The limits of tolerance: before and after Brexit and the German Refugee Crisis

ABSTRACT

This comparative statistical analysis is of tolerance and how it was perceived and experienced by minorities and majorities in the UK and Germany before and after two defining national events that both focused on immigration and national identity. Based on combined sweeps of the European Social Survey (2012–2018), this study applied logistic regression modelling to determine that during the Brexit vote in 2016 and the German Refugee Crisis in 2015, majorities from lower socioeconomic backgrounds were more likely to be intolerant towards minorities, who were more likely to experience discrimination. Majority intolerance and minority perceptions of discrimination increased after both events among less well-off groups. However, majorities from higher socioeconomic backgrounds were less likely to exhibit intolerance towards minorities before these national events and after. This study offers an original perspective on race and ethnic relations at times of national crises over identity and belonging.

KEYWORDS:

• Tolerance
• Brexit
• Germany
• Refugee Crisis
• minorities
• majorities

Introduction

In a climate of growing xenophobia manifested by the resurgence of radical-right parties in Western Europe (Greven 2016), widespread anti-immigration sentiment (Czaika and Di Lillo 2018), and prevailing Islamophobia (Abbas 2019), tolerance appears to be in short supply. As one of the pillars of the liberal democratic creed, tolerance is understood as the rights of groups to participate fully in the political, socio-cultural, and economic spheres of society (Weldon 2006). Due to a general discourse of racialization of immigrants and minorities that pervades Western Europe, the rights of groups who are not native to local populations are increasingly questioned. Furthermore, the presence of these groups is viewed as illegitimate and as a threat to the wellbeing and security of majority populations. Differences in society are seen as risks, meaning that groups exhibiting these differences are seen as risky, all of which adversely impacts on the lives of minority groups. The radical right’s stance on policy issues, particularly immigration, tends to exert multi-level influence within the political system as the popularity of these parties grows (Schain 2006), particularly during times of crisis (Downes and Loveless 2018). The reemergence of nationalist and xenophobic influences through these parties’ exclusivist and inward-looking political outlook constitutes a threat to social cohesion and intergroup relations. Undesirable social changes, such as growing socioeconomic marginalization of parts of society and vilification of “others” by populist forces, lead to the diminishment of the political centre; this in turn polarizes public discourse and allows radicals to push their divisive message into the mainstream (Ali 2015). The impact of these changes is likely to be exacerbated in countries experiencing major socio-political events such as high levels of immigration and profound political transformations.

This study investigates how two social and political developments, in the UK and Germany, impacted on the experiences of minorities and the attitudes of majorities vis-à-vis tolerance in those two countries. The UK and Germany have both recently experienced major socio-political events – the Brexit referendum and the Refugee Crisis, respectively – that have strained existing intergroup relations. These two events occurred approximately at the same time and exhibited common social, political, and economic characteristics. Both Brexit and the Refugee Crisis in Germany were highly politically polarizing, and continue to produce far-reaching ripple effects for majority and minority populations alike. This study, therefore, strives to contribute to our understanding of intergroup relations as a result of unfolding social change and contemporary socio-political development, and how intergroup tensions are differentiated by socioeconomic and minority-majority status.

Migration, xenophobia, and racism

The increasing need for immigration since the end of World War II has led to high demand for research in questions of identity, citizenship and participation in society, but often to shape policy rather than understanding race relations (Bulmer and Solomos 2004). The perceptions of European majorities concerning newcomers were largely based on stereotypes derived from the legacy of colonialism. Orientalist notions of inferiority, primitiveness, and strangeness loomed prominently in how Western Europeans conceived of their new next-door neighbours. However, the economic boom and capital accumulation of Western European states during the 1950s and 1960s would not have been possible without migrant labour from the former colonial territories and European peripheries. These migrant workers were employed in low-skilled sectors that the indigenous populations vacated, primarily in manufacturing. Initial migration policies assumed that migrant workers would return to their countries of origin; however, with an extended period of economic growth and the accompanying high demand for labour despite deindustrialization and globalization, migrant workers began to settle, gaining residence rights and starting families. Subsequently, the economic crisis of the 1970s precipitated cuts to the welfare state and led to high levels of unemployment, with minorities over-represented in the jobless population. The scaling back of redistributive policies was implemented along ideological grounds, amongst them racism. Consequently, the recently settled immigrant populations were increasingly portrayed as a threat to the racial homogeneity of the nation, with immigrant groups being framed as illegitimate competitors for access to scarce resources (Miles 1986).

Newly settled ethnic minorities faced numerous forms of systemic discrimination and disadvantage because of their “otherness” in the eyes of the majority. Beyond the issue of discrimination, ethnic minorities were also pressured to assimilate and adopt the norms and values of the majority population. The onus for integrating was put squarely on the ethnic minorities themselves, with a lack of integration presumed to indicate an unwillingness to participate in society. However, such charges missed the duality of precariousness faced by minority groups, who on the one hand are expected to “be like us”, while on the other hand being discriminated against on the basis of race and viewed as unfit by majority society. Contemporary patterns of racism, discrimination, xenophobia, and anti-immigrant sentiment (particularly Islamophobia) have their roots in the historical differentiation and exclusion stemming from the colonial legacy of Western European states and the rise of modern capitalism (Wilson 1996).

The end of tolerance?

The present-day climate of xenophobia, rooted in racism and the perceived threat from ethnic minorities (especially Muslims) to the wellbeing and security of the nation, fosters the revival of ethnic nationalism and the ascendance of a new radical right. Deregulation and privatization of parts of the public sector, implemented in the spirit of the neoliberal economic transformation that began in the 1980s, have led to profound societal repercussions; these include high levels of social inequality driven by the logic of competition in the “enterprise society” (Lazzarato 2009). Globalization has also been a contributing factor to the electoral success of radical-right parties (Dippel, Gold, and Heblich 2015; Autor et al. 2016; Colantone and Stanig 2018), especially in states that do not provide strong welfare support (Swank and Betz 2003). The reductions to social spending, combined with rapid technological transformation of the economy or a prevailing anomie fuelling a national identity crisis, have left parts of society feeling “left behind” (Ford and Goodwin 2014). The erosion of norms that underpin social coherence leads to a sense of existential uncertainty as societal foundations no longer conform to expectation. The neoliberal economic transformation of the 1980s prioritized competition over solidarity, resulting in a social change towards negative solidarity that has further exacerbated the loss of meaning and the crisis of identity. The costs of aggravating anomie lead social subjects to search for an exit, which may manifest itself in violent attitudes towards others, as seen in the case of xenophobia reinvigorating radical-right forces across the continent (Burgi 2014). Marginalized majorities, mostly white working classes, feel dispossessed by the rapid social change and the lack of opportunities for progress and development, which prompts them to seek scapegoats onto whom they can project their anger. Through populism, and more recently nativism, radical-right parties have managed to exploit that anger and transform it into a viable political force (Vieten and Poynting 2016). Populist nativism appears to be a defining characteristic of these parties, which see the world as divided into “us” and “them”; that is, a notion of “pure people” who are exploited by the “corrupt elites” and “others” who are not the same as “us”, the “ethnically pure natives” (Greven 2016). Numerous factors are contributing to the revival of the radical right in Western Europe (Rydgren 2007) but anti-immigration hostility is often cited as a strong predictor of support for such parties (Ivarsflaten 2008). Ongoing social research in this area continues to indicate an association between socioeconomic background and attitudes towards immigrants in Europe (Mewes and Mau 2013). Typically, groups with higher status, denoting tertiary level education, occupational prestige and high income, are likely to have higher levels of tolerance than those with lower status (Katnik 2002). Thus, the relative status of individuals within a group has a bearing on their attitude towards others. This is especially relevant in the context of majority-minority relations in which power marks the prevailing level of tolerance existing between them (Van Doorn 2014).

Two recent socio-political events are particularly relevant in the context of both growing popular discontent among majorities and the vilification and “otherization” of minorities: the 2016 Brexit referendum in the UK, and the 2015 Refugee Crisis and its impact on Germany in particular. Brexit was largely driven by feelings of nostalgia, articulated by the Leave campaign through slogans such as “take back control”. This sense of melancholia operated on multiple levels, with immigration playing a prominent role. Immigrants were portrayed as the “other” that constituted primarily an economic and security problem, impacting living standards and undermining the cohesion of the British nation. In this way, the Leave campaign strived to obfuscate the discourse on immigration, diverting it away from purely racial connotations while aiming to achieve the same divisive outcome in the minds of the general population, rendering the non-British – meaning non-white non-Christian, or those not subscribing to the “British way of life” – as undesirable elements in their narrowly defined conception of the British community (Virdee and McGeever 2018). The post-2004 surge in immigration to the UK, particularly from Eastern Europe, contributed to unprecedented levels of net migration; this, along with the 2015 Refugee Crisis which preceded the Brexit referendum, compounded anxieties about the changing ethnic and cultural composition of the UK. It also drove support for Brexit, particularly in the communities that experienced high levels of social and cultural change (Goodwin and Milazzo 2017). Moreover, Brexit enjoyed most support amongst communities with a high proportion of pension-age, low-skilled, and/or low-educated citizens – citizens whose profile most resembles the so-called “left-behind” groups, for whom the referendum constituted an outlet for their angst about growing marginalization (Goodwin and Heath 2016). These anxieties were successfully exploited by the populist radical-right, particularly the UK Independence Party (UKIP), which emerged as a significant political force in the debates leading up to the referendum.

Germany, through its decision to adopt an “open door” policy towards refugees fleeing violence and instability in the Middle East, welcomed more than one million asylum seekers during the 2015 Refugee Crisis. While most European countries, particularly the Eastern European states, were closing their borders, Germany exhibited unparalleled solidarity with the refugees, who largely Syrians. Although anti-immigrant and xenophobic sentiments, particularly Islamophobia, were on the rise in Germany, Chancellor Angela Merkel made a politically difficult decision to allow refugees to enter the country. Before the crisis, more than 60% of Germans believed Muslims to be sympathetic to terrorism and more than 50% already believed that there were too many Muslims in Germany. The highest degree of prejudice towards foreigners was noticeable in former East Germany, where the population who had lived under communism did not undergo reconciliation with past racist crimes in the same way as former West Germany (Adam 2015). The decision by Chancellor Merkel led to far-reaching societal and political repercussions for Germany. For instance, between 2014 and 2015, Germany saw a sharp increase in violence perpetrated against refugees, with a disproportionately large number of anti-immigrant and anti-refugee attacks carried out in former East Germany (Bencek and Strasheim 2016). The influx of refugees also adversely impacted the lives of existing, well-established minority communities in Germany, as a climate of heightened ethnic nationalism facilitated the “all foreigners are alike” discourse (Sadeghi 2019). Moreover, the arrival of refugees boosted the profile of the populist radical-right party, Alternative for Germany (Alternative für Deutschland, AfD). The AfD managed to leverage anti-immigrant sentiment in Germany and considerably increased its presence on the German political scene following the 2017 federal elections (Lees 2018).

Methodology

This study aimed to examine the nature of intolerance in the UK and in Germany before and after the Brexit vote and the German Refugee Crisis that occurred approximately at the same time. In what follows, further details are provided about the type and source of the data, as well as the operationalization of dependent and independent variables for the logistic regression models.

Data

This study examines European Social Survey (ESS) data from the UK and Germany, two countries that in the recent past have experienced acute political challenges. Both countries are categorized as high income; around the height of the discussed events in 2015, GNI per capita amounted to $41,550 in the UK and $48,770 in Germany (World Bank 2019), with a GINI coefficient of 0.293 in Germany and 0.360 in the UK, one of the highest in the EU (OECD 2020). These two nations have also experienced high levels of net migration for many consecutive years; between the 2004 EU enlargement and 2018, average net migration per year totalled 251,733 in the UK (Sumption and Vargas-Silva 2020) and 291,321 in Germany (Statistisches Bundesamt n.d.). In both countries, the issue of immigration became a source of political mobilization for populist radical-right parties, with UKIP leading the charge in the UK and AfD capitalizing on Chancellor’s Merkel “open door” policy towards refugees by stirring up anti-immigration sentiments in Germany.

Data for the statistical analysis have been derived from the European Social Survey (ESS). The ESS is a biannual cross-national academic survey that measures changes in attitudes, beliefs, and opinions of populations living across Europe. Four datasets from UK and Germany were utilized for this study, divided into two periods: pre-event (i.e. rounds 6 and 7, conducted in 2012 and 2014 respectively) and post-event (i.e. rounds 8 and 9, conducted in 2016 and 2018 respectively). The sampling strategy employed by ESS is a strict random probability approach with several conditions. Data collection entails face-to-face interviews with respondents above fifteen years of age from different socioeconomic backgrounds. A detailed description of the data collection methodology employed by the ESS is accessible on its website.¹ The ESS ensures reliability and validity of its survey by controlling for several threats such as measurement error, nonresponse bias (with minimum 70% response rate per country), and interviewer behaviour. It is measured equivalently in all countries under consideration, thereby allowing for unique insights into both UK and German populations preceding and following Brexit and the Refugee Crisis.

Dependent variables

This study looks at the experiences of minorities and the attitudes of majorities concerning tolerance. The dataset was divided into minority and majority responses using the question “Do you belong to a minority ethnic group?” The dependent variable for the minorities is based upon yes/no questions concerning perceived discrimination (such as “On what grounds is your group discriminated against?”), with attention to discrimination on racial and ethnic grounds. The dependent variable for the majorities was derived from two questions relating to immigration, namely (i) “To what extent do you think [country] should allow people of the same race or ethnic group as most of [country]’s people to come and live here?” and (ii) “How about people of a different race or ethnic group from most [country] people?”, both measured on a nominal scale from 1 (“Allow many to come and live here”) to 4 (“Allow none”); together, these have been operationalized into the construct “Tolerance” (Cronbach’s alpha α = 0.847). Subsequently, to conduct a logistic regression analysis, the “Tolerance” variable was transformed into a dichotomous variable of either “Tolerance” (1 “Allow many to come and live here” and 2 “Allow some”) or “Intolerance” (3 “Allow a few” and 4 “Allow none”). It is important to note that this analytically derived concept of tolerance is used as a normative benchmark in an attempt to capture changing sentiment, and does not fully contextualize the concepts of race and racialization that are deeply embedded in any operationalization of tolerance (Table 1).

Table 1. Descriptive statistics of dependent and a number of independent variables for majorities and minorities in UK and Germany.

Variables		Majorities %		Minorities %
		UK	DE	UK	DE
Tolerance / Perceived Discrimination	Yes	66.6	89.4	25.6	12.2
	No	33.4	10.6	74.4	87.8
Oesch 5-Class schema	Higher-grade service class	19.3	20.3	21.6	19.3
	Lower-grade service class	17.7	22.4	15.6	10.7
	Small business owners	13.4	9.3	13.3	11.7
	Skilled workers	27.9	34.4	24.8	32.4
	Unskilled workers	21.7	13.6	24.7	25.9
Domicile	A big city	7.2	14.7	29.5	27.8
	Suburbs	19.3	14.3	27.6	13.4
	Small city	47.5	37.8	34.4	39.8
	Rural	26.0	33.2	8.5	19.0
Age	15–19	3.3	6.4	2.9	10.3
	20–24	4.2	5.7	6.1	12.0
	25–29	5.4	6.1	9.6	10.3
	30–39	14.4	12.4	27.5	24.6
	40–54	24.0	26.5	28.4	24.3
	55+	48.7	42.9	25.4	18.5
Gender	Male	44.4	51.1	45.3	55.6
	Female	55.6	48.9	54.7	44.4
Period	Pre-event	47.9	54.2	52.4	43.3
	Post-event	52.1	45.8	47.6	56.7
European unification	Yes	32.6	–	–	–
	No	67.4	–	–	–
Germany	East	–	17.1	–	–
	West	–	82.9	–	–
N		7946	10561	716	633

Source: European Social Survey 2012, 2014, 2016 and 2018.

Independent variables

The statistical models attempt to measure the effect of the socioeconomic position of respondents on their likelihood of belonging to a minority (i.e. one of the groups that are either racially and ethnically discriminated against), and their likelihood of being tolerant or not in the case that they are a member of a majority. In addition, the influence of age, gender, domicile location, and the fixed effect of period on the outcome variable are also measured. The socioeconomic positions here represent an amalgamation of education, income, and employment status outcomes. Unequal socioeconomic positions indicate asymmetrical access to power, privilege, and authority. Socioeconomic position correlates with low educational achievement, poverty, and ill-health, all of which have implications for wider society. The respondent’s socioeconomic position is derived from the Oesch class schema developed by Daniel Oesch (n.d.), the syntax of which is available on the author’s website.² The Oesch class schema is a revised version of the EGP (Erikson–Goldthorpe–Portocarero) schema, previously dominant in European sociology, which takes into consideration widespread occupational trends existing up to the mid-1970s that are characteristic of the era of deindustrialization. The Oesch class schema incorporates trends that have emerged since then and which are typical of post-industrial societies, such as those exhibiting growth of the service sector, an expanded welfare state, and the increased participation of women in the labour market (Oesch 2006). Thus, while the EGP schema is largely rooted in vertical criteria based on advantage coupled with a set of marketable skills that characterize the employment relationship (Erikson and Goldthorpe 1993), the Oesch schema adds a horizontal criterion derived from the concept of “work logic”, which encapsulates differences between occupations (Oesch 2006). In taking into account the variability in the expanding middle class, the concept of work logic entails four distinctions of an employee: the technical work logic, the organization work logic, the interpersonal work logic, and one that includes employers and the self-employed, namely the independent work logic. Each work logic includes a vertical hierarchy that distinguishes between the middle class (professional and associate professional) and the working class (vocational and unskilled) occupations. The resulting class distinction is the 17-class schema which can be collapsed into the 8-class and the 5-class schema, the latter of which has been employed in the statistical models used in this study.

The model for majorities also investigates two additional effects. In the case of the UK, the model includes a variable concerning the respondent’s opinion on further European unification. This “European unification” variable is derived from responses on a Likert scale from 0 (“Unification has already gone too far”) to 10 (“Unification should go further”), which have been recoded into a dichotomous variable with the levels “No” (responses from 0 to 4) and “Yes” (responses from 6 to 10). In the case of Germany, the variable “Germany” is added using the ESS variable “Region”, which is recoded into a dichotomous variable based on the pre-reunification divisions of the German Democratic Republic (East) and Federal Republic of Germany (West), excluding Berlin which has not been included in the analysis.

The statistical model for minorities aims to generate an understanding of how minority groups in both the UK and Germany experience intolerance, understood as their everyday experience with discrimination. Logistic regression was used to ascertain the effect of socioeconomic position of respondents on the likelihood of belonging to a group that has experienced discrimination. The notion of the so-called “left-behind” – that is, marginalized working-class majorities – is employed here. The aim is to examine whether “left-behind” minority populations are more likely to be experiencing discrimination. This model is consequently guided by the following question: How do minorities in the UK and Germany experience intolerance, and how did the events discussed above impact on those experiences? Given the patterns of racial and cultural discrimination affecting ethnic minorities in Western Europe, as well as prevailing anti-immigration sentiments in recent years, this study hypothesizes that:

H₁: The greatest levels of racial and ethnic discrimination are experienced by the least socially mobile segments of minority populations.

H₂: Minorities experienced greater intolerance in the post-event period than pre-event.

The attitudes of majorities vis-à-vis tolerance are analyzed to determine the impact of Brexit and the Refugee Crisis on the attitudes towards newcomers and non-natives. Moreover, the socioeconomic position of majority respondents is utilized to gauge the variability amongst the different classes concerning tolerance. The model for majorities is guided by the following research question: How do majorities perceive tolerance, and how did their attitudes towards tolerance change as a result of the two above-discussed events? Based on the reviewed literature on the subject, it is possible to derive the following hypotheses:

H¹: The more a respondent resembles the economic profile of the so-called ‘left-behind’, the more likely they are to be intolerant.

H²: Majorities show greater intolerance in the post-event period than pre-event.

H³: UK majorities opposed to further European unification are more likely to be intolerant than majorities favoring further unification of Europe.

H⁴: German majorities located in former East Germany are more likely to be intolerant than majorities located in former West Germany.

Results

The results of the logistic regression for minorities in both the UK and Germany are shown in Table 2. For the UK, the results reveal that the four class positions above “unskilled workers” are generally less likely to experience discrimination. The higher-grade service class is 36% less likely to feel racially or ethnically discriminated against than unskilled workers, whereas small business owners are 56% less likely, and skilled workers are 36% less likely. This pattern is identical for the lower-grade service workers, although results for this group are not statistically significant. Accordingly, minority workers who have little bargaining power given their lack of marketable skills are further disadvantaged by intolerance. There is also a visible divide between rural minority respondents and those from the city or suburbs, as the latter are markedly distinguished as having a higher likelihood of feeling discriminated than the former. Furthermore, there is a visible difference in the effect of the variable “age” for younger minority (between 15 and 24 years old) and medium-aged respondents (between 25 and 54 years old): the former appears to have higher odds of being discriminated against than those aged 55+, whereas the latter has lower odds; however, this trend does not reach statistical significance. Minority males are also 28% more likely to report experiencing intolerance than females.

Table 2. Logistic regression (odds ratios) for perceived discrimination in UK and Germany.

Variables	Model 1 – UK	Model 2 – Germany	Model 3 – Country comparison	Model 4 – Country comparison
Oesch 5-class schema: base – unskilled workers
Higher-grade service class	.64* (.007)	.71 (.233)	.70* (.011)	.70* (.009)
Lower-grade service class	.81 (.227)	.64 (.213)	.83 (.208)	.83 (.212)
Small business owners	.44** (.001)	2.50** (.001)	.85 (.256)	.85 (.246)
Skilled workers	.64* (.004)	.91 (.633)	.73* (.008)	.73* (.008)
Domicile: base – rural
A big city	2.16* (.005)	1.44 (.144)	1.88** (.001)	1.87** (.001)
Suburbs	1.96* (.015)	1.38 (.296)	1.66* (.007)	1.66* (.007)
Small city	1.72* (.046)	1.57 (.057)	1.57* (.011)	1.54* (.014)
Age: base – 55+
15–19	1.58 (.229)	1.77 (.171)	1.48 (.137)	1.47 (.150)
20–24	1.29 (.288)	2.03* (.020)	1.38 (.073)	1.36 (.086)
25–29	.79 (.244)	2.62* (.002)	1.12 (.512)	1.09 (.597)
30–39	.82 (.225)	2.99** (.001)	1.20 (.180)	1.19 (.192)
40–54	.98 (.898)	1.39 (.227)	1.05 (.734)	1.02 (.863)
Gender: base – Female
Male	1.28* (.033)	1.28 (.125)	1.27* (.010)	1.27* (.009)
Period: base – Pre-event
Post-event	1.82** (.001)	1.28 (.128)	1.59** (.001)	1.79** (.001)
Country: base – UK
Germany	–	–	.39** (.001)	.49** (.001)
Period * Country	–	–	–	.70 (.061)
Chi-square (df)	x^2(14) = 76.848 p<0.001	x^2(14) = 64.109 p<0.001	x^2(15) = 171.420 p<0.001	x^2(16) = 174.908 p<0.001
Nagelkerke	.06	.07	.08	.08
N	615	549	1164	1164

Note: *p < .05; **p < .001

Source: European Social Survey 2012, 2014, 2016 and 2018.

In Germany, the patterns relating to class positions appear to be, in general, similar to the UK, but with one exception. Minority small business owner respondents in Germany are 150% more likely to experience racial or ethnic discrimination than unskilled workers. This is an interesting outlier that raises questions concerning those workers who occupy the realm of the independent work logic. These workers, despite not being beholden to any corporate structure or needing to search for employment in the labour market, where they face higher chances of being discriminated against than the majority population (Kaas and Manger 2012), are more likely to experience intolerance than unskilled workers who are participating in the labour market. This is in contrast to the results from the UK, where the equivalent group does not report similar concerns to a comparable extent. Similarly to the UK, German minority respondents from cities or suburbs tend to have a greater likelihood of feeling discriminated against than respondents from rural areas, although this pattern is not statistically significant. In addition, German minority respondents between the ages of 15–54 are more likely to experience intolerance than the 55+ category. This is in opposition to the UK, where respondents in the 15–24 age group tend to have a higher likelihood of feeling discriminated, as opposed to the medium-age group, which has lower odds than the 55+ category. Thus, in both countries, it appears that younger minority respondents are particularly prone to experiencing discrimination. In terms of gender, males are more likely to feel racially or ethnically discriminated against than females.

To test hypothesis 1, logistic regression was carried out to ascertain the effect of Oesch 5-class schema on the outcome variable perceived intolerance based on racial and ethnic discrimination, in the UK and in Germany. Both logistic regression models are statistically significant (UK: x²(14) = 76.848, p<0.001, DE: x²(14) = 64.109, p<0.001), and explain 6% (UK) and 7% (DE) of the variance in the outcome variable. In both countries, the tentative results confirm hypothesis 1, whereby the least socially mobile respondents, namely minority working class and unskilled workers, who have little bargaining power stemming from their lack of marketable skills, tend to be more discriminated against than groups from the other social positions included in the statistical analysis. The intolerance that these respondents experience only compounds their marginalization and fuels social exclusion. The advent of the post-industrial economy which these respondents lack the skills to thrive in puts them at an acute disadvantage, which will be difficult to alleviate without adequate social policies.

When analyzing the effect of the Brexit and the Refugee Crisis on the experience of minorities vis-à-vis tolerance, the statistical analysis reveals that in both the UK and Germany, minority groups experienced greater intolerance in the post-event period. For the UK, where the effect is statistically significant, this result amounts to an 82% greater likelihood of belonging to a group that experienced racial or ethnic discrimination than in the pre-event period (see Table 2). For Germany, the results are not statistically significant but point to a similar trend, providing tentative evidence in favour of hypothesis 2. The lived experiences of UK minority groups reveal that anti-immigration attitudes and xenophobic sentiments politicized throughout the Brexit campaign have intensified intergroup tensions in society, thereby adversely impacting on the lives of minority respondents in the post-event period. The climate of antagonism to everything considered to be “different” exposed minorities to greater intolerance. In Germany, a seemingly similar dynamic is at play where latent anti-immigration sentiment operating in the milieu of heightened nationalism was actualized in the light of the unfolding crisis and directed at minority groups, including well-established existing groups. Furthermore, the inter-country comparison models show that minorities in Germany were 61% less likely to experience intolerance than in the UK. The interaction effect between variable “period” and “country” indicates a greater impact of the event in the UK than in Germany on experiences of minorities vis-à-vis intolerance, although this effect is not statistically significant.

An independent-samples t-test (Table 3) was performed per country to ascertain whether the means of the predicted probability of belonging to the group that experienced intolerance were equal in the post-event and pre-event periods, in both the UK and in Germany. Both tests reveal that the degree of intolerance was higher in the post-event period, and in both instances the differences in mean scores were found to be statistically significant, pointing to unequal variations between these groups in the population. It corroborates the results of the logistic regression analysis, in which minority respondents were more likely to experience intolerance in the post-event period.

Table 3. Results of t-test and Descriptive Statistics for Predicted Probabilities and Period in UK & Germany (minorities).

	Post-event			Pre-event
	M	SD	n	M	SD	n	t	p
UK*	.299	.080	300	.196	.057	315	18.331	.001
Germany	.130	.070	304	.107	.064	245	3.948	.001

Note: * unequal variance.

Source: European Social Survey 2012, 2014, 2016 and 2018.

The results of the logistic regression for majorities are shown in Table 4. In the UK, unskilled workers are found to be less tolerant compared to all of the other social positions. But the higher-grade service class is 193% more likely to be tolerant than unskilled workers, whereas the lower-grade service class is 98% more likely. Small business owners and skilled workers were, respectively, 33% and 31% more likely to be tolerant than unskilled workers. In looking at the areas in which respondents lived, those from the bigger cities are 34% more likely to be tolerant than those surveyed in the rural areas, which is similar to respondents from the suburbs (17% more likely) and small cities (13% more likely). There is also a visible relationship between tolerance and age. Younger respondents are found to be more tolerant than older respondents, particularly those between the ages of 15–24. The respondents between the ages of 25–54 are visibly less tolerant than the younger group, but still more likely to be tolerant than the 55+ category.

Table 4. Logistic regression (odds ratios) for tolerance in UK and Germany.

Variables	Model 1 – UK	Model 2 – Germany	Model 3 – Country comparison	Model 4 – Country comparison
Oesch 5-class schema: base – unskilled workers
Higher-grade service class	2.93** (.001)	5.57** (.001)	3.81** (.001)	3.85** (.001)
Lower-grade service class	1.98** (.001)	2.92** (.001)	2.33** (.001)	2.36** (.001)
Small business owners	1.33** (.001)	1.98** (.001)	1.59** (.001)	1.59** (.001)
Skilled workers	1.31** (.001)	1.40** (.001)	1.36** (.001)	1.38** (.001)
Domicile: base – rural
A big city	1.34** (.001)	1.31** (.001)	1.28** (.001)	1.27** (.001)
Suburbs	1.17* (.014)	1.13 (.097)	1.08 (.061)	1.08 (.056)
Small city	1.13* (.018)	.96 (.349)	.98 (.539)	.98 (.531)
Age: base – 55+
15–19	1.91** (.001)	2.85** (.001)	2.99** (.001)	3.04** (.001)
20–24	2.24** (.001)	1.25* (.031)	1.90** (.001)	1.92** (.001)
25–29	1.13 (.193)	.63** (.001)	1.10 (.080)	1.10 (.095)
30–39	1.23** (.001)	.74** (.001)	1.04 (.340)	1.04 (.385)
40–54	1.08 (.134)	.93 (.202)	1.08* (.015)	1.08* (.015)
Gender: base – Female
Male	1.00 (.993)	1.06 (.170)	.98 (.356)	.97 (.325)
Period: base – Pre-event
Post-event	2.13** (.001)	1.04 (.421)	1.58** (.001)	2.04** (.001)
Country: base – UK
Germany	–	–	4.03** (.001)	5.42** (.001)
Period * Country	–	–	–	.52** (.001)
European unification: base – Yes
No	.39** (.001)	–	–	–
Germany: base – East
West	–	2.20** (.001)	–	–
Chi-square (df)	x^2(15) = 1263.177 p<0.001	x^2(15) = 913.684 p<0.001	x^2(15) = 4255.224 p<0.001	x^2(16) = 4398.093 p<0.001
Nagelkerke	.14	.08	.16	.16
N	4977	9279	16947	16947

Note: *p < .05; **p < .001.

Source: European Social Survey 2012, 2014, 2016 and 2018.

In Germany, the difference in tolerance of the social groups included in the model is much more pronounced compared to the UK. Here the higher-grade service class is 457% more likely to be tolerant than unskilled workers. A comparable lean towards tolerance is also true for the lower-grade service class (192% more likely), small business owners (98% more likely), and skilled workers (40% more likely). Similarly, results from Germany indicate a visible trend towards tolerance among respondents from bigger cities, who are 31% more likely to be tolerant than rural majority respondents; this latter group have been found to be more receptive to populist parties such as the AfD (Förtner, Belina, and Naumann 2020). In terms of age, there appears to be a strong divide in attitudes towards tolerance. Younger respondents between 15 and 24 years old were more likely to be tolerant than those aged 55 + . However, respondents aged between 25 and 54 were found to be less tolerant than those aged 55 + . This differs from the UK, where the older generation is distinguished as being less likely to be tolerant than the medium-age category.

As predicted, the results indicate that respondents who most resemble the economic profile of the “left-behind”, namely blue-collar workers, poorly educated, and in a precarious economic position (Ford and Goodwin 2014), are most likely to show intolerance. This assumption has been tested using logistic regression to ascertain the effect of Oesch 5-class schema on the outcome variable tolerance in both the UK and Germany. Both logistic regression models are significant (UK: x²(15) = 1263.177, p<0.001; DE: x²(15) = 913.684, p<0.001), and explain 14% (UK) and 8% (DE) of the variance in the outcome variable. This provides tentative evidence in favour of hypothesis 1 in both countries. The least socially mobile respondents, who stand to lose the most from the accelerating social changes (such as transformation to the knowledge economy) and whose socioeconomic situation renders them particularly vulnerable to crises, are found to be least tolerant of others.

The effect of Brexit and the Refugee Crisis on the attitude of majorities towards minorities reveals that, in both countries, majorities were more likely to be tolerant in post-event periods than in the pre-event period. In the UK, majorities were 113% more likely to be tolerant in the post-event period than in the pre-event; in Germany, for which the results are not statistically significant, the likelihood of belonging to the tolerant group in the post-event period was much lower than in the UK (that is, 4%, but in a similar direction). The inter-country comparison model confirms this trend further: in the post-event period, respondents were 58% more likely to be tolerant than in the pre-event period. These tentative results provide contrary evidence to hypothesis 2, which proposed that the tolerance levels amongst the majorities would be lower after the events than before the events, due to prevailing high levels of anti-immigration sentiment. The model also indicates that German majorities were in general 303% more likely to be tolerant than the UK majorities. Moreover, the effect of interaction between variable “period” and “country” reveals that the Refugee Crisis in Germany had a considerably smaller overall impact on the tolerance of majorities than the Brexit referendum did in the UK. The Brexit referendum was preceded by a heated campaign which is likely to have led to a greater influence on the attitudes of majorities than the relatively sudden onset of the Refugee Crisis in Germany. The comparatively lower individual effect of the variable “period” for Germany nonetheless points to the more drawn our effect of the event, given the lower odds of tolerance in the post-event period for Germany compared to the UK. Brexit appears to have had a greater overall impact as an event, but its effect in the post-event period is more muted as UK majorities were found to be more likely to be tolerant than in the pre-event period.

Table 5 shows the results of a t-test performed to ascertain the relationship between predicted probabilities of belonging to the tolerant group and the variable “period”. The results indicate a significant difference in mean predicted probabilities for both groups under consideration. This underlines the finding that in the pre-event period, the probability of being tolerant was lower than in the post-event period, for both countries. That is, the post-event period is associated with a higher probability of being tolerant, which indicates a small effect of the event on tolerance in both German and UK populations following the Refugee Crisis and the Brexit referendum. It should be noted, though, that the two countries differ greatly in the extent to which this is true, as the mean difference between the two periods for Germany is considerably smaller than the mean difference for the UK. This underscores the more prolonged aftermath of the Refugee Crisis following the peak of the event in contrast with the accumulating impact of the Brexit campaign before the referendum itself took place which influenced levels of hostility to outsiders that declined in intensity following the referendum.

Table 5. Results of t-test and Descriptive Statistics for Predicted Probabilities and Period in UK & Germany (majorities).

	Post-event			Pre-event
	M	SD	n	M	SD	n	t	p
UK*	.755	.108	2427	.576	.135	2628	51.887	.001
Germany*	.896	.067	4353	.880	.074	5013	10.989	.001

Note: * unequal variance.

Source: European Social Survey 2012, 2014, 2016 and 2018.

The two models for majorities also investigate how attitudes towards tolerance are shaped by the place of residence and by opinions concerning European unification (see Table 4). The latter has been included in the model for UK majorities to tap into the central issue of the Brexit referendum. The sample under consideration is skewed against further European unification, with 67.4% of respondents holding this opinion (see Table 1). Opposing further unification is associated with a 61% lower likelihood of tolerance towards outsiders than favouring unification. As such, there is evidence in favour of hypothesis 3, as Euroscepticism is linked to low tolerance. This is in line with other research that has found links between Euroscepticism and concerns over immigration in light of the increasing levels of net migration to the UK since the accession of the country into the union, effectively linking the two issues together (Evans and Mellon 2019). In Germany, on the other hand, there is a visible difference between the Eastern and Western parts of the country, with the latter being 120% more likely to be tolerant. This is in line with the expectations expressed in hypothesis 4, which are based on the enduring legacy of the communist regime, as well as the high degree of prejudice found among residents of former East Germany towards outsiders, particularly Muslims (Adam 2015).

Discussion and concluding thoughts

It is important to appreciate the underlying socioeconomic conditions in the UK and in Germany preceding these events. The growing discourse of anti-immigrant and anti-minority sentiment in the UK context leading up to the events of the Brexit vote was underpinned by the consequences of the austerity policy that came in response to the financial crisis of 2007–2008. This policy had the effect of raising uncertainty and anxiety among majority members of society, who were subsequently misdirected by political elites. It is vital to appreciate that these existing political undercurrents in the UK were influential not just for the result of the Brexit vote but also how the sentiment was instrumentalized by the Vote Leave campaign. The converse is generally true of the situation in Germany, where the response to the financial crisis was a much gentler version of austerity. The German economy was in a relatively healthy position during the middle of the 2010s. The initial positive responses to the arrival of immigrants and refugees, especially from Syria, were therefore in stark contrast to the UK government’s approach. The starting points for both sets of comparisons are very different and so existing political contexts are an important consideration. Such an appreciation can also lead to the view that much of the negative majority sentiment, particularly among poorer majority groups in the UK, may have had more to do with their relative socioeconomic position and the outlook determined by it, as argued by previous studies (Mewes and Mau 2013), rather than deep-seated prejudices; however, the latter cannot be wholly discounted.

These results provide a thought-provoking picture of the views of minority and majority groups in Germany and the UK at a time when both nations faced acute political challenges. In terms of minorities, the results of the analysis point to a trend where the least mobile people, based on their social position within the Oesch class schema, are found to be most susceptible to experiencing intolerance. There is however an outlier: small business owners in Germany deviate from that pattern, and appear to be considerably more likely to experience intolerance than unskilled workers. This is an intriguing finding that poses additional questions as to why this particular class should be more likely to experience racial or ethnic discrimination, and to what extent their work logic plays a role. Exposure to discrimination is troubling regardless of socioeconomic position, as it is associated with negative long-term health effects for members of ethnic minority groups (Carter et al. 2017). However, individuals in lower socioeconomic positions in society are particularly vulnerable to intolerance (Mewes and Mau 2013). The intolerance endured by this minority group carries a threat of further marginalization, thereby adversely impacting their life satisfaction, including diminishing their sense of belonging to their country of residence while increasing identification with their co-ethnic group (Verkuyten 2008). In the climate of growing polarization and hostility to all things different, the alienation experienced by minority groups fuels the cycle of fear and hate, further unravelling the social fabric as majorities constrain minority opportunities to thrive within society. Arguably, to combat the cycle of alienation that minority groups face, comprehensive social support is needed but also wider efforts to tackle intolerance.

A similar pattern regarding socioeconomic position is found for majorities: working-class unskilled workers are shown to be a fringe group, characterized as having the lowest likelihood of espousing a tolerant attitude towards outsiders (Mewes and Mau 2013). This trend is similar in both the UK and Germany. The least socially mobile members of the majority population are the least tolerant, while the least socially mobile minority population are the most likely to experience intolerance. These two groups share certain socioeconomic characteristics that render them particularly vulnerable to accelerating social trends such as rapid technological transformation and globalization. On the one hand, the minority groups face not only a decreasing demand for their blue-collar skillset but also racial and ethnic discrimination, leading to further marginalization. For the majority groups on the other hand, this same economic disadvantage fuels their resentment, which is then vented at the political system and “otherized” members of minority groups. Both groups therefore find themselves struggling in the face of current social trends that put them at risk of economic hardship and social exclusion, while also obliging them to compete with one another for scarce resources. Alienated majorities form an attractive voter base for the radical right; the discontent felt by these majorities is instrumentalized by populist-nativists for political gain, firmly diverting attention from the economic policies that led to their marginalization towards minorities with whom they are left to compete for scarce resources. The economic downturn triggered by the COVID-19 pandemic threatens to increase tensions: with resources becoming scarcer, inter-group competition is heightened. The influence of the radical right on the realm of policy, particularly with regards to immigration, may therefore further adversely impact on the lives of minority groups.

In terms of the impact of Brexit and the Refugee Crisis on experiences of minorities and attitudes of majorities vis-à-vis tolerance, the findings are conflicting. For minorities, the likelihood of being discriminated against in the post-event period was higher, whereas for majorities the post-event period is associated with a higher likelihood of being tolerant. The view of one group on the question of tolerance appears to be entirely disconnected from the view of the other group, particularly in the UK given the climate of heightened anti-immigration sentiment that delivered success to the Leave campaign in the Brexit referendum. It appears that anti-immigration sentiment amongst majorities was higher before each event, particularly Brexit, but declined in the post-event period. These findings are much more pronounced in the UK than in Germany. In Germany, the findings indicate a more balanced pattern throughout the entire period of the study, with a slight shift towards tolerance in the post-event period. The Refugee Crisis thus appears to have had a much smaller overall impact on the attitudes of majorities concerning tolerance, but its effect may be more prolonged given the unanticipated nature of the crisis. Minorities, on the other hand, were more likely to be discriminated against in the post-event periods. This result underscores the detachment between majority and minority groups, especially in the UK: there, the attitude of majorities may have changed by the end of the Brexit campaign, while minorities were left with increased intergroup tensions that rendered them more likely to experience discrimination. In Germany, the impact of the Refugee Crisis is less pronounced, although the trends resemble the situation found in the UK; however, it is notable that majorities in Germany were much more likely to be tolerant than majorities in the UK, despite the UK’s multiculturalist tradition. The notably lower odds of being tolerant for the UK could be attributable to acute anti-immigrant sentiment rooted in the attitude of resentful nationalism, wherein the UK is viewed as being in decline due the impact of outsiders.

In both countries, the attitudes and experiences of the general population vis-à-vis tolerance are largely positive, despite heightened anti-immigration sentiment and a climate of political polarization which the radical right succeeded in turning to their political advantage following the events. However, the situation is noticeably negative among respondents from lower socioeconomic backgrounds. In addition, minorities reported worsening experiences of discrimination in the post-crisis periods. Future research could build on these findings by exploring the dynamics that contribute to higher levels of tolerance among majorities from higher socioeconomic backgrounds relative to those from lower socioeconomic backgrounds.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 See “Methodology Overview” https://www.europeansocialsurvey.org/methodology/

2 See http://people.unil.ch/danieloesch/files/2014/05/Oesch_class_schema_ESS2012_SPSS.txt