Results from the Bivariate and Multivariate Analysis of Homonegativity

Abstract

Our estimates of global homonegativity based on rates of disapproval of homosexual neighbours cover more than 90% of the world's population and are based on research results from 88 countries covert in the World Values Survey and the European Values Study. These data give the population-weighted result of a global percentage of homonegative people of around 55%. We present our research findings on the drivers of homonegativity using correlations, partial correlations and factor analysis, examining the effects of individual global attitudes on secularism, democracy, tolerance and religious particularism, as well as on gender equality, religion, political violence and national resilience on homonegativity, as evidenced in the global attitude data from the World Values Survey. We also analyse the relationship between homophobia and cross-social science data in the countries of the world.

In the following, we will present the results of our quantitative investigations.

5.1 Homonegativity According to the World Values Survey and European Values Survey Data

Our data on homonegativity, i.e., the rejection of homosexual neighbours, are based on the representative data of around 90 percent of the global population. The data are listed in Appendix 3, Appendix 4, and Appendix 5 of this work. Our results suggest that the percentage share of homonegative respondents is 54,9% of the global population. A choropleth map of data from the World Values Survey, 2017–2022, and 2010–2014 again shows the extent of the problem. Rejection rates of gay neighbours are relatively low in the “Global North” and relatively high in the “Global South” and in the (former) Communist countries. Figure 5.1 summarizes the results from Appendix 3.

Fig. 5.1

Homonegativity: Rejection of homosexual neighbours in the world and in the Euro-Mediterranean area. 0 = 0%; 1.0 = 100%

In terms of rejection of homosexual neighbours, the twenty most tolerant countries in the world are Iceland, Denmark, Norway, Sweden, Netherlands, Andorra, Switzerland, United Kingdom, Argentina, France, Germany, Brazil, New Zealand, Australia, Canada, Finland, Austria, Italy, Portugal, and Spain. At the bottom of the list, the most homonegativity countries are Jordan, Burma (Myanmar), Azerbaijan, Zimbabwe, Nigeria, Armenia, Maldives, Egypt, Qatar, Burkina Faso, Morocco, South Korea, Bangladesh, Ghana, Türkiye (Turkey), El Salvador, Rwanda, Uganda, Vietnam and Moldova.

A stark North–South and North-East divide in today's global society is evident.

5.2 A Kuznets Curve of Homonegativity?

Economics Nobel Laureate Simon Kuznets, in 1955, proposed in one of the most influential articles ever written in the social sciences,¹ that economic inequality first increases and later decreases with socio-economic development (Anand et al., 1993; Kuznets, 1955). A very large number of societal problems nowadays are being explained by such a trade-off between development level and a societal process to be explained (Acemoglu et al., 2002; Dinda, 2004), and it is no wonder that also the rejection of homosexual neighbours in the world can be neatly predicted by the UNDP Human Development Index and its square. In the comparative social science literature, there even has been already a suggestion to talk about a “Gender Kuznets Curve” (Eastin et al., 2013). All these attempts start from the common denominator that social crises culminate at the middle-income level. But Ronald F. Inglehart thought all along in his numerous works, also discussed in Chap. 3 of this study, that with growing existential security the acceptancy of homosexuality will increase in a rather linear fashion. But on closer inspection, homonegativity, like economic inequality and social conflict, indeed increase at higher levels of Human Development, only to decrease at higher levels of the UNDP Human Development Index. The Kuznets curve of homonegativity explains none the less than 51.8% of the total variance of the rejection of homosexual neighbours in the world (Fig. 5.2).

Fig. 5.2

Rejection of homosexual neighbours in the world, predicted by the UNDP Human Development Index

5.3 The Bivariate Correlations of Homonegativity at the Individual Level Among the Total Global Population and Among the Major Religious Denominations

After having established this important methodological principle of the Kuznets Curve of Homonegativity, we now begin the round of the presentation of our quantitative bivariate and multivariate results with Table 5.1, which presents the bivariate correlations of homonegativity at the individual level for the world population.

Table 5.1 The bivariate correlations of homonegativity at the individual level among the total global population

The focus here is on the variables of religious salience and religious attendance as explanatory variables, absolutely in line with the research results reported by Janssen and Scheepers (2018) (see Chap. 3, above). Although there are quite significant correlations, none of the correlations shown for the citizens of the world system explains more than 10% of the variance of homonegativity. The conclusion is therefore that at the global level, although there is a statistically significant but quantitatively rather weak correlation between various aspects of religiosity and homonegativity, and nowhere is homonegativity influenced by religiosity by more than 10%. In the following Tables 5.2, 5.3 and 5.4 we now pursue the same question as in Table 5.1 at the separate level of global Roman Catholics, Muslims and Orthodox Christians. Our research results run counter to most assumptions which we discussed above in our Chap. 3 on the hitherto existing studies, and for the Muslim sample, for example, the correlations are even lower than for the global citizens of the world system. Religiosity is not a reliable predictor to explain Muslim homonegativity. This means nothing other than that the homonegativity of Muslims is even less dependent on basic religious convictions than for the other citizens of the world system. A similar statement can also be confirmed for global Orthodoxy. Thus, it is not religious salience, or religious attendance, but socio-cultural traditions that statistically determine homonegativity in those important subsamples of world society (Tables 5.3 and 5.4).

Table 5.2 The bivariate correlations of homonegativity at the individual level among the global Roman Catholics

Table 5.3 The bivariate correlations of homonegativity at the individual level among the global Muslims

Table 5.4 The bivariate correlations of homonegativity at the individual level among the global orthodox

5.4 The Bivariate Correlations of Homonegativity at the Individual Level

In Table 5.5 we now look at the correlations of homonegativity with other phobias for the citizens of the world system. Homonegativity has the strongest statistical correlation with the rejection of neighbours who have AIDS, as well as with the rejection of neighbours who are drug addicts or who have been convicted of crimes. Interestingly, the next highest correlations are observed with the rejection of unmarried couples living together, with the rejection of emotionally unstable people, and with the rejection of alcoholics.

Table 5.5 The bivariate correlations of homonegativity of the global population at the individual level with variables of xenophobia and racism

On the scale of the strength of the determinants of the rejection of homosexuals as neighbours, the rejection of right-wing radical neighbours, the rejection of Jewish neighbours, the rejection of political extremists as neighbours, the rejection of Muslims as neighbours, the rejection of guest workers, the rejection of Christian neighbours and the rejection of immigrants follow. Next in the list is the rejection of Christian neighbours, the rejection of neighbours of a different religion, the rejection of neighbours of a different race, the rejection of Roma neighbours, the rejection of neighbours speaking a different language, the rejection of members of a militant minority and the rejection of Hindu neighbours. This investigation was carried out with the WVS_Longitudinal_1981_2016_Spss_v20180912.sav data set, and our readers will immediately notice that the sample sizes in the chosen WVS data set vary considerably.

Table 5.6 now gives an overview of the significant bivariate correlations of homonegativity at the global level of respondents to the World Values Survey with other key variables of the World Values Survey. As with all correlations based on analyses of the results at the individual level, the correlation coefficients hardly reach more than plus or minus 0.25 and many of the significant correlations are even below or far below in strength; that they are nevertheless significant is due to the large number of observations, some of which exceed 300,000 respondents.

Table 5.6 The bivariate correlations of homonegativity of the global population at the individual level with key variables from the World Values Survey Project

The only two correlations with the rejection of homosexual neighbours that are greater than plus or minus 0.20 are the rejection of the thesis that the only acceptable religion is one's own (working in the direction of homo-positivity) and the rejection of the thesis that politicians who do not believe in God are not suitable for public office (also working in the direction of homo-positivity). Both correlations show that homophobia and an exclusivist interpretation of one's own religion (= religious particularism) are more strongly connected. I leave the further correlations with religiosity and understanding of democracy to the readers of this publication. They speak for themselves. For reasons of the clarity of the presentation, we mention only correlations greater or equal plus minus 0.10. In any case, the correlations are far lower than is often suggested in public discussion.

5.5 Partial Correlations of Homonegativity at the National Level

In the following, we now use the technique of partial correlation analysis, presented in all detail in Tausch et al., (2014). As we have shown in Chap. 3, a considerable part of the research on global values assumes with the great American sociologist Ronald F. Inglehart that values in a society change with the achievement of existential security. To test these effects, in Table 5.7 we have now appropriately held constant the United Nations Human Development Index, which maps existential social security very well, and we used the non-linear formulation presented in Sect. 5.2 of this work, i.e., the Kuznets Curve of Homonegativity. So, we look at the partial correlations of the rejection of a homosexual neighbour in the societies of the world, regardless of whether a society is rich or poor. Restrictive social gender norms, as evidenced by the UNDP’s Gender Social Norms Index (GSNI), which measures how social beliefs obstruct gender equality in areas like politics, work and education. The UNDP GSNI presents stunningly high partial correlations with homonegativity. It is now interesting to note as well that also restrictions against the Jewish population surveyed in the very encompassing study by Fox and Topor, as well as anti-Semitism as expressed in the ADL's anti-Semitism index (ADL 100), correlate very strongly with the rejection of homosexual neighbours, and that the violation of civil and political rights, the proportion of Muslims in the total population, the membership of a country in the Islamic Cooperation and the proportion of Orthodox Christians in the total population correlate strongly with negativity towards homosexuals, holding the society's level of affluence correspondingly constant.

Table 5.7 Partial correlations of homonegativity in the countries of world system with key socio-economic country-level indicators constant: HDI 2018 & HDI (2018)^2, latest edition of the World Values Survey

Continuing with the series of remarkable significant partial correlations, we can clearly conclude that educational indicators, indicators of social globalisation, indicators of population satisfaction as surveyed by Gallup, world-class universities, proportion of women in government, proportion of women in parliament, freedom from corruption, proportion of Protestants in the total population, a high employment rate, and the constitutionality of the social order are the best guarantees that a society is not characterised by homophobia. In our list of significant partial correlations, we should also draw special attention to the correlation of −0.45 with the proportion of Roman Catholics in the total population, as well as, of course, the weight of the variables measuring the realisation of democracy. Table 5.7 is thus presenting strong evidence in the direction of the hypothesis that LGBTQ + rights and the wellbeing of LGBTQ + communities are best guaranteed by the free, democratic and open societies of the Western World.

5.6 Partial Correlations of Homonegativity at the Individual Level of the Inhabitants of the World

Table 5.7 was based on national aggregate data and its results were collected by using evidence about countries at the national, aggregate level. The logic of all this can be summarized as follows: X levels of globalisation of a given country, are associated with Y levels of homonegativity in each country. In Table 5.8 we now look at the partial correlations at the individual level according to the data from the 2010–2014 wave of the World Values Survey. Does the opinion of the respondents of the World Values Survey that for example it is an important characteristic of a democracy that criminals are severely punished, or that in a democracy the army takes over when government is incompetent, lead to higher or lower rates of homonegativity?

Table 5.8 Partial correlations of homonegativity among the global population with key indicators from World Values Survey (2010–2014), constant: Age and Sex and Highest educational level attained

We have chosen this dataset because it tends to provide significantly more religion-specific explanatory variables than the latest available wave of the World Values Survey. As mentioned above, it would be again a miracle if the correlations were greater than plus or minus 0.25. This time we have held constant a respondent's age, gender and the education level in the partial correlations at the individual level. In the result list, reported here, we exclude correlations smaller than plus minus 0.10.

It again can be shown that the highest positive correlations of homophobia occur with variables that involve a very specific and restrictive interpretation of religious traditions, as correctly predicted by Janssen and Scheepers (2018). The highest positive partial correlation of disapproval of a homosexual neighbour is achieved with over 14,000 respondents with the agreement by the respondents to the obligation to veil women. In second place is agreement with the proposition that religious authorities should interpret the laws in a democracy. The round of influencing factors blocking homophobia opens with rejection of polygamy at −0.24 (for more than 17,000 respondents). In second place is the rejection of the opinion that it is better for people with strong religious beliefs to hold public office exclusively, followed by the rejection of sexism against women's activities in the business world and in the world of politics, followed by the rejection of the opinion that one's religion is the only acceptable religion. Rejection of the opinion that politicians who do not believe in God are unfit for public office and rejection of the opinion that a woman must obey are also related to homophobia by more than −0.20.

5.7 Secularism, Democracy, Tolerance and Religious Particularism: Promax Factor Analytical Results for the Explanation of Homonegativity based on the World Values Survey, 2010–2014

We now move to the higher stages of the multivariate analysis of homonegativity. We first investigate how

• Secularism

• Pro-Democracy support

• Religious tolerance, no restrictive gender norms

• Religious particularism.

Affect homonegativity. The database used was WVS_Longitudinal_1981_2016_Spss_v20180912.sav. The data cover around 68% of the global population.

So, this chapter examines the multivariate effects of secularism, pro-democracy attitudes, religious tolerance and non-restrictive gender norms and religious particularism on homonegativity. Table 5.9 first reports the significance criteria of our promax factor analysis. To report insignificant results it is essential to run counter to any standard of quantitative social science. The Kaiser–Meyer–Olkin measure of sampling propensity is well above 0.6, and Bartlett's test is also easily passed. Figure 5.3 shows the Eigenvalues of the factor analysis > 1.0, and four of these factors are suitable for further interpretation according to the so-called scree test, while the last resulting factor with an Eigenvalue > 1.0, which expresses the affirmation of state ownership, is no longer interpretable, since a clear kink in the line of Eigenvalues can only be observed with Factor 4, the factor “religious exclusivity or religious particularism”. Table 5.10 now shows the Eigenvalues and variance shares as well as the cumulative variance shares of the factors used in the final analysis: secularism, pro-democracy, religious tolerance, and religious particularism. Together, these four factors already achieve a total (and high) explained variance of no less than 46%.

Table 5.9 Promax factor analysis of the drivers of homonegativity with data from World Values Survey, Longitudinal_1981_2016: Tests of Significance

Fig. 5.3

Promax factor analysis of the drivers of homonegativity with data from World Values Survey, longitudinal 1981–2016: screetest

Table 5.10 Promax factor analysis of the drivers of homonegativity with data from World Values Survey, longitudinal 1981–2016: eigenvalues and explained variances

As in any factor analysis, the factor loadings of the structural matrix are now used for the substantive interpretation of the results and the reader is referred in this context to Table 5.11. It is emphasised that in any factor analytic study there is a degree of subjectivity in the naming of the factors.

Table 5.11 Promax factor analysis of the drivers of homonegativity with data from World Values Survey, longitudinal 1981–2016: loadings of the factor structure matrix

Rejection of homosexual neighbours is determined by secularism with −0.22, by attitudes towards democracy with −0.06, by religious tolerance and the absence of restrictive gender norms with −0.49, and by the presence of religious particularism with +0.16. We have marked all factor loadings above plus minus 0.5 accordingly in our table.

Table 5.12 now shows the correlations of the factor analytic components. The main determinants of homonegativity appear to be the lack of religious tolerance and restrictive gender norms.

Table 5.12 Promax factor analysis of the drivers of homonegativity with data from World Values Survey, longitudinal 1981–2016: component correlations

Table 5.13 and Fig. 5.4 now present the country results of our analyses.

Table 5.13 Promax factor analysis of the drivers of homonegativity with data from World Values Survey, longitudinal 1981–2016: factor scores

Fig. 5.4

Promax factor analysis of the drivers of homonegativity with data from World Values Survey, 2017–2022: screetest

So, as we already mentioned above, we first report (in Table 5.9) the reliable and satisfactory statistical properties of our model, and all tests of significance wielded good and acceptable results.

Figure 5.3 shows the results of the screetest, suggesting that our choice to interpret only four factors is correct.

As already mentioned, Table 5.10 lists the Eigenvalues and the explained variances of our model. With four interpreted factors, explained total variance is at 46%:

The main result of our investigation with the very inclusive data from the World Values Survey, Longitudinal 1981–2016, listed in Table 5.11 is that the following factors determine homonegativity. We list here, for reasons of the brevity of the presentation, the factors, and the size of their influence:

Religious particularism	0.167
Secularism	−0.223
Religious tolerance, no restrictive gender norms	−0.493

As already hinted at above, in Table 5.12, we show the factor component correlations for the following factors, which we think are important for a future, more thorough understanding of the empirics of global religiously motivated extremism. In Table 5.12, our readers see the correlations between the global citizens’ attitudes on

• Secularism

• Pro-Democracy

• Religious tolerance, no restrictive gender norms

• Religious particularism.

Table 5.13 now lists our country results, based on the promax factor scores, in the alphabetical order of their standard country or territory names.

Both the United States and, of course, Latin American countries such as Mexico, Ecuador, Colombia, Peru, and Brazil, and then, in Europe, Poland, Romania and Türkiye (Turkey), as well as virtually all of Africa and the Middle East, and also the Muslim-majority countries of Southeast Asia and Buddhist Thailand, show higher levels of religiosity, while Sweden, the Netherlands, Belgium, Spain and, of course, the former Soviet Union and especially China are bastions of secularism, as is Australia.

The countries where the population does not sufficiently identify with the support of democracy are to be found in Mexico, Colombia, Ecuador and further to the East on our globe, in South Africa and Nigeria, Jordan, Azerbaijan and Armenia, as well as in Pakistan and Central Asia. And in East Asia and Southeast Asia, the Philippines, South Korea and Malaysia have a very low support rate for democracy.

The areas of lack in religious tolerance and restrictive gender norms are found mainly in Peru; in the African countries for which data are available, in Türkiye (Turkey) and in the Middle East, in Uzbekistan and Pakistan, in India, Thailand, Malaysia and the Philippines, and in some other countries of the former Soviet Union.

Religious particularism is more pronounced in Latin America, Mexico, Peru and Uruguay, also in the Federal Republic of Germany, and in many countries in the Middle East, in Uzbekistan and in Pakistan, South Korea and in the People's Republic of China.

5.8 Gender Equality, Democracy, Religious Salience, Political Violence and National Resilience: Promax Factor Analytical Results for the Explanation of Homonegativity with the Data of the World Values Survey, 2017–2022

In the following multivariate analysis, based on data from the latest wave of the World Values Survey, we analyse the multi-variate relationships between respondents’ attitudes on gender justice, democracy, religion, xenophobia and homophobia, political violence and national resilience. The inclusion of national resilience corresponds well with the World Values Survey item “willingness to defend the country” and “confidence in the Armed Forces” at a time of heightened international tensions, wars and crises. Readers interested in the issues of national resilience are being referred to the recent study Tausch and Neriah (2023), which among others provides readers an in-depth analysis of the problems of the Middle East, the lack of resilience in key countries of the Western alliance, and the multivariate analysis of resilience with World Values Survey data.

Table 5.14 shows the results of the significance tests of our factor analytic procedure, and again the significance tests are fully in line with the research design. Table 5.15 lists the proportions of variance explained by the variables in the model, where, of course, a proportion of variance of 0% is written mathematically as 0.0 and a proportion of variance of 100% is written as 1.0. The two homophobia variables—rejection of homosexual neighbours and rejection of parenthood by homosexual parents—each are explained by more than 50% by our factor analytical model. Our Table 5.16 shows the Eigenvalues of the factor analytical model, and the percentage shares of the explained variance and the cumulative percentage shares of the explained variables of our six factors. The core of the factor analytical results is presented in Table 5.17, which, as before, shows the loadings of the factor structure matrix. In this Table, factor loadings greater than plus or minus 0.5 are again marked accordingly. Table 5.18 shows the results of the correlations between the components of our Promax factor analytic model. Our factors were labelled:

Table 5.14 Promax factor analysis of the drivers of homonegativity with data from World Values Survey, 2017–2022: tests of significance

Table 5.15 Promax factor analysis of the drivers of homonegativity with data from World Values Survey, 2017–2022: extraction (explained variance; 0.0 = 0%; 1.0 = 100%)

Table 5.16 Promax factor analysis of the drivers of homonegativity with data from World Values Survey, 2017–2022: eigenvalues and explained variances

Table 5.17 Promax factor analysis of the drivers of homonegativity with data from World Values Survey, 2017–2022: loadings of the factor structure matrix

Table 5.18 Promax factor analysis of the drivers of homonegativity with data from World Values Survey, 2017–2022: component correlations

• acceptance of gender justice,

• pro-democracy trend,

• religious salience,

• homophobia and xenophobia,

• advocacy of political violence, especially among the younger generation and

• willingness to defend the home country.

The homophobia and xenophobia factor correlates +0.28 with the religious salience factor and +0.13 with the willingness to defend the country (national resilience factor). Ronald Inglehart's assumption (see also Tausch & Neriah, 2023 for a detailed survey of the literature) was that patriotism stems from a social model that is more likely to be found in the poorer societies of the world where traditional gender norms still prevail; homophobia and xenophobia correlate with the acceptance of gender justice at −0.35 and with pro-democracy currents at −0.12.

Table 5.19 shows our sample sizes for our factor analysis.

Table 5.19 Promax factor analysis of the drivers of homonegativity with data from World Values Survey, 2017–2022: factor scores

The ten countries most strongly combining homophobic and xenophobic attitudes in the world are Myanmar, Azerbaijan, Montenegro, Bangladesh, Armenia, Maldives, Belarus, Bulgaria, Ethiopia, Nigeria, Russia, Serbia and Georgia, while the ten societies in the world with the lowest occurrence of homophobic and xenophobic attitudes are Iceland, Uruguay, Sweden, Norway, Denmark, New Zealand, Germany, Canada, Andorra, France, the United Kingdom, the Netherlands and Switzerland. Figure 5.4 now shows the results for acceptance of gender justice based on our factor analysis, and it is clear to see how low is the acceptance of gender justice in the countries of the former Soviet Union, as well as in the other countries of the world with a prolonged rule of a communist party, and unfortunately also in the countries influenced by a widespread “Islamic popular culture”, based on the restrictive reading of the Islamic traditions, quite against the original message of tolerance in Islam (see also, Solomon, 2016; Solomon et al., 2023; Solomon & Tausch, 2020a, b, c, 2021).

Indeed, our factor analysis shows that the pro-democracy factor combines the following variables:

• a democracy is characterised by people choosing their leaders in free elections;

• civil rights defend people's freedoms against oppression;

• democracy is characterised by women having the same rights as men; and

• the importance of democracy per se is emphasised.

The results in some Latin American countries, such as Mexico, Colombia, but also Ecuador and Chile, as well as in Kenya, Kazakhstan, Thailand, Malaysia and the Philippines, are particularly disappointing. According to our research, the most secular societies here are Sweden, the Netherlands, the Czech Republic and Japan.

So, our factors, explaining more than 60% of the total variance are:

• Accepting gender equality

• Pro-Democracy

• Religious salience

• Homophobia/Xenophobia

• Support for political violence among the younger generations

• Willingness to fight for the country.

The statistical properties of our model are listed in Table 5.14.

Table 5.14 lists the explained variances of each variable (0.0 = 0%; 1.0 = 100%).

Table 5.16 lists the Eigenvalues and the explained variances.

Figure 5.4 performs the screetest, suggesting that five factors should be interpreted.

Tables 5.17 and 5.18 list the main factor analytical results. The most important drivers of the Homophobia and Xenophobia factor and the size of their influence on the factor (factor loadings) are

Important child qualities: religious faith	0.176
Democracy: Religious authorities interpret the laws	0.176
Willingness to fight for country	0.108

The most important blocks against the Homophobia and Xenophobia factor and the size of their influence on the factor (factor loadings) are

Men don't make better political leaders than women do	−0.368
Men don't make better business executives than women do	−0.339
Not important in life: Religion	−0.246
University is equally important for a boy and for a girl	−0.212
Importance of democracy	−0.158
Democracy: Women have the same rights as men	−0.118

Our factor component correlations suggest the following very important theoretical connections. The rejection of neighbours: Homosexuals (=homonegativity) is determined by the following factors:

Religious salience	0.335
Willingness to fight for the country	0.198
Support for political violence among the younger generations	−0.001
Pro-democracy	−0.088
Accepting gender equality	−0.308

Homonegativity as the rejection of homosexual parenthood (homosexual couples are not as good parents as other couples) is being determined in the following fashion by the factors:

Religiou salience	0.424
Willingness to fight for the country	0.309
Pro-democracy	−0.061
Support for political violence among the younger generations	−0.223
Accepting gender equality	−0.292

Table 5.19 shows the factor scores for those countries in the world system for which data are available and the number of countries included in the study. At the top of the global homophobia and xenophobia list are Myanmar, Azerbaijan, Montenegro, Bangladesh, Armenia, Maldives, Belarus, Bulgaria, Ethiopia, Nigeria and Russia, Serbia, Georgia, Malaysia and Lithuania.

At this stage of our analysis, we also should emphasize that the 10 countries with the lowest acceptance of gender equality are Pakistan, Burma (Myanmar), Kyrgyzstan, Nigeria, Indonesia, Bangladesh, Libya, Philippines, Russia and Kazakhstan.

The ten countries with the highest acceptancy of gender justice were Norway, Sweden, Iceland, France, Denmark, Puerto Rico, Australia, Germany, Spain and New Zealand.

The ten countries with the lowest support for democracy are Malaysia, Thailand, Mongolia, Mexico, Philippines, Guatemala, Colombia, Kenya, Ecuador and Montenegro.

Support for democracy was strongest in Albania, Germany, Iceland, Denmark, Sweden, Ethiopia, Norway, Greece, Poland and Switzerland.

In times of rising international tensions, we also include—in Table 5.19—our evidence about national resilience, which was recently analysed in extenso by Tausch and Neriah (2023). The drama of the lack of national resilience in key Western countries might in future also affect the willingness of Western societies to stand up for Western tolerance towards the LGBTQ+ communities as an integral part of the Western lifestyle.