.0 -10 HDI

-20 Very High Human Development High Human Development Medium Human Development Low Human Development

Data sourced from the 2015 UNDP HDI report (UNDP 2015) and the 2014 Americas Barometer by LAPOP (www.LapopSurveys.org). LAPOP data based on weighted sample.

145

5.2.2 Religious Climate A commonality across the Americas is that the Christian faith holds the most adherents. Specifically, in a comprehensive demographic study of more than 200 countries the Pew Research Center (Pew Research Center 2011) reports that:

“Of the world’s five major geographic regions, the Americas have both the largest number and the highest proportion of Christians. More than a third of Christians worldwide (37%) live in the Americas, where nearly nine-in-ten people (86%) are Christian (Pew Research Center 2011: 14-15)

However, there is a clear denominational divide. As previously discussed in Chapter 3, there are relatively large shares of Evangelicals in the Commonwealth. Meanwhile, Latin America is historically a Catholic region (Pew Research Center 2014a), owing to their pasts as former Portuguese and Spanish colonies. Although in recent years, many persons have joined Protestant churches – particularly Evangelical Protestant churches – Catholicism currently remains the dominant Christian tradition in these states (Pew Research Center 2014a). As a means of analysing the differences in Evangelical presence across the Americas, the share of Evangelicals across the regional groupings is presented in Figure 5.5. Unfortunately, census information on the share of Evangelicals was unattainable for many countries in the sample. However, the AmericasBarometer survey included a question about religious belonging, which was used to estimate the share of Evangelicals for the region. As expected, the Commonwealth Caribbean has the largest share of persons that identify as Evangelicals. This could provide part of the explanation as to why members of the Commonwealth Caribbean show greater levels of anti-gay prejudice than their geographical neighbours. Recall from the literature review (Chapter 2), several studies have reported that relative to other Christian denominations, Evangelicals tend to be the least tolerant of homosexuality. It follows that citizens of countries with a large Evangelical presence are likely to be less willing to support same- sex marriage than those with a small share of Evangelicals. A scatter plot of national attitudes and the share of Evangelicals in the country is given in Figure 5.6. Consistent with expectations, there is a negative relationship between the share of Evangelicals and approval of same-sex marriage. The correlation coefficient was -0.56.

146

Figure 5.5: Share of persons who identify as Evangelicals (%)

All 19.5

Andean States 9.0

US & Canada 10.2

Southern Cone 11.9

Other Caribbean 15.6 Country Grouping Mexico & Central America 21.4

Commonwealth Caribbean 33.1

0 5 10 15 20 25 30 35

Data sourced from the 2014 AmerciasBarometer survey. Figures based on weighted sample.

Figure 5.6: Evangelicalism and approval of same-sex marriage

80 Canada 70 Uruguay

60 Argentina

50 United States

Chile Sex Sex Marriage

- Brazil Mexico 40 Venezuela Panama Colombia 30 Peru Costa Rica Hondur… Bolivia Dom. Rep. Nicaragua 20 Paraguay Suriname Ecuador El Salvador Trinidad & Tobago Barbados 10 Bahamas

Average Average approval Same of Belize Haiti Guatemala Jamaica Guyana 0 0 10 20 30 40 50 60 -10 Share of Evangelicals (%)

Data sourced from the 2014 AmerciasBarometer survey. Figures based on weighted sample.

147

5.2.3 Laws As noted in the literature review (Chapter 2), a commonly held view is that there is a bi- directional relationship between public attitudes towards homosexual rights. While laws are often a reflection of public opinion, at the same time, they are also socialising agents. So, emphasising homosexual rights could increase public tolerance. The civil rights extended to same-sex couples differ dramatically across the Americas. First, while all states in this sample criminalised same-sex sexual acts at some point in their history, the only five states in the sample that still carry these bans are members of the Commonwealth Caribbean. There are also differences in policies among the states where the bans on same-sex behaviours have been removed. By 2014 (the year the AmericasBarometer was conducted), same-sex marriage was legal in six states, and three either offered all/most of the rights attached to marriage (but not marriage itself) or gave some type of recognition to same-sex relationships (see Table 5.4). Differences in anti-gay prejudice could be due to differences in the rights extended to gays and lesbians in these states.

Table 5.4: Regulation of same-sex behaviours as at 2014

Same-sex marriage Same-sex unions No recognition of Bans on same-sex recognised (in some recognised or some same-sex intimacy parts or nationwide) recognition of same- relationships, but sex relationships in same-sex intimacy law (in some parts or legal nationwide) Argentina Colombia Bahamas Barbados Brazil Costa Rica Bolivia Belize Canada Ecuador Chile Guyana Mexico Dominica Republic Jamaica US El Salvador Trinidad and Tobago Uruguay Guatemala Haiti Honduras Nicaragua Panama Paraguay Peru Suriname Venezuela Note: Information attained from Leveille (2016), Barrientos (2016) and Carroll (2016).

148

Figure 5.7 plots the average attitudes across the legal groupings. As expected, citizens in countries with more liberal policies tend to show higher levels of tolerance. This suggests that the high level of anti-gay prejudice observed in the Commonwealth Caribbean cluster, relative to the other states in the sample, could be due to these individuals being socialised in societies where same-sex behaviours are harshly regulated.

Figure 5.7: Legality of same-sex relations and approval of same-sex marriage 70

60 57.2

sex mariagesex 40 -

30 26.5

20 19.3 Approval of of same Approval

10 10.8

0 Same-sex Recognition of Same-sex Same-sex marriage legal same-sex intimacy legal intimacy banned relationships

Data sourced from the Americas Barometer surveys, Leveille (2016) and Barrientos (2016). Figures based on weighted sample.

5.3 A multi-level analysis of attitudes

5.3.1 Statistical method The previous section hinted that societal context may provide some explanation for the variations in attitudes observed across countries in the Americas. In fact, based on the plots, it would seem logical to assume that the higher level of anti-gay prejudice reported by members of the Commonwealth Caribbean could be largely explained by their larger Evangelical presence, along with the continued criminalisation of intimate homosexual acts, and to a limited extent, their levels of socioeconomic development. This raises the question: are there still differences in attitudes within countries when these factors are

149

taken into account? The previous bivariate correlations only provide a cursory glance at the contextual effects. A more thorough exploration can be attained via a multi-level analysis. As noted in the literature review, public attitudes towards homosexuality are not only influenced by variables at the country level, but also by variables at the individual level. A two-level multilevel model will allow us to discern the variation within countries in the Americas (the individual effects) from variation between these countries (country- level effects). Standard ordinary least squares (OLS) regression models assume that the errors are independent and identically distributed. However, when individuals are nested in countries, they are more likely to be similar to their countrymen than individuals from other countries. This violates the aforementioned OLS assumption. Multi-level models allow us to account for the nesting structure (Snijders and Bosker 2012). The model employed in this study can be written algebraically as:

풀푖푗 = 훼00 + 훽푿푖푗 + 훾풁푗 + 휀푖푗 + 푢0푗

The 푖 and 푗 subscripts represent individual and country-level influences, respectively. 풀푖푗 refers to an individual’s support for same-sex marriage, 푿푖푗 is the set of individual-level predictors, 풁푗 represents the country level predictors and 훼00 is the usual intercept. 휀푖푗 is a vector of individual level error terms, which are assumed to be normally distributed with a mean of 0 and variance of 1. 푢0푗 are the unexplained country effects or the country-level residuals, which are independent and identically distributed.

5.3.2 Description of the variables The dependent variable remains the responses to the survey item measuring responses to the AmericasBarometer survey question about approval of same-sex marriage, and the country level variables remain the same as those used in the previous analyses. At the individual level, the study controls for socioeconomic factors (birth cohort, gender) and the impact of religion (religious identity and religiousness) – see Table 5.5. The individual level predictors chosen were based on the current literature and data provided by the AmericasBarometer survey. Specifically, while the AmericasBarometer survey included items capturing on political ideology and years of education, these questions were not asked in every country. Information on education was not available for US and Canadian respondents, and questions about political ideology were not included in the surveys

150

conducted in the Bahamas, Barbados and Suriname. To maximise the number of countries in the sample, I opted not to include these variables.

Table 5.5: Descriptive statistics for individual level predictor variables Variable Frequency (%)

Sex Men 49.7 Women 50.3

Birth cohort 1990 or later 18.4 1980 24.1 1970 20.7 1960 15.6 1950 11.6 1940 or earlier 9.1 No response 0.6

Religious identity Evangelical 19.6 Catholic 44.2 Mainline Protestant 12.5 Other 9.8 Not religious 11.4 Don’t know 0.9 No response 1.5

Importance of religion Very important 58.6 Rather important 24.6 Not very important 10.0 Not at all important 5.9 Don’t know 0.4 No response 0.5 Note: Data based on weighted sample

Gender is measured via a binary indicator, with men serving as the reference category. The birth cohort variable captures the decade in which the respondent is born. Persons born in the 1990s or later are the youngest cohort and serve as the reference group. The remaining categories are persons born in the: 1980s; 1970s; 1960s; 1950s; 1940s or earlier; and finally, those who did not provide a response to the question. Religious identity is presented as a seven-category variable and captures whether or not

151

the person is (1) Evangelical (reference); (2) Catholic; (3) Mainline Protestant; (4) Other religion; (5) Not religiously affiliated; (6) Don’t know; and (7) No response. The indicator of personal religiosity is based on a survey item that asked respondents about the importance of religion in their lives. Responses to this question are presented as a five- category variable that captures whether or not the individuals reported that (1) religion is very important; (2) rather important; (3) not very important; (4) not at all important; (5) don’t know; and (6) no response. List-wise deletion of missing values of the dependent and independent individual level variables (the exception being age and religion variables – see above descriptions) resulted in an analytical sample of 51,662 respondents from 28 countries.

5.4 Results

The main interest lies in how the variation between countries changes as predictors are added. Hence, the multilevel analysis was done in five steps, and in each step, a new cluster of variables was added to the model. Model 1 has no explanatory variations (often called the random intercept-only model), and serves as the baseline model for comparing the other four models. The intraclass correlation (ICC) for the empty model (Table 5.6) reveals that 26.9 per cent of the variation in attitudes about same-sex marriage can be attributed to differences between nations. Model 2 adds the individual-level variables; Model 3 adds the modernization factor; Model 4 adds the religious climate variable, and Model 5 adds the legal context variable. A likelihood ratio test revealed that, in each case, the change log likelihood was statistically significant at the 5 per cent level of testing. Model 2 in Table 5.6 shows that all individual variables are statistically significant and have the expected effects: men appear less willing to extend marital rights to same- sex couples. Likewise, people from older cohorts are more likely to disapprove of same- sex marriage than individuals from younger cohorts. Across religious identities, Evangelicals reported the lowest approval of same-sex marriage, and persons who assign high importance to religion were also less approving than persons for whom religion is less salient. Though significant, the individual level variables had a modest impact on the cross-country variation. With the addition of the individual level variables, the variance at the country level decreases by 14.2 per cent, which only reduced the ICC by 1.9 percentage points. Hence, after controlling for composition effects, 25.0 per cent of the variation in attitudes can still be attributed to differences between countries.

152

Table 5.6: Multi-level analysis of approval towards same-sex marriage in the Americas

Model 1 Model 2 Model 3 Model 4 Model 5 Intercept 26.926*** 21.272*** 49.951*** 55.944*** 59.015***

Individual Characteristics Gender Men (base) Women 5.912*** 5.910*** 5.910*** 5.916*** Birth Cohort 1990s or later (base) 1980s -3.138*** -3.140*** -3.140*** -3.139*** 1970s -6.236*** -6.239*** -6.239*** -6.236*** 1960s -8.830*** -8.834*** -8.833*** -8.827*** 1950s -9.818*** -9.824*** -9.823*** -9.816*** 1940s or earlier -13.922*** -13.931*** -13.926*** -13.910*** Non-response -9.453*** -9.465*** -9.454*** -9.480*** Religious identity Evangelical (base) Catholic 7.978*** 7.988*** 7.961*** 7.954*** Mainline Protestant 3.718*** 3.716*** 3.714*** 3.709*** Other 6.751*** 6.751*** 6.741*** 6.747*** Not religious 8.441*** 8.451*** 8.446*** 8.456*** Don’t know 3.014* 3.024* 3.023* 3.077* No response 3.544** 3.538** 3.533** 3.545**

153

Table 5.6 cont’d Model 1 Model 2 Model 3 Model 4 Model 5 Importance of religion Very important (base) Rather important 4.609*** 4.603*** 4.600*** 4.607*** Not very important 6.433*** 6.421*** 6.421*** 6.428*** Not at all important 12.976*** 12.966*** 12.966*** 12.978*** Don’t know 3.264 3.260 3.251 3.249 No response 6.994*** 6.991*** 6.983*** 6.976*** Country characteristics HDI Very high human development (base) High human development -29.796*** -25.497*** -12.440*** Medium human development -39.599*** -32.088*** -16.749*** Low human development -49.017*** -52.020*** -32.208*** Religious climate Share of Evangelical -0.553** -0.271* Legal context Same-sex marriage (base) Recognise same-sex relationships -24.406*** Bans on same-sex intimacy removed -24.403*** Same-sex intimacy criminalised -30.539***

Individual variance 952.097*** 902.713*** 902.713*** 902.713*** 902.714*** Country variance 351.235*** 301.213 *** 138.589*** 100.372*** 29.236*** ICC 0.269 0.250 0.133 0.100 0.031 Log likelihood -250,561.641 -249,184.430 -249,173.590 -249,169.100 -249,152.050 Note: ***, **, * and + indicate significance at the 0.1, 1, 5 and 10 per cent levels, respectively

154

Model 3 adds the first contextual variable, HDI, which is statistically significant: compared to persons living in countries with very high human development, persons who live in countries that are classified as having high, medium or low human development are less supportive of homosexual rights. The level of socioeconomic development appears to be a very important factor in explaining country-level differences in attitudes. When this variable is added to the model, the country level variance falls by an additional 54 per cent, which in turn resulted in a significant drop in ICC. In Model 3, the ICC now stands at 0.133, nearly half of what it was in Model 2. Religious climate is introduced in Model 4. This variable was statistically significant, and in line with my a-priori expectation that persons who live in countries with a greater Evangelical presence reported more negative attitudes than those living in areas with smaller shares of Evangelicals. The addition of this variable led to a further reduction in the country level variance (27.6 per cent) and ICC (from 0.133 in Model 3 to 0.100). The final contextual variable (the legal environment) is added in Model 5. Compared to persons living in countries where same-sex marriage is allowed, persons living in states with more restrictions on same-sex behaviours are much less approving of same-sex marriage. Of key interest, the greatest gaps occurred between persons living in states where same-sex marriage is legal, and states where same-sex intimacy is banned (that is, the Commonwealth Caribbean states). With the addition of the legal context variable, the coefficient on the religious climate and socioeconomic development variables decreased significantly, but remained statistically significant. The smaller impact of these variables is not surprising. As noted in the literature review, a commonly held view is that in religious states, religious values are not only transmitted by the church: the legal codes are also infused with religious norms. Similarly, as alluded to in Chapter 3, developed countries are less likely to restrict same-sex behaviours than countries with lower levels of development. Given the interplay between laws, religion, and HDI, the introduction of the legal context variables moderated the effects of religious values and socioeconomic development. With respect to the impact of the legal variables on the country-level variation, by including the information about the laws across these states, the country-level variance falls by 70.9 per cent. In fact, with all the contextual variables in the model, the ICC suggests that only 3.1 per cent of the variation in attitudes towards the civil rights of same-sex couples is still explained by the country in which people live.

155

As noted in Chapter 2, the literature suggests that laws not only shape attitudes in a country, but they are also influenced by them. To investigate the issue of causality, longitudinal data on the Americas would be needed, which to the best of my knowledge is not available. However, similar to Takács and Szalma (2014) and Slenders et al. (2014), I have complemented the multi-level analyses with impact monitoring (see Chapter 2). Using cross-section data from the 2010 and 201437 waves of the AmericasBarometer survey, I compare changes in the average support for same-sex marriage for three types of countries: (1) countries that either legalised same-sex marriage or recognised same-sex relationships before 201038 (n = 7); (2) countries that implemented these laws between 2010 and 2014 (n = 1); (3) countries that are without these laws, but have no bans on same-sex intimacy (n=13); and (4) countries that have laws that prohibit same-sex intimacy (n=4). Table 5.7 shows the average approval of same-sex marriage across countries in the Americas, sorted by the difference in attitudes (labelled diff.). Data were not available for Barbados, the Bahamas, or Brazil in 2010. The largest change in attitudes came from Uruguay, with an average increase of 20.2 points. A subset of countries reported negligible changes (1 point or less). This includes Colombia, Guyana, Peru, Argentina, Venezuela, and Trinidad and Tobago. Finally, some countries had a significant decline in support for same-sex marriage, with Honduras and Belize showing the largest falloffs (8.5 and 9.0 points, respectively). The question then is: was the change in support influenced by the legislation? Table 5.8 hints that the legislation could have played a role. Pairwise comparison tests suggest that in countries where same-sex relationships were recognised before 2010 (group 1) or between 2010 and 2014 (group 2), there was a significant increase in approval for same- sex marriage. On the other hand, there was no significant change in support for same- sex unions in countries where no such legislation was enacted (group 3), or those with bans on same-sex intimacy (group 4). These results suggest that in the Americas, a change in legislation might encourage individuals to become more tolerant. The multi-level model and impact analysis suggests that level of socioeconomic development, the share of Evangelicals, and the legal environment explain much of the variance in attitudes towards same-sex marital rights across the Americas. One cannot help but wonder if controlling for these cross-country differences also affects how the

37 While the AmericasBarometer survey dates back to 2004, the question on same-sex marriage was only introduced in 2010. 38 I focus on when the countries first introduced these laws, irrespective of whether the implementation was only in a few provinces or nationwide. 156

countries are ranked in terms of anti-gay prejudice. Specifically, do members of the Commonwealth Caribbean still show the lowest levels of approval of same-sex marriage after controlling for religion, laws, and HDI? The remaining differences across countries can be explored by looking at the random intercepts for each country.

Table 5.7: Changes in (average) approval of same-sex marriage and legal environment by country 2010-2014 Country 2010 2014 Diff. Same-sex marriage/relationships recognised before 2010 Uruguay 50.5 70.6 20.2 Canada 63.9 71.7 7.8 Mexico 37.8 43.3 5.4 United States 47.4 51.1 3.7 Argentina 57.7 57.5 -0.2 Ecuador 18.4 16.5 -1.9 Colombia 34.4 34.8 0.4 Some legal recognition of same-sex relations between 2010 and 2014 Costa Rica 20.7 28.2 7.5 Same-sex intimacy legal Chile 40.0 47.9 7.9 Paraguay 16.1 20.8 4.7 El Salvador 10.3 13.9 3.7 Dominica Rep. 18.6 22.0 3.5 Panama 22.8 25.0 2.2 Haiti 5.0 6.7 1.6 Peru 26.3 26.4 0.1 Venezuela 57.7 57.5 -0.2 Nicaragua 15.6 14.5 -1.1 Suriname 20.3 17.8 -2.4 Bolivia 24.7 22.2 -2.5 Guatemala 16.5 10.8 -5.7 Honduras 22.6 14.1 -8.5 Bans on same-sex intimacy Jamaica 3.5 5.1 1.5 Guyana 7.2 7.6 0.4 Trinidad and Tobago 15.4 14.4 -1.0 Belize 17.5 8.4 -9.0 Note: (1)***, **, * and + indicates significance at the 0.1, 1, 5 and 10 per cent levels, respectively. (2) Data on attitudes was attained from the AmericasBarometer survey, while information on the legal environment came from Leveille (2016), Barrientos (2016) and Carroll (2016). (3) Figures based on weighted samples.

157

Table 5.8: Differences in approvals of same-sex marriage Average Average Diff. approval approval in 2010 in 2014 Legal Environment 1. Same-sex marriage/relationships recognised before 2010 (n = 7) 42.7 50.5 7.8*** 2. Some legal recognition of same-sex relations between 2010 and 2014 (n = 1) 20.7 28.2 7.5*** 3. Same-sex intimacy legal (n = 13) 20.4 19.7 -0.7 4. Bans on same-sex intimacy (n = 4) 10.8 11.8 1.0 Note: ***, **, * and + indicate significance at the 0.1, 1, 5 and 10 per cent levels, respectively

Figure 5.8 plots the country effects for the final model and the empty model, which was included for comparative purposes. The zero line here represents the intercept presented in Table 5.6 for each model, and the scores are the deviation from this intercept for each country. In layman's terms, a score above the zero-line means the country is more tolerant than conditions would predict. Scores below the zero-line mean that the countries’ tolerance levels are lower than conditions would predict. The country effects are ordered from the bottom (that is, from the smallest values to the largest). In both models, the intercepts between countries vary, but, the intercepts from Model 5 are more narrowly distributed. This is as expected, as the inclusion of the predictors reduced a large portion of the differences in attitudes between the countries As in Figure 5.2, in the empty model, members of the Commonwealth Caribbean (in yellow) are among the least supportive of granting civil rights to same-sex couples. Ranking them by approval, Jamaica ranks last (28th), Guyana at number 27, Belize at 26, the Bahamas at 25, Barbados at 22, and Trinidad and Tobago, the most supportive, comes in at number 18. After controlling for differences in compositional effects, socioeconomic development, the share of Evangelicals, and the legal situation, the rankings change. Belize and the Bahamas are still among the least tolerant, only moving up slightly in rank to 25th and 24th, respectively, Jamaica moves up 15 slots to 13th, Guyana to 11th (16 slots), Trinidad to 10th (8 slots), and, Barbados now has the highest rank at number 6 (up 19 slots). Based on these plots, the high level of anti-gay prejudice observed in the Commonwealth Caribbean relative to other states in the Americas can largely be explained by differences in religious climate, the anti-gay laws, and human development.

158

Figure 5.8: Plot of country effects (Random intercepts) (Commonwealth Caribbean in yellow) Model 1 Model 5

sex sex marriage - 10

-10 Approval of same

-20

-30

Peru

Haiti

Chile

Peru

Haiti

Chile

Brazil

Belize

Costa

Brazil

Belize

Bolivia

Mexico

Canada

Guyana

Jamaica

Bolivia

Panama

Mexico

Ecuador

Canada

Guyana

Uruguay

Jamaica

Panama Bahamas

Paraguay

Ecuador

Colombia

Barbados

Surniame

Uruguay

Honduras

Argentina

Nicaragua

Venezuela

Costa Rica Costa

Bahamas

Paraguay

Guatemala

Colombia

Barbados

Suriname

Honduras

Argentina

El Salvador El

Nicaragua

Venezuela

Guatemala

Dominican

El Salvador El

United States United

Trinidad & Tobago & Trinidad Dominican Republic Dominican Trinidad & Tobago & Trinidad

159

5.4.1 The impact of the age and religion: Does context matter? In the previous chapter, I reported that age, education and religious denomination were not robust predictors of attitudes towards gay rights in Barbados, Guyana, and Trinidad and Tobago. I noted that this was in contrast to studies on the West, where these variables are usually strong predictors of gay rights, even in the presence of other control variables. This finding seemed to suggest some of the findings from the literature on the West may not coextend to Commonwealth Caribbean countries. In fact, based on the extant literature, the combined effects of living in a developing country with a strong Evangelical presence and having anti-gay laws on the books could possibly limit the impact of age and religion. To recap, while Inglehart’s post-materialist thesis tells us that development has a liberalising effect on attitudes, Inglehart (1971) also suggests that the change in values brought about by development may be more evident among the younger cohorts than older cohorts (see Chapter 2). This would suggest that in developed countries, we are likely to see more variation in attitudes by age than in less developed countries. The work by Norris and Inglehart (2011) suggests that in religious countries, religious values and norms are far-reaching. In this environment, individuals who do not subscribe to the main religion are still influenced by the dominant religion in the country (see Chapter 2). Thus, we can expect to find less variation by religious identity in a monopolistic religious environment. Finally, research by Kreitzer et al. (2014) implies that the values expressed by the laws are highly likely to be internalised by persons with a predisposition to approve of the messages being signalled by the law (see Chapter 2). This implies that the impact of age and religion on attitudes may be greater in countries with more progressive laws than those where gay rights are limited. In this subsection, I draw on the extant literature to test if the context in which Caribbeans live could possibly explain the weak age and religious effects found in the previous chapter. Four additional models were estimated, each of which included one of the following interaction terms: (1) HDI and birth cohort; (2) share of Evangelicals and religious identity; (3) legal context and birth cohort; and (4) legal context and religious identity. All the predictors from Model 5 (see Table 5.6) were included as controls. As the interaction terms are not directly interpretable, I opted to calculate and plot the predicted values for ease of interpretation. The output from the multi-level regression is presented in Appendix C. Figure 5.9 presents the effect of birth cohort on approval of same-sex marriage across the varying levels of human development. In line with a-priori expectations, the

160

impact of birth cohort differed with the level of development: cohort differences are larger in developed countries. In countries with very high human development, the predicted value for approval of same-sex marriage is 45.1 (based on a scale of 0 to 100), while the predicted value for someone born in the 1940s or earlier 23.9, yielding a gap of 21.0 points. The gap between the youngest cohort and oldest cohort is much smaller for countries with high human development (13.3) and medium human development (11.7), and countries with low human development (3.0).

Figure 5.9: Approval of same-sex marriage by birth cohorts and level of development

50.0

40.0

30.0

20.0

10.0

Approval of same sex marriage (predicted values) 0.0 Very high human High human Medium human Low human development development development developmen

1990s or later 1940s or ealier

Note: Graphs based on predicted values from adding an interaction of HDI and birth cohort to Model 5 in Table 5.6. See appendix C for the tabulated results.

Also, in line with Inglehart’s thesis, the impact of development on attitudes seem more prominent for younger cohorts. The predicted values for persons born in 1990 or later range from a high of 45.1 points in countries with very high human development to a low of 6.7 in countries with low human development, that is, a gap of 38.4 points. The gap is much smaller for the oldest cohort. For persons born before 1950, the predicted values for approval of same-sex marriage range from 23.9 (very high human development) to 3.7 (low human development), yielding a gap of 20.3 points.

161

Figure 5.10 shows the fitted attitude scores by individual religious identity and religious context, and the results appear consistent with a-priori expectations

Figure 5.10: Approval of same-sex marriage religious identities and Evangelical presence

Evangelical Catholic Mainline Prostestant Other Not religious 40

sex sex marriage (predicted values) -

10 Approval of same

0 LOW EVANGELICAL MEAN HIGH EVANGELICAL PRESENCE PRESENCE

Note: Graphs based on predicted values from adding an interaction of religious identity and share of Evangelicals to Model 5 in Table 5.6. See appendix C for the tabulated results.

In nations characterised as having a low share of Evangelicals (that is, the share of Evangelicals is one standard deviation below the mean), we find large differences in public approval of same-sex marriage across personal religious identities. The predicted values range from a high of 36.2 for persons who are religiously affiliated to a low 17.1 for Evangelicals. However, in countries with a strong Evangelical presence (share of Evangelicals is one standard deviation above the mean), we see less variation in attitudes by religious identity. In this context, the predicted values range between 16.2 (Evangelicals) and 22.8 (not religious). This brings us to a second observation: the predicted values of Evangelicals are roughly the same in both contexts. Evangelicals maintain their conservative attitudes whether they are in a country with a high proportion of Evangelicals or a low proportion of Evangelicals. The impact of a strong Evangelical presence mattered more for non-Evangelicals. Unlike other denominations in the

162

Americas, Evangelicals appear consistent in their teachings and views on same-sex marriage, regardless of the religious culture within which they exist. Finally, I evaluate whether birth cohort and religious identity matter more for residents of countries with progressive laws on homosexual behaviour. These interactions are shown in Figure 5.11 and Figure 5.12. In line with a-priori expectations, differences in attitudes across cohorts are much more substantial in countries with progressive laws (Figure 5.11).

Figure 5.11: Approval of same-sex marriage by birth cohort effects and legal context

1990s or later 1940s ore earlier 60

sex sex marriage (predicted values) - 20

10 Approval of same 0

Note: Graphs based on predicted values from adding an interaction of cohort and legal context toModel 5 in Table 5.6. See appendix C for the tabulated results.

In countries where same-sex marriage was introduced, the difference in attitudes between persons born as part of the youngest cohort and the oldest cohort is 22.9 points. This is significantly higher than the gap observed for countries that only recognise same- sex partnerships (17.8 points), or those that do not recognise same-sex relationships, but where same-sex intimacy is legal (13.6). Finally, the difference in attitudes between the youngest cohort and oldest cohort is smallest in those countries that ban intimate same- sex acts (10.2 points). The results suggest that liberal laws also seem to matter more for the younger cohorts. There is significant variation in the attitudes of persons born in 1990

163

or later across the legal contexts. With the exception of senior persons living in countries where same-sex marriage is legal, the attitudes of older cohorts are roughly on par. In fact, the predictive values suggest that senior persons living in states where same-sex marriage is legal report greater approval of same-sex marriage than persons living in countries where same-sex intimacy is criminalised. Figure 5.12 plots the interaction of religious identity and legal context. Regardless of legal context, religiously unaffiliated individuals show the greatest support for same- sex marriage, followed by Catholics, persons from other religions, and Mainline Protestants. Evangelicals are consistently the least supportive of same-sex marriage. The plot also suggests that variation in attitudes by religious identity is largest in countries where same-sex marriage is legal. The predicted values for approval of same-sex marriage range from 26.9 (Evangelicals) to 56.0 (religiously unaffiliated). As we move to countries with more restrictive policies, the variation in attitudes becomes smaller. In countries that recognise same-sex relationships, the predicted values range from 17.1 to 30.2. In countries where same-sex relationships are not recognised, but same-sex intimacy is legal, the predicted values for approval lie between 15.8 and 24.9. Finally, in states that criminalise same-sex intimacy, the predicted values for approval of same-sex marriage lie between 10.4 and 19.4. An interesting observation is that regardless of religious identity, persons living in countries where same-sex marriage is legal are more approving of same- sex marriage than those living in states where same-sex relationships are not recognised in any form or states where same-sex intimacy is criminalised.

164

Figure 5.12: Approval of same-sex marriage by religious identity and legal context

(predictedvalues)

sexmarriage

- Approvalsame of

165

5.5 Discussion

Attitudes towards same-sex marriage vary widely across the Americas. This chapter attempted to answer the following questions: (1) Is anti-gay prejudice higher in the independent Commonwealth Caribbean than the rest of the Americas? And, (2) To what extent can differences in societal context explain the differences in anti-gay prejudice? The descriptive statistics suggest that the countries tend to vary in the degree to which their citizens approve of same-sex marriage, and the least tolerant region is the Commonwealth Caribbean. A multilevel model was employed to further analyse attitudes towards same-sex marriage. The results are in line with previous research (see Chapter 2), which has suggested that women, younger cohorts, non-church members, and those for whom religion is an important aspect of their lives tend to be more supportive of same-sex marriage than their counterparts. The multilevel analysis also demonstrated that attitudes towards same-sex marriage are largely related to the level of socio-economic development, legal and religious environments. There is evidence of a strong positive relationship between the level of human development and public attitude towards same- sex marriage, and a larger Evangelical presence was associated with lower support for the civil rights of same-sex couples. Progressive regulation of homosexuality was also associated with greater support for same-sex marriage. Of course, it is possible that reforms to laws concerning homosexuality could have been foreshadowed by increased acceptance of homosexuality (reverse causality). Similar to Takács and Szalma (2014) and Slenders et al. (2014), I analysed the trends in attitudes over the period 2010–2014 to test whether the observed association between laws and attitudes were merely reflecting reverse causality. The pairwise comparison test showed that in countries where same-sex relationships are recognised (either before 2010 or between 2010 and 2015) there was a significant increase in support for same-sex marriage, while the remaining countries continued to be characterised by comparatively lower levels of approval. This points to the potential of law reforms to shape public opinions. The greater level of approval witnessed in countries where same-sex relationships are recognised suggests that the policies supportive of gays and lesbians are capable of instituting a new norm, where anti- gay prejudice is unacceptable. This finding is important. In the introduction of this thesis, I noted that policy makers in the Commonwealth Caribbean often justify the retention of the anti-gay laws under the guise of public support for the laws. These results suggest that the ‘wait until society evolves’ approach to law reform that policy makers appear to

166

have adopted may not be the best option. The retention of the laws can possibly slow down or even deter the liberalisation of attitudes towards gay rights. Rather, the empirical analysis suggests that the evolution of attitudes towards homosexuality in the Commonwealth Caribbean could possibly be ‘sped up’ by decriminalising same-sex intimacy. The analysis also revealed that the lion’s share of country variation in attitudes towards gay rights can be explained by the level of socioeconomic development, the aggregate share of Evangelicals, and the legal status of same-sex intimacy. The empty model suggested that 26.9 per cent of the differences in attitudes was explained by the country in which people live. After controlling for individual characteristics, 25 per cent of the variability in attitudes could still be attributed to an individual’s country of residence. However, once HDI, religious tradition, and the legal environment were taken into account, the cross-country variation fell to 3.1 per cent. Additional evidence of the usefulness of the contextual variables was found via an evaluation of the ranking of countries by their citizenry’s reported anti-gay prejudice before and after the contextual variables are controlled for. The intercept plots revealed that once context is taken into account, the positions of Barbados, Jamaica, Guyana and Trinidad and Tobago, change dramatically, each ranking higher on the scale. The main exception to my a-priori expectations is the Bahamas. As discussed in Chapter 3, Bahamas is only country in the Commonwealth Caribbean to decriminalise same-sex intimacy, and in 2014, it had the highest level of human development in this region (see Table 3.2). Yet, public approval of same-sex marriage is lower than in Barbados and Trinidad and Tobago, even before I control for the contextual variables (Figure 5.8). The mismatch between Bahamian attitudes and its socio-legal environment is an interesting puzzle worth further exploration. One cannot help but wonder if the Bahamas is just an exception to the rule, or whether there is something specific about Bahamian society that explains the mismatch. Unfortunately, due to data limitations, this issue could not be explored in this thesis, and could be an area of future research. However, in Chapter 3 of this thesis, I did speculate that tolerance of homosexuality in the Bahamas may be conditional – that is, out of sight, out of mind. It is quite possible that state recognition of same-sex marriage would lead to an increase in the visibility of same-sex couples. While in Chapter 4, I also speculated that residents in Barbados, Guyana and Trinidad may also have a similar stance, these states still criminalise same-sex intimacy. In essence, the possibility of same-sex marriages being

167

legal in these three states in the (very) near future is unlikely. This could lead to leniency in their views. However, with the decriminalisation of lesbianism and buggery in the Bahamas in 1991, same-sex marriage in the Bahamas, and by extension greater visibility of same-sex couples, is a possibility. In fact, in 2013, Chief Justice of the Bahamian Supreme Court Michael Barnet was quoted as saying:

“Bahamian courts will soon have to address the issue of same-sex marriage” (Kaleidoscope Trust and Kaleidoscope Australia Human Rights Foundation 2015)

The possibility that same-sex couples could soon be granted the right to express their love openly could stimulate more negative responses in the Bahamas than in Barbados or Trinidad and Tobago, who have lower levels of human development and stricter regulations on homosexuality. Perhaps the most important finding is that differences in societal contexts did not only lead to differences in the level of anti-gay prejudice observed in the Americas, it also resulted in differences in how attitudes varied by cohorts and religious identities. I found that differences across cohort effects were more prominent in countries with very high levels of development, as well as those with more progressive laws. More than this, differences across religious identities were smallest in countries with a high proportion of Evangelicals and countries where same-sex intimacy is banned. Most interestingly, Evangelicals were the only religious group that were no more or less affected by the religious milieu within which they existed. The results suggest that Evangelical views on same-sex marriage were consistent across religious context. This is somewhat expected. A common feature of Evangelicals across nations is Bible fundamentalism (Moore and Vanneman 2003). This could lead more consistency in teachings across various cultures – at least on the issue of same-sex relationships. Mainline Protestants and Catholics are often less rigid in their interpretations of the Bible (Moore and Vanneman 2003), and this could result in them being more influenced by their local context. The finding that the impact of age and religious identity varies across various contexts is important. As noted in the Introduction, most studies on attitudes towards homosexuals and their rights are largely focused on North American and Western Europe. The fact that we found that the impact of individual correlates vary across different contexts justifies the initial thinking that results based on the West may not be universally valid. In fact, the findings presented in this chapter could be used to explain why we found a lack of impact for some variables (such as education and religious

168

identity) in Chapter 4, when studies on the US and Western Europe often report significant various significant differences in attitudes towards same-sex rights by demographics and religious identity. This lends credence to the view that context may be more important in explaining attitudes than individual level factors. Taken together, the multilevel model results provide evidence that context matters, and that these may be more important than individual level factors. The evidence suggests societal context plays a pivotal role in shaping attitudes in the Commonwealth Caribbean. Specifically, the low level of support is largely a reflection of the region’s levels of socio-economic development, the large Evangelical presence and the continued restrictions on same-sex behaviours.

169

Chapter 6 : ‘Man to man, gyal to gyal…dats wrong’: A qualitative analysis of anti-gay rhetoric in Caribbean [Jamaican] popular culture

“Every man grab a gyal, and every gal grab a man. Man to man, gyal to gyal dats wrong. Scorn dem”. (Vbyz Kartel and Spice 2010).

Thus far, my analysis has focused on the factors predicting attitudes towards gay rights in the Commonwealth Caribbean. Based on Chapter 5, it appears that context matters more than some of the individual-level variables. This finding merits taking a closer look at the context from which these attitudes emerge. As such, this chapter uses a thematic analysis to investigate the policing of homosexuality in reggae/dancehall – music indigenous to Jamaica, but popular throughout the region. An examination of how arguments favouring intolerance are constructed – that is, the overall framing and specific language used – is crucial to understanding how homosexuality may be understood in everyday life. As alluded to in both introductory remarks and Chapter 3, music is often seen an imitation of everyday life. Artists are not isolated from the wider society, and their views are likely to be influenced by the dominant culture in the country. They are also aware of their core audience and in their attempt to maximise profits, they may opt to produce songs that echo public sentiment. In fact, the greatest opposition to homosexuality can be found in the region’s iconic musical genres (dancehall and reggae) Greenberg (2010) noted the following:

“In a country where gay people are routine targets for violence, where the homes of suspected gay people are burned down at night and lesbians frequently confront the threat of rape, where police habitually refuse to intervene in crimes against gay victims and where men do not sit next to each other on a public bus in fear they will accidentally brush up against another man and consequently expose themselves to violent attack, dancehall implacably channels the country’s anti-gay animus” (Greenberg 2010).

A detailed analysis of the music is one avenue that would enable us to understand the context from which these lyrics emerged and to also recognise and examine the ways in

170

which anti-gay prejudice is justified and normalised. Hence, the study seeks to answer the following questions:

1) What are the main arguments used to condemn homosexuality in reggae and dancehall? 2) Are these arguments unique, or are they similar to arguments used to justify anti- gay prejudice elsewhere?

6.1 Music Trends

As a preliminary step to the thematic analysis, I analyse the number of anti-gay songs produced over the last four decades. As alluded to in Chapter 3, the number of anti-gay songs was particularly high in the early 2000s, but in the last decade, the number of songs with anti-gay lyrics declined significantly. In this subsection, I present some potential reasons for the rise and fall of anti-gay songs in reggae and dancehall using Lacoste’s list. In 2003, Christian Lacoste began compiling a list of dancehall/reggae songs with mentions of alternative sexualities (Lacoste 2013). At the time of his writing, the list included over 270 songs, of which 251 songs were verified as containing anti-gay rhetoric39. Each of these songs was recorded between 1978 and 2012. However, this list is wide-ranging and consists of songs solely dedicated to gay bashing (like Boom Bye Bye – see Chapter 3) as well as those not aimed at homosexuals, but which include derogatory references to them. Another important point is that Lacoste’s list is probably not exhaustive. Nonetheless, this list currently stands as the largest public database of reggae/dancehall anti-gay songs. As such, I draw on my edited version of the list to highlight some trends. Of course, given that the list may be incomplete, the trends presented here are more indicative than conclusive. Figure 6.1 provides a time series plot of the edited version of Lacoste’s list. Between 1978 and 1989, eight songs with anti-gay rhetoric were identified. Turning to the early 90s (which also coincides with the rise of dancehall music in the international arena), there is a slight pick-up, with 21 gay-bashing songs being identified between 1990

39 The original list contained over 270 songs. However, I deleted songs that could not be found and verified via an internet search (for instance, Google music, iTunes, Spotify, YouTube, Amazon etc). Also, dancehall/reggae music is now a global phenomenon. Hence, songs not made by Jamaican artists (for instance, Admiral P who is a Norwegian reggae artist) were excluded. 171

and 1995. However, there is a clear structural break in the series from 1996 to 2005: over 190 songs were acknowledged in this time period. This drastic increase in anti-gay lyrics found during this 1996-2005 period largely corresponds with (and maybe not coincidentally) the international controversy surrounding dancehall’s murderous anti-gay lyrics.

Figure 6.1: Time series plot of the edited list of anti-gay songs

30 Stop Murder Music Campaign 25

International 15 controversy begins

0 1980 1985 1990 1995 2000 2005 2010 Note: Original data sourced from Lacoste (2013). The author edited the list (see above).

As alluded to in Chapter 3, the violent anti-gay rants of Jamaican artists caused quite an international stir. Since the late 1990s, artists who lyrically assaulted homosexuals were subjected to intense criticisms from several LGBT organisations, most notably OutRage! and the Gay & Lesbian Alliance Against Defamation (Larcher and Robinson 2009). Based on Figure 6.1, the initial response to these developments seems to have been to produce more songs denouncing homosexuality, an act signalling artists’ refusal to ‘bow’ to the West. As noted by Nelson (2008),

“Jamaicans have always insisted on dancing to the beat of their own drum… Rebelliousness is somewhat of a national virtue, especially in defiance to the United Kingdom and America” (Nelson 2008: 250).

172

However, the situation intensified in July 2004, with the formation of the SMMC, which was initiated by Peter Tatchell (founder of OutRage!). This campaign had a drastic impact on artists: songs were pulled from the radio, major sponsors withdrew support, several appearances and concerts were cancelled across the US, Canada and Europe and some artists even had their US and European visas revoked (Larcher and Robinson 2009). The impact of the SMMC was not limited to the output of the Jamaican music industry: there were spillover effects in other areas of the economy. For instance, Jamaica received a considerable amount of criticism from international non-governmental organisations. Even the more commercial TIME Magazine referred to Jamaica as the ‘most homophobic place on earth’ (Padgette 2006). Cowell (2011) noted that this negative publicity indirectly affected the tourism markets and also impacted the image and international stature of Jamaican businesses that sponsored artists. Given the severe economic repercussions that accompanied the SMMC, the subsequent fall-off of in anti- gay music post-2004 is not surprising (see Figure 6.1). In fact, in an effort to avoid further losses from the SMMC, top reggae promoters brokered a deal with the organisers of the SMMC in 2007. As a part of this deal, artists would sign a ‘Reggae Compassionate Act’ pledging to “respect and uphold the rights of all individuals to live without fear of hatred and violence due to their religion, sexual orientation, race, ethnicity or gender” and to avoid “mak[ing] statements or perform songs that incite hatred or violence against anyone from any community” (Outrage 2007). Organisers of the SMMC agreed to stop campaigning against artists that signed and honoured the act (Outrage 2007). Allegedly, the act has been signed by various reggae and dancehall heavyweights such as Beenie Man, Capleton, Buju Banton, Sizzla and Vbyz Kartel (Freemuse 2007; Jamaica Gleaner 2010; Outrage 2007), though some artists have denied signing it (Wright 2007). The continued low production of violent anti-gay music could also be indicative that tolerance of homosexuality is on the rise. Without comparable data on attitudes towards homosexuality over the last two decades, it is not possible to test whether or not public attitudes have changed in line with the decline in the number of songs with anti- gay lyrics. However, as I mentioned in the introduction, Jamaica successfully hosted its first public gay pride event in 2015, which is remarkable progress and suggests that gradual change is imminent. At the same time, though the number of songs with anti-gay rhetoric seems to on the decline, anti-gay songs from the pre-2005 era still play on many local radio stations across the region, the local rum shops, public transportation and in the clubs. This hints that while it is possible that tolerance could be on the rise, even now,

173

many Caribbeans are either tolerant of or indifferent to anti-gay rhetoric. Thus, the music analysed here is still relevant.

6.2 Thematic analysis of the music

Most research on anti-gay prejudice takes the form of attitudinal research and focuses on the determinants of this prejudice (Jowett 2014). However, while these analyses are informative, certain studies suggest that attitudes should not just be understood as the internal schemas held by individuals. They also represent rhetorical stances taken on controversial matters and various debates on the matter (Billig 1991, 1996). Jowett (2014) notes that by analysing arguments against homosexuality (i.e. by taking a rhetorical approach), we are able to assess how and which resources are drawn upon to justify resistance to homosexuality. Against this backdrop, some researchers have opted to evaluate anti-gay prejudice by investigating the rhetorical legitimation of inequality (Clarke 2001; Ellis and Kitzinger 2002; Hull 2001; Jowett 2014). For example, Clarke (2001) used media documents from the UK and US (95 newspaper and magazine articles and 21 talk shows) to analyse arguments used to oppose lesbian and gay adoption. Ellis and Kitzinger (2002), on the other hand, conducted a thematic analysis of the arguments opposing the proposal to equalise the age of consent for opposite-sex couples and sex between men in the British press (Hansard reports and print media) during the 1990s. Moreover, Hull (2001) and Jowett (2014) examined rhetoric against marriage equality in Hawaii and the UK, respectively. Hull (2001) evaluated 233 letters discussing same-sex marriage in two daily newspapers in Hawaii (the Honolulu Advertiser and the Honolulu Star-Bulletin), and Jowett (2014) critically examined arguments against same-sex marriage using 208 pieces from 11 national UK newspapers (tabloid and broadsheet). A review of the above studies revealed that many of the arguments featured in the media discourses of gay rights in the US and UK largely mimic the theoretical and empirical literature on anti-gay prejudice in the West (see Chapter 2). Specifically, common arguments against gay rights identified in the aforementioned thematic analyses primarily served an expressive function and were also based on concerns regarding the perceived threat of gay rights. Homosexuality was cited as immoral, sinful and unnatural (Clarke 2001; Hull 2001; Ellis and Kitzinger, 2002). The analyses also suggest homosexuality was often presented as a threat to the fabric of society. Specifically, the extension of rights to gays and lesbians was presented as harmful to the vulnerable youth

174

(Clarke 2001; Ellis and Kitzinger 2002), as a threat to health (Ellis and Kitzinger 2002) and religious freedom (Jowett 2014). To complement the analysis conducted in Chapters 4 and 5, I analyse the anti- gay rhetoric found in dancehall and reggae. As noted in the introduction, it is expected that the content and form of the arguments against homosexuality in dancehall and reggae will largely reflect the context in which it is expressed. Thus, the thematic analysis undertaken in this chapter serves to provide additional insight on anti-gay prejudice mainly in Jamaica, and, to a lesser extent, in the rest of the Commonwealth Caribbean. The analysis of the music is based on Lacoste’s list. The data are analysed thematically to identify recurrent arguments against homosexuality. I began with a review of the lyrics in order to obtain a general idea of the content, making detailed notes on any justifications of anti-gay prejudice made in each song. Five recurring arguments were identified:

1. Homosexuality is a violation of gender norms. 2. Homosexuality is a sin. 3. Homosexuality is unnatural. 4. Homosexuality is a threat to the fabric of society. 5. Homosexuality is a foreign concept, and intolerance is part of Jamaican identity.

The first four themes identified in the music are largely in line with some the theoretical and empirical literature on anti-gay prejudice in the West (see Chapter 2) that suggests that (1) anti-gay prejudice serves a value-expressive function: it is a way of affirming religious values and gender role ideologies; (2) beliefs about the (un)naturalness of homosexuality act as a key determinant of prejudice; (3) the perceived threat of lesbians and gay men causes persons to dislike and condemn them. Moreover, many of the above arguments match with the rhetorical tropes used to resist gay rights in the US (Clarke 2001; Hull 2001) and the UK (Clarke 2001; Ellis and Kitzinger 2002; Jowett 2014) media. The fifth theme (homosexuality as alien) has not been identified as a key argument featured in recent Western debates about lesbians and gay men, but such sentiments are often present in public debates about homosexuality in African contexts. Ireland (2013) argues that in Africa, resistance to the West has taken on an anti-gay cast. As Ireland notes,

175

“In parts of Africa, it has been a short step from such judgments for homosexuality to become perceived as the province of corrupt, exploitative political elites in league with the alien and the modern, an ‘un-African’ import or imposition” (Ireland 2013: 54). A few extracts for each argument were collated and are presented in the analysis that follows.

6.2.1 Homosexuality is a violation of gender norms In Chapter 2, I noted that negative attitudes towards homosexuality often serve value-expressive functions, one of which is the promotion of traditional gender role ideologies. In line with this theory, anti-homosexual discourse in music is often couched in the language of gender role ideology (Brown 1999; Hope 2010). Gay men are broadly characterised as ‘feminine’ and lesbians are classed as ‘masculine’ and, as such, are accused of being violators of gender norms. An example of this is Harry Toddler’s dancehall hit, Badman nuh dress like girl (Harry Toddler 1998):

Lyrics Translation “Badman nuh dress like girl Real men don’t dress like girls We nuh bore nose an wi nuh bleach We don’t pierce our noses and we face an don’t bleach our faces We nuh wear drop curl We don’t curl our hair Some freaky bwoy fi stop dress like Gay men need to stop dressing like girl” girls Note: Lyrics taken from http://www.soulrebels.org/. Translation by author.

In this song, Toddler expresses his dislike for men who engage in stereotypical female practices, such as skin bleaching, curling their hair and wearing nose rings, and points out that these are not the ideal behaviours of real men. In a similar fashion, Assassin’s Dem a sissy (Assassin 2007) highlights some of the stereotypical (feminine) characteristics he associates with gay males and assumes that a person displaying such characteristics must be gay:

176

Lyrics Translation “Wah sissy do? What do gay men do? Weh all dem good fah fi chat like dem Well, all they are good for is talking a gal like a girl Go so gwaan, and a carry on, act like Get on, carry on and act like a girl dem a gal They have tattoos on (lower) back Dem have tattoo inna back like dem a like a girl gal You hear they wear dresses like girls A so on yuh hear dem a wear, frack like dem a gal” Note: Lyrics taken from http://genius.com.Translation by author.

Meanwhile, in the song Man Royal (Beenie Man 1996a), Beenie Man generalises lesbians as (unnaturally) masculine, referring to them as Man Royals and labelling their attributes as undesirable:

Lyrics Translation ‘Mi nuh wha nuh Man Royal gal, gal I don’t want any ‘butch’ Gimmi di gal dem wid di real born Give me the girls who act like girls serial …. …. Some girls are walking around with Some gal weh a walk wid all dem bald heads head bawl off As strong as Samson from lifting Strong like a Samson, like a weight weights dem pop of” Note: Lyrics taken from http://genius.com. Translation by author.

An interesting observation is that while lesbians are not immune to lyrical rebukes, the number of songs that focus on female homosexuality pale in comparison to the volume of music that speaks out against gay men (Farquharson 2005). Using the list provided by Lacoste, I investigated the songs whose titles made reference to homosexuals. Of the 45 songs, only two can be identified as making a reference to a lesbian in their titles. The focus on gay men could be due to the fact that dancehall is a male dominated genre. Indeed, the empirical analysis carried out in Chapter 4 suggests that men in Barbados, Guyana and Trinidad and Tobago tend to condemn male homosexuality more harshly than they do female homosexuality. Work by Donna Hope (Hope 2006; Hope 2010) suggests that this is also the case in Jamaica. She further points out that the focus on the homosexual male character in dancehall music is generally a reflection of an underlying narrative in Jamaican society of male superiority and female inferiority. Hope

177

(2006) notes that female homosexuality is incapable of undermining ‘the traditional tenets of Jamaican patriarchy’. For instance, according to Allyn (2012), the male homosexual threatens a key aspect of the ‘true’ Jamaican male identity – which is to dominate and conquer the female. Farquharson (2002) notes that:

“…it is seen as morally wrong, if not politically wrong, for a man to allow himself to be possessed and dominated by another man” (Farquharson 2002:3- 4).

In this vein, we find that the anti-gay lyrics can be largely characterised as a male- to-male discourse, featuring sentiments of disgust or uneasiness with deviations from ‘acceptable’ male behaviours, as showcased previously in the lyrics of artists Harry Toddler and Assassin. Pinnock (2007) opines that in dancehall, male homosexuality as a gender issue is significantly more important than the mere question of the type of sex in which a man engages. The author further contends that in dancehall, the homosexual male is neither male nor female but rather the embodiment of an ‘other’. Along with serving the value-expressive function, the attack on gay men by male artists could simultaneously be serving the social-expressive and defensive functions of sexual prejudice (see Chapter 2). Specifically, given the hostility directed towards gay men in Jamaica and the wider Commonwealth Caribbean, it is plausible that the artist could be using loud and sometimes violent condemnations of men who have sex with men as a means of appealing to his core audience as well as establishing his status as part of the ‘in-group’. The latter can be somewhat seen in Capleton’s song Hang Dem Up (Capleton 1999a) where he sings “Bun batty man only batty man vex, Bun sadamite only sadamite get vex”, which translates to the following: the only persons who have issues with condemnations of gay men and lesbians are gay men and lesbians.

6.2.2 Homosexuality is a sin As discussed in Chapter 3, Christianity is the dominant source of values throughout the Commonwealth Caribbean. In these states, religion infiltrates all aspects of life, even secular music. In addition, a number of Jamaican artists also identify as ‘Rastafarian’, and, as a result, Rastafari themes, imagery and symbolism are often presented in Jamaican music. Some authors even argue that in the early 1970s and 1980s, reggae music served as the key medium through which non-Jamaicans learned about Rastafarianism (Spiker

178

1998) to the extent that many individuals incorrectly accord reggae and Rasta culture as synonyms (Savishinsky 1994). As a side note, Rastafarianism is a religion indigenous to Jamaica and is deeply rooted in the Pan-Africanist teachings of Marcus Garvey. The basic doctrines of Rastafari, as noted by Simpson (1955) are (1) there is one true God (often called Jah) who is black; (2) Christ is black, and Halie Selassie was the second advent of Christ; (3) black people are the true Israelites (God’s chosen people). Rastafarians also draw heavily on the Bible, mainly the Old Testament and the Book of Revelation, when it comes to their beliefs (Veal 2007). Many Rastafarians cite homosexuality as a ‘Babylonian’ sin40 and, like Conservative Christian Protestants, are quick to cite biblical passages such as Genesis 19:5, Leviticus 18:22 and 20:13 and Revelations 21:8 and 22:15 as evidence of Jah’s disgust with homosexuality. Against this backdrop, it is not surprising that the anti-gay content of Jamaican music is often wrapped in biblical cloth. For instance, a popular Old Testament narrative found in the music comes from Genesis 1:28, where God commanded Adam and Eve to ‘be fruitful and multiply and fill the earth’. Some artists use this verse to define sexual relations within the ambit of procreation, and since homosexual sex does not fit into this paradigm, they argue that it goes against God’s divine law. The song Keep it to yourself by Queen Ifrica (Queen Ifrica 2009) exemplifies this view:

Lyrics Translation “Yu fi multiply an replenish di Earth You are to multiply and replenish the An dats why di woman labor inna child earth birth And that is why women labour in child Mi nuh want si mi brother dress up birth inna no skirt” I don’t want to see my brother dressed in a skirt Note: Lyrics taken from http://www.lyricsmode.com. Translation by author.

Another popular biblical story used to condemn homosexuality is the story of Sodom and Gomorrah from Genesis 19. According to the text, these ancient cities were completely consumed by fire and brimstone because of their citizens ‘grievous sins’, among which (many theorise) was that of homosexuality. Hence, in their lyrical treatise against homosexuality, some artists make reference to a divine accountability thesis, often

40 Rastafarians associate the term ‘Babylon’ with oppression, materialism and sensual pleasures; the term often has negative connotations. 179

implying that, like these two cities, God will punish Jamaica if a stance against homosexuality is not taken. A striking example of this is Spragga Benz’s Funny Guy Thing (Spragga Benz 1996):

Lyrics Translation “Wi nah join nuh sodomatical system We will not join a sodomite system Ah brimstone an fire dat ah go bring Brimstone and fire it will bring bring … …. God is going to dismantle their clique Jah Jah ah go dismantle all ah dem And leave them in misery and grief clique Am lef dem in misery and grief” Note: Lyrics taken from http://genius.com. Translation by author.

In line with the logic that these cities were burnt because of the homosexual conduct that had transpired within them, many artists see it fit to ‘Bun Faiya’ on homosexuals [translates to: Burn homosexuals with fire]. While these phrases can sometimes be used as a threat of physical violence, in most instances, the ‘faiya bun’ notions should not be taken literally. As noted by Farquharson (2005), many artists use the term to refer to a spiritual or cleansing fire, possibly referring to God’s judgement or the atonement for sin – that is, burning in hell. Simply put, it is used simply an admonishment on homosexuality, as explained by Capleton (Capleton 2000) in the song More Prophet:

Lyrics Translation “Fire y'all fi use fi bun out confusion You have to use the fire to burn out Bun disillusion and bun temptation confusion Mek dem know di fire is fi di To burn disillusion and temptation purification” Let them know the fire is for purification Note: Lyrics taken from http://genius.com. Translation by author.

With the music so closely intertwined with religious identities, in many cases homonegative statements can be argued to go beyond hate speech and act as the enforcement of social rules with religious authority: it serves the (religious) value expressive function (see Chapter 2). The violent antigay lyrics could be a bi-product of fundamentalist beliefs about homosexuality. Gutzmore (2004) notes that to Christian fundamentalists, homosexuality is a highly grievous sin that the Bible has declared ‘an abomination’ (Ezekiel 16:50), a ‘vile infection’ (Romans 1:26-27), ‘unseemly’ (Roman

180

1:27), ‘unnatural’ (Romans 1:26-27) and a form of ungodliness (Romans 1:28). In Leviticus, the Bible calls for those who engage in male homosexuality to be put to death. Gutzmore notes that since the Bible advocates terminal violence for homosexuals, it is possible that artists may feel somewhat justified in ‘metaphorically’ killing gays. Take for example, the preamble to the song Dat A true by Bounty Killer (Bounty Killer 2004):

Lyrics “What happened to them faggot?! Maggot He that lieth with a next man shall be put to death Quote it from the Bible biblically It's Warlord, I'm killing them spiritually” Note: Lyrics as heard by author

Similar sentiments are offered by Sizzla in Nah Apologise (Sizzla 2005), in which he sings, “Inna de Biblical days, we use to stone dem to death”. 6.2.3 Homosexuality is unnatural The unnaturalness theme is intimately related to the ‘homosexuality is s sin’ theme. Anthony B declares the following in Burn Them (Anthony B 2008):

Lyrics Translation “Fire weh we burn pon corruption We burn corruption with fire Rasta nah bow to di man Rasta will not bow to ‘the man’ Push up onna hand Put up your hands Meh me see you hand Let me see your hands All who promote a woman and a All who promote the union of a woman man” and man …. ….. “It tek a king and a queen It takes a king and a queen Fi show weh Jah love mean” To show what God’s love means … … “Rasta seh Big up mum and dad Rasta says hail [or respect] mum and dad A so we create by Gad” This is how we were created by God Note: Lyrics as heard by author. Translation by author

Here, Anthony B notes that God intended for a man to be paired with a woman, and anything outside of this dynamic is inappropriate. Artists sometimes use variations of the Christian Fundamentalist rhyme ‘God made Adam and Eve, not Adam and Steve’, as evident in Capleton’s song Buggering (Capleton 1993):

181

Lyrics Translation “True god mek other man mek him fi The true God made man for Kim Kim He did not make Man for Jim Never mek other man mek him fi Jim” Note: Lyrics taken from https://sonichits.com. Translation by author.

An interesting observation is that many artists who cite religious morality as the reason for their prejudice against homosexuals are much more flexible with ‘straight’ sexual behaviours that are also condemned by the Bible. Looking at the case of Beenie Man, in his song Bomb and Dynamite (Beenie Man 1994), Beenie presents homosexuality as wrong in the eyes of God and makes several references to Sodom and Gomorrah. Yet, his hit song Nuff Gyal (Beenie Man 1996b) is essentially an ode to promiscuity and fornication, which the Bible also recognises as sins. In Chapter 2, I noted that in there is some selection in terms of which sexual sins are condemned. Some Caribbean academics have linked this selectivity to the Africanisation of Christianity, where slave interpretations of sexual morality were not focused on fornication or adultery but were confined to forms of ‘unnatural’ sex – anal sex, oral sex and homosexual sex (LaFont 2009, 2001). This is prominently displayed in Jamaican music. For instance, in Bun Down Sodom (I Wayne 2011), I-Wayne links any sort of sex outside of vaginal sex as against nature and pagan practices:

Lyrics Translation “Dem a gwaan like dem straight They are acting straight down inna de pit dem sick, but they engage in anal sex, they are Check it out all type a ting dem a kiss sick Nastiness, pagan practice All types of things they kiss …. It is nastiness, a pagan practice fight against nature you have to get … bun tru” When you fight against nature you have to get burned Note: Lyrics as heard by author. Translation by author

In a nutshell, straight sex is showcased as respectable and natural, while other forms of sex are regarded as deplorable. In fact, denouncements of homosexual sex are sometimes accompanied by condemnations of anal and oral sex, even if between heterosexual couples, as suggested by Capleton in In or Out (Capleton 1999b):

182

Lyrics Translation “Sadom and Gomaro, bere batty man Sodom and Gomorrah, a lot of gay Sadom and Gomaro, bere lesbian men Sadom and Gomaro, man ah climb Sodom and Gomorrah, a lot of pon man lesbians Sadom and Gomaro, woman ah climb Sodom and Gomorrah, men have sex pon woman with men Sadom and Gomaro, woman ah yam Sodom and Gomorrah, women have man sex with women Sadom and Gomaro, man ah yam Sodom and Gomorrah, women woman” engage in fellatio Sodom and Gomorrah, men engage in cunnilingus Note: Lyrics taken from http://www.lyricsofsong.com/. Translation by author.

6.2.4 Homosexuality is a threat to the fabric of society Another narrative found in the lyrics is that homosexuality is a ‘threat’ to the fabric of society. For instance, there are concerns about public health, as some artists associate homosexuality with diseases. For instance, in A Nuh Fi Wi Fault (Elephant Man 1998), Elephant Man blames homosexuals for the spread of syphilis in Jamaica. Additionally, in Bomb and Dynamite (Beenie Man 1994), Beenie Man links homosexuality to the AIDS epidemic. There are also concerns that homosexuals are a threat to young people, such as that expressed in Lovindeer’s song Don’t bend down41 (Lovindeer 1987):

Lyrics “Him promise you jewellery, don't bend down Or a bag of money, don't bend down Fancy clothes, don't bend down” ….. “There are some men round the place today Tryin to lead the youth astray But I come to set them on the right track We deal up front and don't back back” Note: Lyrics taken from http://www.lyricsmode.com

Lovindeer is accusing gay men of trying to bribe vulnerable youths into having sex with them, thus linking homosexuality and paedophilia. With the perception that homosexuality presents a threat to Jamaica, it is not surprising that many artists condemn it. As discussed in the literature review, the greater

41 “Don’t bend down” is similar to the popular phrase “Don’t drop the soap”. 183

the extent to which an in-group (in this case, heterosexuals) views the out-group (homosexuals) as a threat (symbolic or realistic), the more negative the attitudes are likely to be. In fact, in an article in the Jamaican Observer, Beenie was quoted as saying the following:

Gay in Jamaica is not like it is in America. It's mostly big men with money going down in the ghetto and turning the local youths so you call that statutory [rape] or child molestation. They convince the youth that they are this way and me know enough youth this way. That's why when it comes to gay murder in Jamaica, it's so vicious. (Livingston 2012)

This raises an interesting question: Are the views about homosexuals voiced in the songs limited to perceptions about Jamaican gays? This is difficult to answer, but most of the lyrics do not specify that they are only valid in the Jamaican context; the references to ‘universals’, like religion and nature, would suggest otherwise.

6.2.5 Homosexuality as a foreign concept and intolerance is part of Jamaican identity Some academic literature has suggested that Jamaicans cling to the idea that homosexuality is a form of colonial contamination (Stychin 1998) and believe it is mainly an imitation of European culture (Farquharson 2002). In fact, LaFont (2001) made the following statement:

“Anti-sodomism is championed as evidence of Jamaica's moral superiority over Western liberal sexual mores. Tolerance of sodomy is seen not only morally reprehensible but also as un-Jamaican -- tarnishing the national image. Anti-sodomism, on the other hand, is regarded as a virtue that Jamaicans willingly share” (Lafont 2001).

Homosexuality is thus situated as a foreign concept that will not be tolerated in Jamaica. For instance, in the song Nah Gwan A Jamaica (Elephant Man 2003) Elephant Man sings:

Lyrics Translation “Certain things wah gwan a foreign, Certain thing that happen abroad Can't gwan a Jamaica Can’t happen in Jamaica Nah support no chi chi” We don’t support gays Note: Lyrics taken from http://www.azlyrics.com. Translation by author.

And, in Mauma Man (Shabba Ranks 1989), Shabba Ranks sings:

184

Lyrics Translation “Batty business can't gwaan inna Gay business (possibly gay rights) can’t Jamdown happen in Jamaica Else man from Jamaica woulda legalize Else guns would have to be legalised in gun Jamaica To kill battyboy would be the greatest Killing gay men would be fun fun” Note: Lyrics as heard by author. Translation by author

Under the premise that homosexuality is ‘alien’, some artists proudly assert their anti-gay stance, which is condemned by a more powerful out-group (the West). Anti-gay prejudice is presented as an expression of their in-group identity and culture and as a defiance of ‘outsider’ social values. For instance, in the song Dat A true (Bounty Killer 2004), Bounty Killer expresses these views:

Lyrics “It's my culture, it's my nature, it's my experiences, It's nothing I'm fighting against That's how I see it That's how I live it That's how I know it, That's how I tell it That's how I show it” Note: Lyrics as heard by author.

And, in Nuh Like (Elephant Man 1999), Elephant Man sings:

Lyrics Translation “We no like gay, we no like gay, We don’t like gays, we don’t like gays Well ah just how Jamaican stay” This is just how Jamaicans are Note: Lyrics taken from http://www.soulrebels.org. Translation by author.

6.2.6 Combinations of themes The chapter thus far has highlighted the main themes present in the music. An interesting feature of the music is that some songs use more than one argument to condemn homosexuality. An exemplar example of this can be found in the song JA don’t like gay by Dr Evil (Dr Evil, 2005):

185

Lyrics “My name is Dr. Evil, I’m trying to find a reason Why, should a next man lie, with a guy who's name is Steven I love pum pum, fat cu cum cum [I love vaginas] I will never, like Trevor I’d prefer wifeing a heifer [I would prefer to marry a heifer] I bought dis A.K. to spray on all gays [Bought gun to shoot all gays] … J.A. [Jamaica] don’t like gay no, J.A. don’t like gay no … You’re into child molesting, gun shots will tear your intestine. Gay life we show no interest in” Note: Lyrics as heard by author. Translations by author

Three themes can be identified in Dr Evils song. The first relates the theme that homosexuality is unnatural or inferior form of sexuality, as the artist points out his disgust for sex between men, and heralds his love for women. He then proceeds to promote the idea that dislike for gays is part of Jamaican identity. Finally, he ends by stating that homosexuals are child molesters, which falls under the ‘homosexuality is a threat to society’ theme. Similarly, in Funny Guy Thing (Spragga Benz 1996), Spragga Benz not only makes references to Sodom and Gomorrah (as alluded to previously in section 6.2.2) but the artist also cites homosexuality as unnatural and implies that gay men ‘go after school boys’.

6.3 Discussion

Homosexuality is largely framed in a negative light in reggae and dancehall music. This chapter examined the ways in which gay men and lesbians are stigmatised in popular music. The condemnation of homosexuality was presented within the ambits of (1) gender role ideologies; (2) religious values; (3) its ‘unnaturalness’; (4) the ‘threat of homosexuality’ and (5) homosexuality being an alien lifestyle and un-Jamaican. In some instances, the arguments were often combined, leading to a complex web of anti-gay rhetoric. The study also sought to determine the extent to which the arguments identified here are similar to those presented to justify anti-gay prejudice in the West. Four of the themes featured in the music have been cited as key arguments against homosexuality in the West (see Chapter 2). Such arguments have also been prominently featured in media

186

discourses about gay rights in the US (Clarke 2001; Hull 2001) and the UK (Clarke 2001; Ellis and Kitzinger 2002; Jowett 2014). The fifth theme, while not prominently featured in recent Western debates about lesbians and gay men, are often used to justify anti-gay prejudice in many African countries (Ireland 2013). A key benefit from the thematic analysis is that we got a general sense of sexual morality in Jamaica. Negative attitudes towards homosexuality seem to be less related to general sexual conservatism. Specifically, promiscuity with women was presented as an acceptable male behaviour, but male homosexuality was not. Given that arguments against homosexuality in the music often echo the sentiment of the church, one would assume persons would condemn all ‘sexual impurities’. As discussed in Chapter 2, this mismatch in attitudes has been tied to definitions of masculinity as well as slave interpretations of Christian ideologies. Based on the lyrics, it is clear that views about the ‘naturalness’ of heterosexuality play a role of skewing views on sexual sins. We also see references to the Biblical command ‘to be fruitful and multiply’. Since same-sex acts do not lead to procreation, homosexuality may be viewed a direct violation of this command, whereas promiscuity with women is not. In addition, the music makes references to the Biblical story of Sodom and Gomorrah as evidence of God’s condemnation of homosexuality. Arguably, some persons may see this as evidence that God condemns homosexuality more than other sexual impurities. A unique finding, at least relative to Western contexts, is that there is no pluralism of expression. For instance, while we often see anti-gay rhetoric presented in Western media, there are probably just as many pro-gay messages coming from the liberal left (Pan et al. 2010). Within the context of music, many pop artists in the West have not only come out in support of equality but have also released pro-gay music. To my knowledge, there is only one reggae artist with pro-gay messages in his music – Mista Majah P. In 2011, Mista Majah P released the first pro-gay reggae album titled Tolerance. He offered this explanation of his intentions:

My music is about tolerance. It shows that reggae music can respect gay and lesbian people. Reggae music used to be about love, peace and unity. Now it is too often about bigotry and violence. I want to bring the music back to its progressive roots. (Tatchell 2015)

It should be noted that Mista Majah P was born in Jamaica but migrated to Canada at a very early age. Arguably, his pro-gay stance could be due to the fact that he was socialised in a society in which people’s stance on homosexuality is much more liberal than it is in

187

Jamaica. Given the reports on the prevalence of hate crimes against LGBT individuals, it seems reasonable to assume that the lyrics of artists socialised in this context go beyond being merely metaphorical: to a large extent, art is imitating life. In fact, the argument that ‘anti-gay sentiment is part of Jamaican identity’ points to how anti-gay sentiment is viewed as part of a cultural package: it is an important element in the construction and maintenance of Jamaican identity. Thus, in the music, only the dominant narrative (i.e. anti-homosexuality) prevails. While the analysis suggests that dancehall and reggae largely echo the sentiments of the church and the laws, is it possible to go a step further and claim that the anti-gay music as a major contributor to anti-gay attitudes in Jamaica and to consumers of the music in other Caribbean states? In the introduction, I noted that researchers often argue that mass media can influence citizens’ learning of issues, as well as the formation or change of their views (Bryant and Zillman 2009). Activists have historically argued that media portrayals of gay men and lesbians have significant impacts on attitudes towards homosexuals (Levina et al. 2000), and there is research in agreement with this argument. For example, experimenting with a national sample of Americans, Jang and Lee (2014) tested whether or not Lady Gaga’s hit song Born This Way (Lady Gaga 2010) primed genetic explanations of sexual orientation. Jang and Lee found that participants who listened to Born This Way placed more emphasis on genetic causation when evaluating gay rights than those who listened to the instrumental version of the song, or those who were not exposed to any stimuli. Alternatively, Levina et al. (2000) focused on the effects of visual media on undergraduate students in Illinois. One group was exposed to an anti- gay video, the second group watched a pro-gay video and the third group was exposed to a neutral video. Attitudes were significantly different, with the attitudes of those who watched the pro-gay video being the most positive, while the views of those who watched the anti-gay video were the most negative. Against this backdrop, it is not surprising that reggae and dancehall – while not the origin of Jamaican homonegativity – have been accused of being a chief medium for the continued ill-treatment of homosexuals in Jamaica (Greenberg 2010; Nelson 2008). In fact, advocates used the perceived link between homophobic violence and dancehall as one of the main reasons to ban dancehall records and Jamaican artists during the SMMC (Stanislas 2014). The most compelling evidence of a link between anti-gay attitudes and the music comes from the work of West and Cowell (2015). These authors found that Jamaicans whose main choice of music was dancehall displayed more negative

188

attitudes, social distance, opposition to rights and negative behaviours towards homosexuals (see Chapter 2). However, since the research was correlational, there is the issue of reverse causality. Arguably, citizens may already possess the necessary tools to construct their own views on homosexuality and may have opted to listen to music that reflects their stance. But, among public opinion scholars, a popular view is that while citizen rely on their own toolkits of moral reasoning to inform their stance on various social and political issues, they also rely on frames borrowed from the media (Potter 2013). To recall, experimental research to date has shown that music, like other forms of media, has the ability to influence public opinions on homosexuality – especially if one constantly listens to the music or if there is a lack of competing frames (in the local media) from which to reference. In this way, the music can be viewed as a cultural bridge between the institutionalised prejudice (from religion and the laws) and anti-gay attitudes at the individual level. The anti-gay stance taken by some artists could make some persons believe that such sentiments make them part of an ‘elite’ group and that their attitude or even aggression towards homosexuals is justified. This in turn could undermine social inhibitions about negative behaviours towards homosexuals. At this juncture, I should point out that while the local media may not offer many messages that challenge the status quo (see Chapter 3), most Caribbean states are largely exposed to Western media, particularly US television, through local cable or satellite television. As is well documented, gay and lesbian characters have been more visible on US television shows in recent years, and they are often portrayed in a positive light (Bond and Compton 2015). While I did not come across any information or studies that would give an indication of the extent to which residents in the Caribbean tune into shows with a pro-gay stance, it seems plausible to assume that a significant subset of the population would at some time be exposed to pro-gay messages via cable or satellite television. It is thus possible that imported media, particularly shows with pro-gay messages, could aid in softening attitudes in the region. At the same time, the messages from the US could be less influential than reggae or dancehall, as they may be seen as ‘foreign’ values and not applicable to Caribbeans. However, this does not completely nullify the impact, and it is quite possible that increased exposure to alternative messages about homosexuality can prime audiences to challenge or at least query the status quo.

189

Chapter 7 : Summary and concluding remarks

Currently, one of the most contested global issues concerns the rights of homosexuals and same-sex couples. In Western countries, the fight for the legality of homosexuality has largely been won, and current discourse of gay rights focus on the marital and adoption rights of same-sex couples. However, the advancement of gay rights in the developing world significantly lags that of the West. Many countries across the globe continue to criminalise consensual sexual acts between persons of the same sex, among which are 11 of the 12 states in the Commonwealth Caribbean. As at July 2016, The Bahamas currently is the only state to have decriminalised same-sex intimacy. However, gays and lesbians are still discriminated against at the institutional level in the Bahamas. Like its neighbours in the Commonwealth Caribbean, same-sex relationships in the Bahamas are not recognised, and there are no laws explicitly prohibiting discrimination on the grounds of sexual orientation in employment, housing, education or health care. Notwithstanding the obvious problems the laws against same-sex intimacy create for LGBT persons living in the region, it should be noted that these laws are currently not used to prosecute persons on the basis of their sexuality. So, if not enforced, why do policy makers opt to maintain the laws, especially in the age where bans on same-sex intimacy are viewed as heinous violations of human rights? The choice to put international relations at risk would suggest the presence of strong domestic politics. Researchers generally agree that public attitudes towards LGBT persons are a key determinant of public policies regulating private intimate acts. Under a theory of legal moralism, if the majority holds a moral value, policy makers may opt to put/keep legislations in place to uphold said value and remain in public favour (Beauchamp 1986; Burstein 1998, 2003). In fact, as shown in Chapter 1, various statements by political elites portray citizens across the Commonwealth Caribbean as largely intolerant of homosexuality and cite public opinion against homosexuality and religious actors as a significant barrier to the repeal of these laws. Yet, while a common view is that these laws are anchored by public support, very few studies have emerged to empirically test whether the attitudes and behaviours of the general population are actually in line with this view. In fact, anecdotal evidence from the region suggests that attitudes towards LGBTI persons may not be as starkly negative as presented. For instance, in Barbados, a self- identified transgender woman often wins the coveted title of ‘Best Flag Person’ in the country’s most popular festival (Chapter 1). Although this may not necessarily mean that 190

attitudes are positive, it does suggest that attitudes are not as stark as presented and that there may be a certain degree of tolerance in some states. Against this backdrop, this study used mixed methods to analyse attitudes towards lesbians and gay men in the Commonwealth Caribbean. The main objectives of the thesis were to determine the level of prejudice in these states using nationally representative data and to examine how individual and societal factors shape these beliefs. The study also evaluates anti-gay rhetoric found in reggae and dancehall – the region’s iconic musical genres – to provide additional insights on how anti-gay prejudice is legitimised in these states. Given the paucity of research on attitudes towards homosexuality, the models were largely informed by the theoretical and empirical literature based on the Global North. This approach is beneficial in that we are able to see the extent to which the findings on the West are applicable to countries with different economic, religious and legal circumstances. In fact, a primary hypothesis in this thesis is that influences from the larger culture may have a more substantial impact on attitudes than individual differences. The primary concern is that the level of development, the anti-gay laws and the strong Evangelical presence in the Commonwealth Caribbean may exert an influence on attitudes that is more powerful than individual characteristics such as age and personal religiosity. In fact, it has been argued that even if some individuals are in the minority and do not hold the same values as others in the society, they are still influenced by the culture at large (Adamczyk and Cheng 2015). This raises questions about the ability of whether the individual level variables found to be significant determinants of anti-gay prejudice in Western contexts also explain anti-gay attitudes in the Commonwealth Caribbean. In this chapter, I summarise the principal findings from the thesis and discusses the related policy implications. I then outline the limitations of the research and discuss potential future research.

7.1 Key findings

As mentioned earlier, a commonly held view is that the laws are anchored by public support. However, my analysis of heterosexual support for the bans on same-sex intimacy in Barbados, Guyana and Trinidad and Tobago revealed that residents in these countries hold mixed views about the criminalisation of homosexuality. Specifically, of the persons who supported the maintenance and enforcement of the laws, over 50 per

191

cent did not support the penalisation of men who have sex with men or women who have sex with women if it is in private. The descriptive statistics suggest that while there is support for the anti-gay laws, it is not as ‘black and white’ as politicians seem to believe. The analysis also revealed that a significant number of individuals seem to believe that the anti-gay laws reflect the morals standards of the society (50 per cent), protect society from certain ills (48 per cent) and protect young persons from abuse (41 per cent). A smaller, albeit still meaningful, portion of individuals believes that the current anti-gay laws reduce the spread of homosexuality (22 per cent). Public support for these laws seems grounded in subscriptions to ill-founded beliefs that homosexuality harms society. Having looked at the prevalence for support for the laws, a key question emanated: what are the determinants of public support? Research on the Global North has found that the individual level, variables such as age, education, gender, religiosity, interpersonal contact with gays and lesbians, beliefs about the origins and homosexuality and attitudes towards other minorities are generally reliable predictors of anti-gay prejudice (Mason and Barr 2006; Haider-Markel and Joslyn 2008; Smith et al. 2009; van den Akker et al. 2013). The quantitative enquiry on the factors influencing support for the anti-gay laws suggests that some of the findings from the research done in Western contexts are applicable to these states, while others are not. The results show that only personal religiosity, interpersonal contact, and beliefs about the origins of homosexuality were significantly related to each of the measures of support for the anti-gay laws. The research suggests that interaction with persons who identify as a lesbian or a gay man encourages tolerance of homosexuality and that those persons who believed that homosexuality was innate reported less support for the anti- gay laws. Greater support for the retention of the anti-gay laws was also found among persons who were actively involved in a religion and whose views on human sexuality were theologically based. Although religious persons tend to be more supportive of the laws, there was no evidence that support varies across religious denominations, which marked some deviation from the current literature. Also surprising was the fact that age and education were not among the strongest predictors of attitudes (that is statistically significant in all the models), as the literature on the West often stresses the importance of these variables in predicting attitudes towards homosexuality. On the one hand, the limited impact of age, religion and education on attitudes towards gay rights could be due to the fact that the issue of gay rights in the

192

Commonwealth Caribbean is still in its infancy stages. Currently, there is a lack of competitive frames on the issue, at least from the main institutions in this society (i.e. laws, religion and media – see Chapter 3). As the issue of gay rights evolves and awareness and activism increases, it is possible that we may see the demographic differences that are similar to those currently being observed in Western societies. I also hypothesised that macro-level factors – such as religious culture, laws and level of socioeconomic development – could be exerting an influence on attitudes that is stronger than that of these personal characteristics. Specifically, in line with the classical research of Inglehart (1971), it was hypothesised that generational differences would be larger in countries with higher levels of development. Also, based on the work of Norris and Inglehart (2011) and Adamcyzk and Pitt (2009), I hypothesised that even if residents are not personally religious, living in a country with a high proportion of the conservative religious majority may conform to the ‘anti-gay’ norm. Finally, while we expect that laws will change with societal needs, laws are not limited to the mechanical application of rules or facts: they are part of a larger socio-political process. Laws help shape norms by prefiguring values, prejudices and interests (Gordon, 1984). However, progressive laws on homosexuality may not only increase tolerance but may also exaggerate the gap between various groups. This is because the message of the laws is likely to be internalised by persons whose characteristics would otherwise predict that they would generally be tolerant of homosexuality (Kreitzer et al. 2014) – most likely, younger persons and persons from less conservative religions. Hence, I hypothesised that generational and religious differences were likely to be more prominent in countries with less restrictive regulations on homosexuality. To test these hypotheses, I carried out a cross-national analysis of attitudes towards same-sex marriage in 28 countries in the Americas, 6 of which were members of the Commonwealth Caribbean. In general, countries with higher levels of development, smaller shares of Evangelicals and more liberal laws on homosexuality were more approving of same-sex marriage. Consistent with a-priori expectations, it was found that context matters. Generational differences were larger in the more developed countries as well as in countries with more liberal laws on homosexuality. The results also suggest that religious differences were also more substantial in countries with smaller proportions of Evangelicals and countries with more progressive laws. Together, these findings suggest that the cultural factors may be more important than individual-level variables such as age and religious identity. They imply that there are boundaries to which age and

193

religious identity can predict attitudes towards homosexuality. Thus, within the Commonwealth Caribbean, we can expect individuals to behave in line with normative expectations, regardless of their personal characteristics, especially when the behaviour is widely sanctioned. While quantitative analyses provided a wealth of information about the factors influencing attitudes towards homosexuality in the Caribbean, it is very likely that my findings are just the tip of the iceberg. To complement these analyses, I conducted a thematic analysis of the anti-gay rhetoric found in reggae and dancehall music. Music, like other forms of media, is influenced by the societal culture. By analysing the music, we gain additional insights on the lenses through which Caribbeans view same-sex relations and how anti-gay prejudice may be normalised. The thematic analysis revealed that homosexuality is viewed as sinful and as a violation of gender norms. Homosexuality was also presented as ‘unnatural’, as a threat towards the fabric of society and as a foreign lifestyle. The manner in which the arguments were shaped was indeed reflective of the context from which they occur. For example, the use of religious morality as justifications for anti-homosexual sentiment is somewhat expected from a society with a conservative religious majority. However, situating homosexuality as a foreign lifestyle is a ‘new’ insight about the drivers of anti-gay prejudice – one that could not be inferred from quantitative analyses. By portraying homosexuality as ‘un-Jamaican’, the rejection of homosexuality is not only about sexual identity; rather, it is also a reflection of Jamaica’s attempts to showcase its sovereignty and reject ‘intrusions’ from the West. To recall, much of the push for equal rights across sexual orientations or to limit anti-gay rhetoric in the music came from the West. It is thus possible that feelings about intrusions from the Global North and anxiety about neo-imperialism could be another factor influencing attitudes towards homosexuality in these states.

7.2 Policy implications

The findings of the study could be useful for developing efficacious policies that could aid in liberalising attitudes towards homosexuality. First, the finding that residents of Barbados, Guyana and Trinidad and Tobago generally support the maintenance of the laws but oppose the government regulation of private sexual acts should be of key interest to gay rights advocates. It implies that residents could be viewing intimate same-sex acts through two lenses – one that is based on their beliefs about or ‘discomfort with’

194

homosexuals and the other based on views about the right to privacy. Thus, when entering into conscience debates about these laws, advocates would be better able to garner support for their removal if they base their arguments on the right of the government to regulate private life. As noted in Chapter 3, it seems that concerns about the right to privacy overshadowed concerns about morality in the Bahamas. It is possible that this could also work in other Caribbean states where residents hold conflicting views about the criminalisation of homosexuality. However, presenting the argument about homosexuality in terms of privacy may not be sufficient to get a majority of persons to support changing the laws. The analysis of the law support items also revealed that many persons are in favour of keeping the laws in place as means of protecting the fabric of Caribbean society. It is possible that ill- founded fears about what law reform could lead to (e.g. deterioration of public health or an increase in incidents of child abuse) could reduce the willingness to extend basic rights to gays and lesbians. In fact, even though some persons may view the regulation of sex between two persons in private as a gross intrusion of privacy, it may be considered a necessary ‘evil’ to protect the fabric of Caribbean society. This possibility thus suggests that activists must work on combating ill-founded beliefs about homosexuality. In light of accurate information about gays and lesbians, persons may no longer perceive them as a threat and may change their attitudes towards homosexuality. Research has shown that homonegativity awareness workshops, panels and campaigns can aid in destigmatising homosexuality (Rye and Meaney 2009). Another avenue for destigmatising homosexuality could come from interpersonal contact. To recap, the quantitative analysis suggested that greater interaction with gay men and lesbians was negatively correlated with support for the anti-gay laws and their use as protectionist measures. However, in the context of the Commonwealth Caribbean where there are no laws preventing discrimination on the basis of sexuality, homosexuals may opt to remain in the closet for fear of condemnation. Thus, to encourage visibility, it could be that legislative efforts to curtail anti-gay bullying would need to be put in place to allow social contact to increase. At the same time, given the history of religious groups in the Caribbean thwarting efforts to reduce discrimination against homosexuality, it seems as though the onus would lie on gays and lesbians and organisations to try to form as many emotional bonds with residents as possible. Perhaps the most important finding of the study was that the introduction of more liberal policies went hand in hand with persons showing more support for same-

195

sex marriage. The main policy implication from the analysis is that the evolution of attitudes towards homosexuality in the Commonwealth Caribbean could possibly be ‘sped up’ by decriminalising same-sex intimacy. However, this outcome is not guaranteed. While the values signalled by policy makers are often internalised by citizens (some to a greater degree than others), there could also be a backlash effect. A backlash could occur if it is believed that governments of the Commonwealth are ‘bowing to international pressures’, and not the will of the electorate. The governments of the Commonwealth Caribbean have encountered intense pressure from the West to remove their anti-gay laws (see Chapter 1). Hence, the perception that the laws are being removed due to coercion from the West could bring about major backlash, particularly given the history of colonisation and the fact that Caribbean states are still attempting to establish their sovereignty. As the thematic analysis implies, ideological intrusions from the West are unlikely to be taken lightly. Thus, to limit the backlash, law change must be (or at least appear to be) home grown. Policy makers must frame any pro-gay policies from a national/regional cultural perspective. A potential solution could be a popular campaign that takes genuine national ownership of the decision to change the law. Also, as noted in Chapter 2, backlash is likely to be more prominent among those whose characteristics would generally predict that they would not agree with the values being expressed by the laws, such as the religiously devout or those with more conservative values. In the Commonwealth Caribbean, there is a large presence of Conservative Protestantism (specifically, Evangelicals), whom the literature identifies to be among the least tolerant of homosexuality in the Christian faith (Burdette et al. 2005). Moreover, results from Chapter 4 and 5 suggest that attitudes seem to conform to the dominant religious majority. This means that there is a high probability of backlash effects if laws are removed. But, at the same time, the analysis presented in Chapter 4 suggests that the residents of Barbados, Guyana and Trinidad and Tobago have a ‘symbolic’ support for the law: they want the laws maintained (possibly due to fears about homosexuality) but do not want private sexual activities penalised. Additionally, in Chapters 4 and 6, we see that the anti-gay prejudice is rooted in ill-founded beliefs about homosexuals, which leads to homosexuality being viewed as a threat to society. Focusing the argument for law reform on whether or not it is the right of the government to regulate of private sexual acts, alongside the implementation of various programmes to

196

break ill-founded beliefs about homosexuality, could help in limiting backlashes in these states. However, even with these measures in place, it is still impossible to say whether or not future attitudes towards homosexuality will become more tolerant. Specifically, the finding that age and education have limited impact on attitudes implies that cohort replacement and the recent rise in the number of persons perusing a tertiary education in the Commonwealth Caribbean may not translate to a shift in attitudes in the near future. At the same time, more attention is being paid to discrimination against sexual minorities, as the LGBT movement is gaining ground in the Caribbean. Whether or not Caribbean attitudes stagnate, evolve or stay at their current level will only be realised in the future.

7.3 Limitations and areas for future research

While the study has provided insight on anti-gay prejudice in the region, it is not without its limitations. For example, the CADRES database only included 3 of the 11 Commonwealth Caribbean countries with instated same-sex intimacy. In spite of the similarities across Commonwealth Caribbean states, it would be careless to assume that the findings from Chapter 4 can be generalised to the entire region. While a better representation of the Commonwealth Caribbean is presented Chapter 5 (6 of the 12 states), the attitudes of individuals living in the Eastern Commonwealth Caribbean (Antigua and Barbuda, Dominica, Grenada, St Kitts and Nevis, St Lucia, and St Vincent and the Grenadines) were not explored in this thesis. Also, the gay rights issue addressed in Chapter 5 was that of same-sex marriage. Currently, the particular gay rights focus in the Commonwealth Caribbean is the decriminalisation of same-sex intimacy, and though attitudes towards this issue may be correlated with that of other gay rights, they are not synonymous. An area for future research could be studying attitudes towards the bans on same-sex intimacy in the remaining 8 Commonwealth Caribbean countries, should such data become publicly available. The analysis of the CADRES data was limited by the large number of respondents who selected ‘Unsure/Prefer not to say’ when asked about their views about the anti-gay laws. In general, the lack of definitive responses to the above items could be reflective of a social desirability bias. While participants in the CADRES surveys were given the option to complete the survey themselves, the primary mode of data collection was via face-to-face interviews. It is possible that some participants chose ‘Unsure/Prefer

197

not to say’ in order to avoid giving socially undesirable responses. The lack of definitive responses could also be due to the survey measure used. As noted by LaMarca (2011), the attitudes, beliefs and opinions in a population generally exist on a vast multidimensional continuum. Hence, some respondents may not have been able to give a firm ‘Yes’ or ‘No’ answer to this issue and may have, by default, opted for ‘Unsure/Prefer not to say’. Perhaps a Likert scale (similar to that used in the AmericasBarometer survey) would have been more appropriate. The analysis in Chapter 4 could have also benefited from more standardised measures of the independent variables. A key example would be the measures of religious participation, where individuals were asked to state their religious participation. Surely, definitions of active and passive are likely to differ across persons, and a standard measure attendance at religious services would have been more fitting. Another possible problem concerns the difference in data collection methods for the AmericasBarometer survey. As mentioned in Chapter 5, data for the US and Canada were collected via web surveys, and face-to-face interviews were used to collect data for countries in Latin America and the Caribbean. While the data analysis was based on the weighted data (which reduces some bias), it is possible that there would be some degree of measurement error in response to the questions about same-sex marriage due to differences in modes of data collection. For instance, responses from in Canada and the US could be less likely to suffer from a social desirability bias than responses from Latin America and the Caribbean, where face-to-face interviews were used. It is likely then that the models did suffer from some degree of bias, though it is not certain whether or not this would significantly affect model results. A general limitation of the study was that many of the survey items used in the thesis were not gender-specific, such as the perceived threat variables in Chapter 4 and the approval of same-sex marriage variable used in Chapter 5. Based on the literature (Chapter 2) and the thematic analysis of anti-gay rhetoric in reggae and dancehall, there is much more criticism of male homosexuality than female homosexuality. There could be differences in the perceived threat associated with gay men and that of lesbians as well as differences in support for marriage between women and marriage between men. Future work could explore whether the perceived usefulness of the laws is indeed limited to the perception of gay men or if it encompasses both gay men and lesbians. It could also investigate whether approval of same-sex marriage differs based on the sex of the couple.

198

It is also important to keep in mind that the empirical analyses presented in this thesis all relied on secondary data. The qualitative component was largely based on Lacoste’s list of anti-gay songs, and this list is not likely to be exhaustive. It is quite possible that there are other major themes in the music that were not presented in Chapter 6. With respect to quantitative enquiries, there is no single set of factors that determines attitudes towards homosexuality. While many of the correlates of anti-gay prejudice have been reviewed in this thesis, there are several other variables that the literature identified as important correlates of anti-gay prejudice (e.g. RWA, SDO, political ideology and traditional values) that could not be included due to data unavailability. Also, the thematic analysis revealed that homosexuality is condemned as it is viewed as a Western import – which may stem from the fact that advocates from the Global North head the current push for equality worldwide. Future work could focus on designing surveys that include measures to capture the impact of the other variables identified in the literature as well as measures on feelings towards ideological and behavioural intrusions from the West. Such work could also test whether the findings presented here hold with the introduction of other variables. Finally, this study primarily focused on public attitudes towards homosexuality and the ways in which those attitudes are justified and legitimised. However, while tolerance by the public engenders a climate that would be more receptive to law reform, attitudes by themselves are insufficient to bring about law reform and increase equality. Lax and Philips (2009) find that in addition to public opinion, the state adoption of policies affecting lesbians and gay rights in the US was also influenced by government liberalism and interest groups. To complement the work presented here, future studies should undertake empirical analyses of the attitudes/values of policy makers and the main interest groups (i.e. the religious organisations – see Chapter 3) in the Commonwealth Caribbean.

7.4 Conclusion

Even with the aforementioned limitations, this study makes a significant contribution to the literature. As mentioned throughout this thesis, much of the empirical work to date focuses on attitudes towards homosexuality within the context of Western societies, where the fight for the legality of homosexuality has already been won. Given the mounting global attention towards gay rights, particularly the removal of bans on

199

same-sex sexual relations, there is a pressing need to investigate attitudes towards homosexuality outside of the Western context and to test whether certain claims about the determinants of anti-gay prejudice are universally valid. Our findings indicate that context matters, and in some cases, may be more important than individual level differences. In fact, a major contribution of this thesis is the contextualisation of the impact of age and religious identity in the multilevel model. While a number of studies applying multilevel analyses to the cross-national study of attitudes towards homosexuality and gay rights (see Chapter 2), to the best of my knowledge, no study has focused on how socioeconomic development, religious culture or legal environment could mitigate the impact of age or religion. By including interaction terms in the multilevel model, the thesis provided insights on how context impacts the determinants of anti-gay prejudice. The evidence generated in the research demonstrates that the higher level of anti-gay attitudes in the Commonwealth Caribbean are a consequence of their levels of development, religious context and a lack of elite led change in law making. The main takeaway is that we cannot solely rely on studies based on the West to inform policies aimed at achieving positive social and legal change for lesbians and gays in countries whose contexts differ substantially. Rather, better parallels can be drawn from countries that have similar societal contexts. This brings us to a second strength of the study, which is that it focused on attitudes towards gay rights in the Commonwealth Caribbean – a region often critiqued for its high levels of anti-gay sentiment, but for which limited empirical studies exist. To the best of my knowledge, this is the first study to quantitatively investigate how both individual level factors and the specific features of Caribbean society affect attitudes towards gay rights in the region. The thesis, then, is a step towards narrowing the gap in the aforementioned literature. It is hoped that the results from this study inspire more research on attitudes towards homosexuals and their respective rights in the Commonwealth Caribbean as well as in many other developing countries. The more we know about anti-gay bias, the more likely efficacious programmes and strategies can be developed and implemented to persuade individuals to reject homosexual prejudice.

200

References ______Adamczyk, Amy, and Yen-hsin Alice Cheng. 2015. "Explaining attitudes about homosexuality in Confucian and non-Confucian nations: Is there a ‘cultural’influence?" Social Science Research 51:276-89. Adamczyk, Amy, Chunrye Kim, and Lauren Paradis. 2015. "Investigating Differences in How the News Media Views Homosexuality Across Nations: An Analysis of the United States, South Africa, and Uganda." Sociological Forum 30 (4):1038-58. Adamczyk, Amy, and Cassady Pitt. 2009. "Shaping attitudes about homosexuality: The role of religion and cultural context." Social Science Research 38 (2):338-51. AIDS-Free World. 2010. "The Unnatural Connexion: Creating Social Conflict Through Legal Tools ". Georgetown, Guyana: AIDS-Free World. Alden, Helena L, and Karen F Parker. 2005. "Gender role ideology, homophobia and hate crime: Linking attitudes to macro-level anti-gay and lesbian hate crimes." Deviant Behavior 26 (4):321-43. Alexander, M Jacqui. 1994. "Not just (any) body can be a citizen: The politics of law, sexuality and postcoloniality in Trinidad and Tobago and the Bahamas." Feminist Review 48 (1):5-23. Allport, Gordon Willard. 1954. The nature of prejudice. New York, US: Addison-Wesley. Allyn, Angela R. 2012. "Homophobia in Jamaica: A Study of Cultural Heterosexism in Praxis." Available at http://ssrn.com/abstract=2097180. [cited on June 15, 2015]. Altemeyer, Bob. 1981. Right-wing authoritarianism. Canada: University of Manitoba Press. ———. 2002. "Changes in attitudes toward homosexuals." Journal of Homosexuality 42 (2):63-75. Altemeyer, Bob, and Bruce Hunsberger. 1992. "Authoritarianism, religious fundamentalism, quest, and prejudice." The International Journal for the Psychology of Religion 2 (2):113-33. Andersen, Robert, and Tina Fetner. 2008a. "Cohort differences in tolerance of homosexuality attitudinal change in Canada and the United States, 1981–2000." Public Opinion Quarterly 72 (2):311-30. ———. 2008b. "Economic inequality and intolerance: attitudes toward homosexuality in 35 democracies." American Journal of Political Science 52 (4):942-58. Anderssen, Norman. 2002. "Does contact with lesbians and gays lead to friendlier attitudes? A two year longitudinal study." Journal of Community & Applied Social Psychology 12 (2):124-36. Anthony B. 2008. "Burn Dem." In Culture Uprising Vol. 1: Yellow Moon. Aosved, Allison C, and Patricia J Long. 2006. "Co-occurrence of rape myth acceptance, sexism, racism, homophobia, ageism, classism, and religious intolerance." Sex Roles 55 (7-8):481-92. Assassin. 2007. "Dem A Sissy." In Strictly the Best Vol. 37: VP Records. Lyrics taken from http://genius.com/Agent-sasco-assassin-dem-a-sissy-lyrics [cited May 05, 2015]. Atluri, Tara L. 2001. When the closet is a region: homophobia, heterosexism and nationalism in the Commonwealth Caribbean. Barbados: Centre for Gender and Development Studies, University of the West Indies, Cave Hill. Baboolal, Yvonne. 2012. "PM promises rights for gays in gender policy." The Trinidad and Tobago Guardian, December 18. ———. 2013. "Coudray: Gay rights out of proposed gender policy." The Trinidad & Tobago Guardian, May 18.

201

Badgett, M.V. Lee, Sheila Nezhad, Kees Waaldijk, and Yana van der Meulen Rodgers. 2014. "The relationship between LGBT inclusion and economic development: An analysis of emerging economies." Los Angeles, CA: The Williams Institute & USAID. Badinter, Elisabeth, and Lydia Davis. 1997. XY: On masculine identity. New York, US: Columbia University Press. Barrientos, Jaime. 2016. "The social and legal status of gay men, lesbians and transgender persons, and discrimination against these populations in Latin America." Sexualidad, Salud y Sociedad (Rio de Janeiro) (22):331-54. Barro, Robert J, and Rachel McCleary. 2003. "Religion and economic growth." National Bureau of Economic Research. Barth, Jay, and Janine Parry. 2009. "2> 1+ 1? The Impact of Contact with Gay and Lesbian Couples on Attitudes about Gays/Lesbians and Gay‐Related Policies." Politics & Policy 37 (1):31-50. Baunach, Dawn M, Elisabeth O Burgess, and Courtney S Muse. 2009. "Southern (dis) comfort: Sexual prejudice and contact with gay men and lesbians in the south." Sociological Spectrum 30 (1):30-64. Baunach, Dawn Michelle. 2012. "Changing same-sex marriage attitudes in America from 1988 through 2010." Public Opinion Quarterly 76 (2):364-78. BBC. 2011. Cameron threat to dock some UK aid to anti-gay nations http://www.bbc.co.uk/news/uk-15511081 [cited March 28 2014]. Bean, Lydia, and Brandon C Martinez. 2014. "Evangelical ambivalence toward gays and lesbians." Sociology of Religion 75 (3):395-417. Beatty, Kathleen Murphy, and B Oliver Walter. 1988. "Fundamentalists, evangelicals, and politics." American Politics Research 16 (1):43-59. Beauchamp, Dan E. 1986. "Morality and the health of the body politic." Hastings Center Report 16 (6):30-6. Becker, Amy B. 2012. "Determinants of public support for same-sex marriage: Generational cohorts, social contact, and shifting attitudes." International Journal of Public Opinion Research 24 (4):524-33. Becker, Amy B, and Dietram A Scheufele. 2009. "Moral Politicking Public Attitudes toward Gay Marriage in an Election Context." The International Journal of Press/Politics 14 (2):186-211. ———. 2011. "New Voters, New Outlook? Predispositions, Social Networks, and the Changing Politics of Gay Civil Rights*." Social Science Quarterly 92 (2):324-45. Beenie Man. 1994. "Bomb & Dynamite." In Defend It: VP Records. ———. 1996a. "Man Royal." In Maestro: VP Records & Greensleves. Lyrics taken from http://genius.com/Beenie-man-man-royal-lyrics [cited May 15, 2015]. ———. 1996b. "Nuff Gyal." In Maestro: VP Records & Greensleves. Bhugra, Dinesh. 1987. "Homophobia: A review of the literature." Sexual and Marital Therapy 2 (2):169-77. Billig, Michael. 1991. Ideology and opinions: Studies in rhetorical psychology. London, UK: Sage. ———. 1996. Arguing and thinking: A rhetorical approach to social psychology. Cambridge, UK: Cambridge University Press. Bilz, Kenworthey, and Janice Nadler. 2009. "Law, psychology, and morality." Psychology of Learning and Motivation 50:101-31. Bisnauth, Dale A. 1996. History of Religions in the Caribbean. Kingston, Jamaica: Kingston Publishes Limited. Bond, Bradley J., and Benjamin L. Compton. 2015. "Gay On-Screen: The Relationship Between Exposure to Gay Characters on Television and Heterosexual Audiences'

202

Endorsement of Gay Equality." Journal of Broadcasting & Electronic Media 59 (4):717-32. Borooah, Vani K, and John Mangan. 2007. "Love thy neighbour: how much bigotry is there in western countries?" Kyklos 60 (3):295-317. Boseley, Sarah. 2012. "Jamaican gay rights activists hopeful of repealing anti- homosexuality law." The Guardian, 10 February. Bounty Killer. 2004. "Dat A True." Scare Dem. Brambilla, Marco, and David A Butz. 2013. "Intergroup threat and outgroup attitudes: Macro-level symbolic threat increases prejudice against gay men." Social Psychology 44 (5):311. Brewer, Paul R., Joseph Graf, and Lars Willnat. 2003. "Priming or Framing: Media Influence on Attitudes Toward Foreign Countries." Gazette 65 (6):493-508. Briguglio, Lino. 1995. "Small island developing states and their economic vulnerabilities." World Development 23 (9):1615-32. Brown, Jarret. 1999. "Masculinity and Dancehall." Caribbean Quarterly 45 (1):1-16. Bryant, Jennings, and Dolf Zillman. 2009. "A Retrospective and Prospective Look at Media Effects." In The SAGE Handbook of Media Processes and Effects, ed. R. Nabi and M. B. Oliver. California, United States: Sage Publishers. Buju Banton. 1993. "Boom Bye Bye." In Strictly the Best: VP Records. Lyrics taken from http://www.lyricsmode.com/lyrics/b/buju_banton/boom_bye_bye.html [cited May 06, 2015]. Burdette, Amy M, Christopher G Ellison, and Terrence D Hill. 2005. "Conservative protestantism and tolerance toward homosexuals: An examination of potential mechanisms." Sociological Inquiry 75 (2):177-96. Burstein, Paul. 1998. "Bringing the public back in: Should sociologists consider the impact of public opinion on public policy?" Social Forces 77 (1):27-62. ———. 2003. "The impact of public opinion on public policy: A review and an agenda." Political Research Quarterly 56 (1):29-40. Cadge, Wendy, Laura R Olson, and Christopher Wildeman. 2008. "How denominational resources influence debate about homosexuality in mainline protestant congregations." Sociology of Religion 69 (2):187-207. Cameron, A. Collin, and Pravin K. Trivedi. 2010. Microeconometrics Using Stata, Revised Edition. Texas, USA: Stata Press. Campbell, Carl. 2011. "Education, Religion and Culture." In General History of the Caribbean, ed. K. O. Laurence. Paris; London: UNESCO Publishing and Macmillan Publishers. Campbell, Colin, and Jonathan Horowitz. 2016. "Does College Influence Sociopolitical Attitudes?" Sociology of Education 89 (1):40-58. Canadian HIV/AIDS Legal Network, and AIDS-Free World. 2015. "Jamaican gay activist launches constitutional challenge." Toronto, Canada: Canadian HIV/AIDS Legal Network. ———. 2016. "Jamaican Supreme Court admits religious groups as interested parties in challenge to anti-sodomy law." Toronto, Canada: Canadian HIV/AIDS Legal Network. Capleton. 1993. "Buggering." In Strictly The Best 11: VP Records. Lyrics taken from https://sonichits.com/video/Capleton/Buggering [cited May 02, 2015]. ———. 1999a. "Hang Dem Up." Fat Eyes Records. ———. 1999b. "In or Out." In Planet Reggae: The Beginning: VP Records. Lyrics taken from http://www.lyricsofsong.com/lyrics/capleton-in-or-out-lyrics- _krknlm.html [cited May 05, 2015].

203

———. 2000. "More Prophet." In More Fire: VP Records. Lyrics taken from http://genius.com/Capleton-more-prophet-lyrics [cited May 05, 2015]. Caribbean360. 2006. Call for Americans to ban homophobic Caribbean destinations https://www.caribbean360.com/travel/call-for-americans-to-ban-homophobic- caribbean-destinations [cited May 05 2014]. Caribbean Development Research Services. 2013a. "Attitudes towards homosexuals in Barbados (2013)." Caribbean Development Research Services, Barbados: Caribbean Development Research Services. ———. 2013b. "Attitudes towards homosexuals in Guyana (2013)." Caribbean Development Research Services, Barbados: Caribbean Development Research Services. ———. 2013c. "Attitudes towards homosexuals in Trinindad and Tobago (2013)." Caribbean Development Research Services, Barbados: Caribbean Development Research Services. Carmines, Edward G, and James Woods. 2002. "The role of party activists in the evolution of the abortion issue." Political Behavior 24 (4):361-77. Carroll, Aengus. 2016. "State Sponsored Homophobia 2016: A world survey of sexual orientation laws: criminalisation, protection and recognition ". Geneva, Switzerland: International Lesbian, Gay, Bisexual, Trans and Intersex Association. Chadee, Derek, Chezelle Joseph, Claire Peters, Vandana Sankar, Nisha Nair, and Jannel Philip. 2013. "Religiosity and attitudes towards homosexuals in a Caribbean environment." Social and Economic Studies 62 (1):1-28. Chen, Daniel L, and Susan Yeh. 2014. "The construction of morals." Journal of Economic Behavior & Organization 104:84-105. CHRI. 2011. Criminalised States [http://www.humanrightsinitiative.org/london/lgbt_rights/caribbean.pdf [cited April 05 2013]. Clarke, Victoria. 2001. "What about the children? Arguments against lesbian and gay parenting." Women's Studies International Forum 24 (5):555-70. Clements, Ben, and Clive D Field. 2014. "Public opinion toward homosexuality and gay rights in Great Britain." Public Opinion Quarterly 78 (2):523-47. Collier, Kate L, Stacey S Horn, Henny MW Bos, and Theo GM Sandfort. 2015. "Attitudes toward lesbians and gays among American and Dutch adolescents." The Journal of Sex Research 52 (2):140-50. Comstock, Gary David. 1991. Violence against lesbians and gay men. New York: Columbia University Press. Connell, R.W. 1987. Gender and power: Society, the person and sexual politics. UK: Polity Press & Blackwell Publishers. ———. 2005. Masculinities. Second Edition ed. California, US: University of California Press. Cooper, Carolyn. 1994. ""Lyrical Gun": Metaphor and Role Play in Jamaican Dancehall Culture." The Massachusetts Review 35 (3/4):429-47. ———. 2004. Sound clash: Jamaican dancehall culture at large. New York: Palgrave Macmillan. Copolov, Carly, and Ann Knowles. 2016. "Support for gay men: an Australian study." Culture, Health & Sexuality 18 (2):211-20. Corrales, Javier. 2015. "The Politics of LGBT Rights in Latin America and the Caribbean: Research Agendas." European Review of Latin American and Caribbean Studies 100:53- 62. Corrêa, Sonia, Sonia Correa, Rosalind Petchesky, and Richard Parker. 2008. Sexuality, health and human rights. London: Routledge.

204

Cowell, Noel M. 2011. "Public Discourse, Popular Culture and Attitudes Towards Homosexuals in Jamaica." Social and Economic Studies 60 (1):31-60. Cowell, Noel M, and Tanzia S Saunders. 2011. "Exploring heteronormativity in the public discourse of Jamaican legislators." Sexuality & Culture 15 (4):315-31. Criminal Law Amendement Act. 1885. The Criminal Law Amendment Act, 1885. With introduction, commentary and forms of indictments https://archive.org/details/criminallawamen00bodkgoog [cited May 20, 2016] Crockett, Alasdair, and David Voas. 2003. "A divergence of views: Attitude change and the religious crisis over homosexuality." Sociological Research Online 8 (4). Daily Mail Online. 2007. Elton John could turn us gay, says archdeacon. http://www.dailymail.co.uk/tvshowbiz/article-442724/Elton-John-turn-gay- says-archdeacon.html [cited January 28 2014]. Davies, Michelle. 2004. "Correlates of negative attitudes toward gay men: Sexism, male role norms, and male sexuality." Journal of Sex Research 41 (3):259-66. Davis, Nancy J, and Robert V Robinson. 1999. "Religious cosmologies, individualism and politics in Italy." Journal for the Scientific Study of Religion 38 (3):339-53. Davis, Nick. 2015. "Jamaica's first public gay pride event a symbol of change: 'It felt liberating'." The Guardian, August 10. Day, Aaron. 2014. Trinidad and Tobago: Prime Minister says gay rights is 'not legally possible' http://www.pinknews.co.uk/2014/09/26/trinidad-and-tobago-prime-minister- says-gay-rights-is-not-legally-possible/ [cited April 05 2015]. Djupe, Paul A, and Christopher P Gilbert. 2003. The prophetic pulpit: Clergy, churches, and communities in American politics: Rowman & Littlefield. Douglas, Claude J. 2007. Homosexuality in the Caribbean: Crawling out of the closet. Grenada: Maryzoon Press. Dr. Evil. 2005, “J.A. Nuh Like Gay”. Coppershot. Duffy, Bobby. n.d. "Gender roles and homosexuality." In Ipsos MORI Generations. http://www.ipsos-mori-generations.com/gender-roles-and-homosexuality [cited June 04 2016]. Duran, Khalid. 1993. "Homosexuality and Islam." In Homosexuality and world religions, ed. A. Swidler. Valley Forge, PA: Trinity Press International. Ekehammar, Bo, Nazar Akrami, Magnus Gylje, and Ingrid Zakrisson. 2004. "What matters most to prejudice: Big Five personality, Social Dominance Orientation, or Right‐Wing Authoritarianism?" European Journal of Personality 18 (6):463-82. Elephant Man. 1998. "A Nuh Fi We Fault." In Reggae Hits Vol. 24: Jet Star Records. ———. 1999. "Nuh Like." In Face Off Vol. 2: VP Records. Lyrics taken from http://www.soulrebels.org/dancehall/u_lyrics_nuh2.htm [cited May 04, 2015]. ———. 2003. "Nah Gwan A Jamaica." In Good 2 Go: VP Records. Lyrics taken from http://www.azlyrics.com/lyrics/elephantman/nahgwanajamaica.html [cited May 10, 2015]. Ellickson, Robert C. 2001. "The market for social norms." American Law and Economics Review 3 (1):1-49. Ellis, Sonja J, and Celia Kitzinger. 2002. "Denying equality: An analysis of arguments against lowering the age of consent for sex between men." Journal of Community & Applied Social Psychology 12 (3):167-80. Equality Network. n.d. Caribbean and the Americas. http://www.equality- network.org/caribbean-and-americas/ [cited June 06 2016]. Evans, MDR, and Jonathan Kelley. 2004a. "Attitudes towards homosexuality in 29 nations." In Australian Economy and Society 2002 - Religion, Morality and Public Policy in International Perspective, 1984-2002, ed. M. Evans and J. Kelley. Sydney, Australia: The Federation Press.

205

———. 2004b. "Religious belief in Australia." In Australian Economy and Society 2002 - Religion, Morality and Public Policy in International Perspective, 1984-2002, ed. M. Evans and J. Kelley. Sydney, Australia: The Federation Press. Fairbanks, David. 1977. "Religious forces and "morality" policies in the American states." The Western Political Quarterly 30 (3):411-7. Farquharson, Joseph T. 2002. "Subversive discourse and covert self-afﬁrmation through humour." Ms. University of Cambridge. ———. 2005. "Faiya—bon The socio-pragmatics of homophobia in Jamaican (Dancehall) culture." In Politeness and face in Caribbean Creoles, ed. S. Muhleisen and B. Migge. Amsterdam, Netherlands: John Benjamins Publishing Company. Ficarrotto, Thomas J. 1990. "Racism, sexism, and erotophobia: Attitudes of heterosexuals toward homosexuals." Journal of Homosexuality 19 (1):111-6. Finke, Roger and Amy Adamczyk. 2008. Cross-national moral beliefs: The Influence of National Religious Context. Sociological Quarterly, 49: 617–652. Flores, Andrew R, and Scott Barclay. 2016. "Backlash, Consensus, Legitimacy, or Polarization The Effect of Same-Sex Marriage Policy on Mass Attitudes." Political Research Quarterly 69 (1):43-56. Foose, Ethan, Jeremy Freese and Neil Cross. 2014. Political liberalism and graduate school attendance: A longitudinal analysis. In Professors and their Politics, ed. N. Gross and S. Simmons. Maryland, USL John Hupkins University Press. Frank, David John, and Elizabeth H. McEneaney. 1999. "The Individualization of Society and the Liberalization of State Policies on Same-Sex Sexual Relations, 1984– 1995." Social Forces 77 (3):911-43. Franklin, Karen. 2000. "Antigay behaviors among young adults prevalence, patterns, and motivators in a noncriminal population." Journal of Interpersonal Violence 15 (4):339- 62. Freemuse. 2007. Dancehall star signs Reggae Compassionate Act. http://freemuse.org/archives/997 [cited May 06 2016]. Frias-Navarro, Dolores, Hector Monterde-i-Bort, Marcos Pascual-Soler, and Laura Badenes-Ribera. 2015. "Etiology of homosexuality and attitudes toward same-sex parenting: A randomized study." Journal of Sex Research 52 (2):151-61. Friedrichs, Ellen. n.d. Homophobia in Music http://gayteens.about.com/od/glbtteenlifestyle/ss/homophobiamusic.htm [cited February 13 2015]. Gaskins, Joseph. 2013. "‘Buggery’ and the Commonwealth Caribbean: a comparative examination of the Bahamas, Jamaica, and Trinidad and Tobago " In Human Rights, Sexual Orientation and Gender Identity in The Commonwealth: Struggles for Decriminalisation and Change ed. C. Lennox and M. Waites. London: School of Advanced Study, University of London. Gato, Jorge, and Anne Marie Fontaine. 2016. "Attitudes toward adoption by same-sex couples: Effects of gender of the participant, sexual orientation of the couple, and gender of the child." Journal of GLBT Family Studies 12 (1):46-67. Genrich, Gillian L, and Brader A Brathwaite. 2005. "Response of religious groups to HIV/AIDS as a sexually transmitted infection in Trinidad." BMC Public Health 5 (121):1-12. Gerhards, Jürgen. 2010. "Non-Discrimination towards Homosexuality The European Union’s Policy and Citizens’ Attitudes towards Homosexuality in 27 European Countries." International Sociology 25 (1):5-28. Gilbert, Christopher P. 1993. The impact of churches on political behavior: An empirical study. Westport CN, USA: Greenwood Press.

206

Gordon, Demoya. 2006. "Policing Sexuality in America’s Churches: Theological, Congregational, and Political Influences on Regulation of Same-Sex Behavior." In Sociology Honors Projects Paper 5. Gordon, Robert W. 1984. "Critical legal histories." Stanford Law Review 36 (1):57-125. Greenberg, Anna. 2000. "The church and the revitalization of politics and community." Political Science Quarterly 115 (3):377-94. Greenberg, Ilan. 2010. "Murder Music." In Guernica. https://www.guernicamag.com/features/greenberg_12_1_10/ [cited February 28 2015]. Greene, William. 2012. Econometric Analysis. 7th ed. Essex, England: Pearson Education Limited. Gromer, Jill M, Michael H Campbell, Tomi Gomory, and Donna M Maynard. 2013. "Sexual Prejudice Among Barbadian University Students." Journal of Gay & Lesbian Social Services 25 (4):399-419. Gross, Neil, and Ethan Fosse. 2012. "Why are professors liberal?." Theory and Society 41(2): 127-168. Guittar, Nicholas A, and Heili Pals. 2014. "Intersecting gender with race and religiosity: Do unique social categories explain attitudes toward homosexuality?" Current Sociology 62 (1):41-62. Gupta, Alok, and Scott Long. 2008. "This alien legacy: the origins of" sodomy" laws in British colonialism." New York, NY: Human Rights Watch. Gusmano, Michael K, Mark Schlesinger, and Tracey Thomas. 2002. "Policy feedback and public opinion: the role of employer responsibility in social policy." Journal of Health Politics, Policy and Law 27 (5):731-72. Gutzmore, Cecil. 2004. "Casting the first stone!: Policing of homo/sexuality in Jamaican popular culture." Interventions 6 (1):118-34. Gwartney, Patricia A., and Daniel S. Schwartz. 2016. "You May Kiss the Groom: Americans’ Attitudes Toward Same-Sex Marriage." In Understanding Inequality: Social Costs and Benefits, ed. A. Machin and N. Stehr. Wiesbaden: Springer Fachmedien Wiesbaden. Haas, Hayley, and Fred Cowell. 2011. "The Impact of Laws Criminalising Same-Sex Sexual Conduct in the Commonwealth." London, UK: Commonwealth Human Rights Initiative. Haider-Markel, Donald P, and Mark R Joslyn. 2008. "Beliefs About the Origins of Homosexuality and Support For Gay Rights An Empirical Test of Attribution Theory." Public Opinion Quarterly 72 (2):291-310. Hamilton, Conrad. 2011. "Portia promises to review buggery law." The Jamaica Observer, December 21. Han, Enze, and Joseph O'Mahoney. 2014. "British colonialism and the criminalization of homosexuality." Cambridge Review of International Affairs 27 (2):268-88. Hans, Jason D, Megan Kersey, and Claire Kimberly. 2012. "Self-Perceived Origins of Attitudes Toward Homosexuality." Journal of Homosexuality 59 (1):4-17. Harry Toddler. 1998. "Bad Man Nuh Dress Like Girl." In Reggae Gold 1998: VP Records. Lyrics taken from http://www.soulrebels.org/dancehall/u_lyrics_badman3 [cited April 21, 2015]. Hart-Brinson, Peter. 2014. "Discourse of generations: The influence of cohort, period and ideology in Americans’ talk about same-sex marriage." American Journal of Cultural Sociology 2(2): 221-252. Hayes, Bernadette C, and Lizanne Dowds. 2015. "Religion and Attitudes Towards Gay Rights in Northern Ireland: The God Gap Revisited." In The Changing World

207

Religion Map: Sacred Places Identities, Practices and Politics, ed. S. D. Brunn and D. A. Gibreath. New York, US: Springer. Henley, Nancy M, and Fred Pincus. 1978. "Interrelationship of sexist, racist, and antihomosexual attitudes." Psychological Reports 42 (1):83-90. Henshaw, Alexis Leanna. 2014. "Geographies of Tolerance: Human Development, Heteronormativity, and Religion." Sexuality & Culture 18 (4):959-76. Herek, Gregory M. 1990. "The Context of Anti-Gay Violence Notes on Cultural and Psychological Heterosexism." Journal of Interpersonal Violence 5 (3):316-33. ———. 1991. "Stigma, Prejudice, and Violence Against Lesbians and Gay Men." In Homosexuality: Research implications for public policy ed. J. C. Gonsiorek and J. D. Weinrich. Newbury Park, CA: SAGE. ———. 2000a. "Sexual Prejudice and Gender: Do Heterosexuals' Attitudes Toward Lesbians and Gay Men Differ?" Journal of Social Issues 56 (2):251-66. ———. 2000b. "The psychology of sexual prejudice." Current Directions in Psychological Science 9 (1):19-22. ———. 2002. "Gender gaps in public opinion about lesbians and gay men." Public Opinion Quarterly 66 (1):40-66. ———. 2004. "Beyond “homophobia”: Thinking about sexual prejudice and stigma in the twenty-first century." Sexuality Research & Social Policy 1 (2):6-24. ———. 2012. Facts About Homosexuality and Child Molestation http://facultysites.dss.ucdavis.edu/~gmherek/rainbow/html/facts_molestation .html [cited January 05 2016]. Herek, Gregory M, and John P Capitanio. 1995. "Black heterosexuals’ attitudes toward lesbians and gay men in the United States." Journal of Sex Research 32 (2):95-105. ———. 1996. "" Some of my best friends": Intergroup contact, concealable stigma, and heterosexuals' attitudes toward gay men and lesbians." Personality and Social Psychology Bulletin 22:412-24. ———. 1999a. "AIDS stigma and sexual prejudice." American Behavioral Scientist 42 (7):1130-47. ———. 1999b. "Sex differences in how heterosexuals think about lesbians and gay men: Evidence from survey context effects." Journal of Sex Research 36 (4):348-60. Herek, Gregory M, and Eric K Glunt. 1993. "Interpersonal contact and heterosexuals’ attitudes toward gay men: Results from a national survey." Journal of Sex Research 30 (3):239-44. Hicks, Gary R, and Tien-Tsung Lee. 2006. "Public attitudes toward gays and lesbians: Trends and predictors." Journal of Homosexuality 51 (2):57-77. Hindu American Foundation. 2013. "Hindus in South Asia and the Diaspora: A Survey of Human Rights 2013." Hindu American Foundation. Hobolt, Sara Binzer, and Robert Klemmemsen. 2005. "Responsive government? Public opinion and government policy preferences in Britain and Denmark." Political Studies 53 (2):379-402. Hooghe, Marc, Ellen Claes, Allison Harell, Ellen Quintelier, and Yves Dejaeghere. 2010. "Anti-gay sentiment among adolescents in Belgium and Canada: A comparative investigation into the role of gender and religion." Journal of Homosexuality 57 (3):384-400. Hooghe, Marc, and Cecil Meeusen. 2013. "Is same-sex marriage legislation related to attitudes toward homosexuality?" Sexuality Research and Social Policy 10 (4):258-68. Hope, Dona. 2010. Man Vibes: Masculinities in the Jamaican Dancehall. Kingston, Jamaica: Ian Randle Publishers. Hope, Donna. 2006. Inna Di Dancehall: Popular Culture and the Politics of Identity in Jamaica. Jamaica: University of the West Indies Press.

208

Hough, Nicole A. 2004. "Sodomy and Prostitution: Laws Protecting the Fabric of Society." Pierce Law Review 3:101-24. Huckfeldt, Robert. 1986. Politics in context: Assimilation and conflict in urban neighborhoods. New York, US: Algora Publishing. Hull, Kathleen E. 2001. "The political limits of the rights frame: The case of same-sex marriage in Hawaii." Sociological Perspectives 44 (2):207-32. Human Dignity Trust. 2015. "The criminalisation of consensual same-sex sexual relations across the Commonwealth - Developments and opportunities." http://www.humandignitytrust.org/pages/COUNTRY%20INFO/Commonw ealth%20Countries: The Human Dignity Trust. Huntington, Samuel P. 1993. "The clash of civilizations?" Foreign affairs, 22-49. I Wayne. 2011. "Burn Down Sodom." In Life Teachings: VP Records. IGLHRC. 2011. "Human Rights Violations of Lesbian, Gay, Bisexual, and Transgender (LGBT) People in Jamaica: A Shadow Report." Geneva: International Gay and Lesbian Human Rights Commission. ———. 2012. "Human Rights Violations of Lesbian, Bisexual, and Transgender (LBT) People in Guyana: A Shadow Report." New York: International Gay and Lesbian Human Rights Commission. IGLHRC, and United and Strong. 2015. "Homophobia & Transphobia in Caribbean Media: A baseline study from Belize, Grenada, Guyana, Jamaica and Saint Lucia." New York, US: International Gay and Lesbian Human Rights Commission. Inglehart, Ronald. 1971. "The silent revolution in Europe: Intergenerational change in post-industrial societies." American Political Science Review 65 (4):991-1017. ———. 1990. Culture shift in advanced industrial society. Princeton, New Jersey US: Princeton University Press. ———. 1997. Modernization and postmodernization: Cultural, economic, and political change in 43 societies. Princeton, New Jersey US: Princeton University Press. ———. 2006. "Changing Norms: Existential Security Leads to Growing Acceptance of Out-Groups." WZB Mitteilungen 113:26-9. Inglehart, Ronald, and Wayne E Baker. 2000. "Modernization, cultural change, and the persistence of traditional values." American Sociological Review 65 (1):19-51. Inglehart, Ronald, and Christian Welzel. 2005. Modernization, cultural change, and democracy: The human development sequence. New York, US: Cambridge University Press. Ireland, Patrick R. 2013. "A Macro-Level Analysis of the Scope, Causes, and Consequences of Homophobia in Africa." African Studies Review 56 (2):47-66. Jackman, Mahalia. 2014. "Output volatility and tourism specialization in small island developing states." Tourism Economics 20 (3):527-44. Jackman, Mahalia, Troy Lorde, Shane Lowe, and Antonio Alleyne. 2011. "Evaluating tourism competitiveness of small island developing states: a revealed comparative advantage approach." Anatolia 22 (3):350-60. Jackman, Mahalia, and Simon Naitram. 2015. "Nowcasting Tourist Arrivals to Barbados: Just Google It!" Tourism Economics 21 (6):1309-13. Jamaica Gleaner. 2010. "J-FLAG applauds Kartel on signing Compassionate Act." Jamaica Gleaner. ———. 2013. "Another challenge to Buggery laws." Jamaica Gleaner, May 26. Jang, S Mo, and Hoon Lee. 2014. "When pop music meets a political issue: Examining how “Born This Way” influences attitudes toward gays and gay rights policies." Journal of Broadcasting & Electronic Media 58 (1):114-30. Jaspers, Eva, Marcel Lubbers, and Nan Dirk De Graaf. 2007. "‘Horrors of Holland’: explaining attitude change towards euthanasia and homosexuals in the

209

Netherlands, 1970–1998." International Journal of Public Opinion Research 19 (4):451- 73. Jelen, Ted G. 1992. "Political Christianity: A contextual analysis." American Journal of Political Science:692-714. Jowett, Adam. 2014. "‘But if you legalise same-sex marriage…’: Arguments against marriage equality in the British press." Feminism & Psychology 24 (1):37-55. Jäckle, Sebastian, and Georg Wenzelburger. 2015. "Religion, Religiosity, and the Attitudes Toward Homosexuality—A Multilevel Analysis of 79 Countries." Journal of homosexuality 62 (2):207-41. Kaleidoscope Trust. 2012. Trinidad PM promises to 'end all discrimination' against LGBT people. http://kaleidoscopetrust.com/news/2 [cited January 21 2016]. ———. 2014. "Generation change: LGBT youth from around the Caribbean speak out." UK: The Kaleidoscope Trust. Kaleidoscope Trust, and Kaleidoscope Australia Human Rights Foundation. 2015. "Speaking out 2015: The rights of LGBTI people across the Commonwealth Caribbean." UK: Kaleidoscope Trust. Keleher, Alison, and Eric RAN Smith. 2012. "Growing support for gay and lesbian equality since 1990." Journal of Homosexuality 59 (9):1307-26. Kelley, Jonathan, Mariah DR Evans, and Bruce Headey. 1993. "Moral reasoning and political conflict: The abortion controversy." British Journal of Sociology. 44 (4):589- 612. Kellstedt, Lyman A, and John C Green. 1993. "Religion, the neglected variable: An agenda for future research on religion and political behavior." In Rediscovering the religious factor in American Politics, ed. D. Leege and L. Kellstedt: ME Sharpe Inc,. Kempadoo, Kamala. 2009. "Caribbean Sexuality: Mapping the Field." Caribbean Review of Gender Studies (3):1-24. Kiesnoski, Kenneth. 2005. "How Homophobic Is the Caribbean?" Out Traveler, 38-39. Kimmel, Michael S. 1994. "Masculinity as homophobia: Fear, shame, and silence in the construction of gender identity." In Theorizing Masculinities, ed. H. Brod and M. Kaufman. London, UK: Sage Publications Kite, Mary E. 1984. "Sex differences in attitudes toward homosexuals: A meta-analytic review." Journal of Homosexuality 10 (1-2):69-81. Kite, Mary E, and Kay Deaux. 1987. "Gender belief systems: Homosexuality and the implicit inversion theory." Psychology of Women Quarterly 11 (1):83-96. Kite, Mary E, and Bernard E Whitley Jr. 1996. "Sex Differences in Attitudes Toward Homosexual Persons, Behaviors, and Civil Rights A Meta-Analysis." Personality and Social Psychology Bulletin 22 (4):336-53. Kort, Joe. 2012. "Homosexuality and pedophilia: The false link." In Huffington Post - The Blog. http://www.huffingtonpost.com/joe-kort-phd/homosexuality-and- pedophi_b_1932622.html. Kreitzer, Rebecca J, Allison J Hamilton, and Caroline J Tolbert. 2014. "Does Policy Adoption Change Opinions on Minority Rights? The Effects of Legalizing Same- Sex Marriage." Political Research Quarterly 67 (4):795-808. Kuhar, Roman. 2013. "Introduction to the issue:“In the Name of Hate: Homophobia as a Value”." Southeastern Europe 37 (1):1-16. Kuntz, Anabel, Eldad Davidov, Shalom H Schwartz, and Peter Schmidt. 2015. "Human values, legal regulation, and approval of homosexuality in Europe: A cross‐ country comparison." European Journal of Social Psychology 45 (1):120-34. Kuyper, Lisette, Jurjen Iedema, and Saskia Keuzenkamp. 2013. Towards tolerance. Exploring changes and explaining differences in attitudes towards homosexuality in Europe: Sociaal en Cultureel Planbureau.

210

Lacoste, Christian. 2013. Murder inna dancehall http://www.soulrebels.org/dancehall.htm [cited December 20 2014]. Lady Gaga. 2010. "Both this way." In Both This Way, ed. S. Germanotta and J. Laursen. US: Streamline, Kon Live and Interscope. LaFont, Suzanne. 2001. "Very straight sex: The development of sexual mores in Jamaica." Journal of Colonialism and Colonial History 2 (3). ———. 2009. "Not Quite Redemption Song: LGBT-Hate in Jamaica." In Homophobias: Lust and Loathing across Time and Space, ed. D. A. B. Murray. United States of America: Duke University Press. LaMar, Lisa, and Mary Kite. 1998. "Sex differences in attitudes toward gay men and lesbians: A multidimensional perspective." Journal of Sex Research 35 (2):189-96. LaMarca, Nicole. 2011. "The Likert Scale: Advantages and Disadvantages." In Field Research in Organizational Psychology. https://psyc450.wordpress.com/2011/12/05/the-likert-scale-advantages-and- disadvantages/. Larcher, Akim Ade, and Colin Robinson. 2009. "Fighting 'Murder Music': Activist Reflections." Caribbean Review of Gender Studies (3):1-12. Latin American Public Opinion Project. 2014. "AmericasBarometer 2014 Sample Design." http://www.vanderbilt.edu/lapop/ab2014.php: Vanderbilt University. ———. n.d. "AmericasBarometer Weighting Scheme." http://www.vanderbilt.edu/lapop/docs/AmericasBarometer_weighting_schem e_all_years_of_AB_v2.pdf: Vanderbilt University. Lax, Jeffrey, and Justin Phillips. 2009. "Gay Rights in the States: Public Opinion and Policy Responsiveness." American Political Science Review 103 (3):367-86. Layman, Geoffrey. 2001. The great divide: Religious and cultural conflict in American party politics: Columbia University Press. Laythe, Brian, Deborah Finkel, and Lee A Kirkpatrick. 2001. "Predicting prejudice from religious fundamentalism and right‐wing authoritarianism: a multiple‐regression approach." Journal for the Scientific Study of Religion 40 (1):1-10. Lee, Tien-Tsung, and Gary R Hicks. 2011. "An analysis of factors affecting attitudes toward same-sex Marriage: Do the Media Matter?" Journal of Homosexuality 58 (10):1391-408. Leslie, Christopher R. 2000. "Creating criminals: the injuries inflicted by unenforced sodomy laws." Harvard Civil Rights-Civil Liberties Law Review 35:103-81. Leveille, Dan. 2016. Equaldex http://www.equaldex.com/ [cited May 06 2016]. Levina, Marina, Craig R Waldo, and Louise F Fitzgerald. 2000. "We're here, we're queer, we're on TV: The effects of visual media on heterosexuals' attitudes toward gay men and lesbians." Journal of Applied Social Psychology 30 (4):738-58. Lewis, Gregory B. 2009. "Does believing homosexuality is innate increase support for gay rights?" Policy Studies Journal 37 (4):669-93. ———. 2011. "The friends and family plan: Contact with gays and support for gay rights." Policy Studies Journal 39 (2):217-38. Lewis, Gregory B, and Seong Soo Oh. 2008. "Public opinion and state action on same- sex marriage." State & Local Government Review 40 (1):42-53. Lewis, Linden. 2003. "Caribbean Masculinity: Unpacking the Narrative." In The Culture of Gender and Sexuality in the Caribbean, ed. L. Lewis. Florida: University Press of Florida. Lin, Kai, Deeanna M Button, Mingyue Su, and Sishi Chen. 2016. "Chinese College Students’ Attitudes Toward Homosexuality: Exploring the Effects of Traditional Culture and Modernizing Factors." Sexuality Research and Social Policy 13 (2):158- 72.

211

Livingston, Ceclia. 2012. "Beenie Man explains gay apology." The Jamaica Observer, August 03. Loftus, Jeni. 2001. "America's liberalization in attitudes toward homosexuality, 1973 to 1998." American Sociological Review: 66 (5):762-82. Lovindeer. 1987. "Don't Bend Down." The Sound Of Jamaica Records. Lyrics taken from http://www.lyricsmode.com/lyrics/l/lloyd_lovindeer/dont_bend_down.html [cite May 10, 2015]. Maiorana, Andre, Greg Rebchook, Nadine Kassie, and Janet J Myers. 2013. "On Being Gay in Barbados:“Bullers” and “Battyboys” and their HIV Risk in a Societal Context of Stigma." Journal of Homosexuality 60 (7):984-1010. Maitland, Richie. 2013. "On Religion and Rights in the Caribbean." In Groundation Grenada. http://groundationgrenada.com/2013/10/07/on-religion-and-rights- in-the-caribbean/#. [cited May 04, 2014] Marshall, Luigi. 2011. "Lashley Firm on Homosexuality." Nation News, December 02. Mason, Gail, and Mischa Barr. 2006. "Attitudes towards homosexuality: A literature review." http://sydney.edu.au/law/criminology/ahcn/docs_pdfs/Mason_Barr_2006.pd f: The University of Sydney. McAdams, Richard H. 2000a. "A focal point theory of expressive law." Virginia Law Review 86:1649-729. ———. 2000b. "An attitudinal theory of expressive law." Oregon Law Review 79:339-90. McCartney, Juan. 2014. "Death threats over GB gay pride event." The Nassau Guardian, September 01. McCleary, Rachel M, and Robert J Barro. 2006. "Religion and political economy in an international panel." Journal for the Scientific Study of Religion 45 (2):149-75. McPherson, Miller, Lynn Smith-Lovin, and James M Cook. 2001. "Birds of a feather: Homophily in social networks." Annual Review of Sociology:415-44. Moore, Carla. 2014. Wah eye ne see heart nuh leap: Queer marronage in the Jamaican dancehall, Department of Gender Studies, Queen's University, Ontario, Canada. Moore, Laura M, and Reeve Vanneman. 2003. "Context matters: Effects of the proportion of fundamentalists on gender attitudes." Social Forces 82 (1):115-39. Morgan, Joan. 1993. "No Apologies, No Regrets." VIBE, 75-82. Murray, David AB. 2009. "Bajan queens, nebulous scenes: Sexual diversity in Barbados." Caribbean Review of Gender Studies 3:1-20. ———. 2012. Flaming souls: homosexuality, homophobia, and social change in Barbados. Toronto, Canada: University of Toronto Press. Naidu, Janet. 2007. "Retention and Transculturation of Hinduism in the Caribbean." Guyana Journal. Naipaul, Chamanlall. 2003. "Parliament passes fundamental Rights Bill - puts aside 'sexual orientation' measure." Guyana Chronicle, July 25. Nationwide. 2014. "Holness has no problem with gays." In Impact. Jamaica: Nationwide. Nelson, Camille. 2008. "Lyrical Assault: Dancehall Versus the Cultural Imperialism of the North-West." Southern California Interdisciplinary Law Journal 17:231-78. Niedwiecki, Anthony. 2013. "Save Our Children: Overcoming the Narrative that Gays and Lesbians are Harmful to Children." Duke Journal of Gender Law & Policy 21:125-175. Norris, Pippa, and Ronald Inglehart. 2011. Sacred and secular: Religion and politics worldwide. US: Cambridge University Press.

212

Ohlander, Julianne, Jeanne Batalova, and Judith Treas. 2005. "Explaining educational influences on attitudes toward homosexual relations." Social Science Research 34 (4):781-99. Oldendick, Robert W. 2002. "The Role of Public Opinion in Policy and Practice." Public Policy & Practice 2 (1):30-35. Olson, Laura R, Wendy Cadge, and James T Harrison. 2006. "Religion and Public Opinion about Same‐Sex Marriage." Social Science Quarterly 87 (2):340-60. Outrage. 2007. "Reggae stars renounce homophobia." http://outrage.org.uk/2007/06/reggae-stars-renounce-homophobia/: Outrage. Overby, L Marvin, and Jay Barth. 2002. "Contact, community context, and public attitudes toward gay men and lesbians." Polity 34 (4):433-56. Padgette, Tim. 2006. "The Most Homophobic Place on Earth?" Time. Pan, Po-Lin, Juan Meng, and Shuhua Zhou. 2010. "Morality or equality? Ideological framing in news coverage of gay marriage legitimization." The Social Science Journal 47 (3):630-45. Perry, Samuel L. 2013. "Are interracial daters more supportive of same-sex unions?" The Social Science Journal 50 (2):252-6. Petersen, Jennifer L, and Janet Shibley Hyde. 2010. "A meta-analytic review of research on gender differences in sexuality, 1993–2007." Psychological bulletin 136 (1):21. ———. 2011. "Gender differences in sexual attitudes and behaviors: A review of meta- analytic results and large datasets." Journal of Sex Rsesearch 48 (2-3):149-65. Pettigrew, Thomas F, and Linda R Tropp. 2006. "A meta-analytic test of intergroup contact theory." Journal of Personality and Social Psychology 90 (5):751. Pettys, Todd E. 2010. "Sodom's Shadow: The Uncertain Line Between Public and Private Morality." Hastings Law Journal 61 (5):1161-216. Pew Research Center. 2011. "Global Christianity - A report on the size and distribution of the world’s Christian population." In The Pew Forum on Religious and Public Life. http://www.pewforum.org/files/2011/12/Christianity-fullreport-web.pdf: Pew Research Center. ———.2013. Global Acceptance of Homosexuality. http://www.pewglobal.org/2013/06/04/global-acceptance-of-homosexuality/ [cited March 04, 2016]. ———. 2014a. "Religion in Latin America: Widespread Change in a Historically Catholic Region." http://www.pewforum.org/2014/11/13/religion-in-latin- america/#religious-affiliations-of-latin-americans-and-u-s-hispanics: Pew Research Center. ———. 2014b. "The Global Divide on Homosexuality." In Pew Global Attitudes Project. http://www.pewglobal.org/files/2014/05/Pew-Global-Attitudes- Homosexuality-Report-REVISED-MAY-27-2014.pdf: Pew Research Center. ———. 2015. ‘The Whys and Hows of Generations Research” http://www.people- press.org/2015/09/03/the-whys-and-hows-of-generations-research/#fn- 20058600-2 : Pew Reasearch Center Phelan, Jo, Bruce G. Link, Ann Stueve, and Robert E. Moore. 1995. "Education, Social Liberalism, and Economic Conservatism: Attitudes Toward Homeless People." American Sociological Review 60(1): 126-40. Phillip, Danny K, and Traci S Williams. 2013. "Bullermen and Zamis: The LGB West Indian's Perspective." Journal of LGBT Issues in Counseling 7 (2):143-53. Piazza, Jared. 2012. "“If You Love Me Keep My Commandments”: Religiosity Increases Preference for Rule-Based Moral Arguments." International Journal for the Psychology of Religion 22 (4):285-302.

213

Pinnock, Agostinho MN. 2007. “'A Ghetto Education Is Basic':(Jamaican) Dancehall Masculinities As Counter-Culture." Journal of Pan African Studies 1(9):47-84. Potter, W. James. 2013. Media Effects. California, US: Sage. Pratto, Felicia, Jim Sidanius, Lisa M Stallworth, and Bertram F Malle. 1994. "Social dominance orientation: A personality variable predicting social and political attitudes." Journal of Personality and Social psychology 67 (4):741. Queen Ifrica. 2009. "Keep it to Yourself." In Montego Bay. US: VP Records. http://www.lyricsmode.com/lyrics/q/queen_ifrica/keep_it_to_yourself.html [cited June 01, 2015]. Radloff, Timothy D. 2007. “Measuring the Impact of Higher Education on Racial Prejudice and Opposition to Race-Based Policy”. New York Sociologist. 2(1): 1-26. Reding, Andrew. 2003. "Sexual Orientation and Human Rights in the Americas." In World Policy Reports. New York, USA: World Policy Institute. Reynolds, Jerome. 2013. "Buggery law conscience vote for Parliament soon." Jamaica Gleaner, June 6. Roberts, George W, and Joycelin Byrne. 1966. "Summary statistics on indenture and associated migration affecting the West Indies, 1834–1918." Population Studies 20 (1):125-34. Roberts, Scott. 2014. Lawyer: Jamaican Prime Minister has betrayed gay voters by refusing to lift buggery law. http://www.pinknews.co.uk/2014/05/19/lawyer-jamaican-prime- minister-has-betrayed-gay-voters-by-refusing-to-lift-buggery-law/ [cited June 20 2015]. Robinson, Tracy. 2009. "Authorized sex: Same-Sex Sexuality and Law in the Caribbean." In Sexuality, Social Exclusion and Human Rights: Vulnerability in the Caribbean Context of HIV, ed. C. Barrow, M. d. Bruin and R. Carr. Kingston, Jamaica: Ian Randle Publishers. Roozen, David A, William McKinney, and Jackson W Carroll. 1984. Varieties of religious presence: Mission in public life. New York: Pilgrim Press. Rowatt, Wade C, Jordan LaBouff, Megan Johnson, Paul Froese, and Jo-Ann Tsang. 2009. "Associations among religiousness, social attitudes, and prejudice in a national random sample of American adults." Psychology of Religion and Spirituality 1 (1):14. Rutledge, Scott Edward, and Neil Abell. 2005. "Awareness, acceptance, and action: An emerging framework for understanding AIDS stigmatizing attitudes among community leaders in Barbados." AIDS Patient Care & STDs 19 (3):186-99. Rye, BJ, and Glenn J Meaney. 2009. "Impact of a homonegativity awareness workshop on attitudes toward homosexuality." Journal of Homosexuality 56 (1):31-55. Sakalli, Nuray. 2002a. "Application of the attribution-value model of prejudice to homosexuality." The Journal of Social Psychology 142 (2):264-71. ———. 2002b. "The relationship between sexism and attitudes toward homosexuality in a sample of Turkish college students." Journal of Homosexuality 42 (3):53-64. Salter, Anna. 2003. Predators: Pedophiles, rapists and other sex offenders. New York, US: Basic Books. Saunders, Patricia J. 2003. "Is not everything good to eat, good to talk: Sexual economy and dancehall music in the global marketplace." Small Axe 7 (1):95-115. Savishinsky, Neil J. 1994. "Transnational popular culture and the global spread of the Jamaican Rastafarian movement." NWIG: New West Indian Guide 68 (3/4):259-81. Scheepers, Peer, Manfred Te Grotenhuis, and Frans Van Der Slik. 2002. "Education, religiosity and moral attitudes: explaining cross-national effect differences." Sociology of Religion 63 (2):157-76.

214

Scherr, Sonia. 2009. Why is iTunes Selling White Supremacy http://www.alternet.org/story/142311/why_is_itunes_selling_white_supremac y [cited May 8 2015]. Scheufele, Dietram A, and David Tewksbury. 2007. "Framing, agenda setting, and priming: The evolution of three media effects models." Journal of Communication 57 (1):9-20. Schleifer, Rebecca. 2004. "Hated to Death: Homophobia, Violence and Jamaica's HIV/AIDS Epidemic." In Human Rights Watch. http://www.hrw.org/sites/default/files/reports/jamaica1104.pdf. Schulte, Lisa J, and Juan Battle. 2004. "The relative importance of ethnicity and religion in predicting attitudes towards gays and lesbians." Journal of Homosexuality 47 (2):127-42. Schwartz, Joseph. 2010. "Investigating differences in public support for gay rights issues." Journal of Homosexuality 57 (6):748-59. Shabba Ranks. 1989. "Mauma Man." In Best Baby Father: VP Records. Sherkat, Darren E, Kylan Mattias De Vries, and Stacia Creek. 2010. "Race, Religion, and Opposition to Same‐Sex Marriage." Social Science Quarterly 91 (1):80-98. Sherkat, Darren E, Melissa Powell-Williams, Gregory Maddox, and Kylan Mattias De Vries. 2011. "Religion, politics, and support for same-sex marriage in the United States, 1988–2008." Social Science Research 40 (1):167-80. Siker, Jeffrey S. 2007. Homosexuality and religion: An encyclopedia: Greenwood Publishing Group. Singaravelou. 1990. "Indians in the French Overseas Departments: Guadeloupe, Martinique, Reunion." In South Asians Overseas: Migration and Ethnicity, ed. C. larke, C. Peach and S. Vertovec. Cambridge: Cambridge University Press. Sizzla. 2005. "Nah Apologize." Dark Side Records. Skipworth, Sue Ann, Andrew Garner, and Bryan J Dettrey. 2010. "Limitations of the contact hypothesis: Heterogeneity in the contact effect on attitudes toward gay rights." Politics & Policy 38 (5):887-906. Skyers, Javene. 2014. "Thousands rally against tossing out buggery act; shout for clean, righteous living." The Jamaica Observer, June 30. Slenders, Susanne, Inge Sieben, and Ellen Verbakel. 2014. "Tolerance towards homosexuality in Europe: Population composition, economic affluence, religiosity, same-sex union legislation and HIV rates as explanations for country differences." International Sociology 29 (4):348-67. Smith, Charlene, and Ryan Kosobucki. 2011. "Homophobia in the Caribbean: Jamaica." LSD Journal 1:1-55. Smith, Ian P. 2012. "An introduction to social dominance orientation." In Moving Psychology. http://sites.psu.edu/movingpsychology/2012/09/06/an- introduction-to-social-dominance-orientation/. Smith, Sara J, Amber M Axelton, and Donald A Saucier. 2009. "The effects of contact on sexual prejudice: A meta-analysis." Sex Roles 61 (3-4):178-91. Snijders, Tom AB, and Roel J Bosker. 2012. Multilevel analysis: An introduction to basic and advanced multilevel modelling. 2nd ed. London, UK: Sage. Spartacus World. 2016. Gay Travel Index 2016. Spartacus - International Gay Guide 2016 http://www.spartacusworld.com/ [cited June 28 2016]. Spiker, D. Chad. 1998. Reggae as Social Change: The Spread of Rastafarianism https://debate.uvm.edu/dreadlibrary/spiker.html [cited February 02 2015]. Spragga Benz. 1996. "Funny Guy Thing." In Ragga Ragga Ragga 7. East Coast. Lyrics taken from http://genius.com/Spragga-benz-no-funny-guy-thing-lyrics [cited May 8, 2015].

215

Srivastava, Pratima Ramful. 2013. Recreational drug consumption: An economic perspective. Edited by H. E. Frech III and P. Zweifel. Vol. 11. New York, US: Springer. Stabroek News. 2003. "Religious groups say no to sexual orientation in human rights bill." Stabroek News, July 9. Stanislas, Perry. 2014. "Policing violent homophobia in the Caribbean and the British Caribbean diaspora: Postcolonial discourses and the limits of postmodernity." Interventions 16 (1):135-56. Stanley, Sonja Niah. 2005. "'Dis slackness ting': A dichotomizing master narrative in Jamaican dancehall." Caribbean quarterly 51 (3/4):55-76. Stephan, Walter G, and Cookie W Stephan. 2000. "An integrated threat theory of prejudice." In Reducing Prejudice and Discrimination, ed. S. Oskamp. Mahwah, New Jersey: Earlbaum. Stewart, Colin. 2014. "Fear for family safety dissolves lawsuit vs anti-gay law." In Erasing 76 crimes. https://76crimes.com/2014/08/31/fear-for-family-safety-dissolves- lawsuit-vs-anti-gay-law/. ———. 2016. "Caribbean nations inch their way towards LGBTI rights." In Erasing 76 crimes. https://76crimes.com/2016/04/27/caribbean-nations-inch-their-way- toward-lgbti-rights/. Storm, Ingrid. 2009. "Halfway to heaven: Four types of fuzzy fidelity in Europe." Journal for the Scientific Study of Religion 48 (4):702-18. Stubager, Rune. 2008. "Education effects on authoritarian–libertarian values: a question of socialization." The British Journal of Sociology 59(2): 327-350. Stychin, Carl Franklin. 1998. A nation by rights: National cultures, sexual identity politics, and the discourse of rights: Temple University Press. Takács, Judit, and Ivett Szalma. 2011. "Homophobia and same-sex partnership legislation in Europe." Equality, Diversity and Inclusion: An International Journal 30 (5):356-78. Tatchell, Peter. 2015. "Jamaica homophobia: Hail reggae singer Mista Majah P's lone crusade against 'batty bwoy' hatred." In International Business Times. http://www.ibtimes.co.uk/jamaica-homophobia-hail-reggae-singer-mista- majah-ps-lone-crusade-against-batty-bwoy-hatred-1494556: IBT Media. Taylor, Jocelyn. 2014. Perceived HIV-Related Stigma and Homophobia within a Nationally Representative Sample. Paper read at 2014 National Conference on Health Communication, Marketing, and Media (August 19-21). The Nassau Guardian. 2014. "The LGBT debate: A historic perspective." The Nassau Guardian, March 17. Tilley, James, and Geoffrey Evans. 2014. "Ageing and generational effects on vote choice: Combining cross-sectional and panel data to estimate APC effects." Electoral Studies, 33: 19-27. Tinker, Hugh. 1993. A New System of Slavery: The Export of Indian Labour Overseas 1830- 1920. London: Hansib Publishing Limited. Tomlinson, Maurice. 2016. "Losing to win at the Caribbean Court of Justice." In Canadian HIV/AIDS Legal Network Blog. http://www.aidslaw.ca/site/losing-to-win-at- the-caribbean-court-of-justice/: Canadian HIV/AIDS Legal Network. Tygart, CE. 2000. "Genetic causation attribution and public support of gay rights." International Journal of Public Opinion Research 12 (3):259-75. UN Human Rights Committee. 2009. "Consideration of reports submitted by States parties under Article 40 of the Covenant - International Convenant on Civil and Political Rights (Barbados): Information received from Barbados on the implementation of the concluding observations of the Human Rights Committee (CCPR/C/BRB/CO/3)." http://www.refworld.org/docid/4a9fb80f2.html: United Nations.

216

UN Human Rights Council. 2010. "National report submitted in accordance with paragraph 15(a) of the annex to Human Rights Council resolution 5/1 - Saint Lucia." http://www.upr- info.org/sites/default/files/document/saint_lucia/session_10_- _january_2011/a_hrc_wg-6_10_lca_1_e.pdf: United Nations. ———. 2011. "Report of the Working Group on the Universal Periodic Review - Saint Vincent and the Grenadines: Views on conclusions and/or recommendations, voluntary commitments and replies presented by the State under review." http://www2.ohchr.org/english/bodies/hrcouncil/docs/18session/A.HRC.18. 15.ADD.1_en.pdf: United Nations. UNAIDS. 2010. "The Status of HIV in the Caribbean." Port of Spain, Trinidad and Tobago: UNAIDS Caribbean Regional Support Team. UNDP. 2015. "Human Development Report 2015: Work for Human Development." In Global Human Development Reports. http://hdr.undp.org/sites/default/files/2015_human_development_report.pdf: United Nations Development Programme. van den Akker, Hanneke, Rozemarijn van der Ploeg, and Peer Scheepers. 2013. "Disapproval of homosexuality: Comparative research on individual and national determinants of disapproval of homosexuality in 20 European countries." International Journal of Public Opinion Research 25 (1):64-86. van der Veer, Peter, and Steven Vertovec. 1991. "Brahmanism abroad: On Caribbean Hinduism as an ethnic religion." Ethnology 30 (2):149-66. Vbyz Kartel, and Spice. 2010. "Romping Shop." In Pon Di Gaza 2.0: VP Records. Veal, Michael. 2007. Dub: soundscapes and shattered songs in Jamaican reggae. Middletown, CT, USA: Wesleyan University Press. Vilaythong, T, Nicole M Lindner, and Brian A Nosek. 2010. "“Do unto others”: Effects of priming the Golden Rule on Buddhists’ and Christians’ attitudes toward gay people." Journal for the Scientific Study of Religion 49 (3):494-506. Vincent, Wilson, John L Peterson, and Dominic J Parrott. 2009. "Differences in African American and White women’s attitudes toward lesbians and gay men." Sex Roles 61 (9-10):599-606. Voas, David, and Alasdair Crockett. 2005. "Religion in Britain: Neither believing nor belonging." Sociology 39 (1):11-28. Voas, David, and Siobhan McAndrew. 2012. "Three puzzles of non-religion in Britain." Journal of Contemporary Religion 27 (1):29-48. Waites, Matthew. 1998. "Sexual citizens: legislating the age of consent in Britain." In Politics of Sexuality: Identity, Gender, Citizenship, ed. T. Carver and V. Mottier. London, UK: Routledge. Wald, Kenneth D, Dennis E Owen, and Samuel S Hill. 1988. "Churches as political communities." American Political Science Review 82 (2):531-48. Walls, N Eugene. 2010. "Religion and support for same-sex marriage: Implications from the literature." Journal of Gay & Lesbian Social Services 22 (1-2):112-31. Weber, Max. 1958. The Protestant Ethic and the Spirit of Capitalism. New York: Allen and Unwin. Weiner, Bernard. 1979. "A theory of motivation for some classroom experiences." Journal of Educational Psychology 71 (1):3-25. Weiner, Bernard, Raymond P Perry, and Jamie Magnusson. 1988. "An attributional analysis of reactions to stigmas." Journal of Personality and Social Psychology 55 (5):738- 748.

217

West, Donald J, and Andrea Woelke. 2002. "England." In Sociolegal control of homosexuality: A multi-nation comparison, ed. D. J. West and R. Green. United States: Springer Science & Business Media. West, Keon, and Noel M Cowell. 2015. "Predictors of prejudice against lesbians and gay men in Jamaica." Journal of Sex Research 52 (3):1-10. West, Keon, and Miles Hewstone. 2012. "Culture and contact in the promotion and reduction of anti-gay prejudice: Evidence from Jamaica and Britain." Journal of Homosexuality 59 (1):44-66. White, Ruth C, and Robert Carr. 2005. "Homosexuality and HIV/AIDS stigma in Jamaica." Culture, Health & Sexuality 7 (4):347-59. Whitehead, Andrew L. 2010. "Sacred Rites and Civil Rights: Religion's Effect on Attitudes Toward Same‐Sex Unions and the Perceived Cause of Homosexuality." Social Science Quarterly 91 (1):63-79. ———. 2013. "Religious organizations and homosexuality: The acceptance of gays and lesbians in American congregations." Review of Religious Research 55 (2):297-317. Whitehead, Andrew L, and Joseph O Baker. 2012. "Homosexuality, Religion, and Science: Moral Authority and the Persistence of Negative Attitudes." Sociological Inquiry 82 (4):487-509. Whitehead, Andrew L, and Samuel L Perry. 2016. "Religion and Support for Adoption by Same-Sex Couples The Relative Effects of Religious Tradition, Practices, and Beliefs." Journal of Family Issues 37 (6):789-813. Whitley Jr, Bernard E. 1990. "The relationship of heterosexuals' attributions for the causes of homosexuality to attitudes toward lesbians and gay men." Personality and Social Psychology Bulletin 16 (2):369-77. ———. 1999. "Right-wing authoritarianism, social dominance orientation, and prejudice." Journal of Personality and Social Psychology 77 (1):126-134. ———. 2001. "Gender-role variables and attitudes toward homosexuality." Sex Roles 45 (11-12):691-721. ———. 2009. "Religiosity and attitudes toward lesbians and gay men: A meta-analysis." International Journal for the Psychology of Religion 19 (1):21-38. Whitley Jr, Bernard E, and Sarah E Lee. 2000. "The Relationship of Authoritarianism and Related Constructs to Attitudes Toward Homosexuality." Journal of Applied Social Psychology 30 (1):144-70. Whitley Jr, Bernard E, and Stefanía Ægisdóttir. 2000. "The gender belief system, authoritarianism, social dominance orientation, and heterosexuals' attitudes toward lesbians and gay men." Sex Roles 42 (11-12):947-67. Wiley, Tisha RA, and Bette L Bottoms. 2013. "Attitudinal and individual differences influence perceptions of mock child sexual assault cases involving gay defendants." Journal of Homosexuality 60 (5):734-49. Wilson, Paul R. 1971 Sexual Dilemma. Brisbane, Australia: University of Queensland Press. Wood, Peter B, and John P Bartkowski. 2004. "Attribution Style and Public Policy Attitudes Toward Gay Rights." Social Science Quarterly 85 (1):58-74. Woolfalk, Miya. 2012. "Sermons Aren’t Explicitly Political but Clergy Are:Political Cue- Giving in Sermons by U.S. Christian Clergy ". World Bank. 2016. "The World Bank EdStats (Education Statistics)." http://data.worldbank.org/data-catalog/ed-stats: The World Bank. Worrell, DeLisle. 2012. "Policies for Stabilization and Growth in Small Very Open Economies." In Group of 30 Occasional Working Papers 85. Washington, DC, USA: Group of Thirty.

218

Worrell, DeLisle, Shane Lowe, and Simon Naitram. 2013. "Growth Forecasts for Foreign Exchange Constrained Economies." Journal of Business, Finance and Economics in Emerging Economies 8 (1):1-25. Wright, Keril. 2007. "Beenie Man denies signing deal with gay group." The Jamaica Observer, July 22. WTTC. 2016. "Travel & Tourism: Economic Impact 2016 Caribbean." http://www.wttc.org/research/economic-research/economic-impact- analysis/regional-reports/. Yeh, Susan. 2012. "Laws and Social Norms: Unintended Consequences of Obesity Laws." University of Cincinnati Law Review. 81:173-223. Yen, Steven T, and Ernest M Zampelli. 2016. "Religiousness and Support for Same‐Sex Marriage: An Endogenous Treatment Approach." Forthcoming in Social Science Quarterly. Štulhofer, Aleksander, and Ivan Rimac. 2009. "Determinants of homonegativity in Europe." Journal of Sex Research 46 (1):24-32.

219

Appendix A – Appendix to Chapter 3 ______In this Appendix, I present the anti-gay laws of Antigua and Barbuda; Barbados; Belize; Dominica; Grenada; Jamaica; St Kitts and Nevis; St Vincent and the Grenadines; and Trinidad and Tobago. Information obtained from Human Dignity Trust (2015) and Carroll (2016).

Antigua and Barbuda

Sexual Offences Act 1995 (Act No. 9)

Buggery Article 12 1) “A person who commits buggery is guilty of an offence and is liable on conviction to imprisonment – a) for life, if committed by an adult on a minor; b) for fifteen years, if committed by an adult on another adult; c) for five years, if committed by a minor”

2) “In this section "buggery" means sexual intercourse per mum by a male person with a male person or by a male person with a female person.”

Serious Article 15 1) “A person who commits an act of serious or towards Indecency another is guilty of an offence and is liable on conviction to imprisonment – a) for ten years, if committed on or towards a minor under sixteen years of age; b) for five years, if committed an or towards a person sixteen years of age of more.”

2) “Subsection (1) does not apply to an act of serious indecency committed in private between – a) a husband and his wife; or

220

b) a male person and a female person each of whom is sixteen years of age or more; both of whom consent to the commission of the act.”

3) “An act of "serious indecency" is an act, other than sexual intercourse (whether natural or unnatural), by a person involving the use of the genital organ for the purpose of arousing or gratifying sexual desire.”

Barbados

Sexual Offences Act 1992, Chapter 154

Buggery Section 9 “Any person who commits buggery is guilty of an offence and is liable on conviction on indictment to imprisonment for life.”

Serious Section 12 1) “A person who commits an act of serious indecency on Indecency or towards another or incites another to commit that act with the person or with another person is guilty of an offence and, if committed on or towards a person 16 years of age or more or if the person incited is of 16 years of age or more, is liable on conviction to imprisonment for a term of 10 years.”

2) “A person who commits an act of serious indecency with or towards a child under the age of 16 or incites the child under that age to such an act with him or another, is guilty of an offence and is liable on conviction to imprisonment for a term of 15 years.”

3) “An act of “serious indecency” is an act, whether natural or unnatural by a person involving the use of the

221

genital organs for the purpose of arousing or gratifying sexual desire.”

Belize

Criminal Code, Chapter 101 (Revised Edition, 2000)

Buggery Section 53 “Every person who has carnal intercourse against the order of nature with any person or animal shall be liable to imprisonment for ten years.”

Immigration Act, Chapter 156 (Revised Edition, 2000)

Categories of Section 5 -1e “Any prostitute or homosexual or any person who may prohibited be living on or receiving or may have been living on or immigrants receiving the proceeds of prostitution or homosexual behaviour”

Dominica

Sexual Offences Act 1998

Gross Section 14 1) “Any person who commits an act of gross indecency Indecency with another person is guilty of an offence and liable on conviction to imprisonment for five years.”

2) “Subsection (1) does not apply to an act of gross indecency committed in private between an adult male person and an adult female person, both of whom consent.”

3) “For the purposes of subsection (2) – a) an act shall be deemed not to have been committed in private if it is committed in a public place; and

222

b) a person shall be deemed not to consent to the commission of such an act if – i) the consent is extorted by force, threats or fear of bodily harm or is obtained by false and fraudulent representations as to the nature of the act; ii) the consent is induced by the application or administration of any drug, matter or thing with intent to intoxicate or stupefy the person; or iii) that person is, and the other party to the act knows or has good reason to believe that the person is suffering from a mental disorder.

4) “In this section “gross indecency” is an act other than sexual intercourse (whether natural or unnatural) by a person involving the use of genital organs for the purpose of arousing or gratifying sexual desire.”

Buggery Section 16 1) “A person who commits buggery is guilty of an offence and liable on conviction to imprisonment for – a) twenty-five years, if committed by an adult on a minor; b) ten years, if committed by an adult on another adult; c) or five years, if committed by a minor; and, if the Court thinks it fit, the Court may order that the convicted person be admitted to a psychiatric hospital for treatment.”

Buggery Section 16 2) “Any person who attempts to commit the offence of buggery, or is guilty of an assault with the intent to commit the same is guilty of an offence and liable to imprisonment for four years and, if the Court thinks it

223

fit, the Court may order that the convicted person be admitted to the psychiatric hospital for treatment.”

3) “In this section “buggery” means sexual intercourse per anum by a male person with a male person or by a male person with a female person.”

Grenada

Criminal Code of 1987 (as amended in 1993)

Unnatural Article 431 “If any two persons are guilty of unnatural connexion, or if Connexion any person is guilty of an unnatural connexion with any animal, every such person shall be liable to imprisonment for ten years.”

Guyana

Criminal Law (Offences) Act

Committing Section 352 “Any male person, who in public or private, commits, or acts of is a party to the commission, or procures or attempts to gross procure the commission, by any male person, of an act of indecency gross indecency with any other male person shall be guilty of misdemeanour and liable to imprisonment for two years”.

Attempts to Section 353 “Everyone who – commit a) attempts to commit buggery; or unnatural b) assaults any person with the intention to commit offences buggery; or c) being a male, indecently assaults any other male person,

224

shall be guilty of felony and liable to imprisonment for ten years”

Buggery Section 354 “Everyone who commits buggery, either with a human being or with any other living creature, shall be guilty of felony and liable to imprisonment for life”

Jamaica

Offences against the Person Act, 1864

Unnatural Article 76 “Whosoever shall be convicted of the abominable crime of Crime buggery [anal intercourse] committed either with mankind or with any animal, shall be liable to be imprisoned and kept to hard labour for a term not exceeding ten years”.

Attempt Article 77 “Whosoever shall attempt to commit the said abominable crime, or shall be guilty of any assault with intent to commit the same, or of any indecent assault upon any male person, shall be guilty of a misdemeanour, and being convicted thereof shall be liable to be imprisoned for a term not exceeding seven years, with or without hard labour”.

Outrages Article 79 “Any male person who, in public or private, commits, or is a on party to the commission of, or procures or attempts to procure Decency the commission by any male person of, any act of gross indecency with another male person, shall be guilty of a misdemeanor, and being convicted thereof shall be liable at the discretion of the court to be imprisoned for a term not exceeding 2 years, with or without hard labour”.

St Kitts and Nevis

Offences Against the Person Act 1986, Chapter 4.21 (revised edition 2002)

225

Sodomy Section 56 “Any person who is convicted of the abominable crime of and buggery, committed either with mankind or with any animal, Bestiality shall be liable to be imprisoned for a term not exceeding ten years, with or without hard labour”

Attempt Section 57 “Any person who attempts to commit the said abominable to crime, or is guilty of any assault with intent to commit the Commit same, or of any indecent assault upon any male person, an commits a misdemeanour, and, on conviction, shall be liable infamous to be imprisoned for a term not exceeding four years, with crime or without hard labour.”

St. Lucia Criminal Code, No. 9 of 2004

Gross Section 132 1) “Any person who commits an act of gross indecency with indecency another person commits an offence and is liable on conviction on indictment to imprisonment for ten years or on summary conviction to five years.”

2) “Subsection (1) does not apply to an act of gross indecency committed in private between an adult male person and an adult female person, both of whom consent.”

3) “For the purposes of subsection (2) — a) an act shall be deemed not to have been committed in private if it is committed in a public place; and b) a person shall be deemed not to consent to the commission of such an act if — i) the consent is extorted by force, threats or fear of bodily harm or is obtained by false and fraudulent representations as to the nature of the act;

226

ii) the consent is induced by the application or administration of any drug, matter or thing with intent to intoxicate or stupefy the person; or iii) that person is, and the other party to the act knows or has good reason to believe that the person is suffering from a mental disorder 4) In this section “gross indecency” is an act other than sexual intercourse (whether natural or unnatural) by a person involving the use of the genital organs for the purpose of arousing or gratifying sexual desire.

Buggery Section 133 1) “A person who commits buggery commits an offence and is liable on conviction on indictment to imprisonment for — a) life, if committed with force and without the consent of the other person; b) ten years, in any other case.”

2) “Any person who attempts to commit buggery, or commits an assault with intent to commit buggery, commits an offence and is liable to imprisonment for five years.”

3) “In this section “buggery” means sexual intercourse per anus by a male person with another male person.”

St Vincent and the Grenadines

Criminal Code (revised 1990)

Buggery Section 146 “Any person who – a) commits buggery with any other person; b) commits buggery with an animal; or

227

c) permits any person to commit buggery with him or her; is guilty of an offence and liable to imprisonment for ten years.”

Gross Section 133 “Any person, who in public or private, commits an act of Indecency gross indecency with another person of the same sex, or procures or attempts to procure another person of the same sex to commit an act of gross indecency with him or her, is guilty of an offence and liable to imprisonment for five years.”

Trinidad and Tobago

Sexual Offences Act 1986 (Consolidated Version, 2000) Buggery Section 13 1) “A person who commits buggery is guilty of an offence and is liable on conviction to imprisonment— a) if committed by an adult on a minor, for life; b) if committed by an adult on another adult, for twenty- five years; c) if committed by a minor, for five years.”

2) “In this section “buggery” means sexual intercourse per anum by a male person with a male person or by a male person with a female person.”

Serious Section 16 1) “A person who commits an act of serious indecency on Indecency or towards another is guilty of an offence and is liable on conviction to imprisonment— a) if committed on or towards a minor under sixteen years of age for ten years for a first offence and to imprisonment for fifteen years for a subsequent offence;

228

b) if committed on or towards a person sixteen years of age or more for five years.”

2) “Subsection (1) does not apply to an act of serious indecency committed in private between— a) a husband and his wife; or b) a male person and a female person each of whom is sixteen years of age or more, both of whom consent to the commission of the act.

3) “An act of “serious indecency” is an act, other than sexual intercourse (whether natural or unnatural), by a person involving the use of the genital organ for the purpose of arousing or gratifying sexual desire.”

229

Appendix B – Appendix to Chapter 4 ______

Sexual orientation of CADRES survey respondents by country

Figure B1: Responses to question about sexual orientation by country of residence

1.6% 1.8%

BARBADOS 90.3% 6.4%

3.2% 3.9%

GUYANA 76.4% 16.5%

1.6% 1.8%

TRINIDAD & 90.3% TOBAGO 6.4%

Heterosexual Gay/lesbian Bisexual Prefer not to say/no response

Missing values

Table B1: Missing values for dependent variables

Missing Missing Missing Barbados Guyana Trinidad and Total Missing (%) (%) Tobago (%) (%) Maintain the laws 5 (0.7) 124 (15.7) 5 (0.6) 134 (5.6) Enforce the laws 6 (0.9) 124 (15.7) 4 (0.4) 134 (5.6) Penalise men in private 7 (1.0) 22 (2.8) 9 (1.0) 38 (1.6) Penalise women in private 10 (1.5) 21 (2.7) 8 (0.9) 44 (1.8) Laws express moral standards 10 (1.5) 140 (17.7) 12 (1.3) 162 (6.8) Laws stop spread of homosexuality 15 (2.2) 145 (18.4) 14 (1.5) 174 (7.3) Laws important from health 18 (2.6) 145 (18.5) 15 (1.7) 178 (7.5) perspective Laws protect young people from 18 (2.6) 143 (18.1) 13 (1.4) 174 (7.3) abuse

230

‘Discrimination against minorities’ variable

As mentioned in Chapter 5, the CADRES survey asked respondents whether or not it was acceptable to treat persons differently on the basis of (1) Race; (2) Gender; (3) Religion and (4) Age. These variables were all dichotomised such that a “Yes” response was coded as 1, and the remaining responses (No or Unsure/Won’t Say) were given values of 0. The tetrachoric correlations are shown in Table B2.

Table B2: Tetrachoric correlations

Race Gender Religion Age Race 1.000 Gender 0.854 1.000 Religion 0.857 0.820 1.000 Age 0.702 0.860 0.739 1.000 Note: ***, **, * and + indicates significance at the 0.1, 1, 5 and 10 per cent levels, respectively

A two parameter model Item response theory is then used to estimate the relationship between the latent trait of interest and the above items. Table B3 presents the estimate of the item discrimination parameters (labelled IDP) and the difficulty estimate. Table B4 shows the goodness of fit statistics

Table B3: Item discrimination parameter and difficulty estimates

Coefficient p-values Item Discrimination Race 1.665 0.000 Gender 3.058 0.000 Religion 2.723 0.000 Age 2.719 0.000 Item Difficulties Race 1.521 0.000 Gender 1.267 0.000 Religion 0.982 0.000 Age 0.970 0.000

Table B4: Goodness of fit statistics

Value Baseline comparison Comparative fit index (CFI) 0.975 Tucker Lex index 0.988 Error Root mean squared error of approximation (RMSEA) 0.041 Probability that RMSEA ≤0.05 0.102

231

Appendix C – Appendix to Chapter 5 ______This Appendix presents the results from the interacting the contextual variables with the cohort and religious identity variables.

HDI and cohort

Table C 1: Interacting HDI with birth cohort Coefficients Intercept 61.208***

Individual Characteristics Gender Men (base) Women 5.904*** Birth Cohort 1990s or later (base) 1980s -1.343 1970s -8.180*** 1960s -12.164*** 1950s -11.990*** 1940s or earlier -21.193*** Non-response -14.077+ Religious identity Evangelical (base) Catholic 7.944*** Mainline Protestant 3.778*** Other 6.715*** Not religious 8.400*** Don’t know 3.025* Non-response 3.486** Importance of religion Very important (base) Rather important 4.632*** Not very important 6.438*** Not at all important 12.815*** Don’t know 3.343 No response 6.927*** Country characteristics HDI Very high human development (base) High human development -14.871*** Medium human development -18.783*** Low human development -38.441***

232

Table C 1 cont’d Religious climate Share of Evangelicals -0.273* Legal context Same-sex marriage (base) Recognise same-sex relationships -24.279*** Bans on same-sex intimacy removed -24.228*** Same-sex intimacy criminalised -30.384*** Interactions HDI × Birth Cohort High human development × 1980s -1.966 High human development × 1970s 2.372 High human development × 1960s 3.657* High human development × 1950s 2.182 High human development × 1940s or earlier 7.881*** High human development × No response 4.213 Medium human development × 1980s -2.269 Medium human development × 1970s -0.641 Medium human development × 1960s 3.108+ Medium human development × 1950s 1.859 Medium human development × 1940s or earlier 9.540** Medium human development × No response 9.590 Low human development × 1980s 0.051 Low human development × 1970s 7.033* Low human development × 1960s 7.706* Low human development × 1950s 9.264** Low human development × 1940s or earlier 18.149*** Low human development × No response (empty)

Log likelihood -249,116.350 Note: ***, **, * and + indicate significance at the 0.1, 1, 5 and 10 per cent levels, respectively

Table C2: Significance of cohort effects by level of human development H0: Cohort effects are jointly insignificant

p-value Very high human development 0.000 High human development 0.000 Medium human development 0.000 Low human development 0.624

233

Religious climate and religious identity

Table C3: Interacting religious climate with religious identity

Coefficients Intercept 52.267***

Individual Characteristics Gender Men (base) Women 5.814*** Birth Cohort 1990s or later (base) 1980s -3.164*** 1970s -6.274*** 1960s -8.860*** 1950s -9.932*** 1940s or earlier -14.029*** Non-response -9.362*** Religious identity Evangelical (base) Catholic 15.221*** Mainline Protestant 9.895*** Other 11.591*** Not religious 22.611*** Don’t know 13.343*** No response 9.935*** Importance of religion Very important (base) Rather important 4.555*** Not very important 6.247*** Not at all important 11.602*** Don’t know 2.198 No response 6.583*** Country characteristics HDI Very high human development (base) High human development -12.550*** Medium human development -16.730*** Low human development -32.684*** Religious climate Share of Evangelicals -0.037 Legal context Same-sex marriage (base) Recognise same-sex relationships -24.201*** Bans on same-sex intimacy removed -23.860*** Same-sex intimacy criminalised -29.993***

234

Table C3 cont’d Interactions Religious climate × Religious identity Catholic -0.281*** Mainline Protestant 0.222*** Other -0.166** Not religious -0.585*** Don’t know -0.383** No response -0.218*

Log likelihood -249,042.880 Note: ***, **, * and + indicate significance at the 0.1, 1, 5 and 10 per cent levels, respectively

Table C4: Significance of religious identity variable by religious climate H0: Religious identities are jointly insignificant

p-value Low share of Evangelicals (1 s.d. above mean) 0.000 Regional average 0.000 High share of Evangelicals (1 s.d. below mean) 0.000

235

Legal context and cohort

Table C5: Interacting legal context with birth cohort

Coefficients Intercept 61.678***

Individual Characteristics Gender Men (base) Women 5.933*** Birth Cohort 1990s or later (base) 1980s -2.076+ 1970s -9.496*** 1960s -11.493*** 1950s -13.498*** 1940s or earlier -20.874*** Non-response -12.568+ Religious identity Evangelical (base) Catholic 8.012*** Mainline Protestant 3.642*** Other 6.686*** Not religious 8.461*** Don’t know 3.318* No response 3.708** Importance of religion Very important (base) Rather important 4.634*** Not very important 6.491*** Not at all important 12.766*** Don’t know 3.219 No response 6.989*** Country characteristics HDI Very high human development (base) High human development -12.630*** Medium human development -16.873*** Low human development -32.273*** Religious climate Share of Evangelicals -0.268* Legal context Same-sex marriage (base) Recognise same-sex relationships -22.271*** Bans on same-sex intimacy removed -27.633*** Same-sex intimacy criminalised -34.735***

236

Table C5 cont’d Interactions Recognise same-sex relationships × 1980s -5.011** Recognise same-sex relationships × 1970s -0.950 Recognise same-sex relationships × 1960s -3.355+ Recognise same-sex relationships × 1950s -6.096** Recognise same-sex relationships × 1940s or earlier 0.228 Recognise same-sex relationships × No response -4.603 Bans on same-sex intimacy removed × 1980s -0.688 Bans on same-sex intimacy removed × 1970s 4.018** Bans on same-sex intimacy removed × 1960s 3.418* Bans on same-sex intimacy removed × 1950s 4.584** Bans on same-sex intimacy removed × 1940s or earlier 10.453*** Bans on same-sex intimacy removed × No response 3.162 Same-sex intimacy criminalised × 1980s -1.339 Same-sex intimacy criminalised × 1970s 4.795** Same-sex intimacy criminalised × 1960s 4.430** Same-sex intimacy criminalised × 1950s 7.474*** Same-sex intimacy criminalised × 1940s or earlier 16.426*** Same-sex intimacy criminalised × No response 8.012

Log likelihood 249,052.850 Note: ***, **, * and + indicate significance at the 0.1, 1, 5 and 10 per cent levels, respectively

Table C6: Significance of cohort effects by legal context H0: Cohort effects are jointly insignificant

p-value Same-sex marriage 0.000 Recognise same-sex relationships 0.000 Bans on same-sex intimacy removed 0.000 Same-sex intimacy criminalised 0.000

237

Legal context and religious identity

Table C7: Interacting legal context with religious identity

Coefficients Intercept 46.051***

Individual Characteristics Gender Men (base) Women 5.787*** Birth Cohort 1990s or later (base) 1980s -3.252*** 1970s -6.326*** 1960s -8.899*** 1950s -10.023*** 1940s or earlier -14.076*** Non-response -9.546*** Religious identity Evangelical (base) Catholic 22.534*** Mainline Protestant 13.340*** Other 15.110*** Not religious 29.125*** Don’t know 20.385*** No response 14.634*** Importance of religion Very important (base) Rather important 4.507*** Not very important 6.263*** Not at all important 11.359*** Don’t know 3.147 No response 6.822*** Country characteristics HDI Very high human development (base) High human development -12.906*** Medium human development -17.384*** Low human development -32.565*** Religious climate Share of Evangelicals -0.267* Legal context Same-sex marriage (base) Recognise same-sex relationships -16.060*** Bans on same-sex intimacy removed -9.809** Same-sex intimacy criminalised -15.055**

238

Table C7 cont’d Interactions Recognise same-sex relationships × Catholics -8.874*** Recognise same-sex relationships × Mainline Protestant -9.712** Recognise same-sex relationships × Other -5.889+ Recognise same-sex relationships × Not religious -9.717*** Recognise same-sex relationships × Don’t know -4.142 Recognise same-sex relationships × No response -4.386 Bans on same-sex intimacy removed × Catholic -16.074*** Bans on same-sex intimacy removed × Mainline Protestant -10.904*** Bans on same-sex intimacy removed × Other -7.529*** Bans on same-sex intimacy removed × Not religious -21.307*** Bans on same-sex intimacy removed × Don’t know -18.587*** Bans on same-sex intimacy removed × No response -9.422** Same-sex intimacy criminalised × Catholics -16.117*** Same-sex intimacy criminalised × Mainline Protestant -9.901*** Same-sex intimacy criminalised × Other -10.281*** Same-sex intimacy criminalised × Not religious -30.534*** Same-sex intimacy criminalised × Don’t know -20.655*** Same-sex intimacy criminalised × No response -14.253***

Log likelihood 248,918.4500 Note: ***, **, * and + indicate significance at the 0.1, 1, 5 and 10 per cent levels, respectively

Table C8: Significance of religious identity by legal context H0: Religious identities are jointly insignificant

p-value Same-sex marriage 0.000 Recognise same-sex relationships 0.000 Bans on same-sex intimacy removed 0.000 Same-sex intimacy criminalised 0.000

239