IS IT THE ECONOMY?
How economic factors affect
anti-immigration sentiments and support for
authoritarian populism in Europe
AUTHOR JELLE VAN DE SANDE S2089866 DATE 01-10-2020 ADVISOR A. AFONSO SECOND READER O. VAN VLIET STUDY PUBLIC ADMINISTRATION ECONOMICS & GOVERNANCE LEIDEN UNIVERSITY ABSTRACT
This study is an attempt to determine if and how economic factors influence anti-immigration sentiments and support for authoritarian populism in Europe. Data from the European Social Survey and the OECD are used to conduct analyses on a level and a macro-level. The
micro-level results show that citizens with a relatively bad economic situation are more likely to have anti-immigration sentiments and support authoritarian populists. Subjective factors seem to be the most important determinant. The results on a macro-level are less clear, which might be caused by the supply-side.
1 CONTENTS Introduction ... 3 Theoretical framework ... 6 Literature study ... 6 Cultural Backlash... 7 Immigration ... 9 The economy ... 11 Conceptualization ... 15 Dependent variables... 16 Independent variables ... 16 Control variables ... 17 Methodology ... 19 Operationalization ... 19 Dependent variables... 19 Independent variables ... 24 Control variables ... 25
Reliability and validity ... 28
Descriptive statistics ... 32 Results ... 38 Hypothesis 1 ... 38 Hypothesis 2 ... 48 Conclusion ... 52 References ... 54
2
TABLES
1. Conceptualization of independent variables. ... 17
2. Operationalization of the dependent variables ... 22
3. Classification of authoritarian populist parties based in Europe ... 23
4. Operationalization of the independent variables ... 25
5. Operationalization of the control variables ... 27
6. Equations of hypotheses and regressions ... 31
7. Descriptive statistics of the variables used in the micro-level analysis ... 32
8. Correlation matrix of the variables used in the micro-level analysis ... 33
9. Descriptive statistics of the variables used in the macro-level analysis ... 34
10. Correlation matrix of the variables used in the macro-level analysis ... 35
11. Linear regression analysis of the influence of objective economic factors on immigration sentiments (micro-level) ... 38
12. Linear regression analysis of the influence of objective economic factors on immigration sentiments (macro-level) ... 40
13. Linear regression analysis of the influence of subjective economic factors on immigration sentiments (micro-level) ... 42
15. Linear regression analysis of the influence of objective and subjective economic factors on immigration sentiments (micro-level) ... 44
16. Linear regression analysis of the influence of objective and subjective economic factors on immigration sentiments (macro-level) ... 45
17. Logistic regression analysis of the influence of objective and subjective economic factors on support for authoritarian populist parties (micro-level) ... 48
18. Linear regression analysis of the influence of objective and subjective economic factors on support for authoritarian populist parties (macro-level) ... 51
FIGURES 1. Conceptual model ... 15
2. Histogram with frequencies to test for normal distribution (micro-level analysis) ... 29
3. Kernel density estimate to test for normality of residuals (macro-level analysis) ... 29
4. Average scores for the dependent variables per year for each country ... 36
5. Average scores for the economic variables per year for each country ... 37
3
INTRODUCTION
The support for authoritarian populism in Europe is growing. It is the only relatively young political party family that has been able to compete with more traditional political families in all parts of the European continent (Mudde, 2007). On the 23rd of June 2016, a majority of the British electorate decided that the United Kingdom should leave the European Union in a historic referendum. Just a few months later, on the 8th of November, Mr. Trump won the presidential election in the United States. Both events were not expected by most polls and experts, and were maybe even considered to be impossible (Norris & Inglehart, 2019). However, these events were no coincidence and are instead part of a longer trend of rising authoritarian populism and anti-immigration sentiments in Western countries. The vote share for populist parties in western societies has increased from 9.9 percent in the 1990s to 11.4 percent in the 2000s and 12.4 percent in the 2010s (Norris & Inglehart, 2019). Authoritarian populists often use problems with the economy and immigration as narratives in their campaigns. Then-candidate Trump blamed immigrants for stealing jobs (Kohn, 2016), and populist Wilders blamed the Dutch prime-minister for destroying The Netherlands because of retrenchments during the economic crisis (Tweede Kamer, 2012). Mr. Orban, Prime Minister of Hungary, took advantage of the ‘welcoming culture’ of other European leaders (Mudde, 2016). It seems likely to believe that the narratives have helped these politicians in their elections, but at the same time, they are mainly temporarily. Refugee crises and economic depressions are both events that may occur occasionally, but they cannot explain the longitudinal trend that has been going on for decades (Mudde, 2007, 2016; Norris & Inglehart, 2019).
According to populism, the legitimacy of democratic authorities comes from the voice of the people (Norris & Inglehart, 2019). The view of the majority outweighs the protection of minorities and is more important than the opinion of experts, legal authorities, scientists, and politicians. Populists separate the society into two groups: the pure people and the corrupt elite, us, and them (Mudde 2007, 2016). Not all populists endorse authoritarianism, but a combination of the two often occurs. Authoritarianism is a psychological concept that initially refers to favoring obedience over liberty (Norris & Inglehart, 2019). Today, politicians with authoritarian values put emphasis on three components: conformity to group conventions and traditional customs, the security and accepted norms of the group, and loyalty to the group and its leader. Citizens with authoritarian values are less open to new ideas and less tolerant of people from other groups. While libertarian populists favor social liberalism, individualism and tolerance towards multiculturalism, authoritarian populists prioritize social conservatism, strong leaders, and social stability (Norris & Inglehart, 2019, p. 78). Another central element of authoritarian populism is nativism (Mudde, 2007, pp. 18-20). Nativists hold that members of the native group should exclusively inhabit the state and that all persons and ideas that cannot be considered as native are threatening the homogenous nation-state. The rise of anti-immigration parties in European countries can be seen as almost equal to the growth of
4 authoritarian populism in Europe, as there is a broad consensus that hostility toward immigration, multiculturalism, and minority rights are defining features of authoritarian populist parties (Norris & Inglehart, 2019). All successful Western European populist parties have used immigration problems in their election campaigns (Ivarsflaten, 2008), and survey evidence indicates that hostile attitudes toward outsiders trigger an authoritarian reflex that has links with in-group conformity or xenophobia (Norris & Inglehart, 2019).
Studies on the causes of support for authoritarian populism and anti-immigrant sentiments seem to be more relevant than they have ever been. Before the election of Mr. Trump and the referendum about Brexit, Mudde (2007) concluded that there were yet only limited effects of the populist right on policies despite their increasing support. However, he also stated that an increase of their impact in the future was not unlikely due to the tabloidization of the political discourse, the learning curves of right-wing populists, and the consequences of economic depressions. He turned out to be right. Examining why voters cast their ballot for these types of parties gives us not only a better understanding of voters’ motivations, but it also brings the possibility to predict better whether a populist party has a chance to win a particular election. Experts were often wrong when trying to predict results for authoritarian populist parties (Norris & Inglehart, 2019). Furthermore, knowing which concerns make people vote in favor of authoritarian populists can be helpful for policymakers. Or, as Mudde (2016) describes it: “Today’s floundering, hollowed-out mainstream European parties and the entrenched elite who guide them will have to respond with far more dexterity and creativity than they have shown in recent decades” (p. 25).
There are already many studies on authoritarian populism. Most of them were cross-sectional (Mudde, 2007; Hatton, 2016), which means that they examined one point in time, often being a particular election in Western Europe or the United States. Autor et al. (2017) and Becker, Fetzer, and Novy (2017) compared the results of different regions in respectively the US election and the referendum about Brexit. They used specific characteristics of regions with a high number of Republican or Leave-votes to explain why citizens voted for authoritarian populist ideas. There are comparable studies on an individual level. These studies are unquestionably relevant, but as Hatton (2016) describes, they are only able to tell which citizens oppose immigration and vote for authoritarian populists. They do not tell us how their opinions change over time and why. The effects of macro-economic factors are, for example, not examined. Furthermore, most of the studies do not have attention for developments in eastern Europe (Mudde, 2007), while events in especially Hungary and Poland make the eastern part of the continent relevant as well (Norris & Inglehart, 2019). The EU memberships of countries in Eastern Europe strengthens this relevance, as they have made them more comparable to western Europe (Mudde, 2007).
5 There have been studies that do try to answer these questions. Hatton (2016) has conducted an analysis with European Social Survey data over a more extended period, and Norris and Inglehart (2019) have also conducted some analyses in their book Cultural Backlash in an attempt to explain longitudinal trends. This study is different on several points. First of all, it will conduct analyses on both a micro-level and a macro-level. Citizens’ characteristics and points of view are used to examine what the most critical determinants for their voting behavior are, while there will also be a longitudinal study comparing countries to determine which macro-environmental factors affect support for anti-immigration sentiments and authoritarian populist parties. Furthermore, this study uses data of the European Social Survey (2018b) that were released just before the analysis, making it possible to use data that other studies cannot have used.
After all, it is possible to conclude that there is enough relevance for a study with the following research question:
How do economic factors influence anti-immigration sentiments and support for authoritarian populist parties in Europe?
6
THEORETICAL FRAMEWORK
LITERATURE STUDY
The rise of anti-immigration sentiments and authoritarian populists seem to be correlated. Most studies combine the two in a concept that’s often called the populist radical right (Mudde, 2007). However, there might also be differences between authoritarian populism and anti-immigration sentiments. It is possible that an authoritarian populist party does not have a negative attitude toward immigration, and citizens who vote for anti-immigrant authoritarian populists might not base their vote on migration policies. Furthermore, there might be political parties defending strict migration policies, while they are not authoritarian populist. For these reasons, this study makes a distinction between authoritarian populism and anti-immigrant views. Thus, during this literature study, modest distinctions between these two concepts may be found.
Mudde (2007) adds a few cautionary comments about the concept of populist radical rightists that are important to mention before coming to an overview of the available literature. First, populist radical rightists are nationalists, although not all nationalists necessarily are populist radical rightists. Second, they do not adhere to a moderate form of extreme right ideologies like fascism and national socialism, as they usually are democratic. Third, the broader radical right might include non-populist values, and therefore it is essential to distinguish the populist radical right and the non-populist radical right.
Studies on the concept of authoritarian populism have almost exclusively focused on the demand-side for an extended period. The thesis of normal pathology suggested that the level of support for the populist radical right is marginal under normal circumstances and that the support could rise during a socioeconomic or sociodemographic crisis, but this thesis gets no support from empirical evidence (Mudde, 2007, pp. 296-297). Authoritarian populist parties are more widely supported than is often suggested, and their discourses are comparable to but just more radical than those of established parties in terms of nativism, authoritarianism, and populism. So, Mudde (2007) defines the populist radical right as pathological normalcy: a radicalized version of mainstream ideas that does not necessarily need an explanation from new demand-side theories. There is a relatively high demand for populist radical right parties in all liberal democracies, and the question is not why citizens support authoritarian populist values, but why they do not always vote for such parties. Every European country has the potential for a strong populist radical right movement, but it does not flourish in all elections. According to Mudde (2007), the explanation does not mainly come from objective demand-side factors like the economy or immigration but should primarily be found on the supply-side. Political parties and media decide how an economic crisis or terrorist attack is ‘framed,’ and the populist radical right uses certain events to attract support, primarily when established parties can be blamed (Mudde, 2007). After the so-called electoral breakthrough of a populist radical right party, the internal supply-side becomes more critical. The political opportunity
7 structure has less importance and is replaced by how the party functions in an attempt to maintain the support of hardcore voters while attracting new soft voters. One of the crucial factors during this phase is the presence of a charismatic leader. Although the analyses of this study mainly focus on the demand-side, this study is also an attempt to combine both sides by using examples and discussing theories on how certain events can be used by authoritarian populists to gain support. Furthermore, there is attention for the demand-side when the conceptual framework is composed.
In theories about the rising support for authoritarian populism, there is a distinction between different causes like culture, economic circumstances, and immigration. Some short-term factors like the economic situation of a country or individual and the threat of (possible) refugee crises can lead to a temporary increase of support, while long-term cultural and economic trends are not likely to change quickly (Norris & Inglehart, 2019). There is a comparable distinction in the literature about negative attitudes toward immigration. Dustmann and Preston (2007) distinguish labor market concerns, welfare concerns, and racial or cultural concerns as possible causes of rising opposition against immigration. A distinction must be made between economic circumstances, policies, and interpretations in general, and those that are specifically related to immigrants. The latter will be discussed in the immigration paragraph of this literature study, while the impact of general economic factors is examined in the paragraph about the economy. This is also the case for cultural theories. While general cultural trends will be discussed in the cultural backlash part, the cultural consequences of immigration are discussed in the immigration part. Mudde (2007) adds the effects of globalization in his book. These effects can, however, all be categorized as economic, immigration-related, or cultural. Thus, the categories as described by Norris and Inglehart (2019), will be used as the thread in this literature study, in which all other literature are examined as well.
CULTURAL BACKLASH
In their book called Cultural Backlash, Norris and Inglehart (2019) use their cultural backlash theory to explain how the support for the authoritarian populist movement has grown. This theory starts with the increase of economic and physical security in European societies that started after the second world war. The increase of existential security has led to a silent revolution: the shift in values between generations from economic and physical security to post-materialist values like free choice and self-expression. This increasing attention for post-materialist values is part of a shift that led to more attention for ideals like peace, environmental protection, gender equality, and the protection of ethnic or racial minorities. This shift in opinion exists in most secure high-income societies, but at the same time, there is still a substantial minority of the population that endorses traditional social norms and moral beliefs founded on faith, family, and the nation-state. Especially for older cohorts, the gap between the norms of the world into which they
8 were born and the world in which they live now is growing (Norris & Inglehart, 2019). The replacement of generations is one of the causes of the shift in values. However, other long-term trends coming with this replacement, like the rising educational levels, the growth of ethnic diversity, gender equality, and urbanization, also contributed to the change. Besides, time-related effects like economic insecurity, perceived risks of terrorism, and immigrant flows influence cultural values.
But, how does this evolution of cultural values explain the increasing support for authoritarian populism? On the demand side, the tipping point from the socially conservative to the socially liberal values might have led to a backlash among the once-dominant conservative group (Norris & Inglehart, 2019). Leadership appeals and media cues on the supply side activate the latent authoritarian attitude among social conservatives. The activation of this group can lead to the rise of authoritarian-populist parties, mainly because younger, socially progressive citizens are less likely to vote.
Cultural globalization is an element of the cultural backlash theory. Although studies consider the anti-globalization movement to be more left-wing or progressive and even though globalization is not an essential issue in the ideology of the radical right, the most successful opponents of globalization are the populist radical right parties (Mudde, 2007). Part of this opposition is related to anti-immigration sentiments and the opposition toward the globalizing economy. Other paragraphs will describe these components of globalization, but there is also a more cultural and political component of globalization that this part discusses. According to Mudde (2007, p. 193), the populist radical right fears the so-called ‘New World Order,’ which is characterized by increasing cooperation between countries. The European Union is, for example, not just an economic project but also a political project that is seeking for “commitment to progress towards an ever-closer union among the peoples and Member States of the European Community” (European Council, 1983, p. 25). Negative attitudes toward this commitment might cause citizens with nationalist attitudes to vote for authoritarian populist parties. Furthermore, globalization has caused the so-called Americanization of European countries, symbolized by American television programs and fast-food chains (Mudde, 2007, pp. 190-192). According to the populist radical right, globalization has led to the homogenization of cultures.
Norris and Inglehart (2019) use datasets of the European Social Survey, World Values Survey, Eurobarometer, and Gallup Polls in an attempt to find support for their cultural backlash theory. They conclude that there is indeed a big difference in values between generational cohorts. Authoritarianism and populism are supported most by the interwar generation, while there is substantially less support for these tendencies in the millennial generation. So, Norris and Inglehart (2019) conclude that a silent revolution in the younger generation has undoubtedly occurred and that this has catalyzed a cultural backlash that took place primarily in the Interwar and Baby Boomer generations.
9
IMMIGRATION
This paragraph is an attempt to examine the literature about if and how the number of immigrants coming into a country influences citizens’ attitudes toward immigration. According to Ivarsflaten (2008), all successful right-wing populist parties in western European elections around 2002 have used immigration issues to mobilize support, while there are successful populist parties that did not use economic hardships or political elitism in their campaigns. So, the use of problems with immigration is the unifying factor of all successful populist right parties. But why is this the case? As discussed, the cultural backlash theory describes how many socially conservative people feel threatened by rapid cultural change (Norris & Inglehart, 2019). This change is not just caused by economic prosperity that shifted the attention of younger cohorts to post-materialist values, but also by growing ethnic diversity. The proportion of the population that is born in a foreign country has increased for a long time in most countries, but a rapid influx of immigrants in western Europe and Scandinavia, especially in the recent refugee crisis, has caused more discussions about welfare, social cohesion and religious values across the continent. Terrorist attacks in several European cities strengthen these discussions. Norris and Inglehart (2019) make a distinction between three mechanisms that describe how immigration causes support for authoritarian populists: instrumental concerns, cultural anxieties, and fear of terrorism. The next part of this paragraph discusses these mechanisms in more detail.
Instrumental or interest-based theories of immigration describe how competition over scarce resources like housing and social benefits trigger opposition to immigration (Citrin et al., 1997; Dustman & Preston, 2007; Norris & Inglehart, 2019). In the Netherlands, the Party for Freedom (PVV) argued for instance that immigrants were prioritized over Dutch citizens in the case of the allocation of scarce social housing and the far-right party proposed a motion in parliament to stop the so-called discrimination of the Dutch people (Kops & Wilders, 2018). Anti-immigrant populists blame immigrants for free-riding and argue that the government should serve the interests of native-born people over those of immigrants (Norris & Inglehart, 2019). Immigration might not only be a threat to the welfare system, but it might also affect the position of native people in the labor market. It is likely to believe that low-skilled workers are more vulnerable to competition from immigrants. For example, then-candidate Trump successfully used this narrative during his campaign, arguing that immigrants were stealing the jobs of native Americans (Kohn, 2016). According to the conservative-leaning polling group of Rasmussen, 51 percent of the people believed that illegal immigrants were taking jobs away from native citizens. However, the narrative turned out to be untrue to a great extent, as skilled immigrant workers are taking on different jobs than low-skilled native-born workers (Enchautegui, 2015; Kohn, 2016). Nevertheless, the narrative seemed to work: survey data do show that especially low-skilled workers are more sensitive to the narrative, as they are less likely to believe that immigrants influence the economy in a positive way (Hatton, 2016). The
interest-10 based theory implicates that unfavorable economic conditions of individuals reflect negative sentiments about immigrants. The question is whether this also means that there is less support for immigration during economic downturns. The economic paragraph of this literature study will discuss this question.
Apart from economic threads, native-born people might also feel cultural anxieties that cause negative attitudes toward immigrants (Norris & Inglehart, 2019). This mechanism is part of the cultural backlash theory, as it threatens the traditional social values of a previously dominant group. It is likely that this backlash occurs mainly in isolated and homogeneous rural areas that have little experience with welcoming immigrants (Norris & Inglehart, 2019). Besides, the occurrence of these anxieties depends on the educational level of immigrants, humanitarian considerations like sympathy for refugees fleeing war, and the religious faith of the immigrants. Citizens feel especially threatened by immigrants who are ethnically more distant from most of the population (Dustmann & Preston, 2007). The correlation between the educational level of citizens and their opinion toward immigration might not just be caused by economic threads, as it is also possible that this group feels more threatened by immigrants on cultural grounds (Dustmann & Preston, 2007).
Finally, terrorist attacks can be an explanation for a causal relationship between immigration and anti-immigrant sentiments (Norris & Inglehart, 2019). In the 21st century, terrorist events claimed by Islamic groups occurred in a lot of European countries. These events raised fears that Muslim immigrants were a threat to the security of European societies. There are examples of politicians exploiting terrorist threads after all attacks in Europe. Nigel Farage of the UK Independence Party blamed multiculturalism and immigration policies for the Westminster attack in 2017 (Oppenheim, 2017). He did not mention that most of the victims were foreigners as well. After terrorist attacks that took the lives of 130 people in Paris, far-right leader Marine Le Pen of the Front National party held her government responsible too and advocated for stronger border controls (Chrisafis, 2015). The Front National raised in poll ratings after the attack. As a consequence of attacks in Europe, Islamophobia has grown, and authoritarian values were strengthened (Norris & Inglehart, 2019).
After all, populists have used the immigration issue by depicting immigrants as criminals, terrorists, or freeloaders. The question is however how important this narrative is. The banking crisis that started in 2007 caused concerns about unemployment in Europe, which started falling again from mid-2013, while concerns about immigration rose from mid-2013 until late 2015 (Norris & Inglehart, 2019). However, almost all the time, Europeans saw unemployment as a more urgent problem facing their country, according to Eurobarometer data. It is debatable whether this means that citizens feel less threatened by immigrants. As discussed, unemployment and immigration might be correlated or at least seen as correlated by the electorate.
11 The discussion showed how economic, cultural, and security concerns could cause negative views toward immigrants. However, do these concerns rise when there are more immigrants in a country? A longitudinal study of Hatton (2016) on European Social Survey data does find proof for this: the author showed that there is a relation between the share of immigrants in a population and negative attitudes toward immigration. The refugee crisis in Europe that started in 2015 also seems to have increased support for authoritarian populists in Europe who successfully framed the crisis as a security issue (Mudde, 2016).
THE ECONOMY
The economy has changed over time. Globalization of labor, trade, and goods, together with liberalization and deregulation, have decreased the job security of chiefly unskilled workers and have caused a growing gap between the rich and the poor (Rodrik, 2018; Norris & Inglehart, 2019). While the so-called ‘winners’ from globalization profit from more trading and job opportunities, the ‘losers’ risk losing their jobs to low-wage countries. The economic grievance theory argues that the ‘losers’ from globalization in advanced economies are more likely to support authoritarian populist parties (Norris & Inglehart, 2019, pp. 132-174). They blame their national governments and social democratic parties for their increased risk of being unemployed and stagnant wages. It is possible to test this theory on different levels. While the macro-level might help us to explain why there is a specific rise of support for authoritarian populist and anti-immigrant parties, the microlevel is also needed to explain why an individual has specific ideas and values about migration and populism, and how these values and ideas are related to other factors, like cultural ones (Norris & Inglehart, 2019). This study will discuss both levels.
Recent within-country studies on the Brexit referendum in the United Kingdom and the presidential election in the United States examined whether the populist-vote share was higher in regions that are more vulnerable to globalization. Analyses show that the vulnerability to cheaper Chinese imports turned out to influence the US presidential election in favor of the Republican party (Autor et al., 2017). If the trade exposure from China had been 50 percent lower, Democratic candidates would have won battleground states like Michigan, Wisconsin, and Pennsylvania, which could have made the difference in the electoral college. According to Becker, Fetzer, and Novy (2017), the percentage of the British population voting in favor of Brexit is not higher in regions that are exposed more to the European Union when it comes to trade and migration. However, the authors did conclude that the share of leave-voters was higher in regions with characteristics like manufacturing employment, low incomes, and high unemployment. This suggests that leave-voters might not have voted for leaving the European Union because they blamed the EU for their economic insecurity. Instead, their economic insecurity as the consequence of broader globalization made them vulnerable to populist rhetoric. This theory seems to be supported by the study of Calantone and
12 Stanig (2018), who compared the leave share in local labor markets and their vulnerability to trade from China: there is more support for Brexit in economic zones that are more exposed to this trade.
We can conclude that macro-level analyses that compare regions or countries often show support for the economic grievance theory. However, these analyses bring the risk of overestimating the effects of economic factors, while cultural factors might play a more critical role (Dustmann & Preston, 2007; Norris & Inglehart, 2019). Norris and Inglehart (2019) use individual data of the European Social Survey to examine whether authoritarian populists are supported most by voters who have low incomes, are blue-collar workers in manufacturing industries, or depend on welfare benefits. They also control for citizens’ personal valuation of the current macro-economic situation and their perceptions about their financial situation, although they do believe that the objective data will give the most reliable results, as subjective attitudes might be vulnerable to endogeneity problems like reversed causality. The authors conclude that individuals who have low incomes and work in manufacturing sectors are indeed more likely to support authoritarianism and populism. This is also the case for respondents with relatively low subjective income security and dissatisfaction with the current state of the economy. However, the prediction of authoritarian values based on economic indicators has proven to be less important than the prediction based on generational cohorts. This implies that the previously discussed cultural factors are of more importance, and it suggests that support for authoritarianism will endure, even in times of economic prosperity. The opposite is the case for populist attitudes, for which subjective economic characteristics turn out to be slightly more important than a birth cohort. Especially the feelings of respondents about the current state of the national economy are a strong predictor of political mistrust (Norris & Inglehart, 2019). This conclusion confirms a study of Magalhães (2017), who used European Social Survey data to show that objective macroeconomic performances are a significant predictor for the confidence of people in political institutions, mainly in countries where the quality of the government is low.
The study of literature shows much support for the economic grievance theory. A problem in most studies seems to be the concept of time. Cross-sectional studies can predict that individuals who are part of the ‘losers’ group are more likely to support authoritarian populism and anti-immigration parties. They have also been able to explain why there is more support for authoritarian populist values in specific regions and countries, mainly because the share of ‘losers’ living there is substantially higher. However, this does not automatically mean that the value change is related to the current economic situation. Instead, the ‘losers’ group may have specific values that are unrelated to the economy, which make them vote for authoritarian populists. As mentioned in the introduction, it is essential to examine whether people feel more threatened by immigrants in times of economic depression, but there is only a limited number of studies conducted on this (Hatton, 2016). Magalhães (2017) did examine different years and concluded that the economic situation over time does have some influence. Norris and Inglehart (2019) combine their
13 backlash theory about cohorts with period-effects, arguing that even the oldest cohorts shift toward post-materialist values in times of economic prosperity, but shift back toward post-materialist values during economic downturns. So, it is crucial to not only use the individual’s or region’s economic situation at a particular time but take the economic fluctuations on a more extended period into account as well. For this reason, this study will conduct both a cross-sectional analysis on a micro-level and a longitudinal study on a macro-level.
Hatton (2016) has used European Social Survey data to examine how the most recent economic recession in Europe affected citizens’ attitudes toward immigration. The author tested for these individual data with macro-level variables. As already mentioned in the paragraph about the consequences of immigration, results showed that the share of immigrants in the population and the share of welfare benefits in GDP were the most important variables influencing the public opinion toward immigration over time. The effect of social benefits reflects concerns about increasing budget deficits, to which welfare state concerns are highly correlated. The author explains that there is a possibility that the strength of the effect depends on specific types of social expenditures and that it is likely that the share of cash transfers matters. There is no significant effect on the unemployment rate, except for respondents' opinions about whether immigrants positively affect the economy. So, Hatton (2016) concludes that his results are inconsistent with the, on the face of it, logical view that the recession has caused a backlash against immigration and that the economic crisis was a significant cause of the rise of authoritarian populist parties. While cross-sectional results argue that the opinion of citizens varies because of characteristics like the individual’s labor market position, Hatton (2016) argues that there is relatively little variation in the response of macro-economic changes among individuals with different characteristics.
As discussed, it is possible that citizens have more negative attitudes toward immigrants during economic hardships, and that these attitudes lead to their votes for authoritarian populists. There is also a possibility that voters choose authoritarian populists during economic recessions because they blame the (non-authoritarian) government. But what are the economic policies of authoritarian populists? Studies often presume that authoritarian populists trust the market more than the state, and neoliberal economic policies are, therefore, an essential feature of these parties (Mudde, 2007, pp. 119-137). There is however a lack of evidence for this presumption. Mudde (2007) concludes that the ideologies of the most important representatives of the European radical right on the economy cannot be considered as neoliberal and are more comparable to the economic ideologies of the Christian Democrats in Europe.
Furthermore, the economic program of the radical right is only a way to put their core ideas about issues like immigration into practice. So, economic policies do not have priority for radical-right parties or their voters (Mudde, 2007). However, a more recent analysis of Afonso and Rennwald (2018) on the economic policies of radical right parties showed that the attention for the welfare state of some parties has
14 increased. To be eligible for coalitions or to normalize their image, radical rightists were forced to clarify their stance on welfare. Voter data showed that the economic crisis of 2008 increased the attention for economic policies and made it also necessary for all parties to take a clear stance on welfare (Afonso & Rennwald, 2018). While the position of traditional party families on the welfare state is clear, this is not the case for the radical right. Their voters are more and more part of the working class, which originally supports welfare programs, but the welfare policies that are proposed vary. The economic crisis and the increasing support from the working class have compelled most radical right parties to change their position in favor of the welfare state. If the issue of redistribution is more salient, radical right parties turn out to be more left-wing. This is also the case for social-democratic and radical left parties, but not for other right-wing parties (Afonso & Rennwald, 2018). The increasing support from authoritarian populists for the welfare state is remarkable, given the fact that there are more anti-immigration sentiments in countries that spend more money on particular welfare programs (Hatton, 2016). The type of welfare program might have an essential role in this.
15 CONCEPTUALIZATION
After having discussed the possible causes of the rise of authoritarian populist parties and anti-immigration perceptions in Europe, it is possible to come to a conceptual framework that will be tested. The framework focuses on the economic factors that might influence the dependent variables. The other possible causes discussed in the literature study will be the control variables in the analyses.
This research analyses two dependent variables, which are likely to be highly correlated but might also give substantially different results. This distinction makes it possible to explain the support for authoritarian populism and anti-immigrant sentiments in more depth. As described in the literature review, a negative economic perception can lead to a negative perception of immigration, which might lead to support for authoritarian populism, but it is also possible that there is a direct relationship between the economic situation and authoritarian populism. Besides, anti-immigration sentiments do not automatically have to lead to support for authoritarian populism, as voters might base their vote on other issues. They might also vote for parties which propose anti-immigration policies but are not authoritarian populist. Furthermore, the demand-side is not the only aspect that determines support for authoritarian populists. As discussed, the attraction by parties can play an essential role in the supply-side (Mudde, 2007). It seems likely that the supply-side is less strongly related to anti-immigration sentiments.
Hypothesis 1: Anti-immigration sentiments rise during economic hardships.
Hypothesis 2: The support for authoritarian populist parties rises during economic hardships.
Figure 1. Conceptual model.
Note. All causal relations are hypothesized as positive. Objective economic situation Subjective economic situation Anti-immigration sentiments Authoritarian populism
16
DEPENDENT VARIABLES
So, anti-immigration sentiments and support for authoritarian populism are the dependent variables in this conceptual model. Along with the independent variables, the dependent variables will be analyzed on both a micro-level and a macro-level.
This study defines anti-immigration sentiments by citizens’ negative views on immigration into their country in the past, present, and future. This means that the negative sentiments are not only defined by the opinion on the inflow of migrants at the moment but also about the opinion on the attendance of foreign-born people who have migrated earlier and the opinion on possible immigration in the future. As there is no clear national conceptualization possible for this variable, the average opinion of a country on specific moments in time is used for the analysis between countries.
The concept of authoritarian populism is already defined at the beginning of this theoretical framework. As discussed, this authoritarian populism has three main components: conformity to group conventions and traditional customs, the security of (the accepted norms of) the group, and loyalty to the group and its leader (Norris & Inglehart, 2019). So, authoritarian populists are social conservatives who are less open to new ideas and tolerant to people of other groups. The classification of these parties is essential and takes place in the operationalization.
INDEPENDENT VARIABLES
This study uses four different ways to conceptualize the economic situation. When someone’s current economic situation is right, this does not necessarily mean that the person is positive about the economic situation. A person might still feel anger about the past or be skeptical about the future, and the perception of the current situation might not be the same as the situation objectively is. Furthermore, voters might not only look at their own economic situation but also at the situation of others. That is why it is also necessary to analyze a subjective microeconomic level: the individuals’ perception of the current economic situation. The use of citizens’ own opinions about the economic situation might be vulnerable to reversed causality, as their political support or opposition toward their government might positively or negatively affect their view on the economy (Norris & Inglehart, 2019). This makes it necessary also to use objective economic variables. This distinction between the subjective and objective economic situation can also be made on a national level, as citizen’s perception of the current national economy might not be the same as the actual economic situation.
17 Table 1: Conceptualization of independent variables.
Micro-level (cross-sectional)
Macro-level
(longitudinal & cross-sectional) Subjective Citizen’s perception of economic situation Average perception of national economy Objective Financial situation of citizen Performance of national economy
CONTROL VARIABLES
The literature study has made clear that economic factors are not the only determinants of support for anti-immigration sentiments and authoritarian populists. Other factors that were discussed are part of the analysis as control variables. As discussed, Norris and Inglehart (2019) describe how a cultural backlash occurred as a response to the post-materialist silent revolution. Generational replacement and other trends like rising educational levels, the growth of ethnic diversity, gender equality, and urbanization have caused this revolution. This research will define the cultural backlash as a conservative trend among mainly older once-dominant generational cohorts, which consists of strong opposition against the silent revolution.
So, this trend can be analyzed on an individual level by looking at someone’s generational cohort and an individual’s opposition against the post-materialist values of the silent revolution. On a macro-level, it is possible to use these concepts as well, although they might be less straightforward. The attendance of strong opposition against post-materialist values in a country might be a determining factor for anti-immigration sentiments and support for authoritarian populism. However, if the cultural backlash theory is valid, there should also be support for post-materialist values in a country with a high presence of the dependent variables. In some countries, the generational replacement might be faster than in others. A fast replacement means that there are less old, often more conservative people. So, if there is less support for authoritarian populism and anti-immigration sentiments in countries where there are fewer old citizens, there is support for the cultural backlash theory. Thus, strong opposition against post-modern values in a country and generational replacement are used as conceptualizations of the cultural backlash variables on a national level. However, when jumping to conclusions, it needs to be considered that the individual variables of the cultural backlash concept are probably more valid.
Immigration is also a control variable in the analysis. It is defined in two different ways: the current inflow of immigrants in the country and the current percentage of foreign-born people in the country. This research will use both conceptualizations, as people’s opinions might not be caused by the current inflow and current policies, but also by inflow that has occurred in the past. On a macro-level, immigration is defined as the objective number of immigrants that are coming into the country or the objective percentage
18 of foreign-born people. Some studies examine the effects of only non-western immigrants instead of the full share of immigrants. However, the evidence does not show that the effect of immigrants over time is more substantial if there are more non-western immigrants in a country (Hatton, 2016). It would have been desirable to use respondents’ estimations of the number of immigrants in the individual analysis as these estimations might substantially differ from the objective numbers. According to Norris and Inglehart (2019), citizens are likely to overestimate the number of migrants. Unfortunately, these data are not available for most years, as will be mentioned in the operationalization.
As discussed, the educational level of citizens is one of the essential determinants for citizens’ opinions on immigration. Low-skilled or manual workers are more likely to have negative attitudes toward immigrants (Dustmann & Preston, 2007; Hatton, 2016) and are more likely to support authoritarian populists (Norris & Inglehart, 2019). Feelings of job insecurity can be one of the explanations of this mechanism, which is part of the dependent variables in the analysis. However, it is also possible that the characteristics of low-skilled groups make them feel threatened more by immigrants in other fields, like culture. This makes it necessary to add the educational level of citizens as a control variable in both the national and individual analyses.
Finally, Norris and Inglehart (2019) show that age, gender, urbanization, religiosity, parenthood, and marital status are also significant determining factors for the chance that an individual supports authoritarian values. Women are less likely to support authoritarian parties (Norris & Inglehart, 2019). While the values of men and women toward the radical right ideology are primarily the same, their populist radical right actions differ (Mudde, 2007, pp. 90-118). The different socialization of men and women is a possible explanation. Women are more likely to vote for established conservative parties because they do not like new and extreme parties. A comparable mechanism appears when the effects of religiosity are examined (Mudde, 2007 pp. 85-86). If religious authorities denounce authoritarian populist or anti-immigrant parties, it is unlikely that they will gain much support from religious voters. Examples of this exist in France and Germany, and there is further proof for this mechanism where the populist radical right is supported by some religious leaders, like Poland and Slovakia. Overall, religious Europeans seem to be more likely to support authoritarian populists (Norris & Inglehart, 2019). Married or divorced citizens and citizens with minor children are also more likely to vote in favor of authoritarian populists. Hatton (2016) uses comparable control variables to predict the opinion of individuals toward immigration and adds variables that show whether individuals themselves were born in another country or are part of an ethnic minority. So, whether a respondent is a native citizen or not is also a control variable in the analysis.
19
METHODOLOGY
OPERATIONALIZATION
Data from the European Social Survey (ESS) and the Organization for Economic Cooperation and Development (OECD) are used to operationalize all variables. The most recent data published by the ESS were collected in 2018 and published in the fall of 2019. These data are used for the cross-sectional analyses on a micro-level. The national analyses are conducted with data from all countries that participated in at least two of the surveys, and which have data available on the relevant issues in the datasets of the OECD. These 26 countries are Austria, Belgium, Czechia, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Luxembourg, the Netherlands, Norway, Poland, Portugal, Slovakia, Slovenia, Spain, Sweden, Switzerland, Turkey, and the United Kingdom. For the longitudinal analysis between countries, data are used from all nine rounds of the ESS survey, which started in 2002. This leads to a total of 416 observations. However, in most analyses, the number of observations will be substantially lower, given the fact that the ESS is only conducted once every two years and considering that there are countries that did not participate in all ESS rounds.
For the micro-level analysis, the number of countries is more limited as the ESS had only published data of 16 at the moment of the conduction of the analysis. These countries are Austria, Belgium, Czechia, Germany, Estonia, Finland, France, Hungary, Ireland, Italy, The Netherlands, Norway, Poland, Slovenia, Switzerland, and the United Kingdom. The ESS has drafted a specification to ensure the accuracy of the data in countries, which will be discussed in more detail in the reliability and validity paragraph of this methodology (European Social Survey, 2018a).
DEPENDENT VARIABLES
Anti-immigration sentiments are the first variable to be operationalized. Both on a macro- and micro-level, this is done by using questions of the ESS. The survey asked respondents whether immigrants positively or negatively affect the country’s economy, if they undermine or enrich the cultural life, and whether they make the country a better place to live or not (European Social Survey, 2018b). An analysis shows that the answers to these questions have a high degree of intern consistency. The Cronbach’s Alpha is 0.891 on a macro-level and 0.881 on a micro-level, where a value of at least 0.7 is necessary to be able to speak of consistency (Salkind, 2014). The intern consistency makes it possible to not only use the questions separated but also combine them into one variable. The scale of these variables will be zero to ten, where a low value relates to a negative view of immigration. The average of a country is calculated for each year to be able to use these variables on a national level.
20 The other dependent variable is the support for authoritarian populist parties in a country. It is less simple to operationalize this variable straightforwardly for two reasons. The first problem has to do with the different electoral systems in European countries. As Norris and Inglehart (2019, p. 53) summarize, many studies have proved that the electoral system has an essential impact on the number of parties that wins seats. Minor parties of which the support is spatially dispersed, like authoritarian populist ones, are not likely to gain seats in majoritarian systems. In systems with proportional representation, relatively high district magnitude, and a relatively low threshold, they are more likely to win a substantial number of seats. A clear example of this can be found in the general election of 2015 in the United Kingdom, where the UK Independence Party (UKIP) gained only 0.2 of the seats in the House of Commons, while 12.6 percent of the British electorate voted for UKIP (Hawkins, Keen & Nakatudde, 2015). This makes it impossible to use the percentage of seats in the national parliament as a variable. The percentage of national votes is also considered to be not suitable, as voters are likely to adapt their votes to the electoral system. If they live in a district in which the chances for their favorite party are limited, they might choose for their second-best option. So, voting results cannot be used as the operationalization of support for authoritarian populism. This assumption is strengthened by the fact that elections are only held once in a few years, which would mean that the use of this variable would lead to a substantial loss of observations. As a solution, the European Social Survey (2018b) will be used to operationalize the support for authoritarian populist parties. The survey asks the respondent toward which party he or she feels closest. In this way, the problem of the electoral system is solved to a large extent.
This leads to the second problem of operationalizing the support for authoritarian populist parties: the classification of parties. It is debatable which parties are authoritarian populist and which are not. The categorization of political parties is often too simplistic but is, at the same time, required to be able to conduct analyses (Norris & Inglehart, 2019, p. 216). Mudde (2017, pp. 58-59) concludes that there are parties that are often not recognized as populist radical right, while they do meet the conditions. At the same time, some parties are often perceived to be populist radical right, while they are not. Some parties are right populist, but not radical, while others are radical right, but not populist. Furthermore, Mudde (2007) excludes extreme right parties because he considers them to be undemocratic. Studies can use party affiliations, party names, rhetorical and discourse analyses, party elites, party activists, party supporters, and issue positions to determine the position of a party (Norris & Inglehart, 2019, pp. 223-229). Most studies choose for the last option and often use the Comparative Manifesto Project (CMP) to determine the issue salience and issue positions of parties. These studies often used the left-right cleavage, but this cleavage has lost power as it does not reflect all modern cultural cleavages. Besides, Norris and Inglehart (2019) argue that there are questions about whether the CMP can capture the core characteristics of populism. For these reasons, the authors have used data from the Chapel Hil Expert Survey in which experts
21 are asked to estimate the ideology and policies of parties in their country and supplemented these by data from the CMP. Norris and Inglehart (2019, pp. 229-247) use these data to place parties on a three-dimensional scale with the dimensions of populism-pluralism, authoritarianism-libertarianism, and the left-right cleavage. This study has mainly used the definitions of Norris and Inglehart (2019) to explain the concepts of authoritarian populism and will also use the classification of these authors to decide which parties are authoritarian populist.
It is problematic that these authors base their classification on data from a period between 2000 and 2015. They emphasize that their categorization is only a snapshot, as parties’ positions can change (Norris & Inglehart, 2019). Furthermore, some new parties that might be considered authoritarian populist are not categorized. The Forum for Democracy (FvD) in the Netherlands is an example of such a party. The party gained only 2 out of 150 parliamentary seats in the general election of 2017 but became the biggest party in the provincial elections of 2019, which indirectly determine the composition of the Dutch Senate (Schaart, 2019). The party is known for its opposition against immigration, the traditional media, and the current political elite. It wants to put the power back to the people by organizing referendums, for example, about the Dutch membership of the European Union. The Freedom and Direct Democracy (SPD) party in Czechia was founded in 2015 and is therefore also not categorized by Norris and Inglehart (2019). Authoritarian populist politicians like Ms. Le Pen and Mr. Wilders came to Prague to support the SPD in the campaign for the 2019 European Parliament election (Mortkowitz & Bauerova, 2019). This party is also Euroskeptic, anti-immigration, and attacks the political elites. The FvD and SPD both seem to meet the characteristics of authoritarian populist parties and are, therefore, added in this study.
The recent founded Brexit Party in the United Kingdom does not need to be added, as the foundation of this party took place after the conduction of the European Social Survey (2018b). A noticeable party to discuss is the Conservative People’s Party of Estonia (EKRE). Norris and Inglehart (2019) do not categorize the EKRE as an authoritarian populist party. However, the party is a member of the European Parliament group Identity and Democracy that further consists of parties that are all on the list of authoritarian populist parties (ID-Party, 2019; Norris & Inglehart, 2019). A recent study of Petsinis (2019) tried to determine whether the EKRE is actually part of the European radical right. The author argues that the current classification schemes should be updated, as domestic ethnopolitics, regional geopolitics, and anti-immigration politics are the main characteristics of the radical right in Central and Eastern Europe. He does, however, argue that the EKRE is part of the populist radical right (Petsinis, 2019), and therefore this party is added to the list of authoritarian populist parties.
Table 3 contains a full list of all authoritarian populist parties. The Five Star Movement in Italy, the Coalition of the Radical Left in Greece, and Podemos in Spain are not listed because Norris and Inglehart (2019, p. 237) define them as libertarian populist parties. It is noted that not all countries have authoritarian
22 populist parties in their political system: Iceland, Ireland, Portugal, and Spain. The first three countries do not seem to have any parties that qualify for being authoritarian populist. Spain’s Vox party might, by some, be seen as authoritarian populist. The party is indeed far-right but much more nationalistic than populist (Ferreira, 2019), and therefore not included as an authoritarian populist party.
Table 2. Operationalization of the dependent variables.
Variables Micro-level (cross-sectional) Macro-level (longitudinal & cross-sectional) 1. Immigration sentiments Average of three answers (bad 0- good 10) Average of three answers (0-10, average)
1.1: Immigrants on
economy Immigration bad or good for country’s economy (0-10)
Immigration bad or good for country’s economy (0-10, average) 1.2: Immigrants on
cultural life
Immigration undermines or enriches cultural life of country (0-10)
Immigration undermines or enriches cultural life of country (0-10, average)
1.3: Immigrants on place to live
Immigration makes country worse or better place to live (0-10)
Immigration makes country worse or better place to live (0-10, average)
2.
Support for
authoritarian populist parties
Feels closest to authoritarian populist party (0-1)
% of respondents that feels closest to authoritarian populist party Note. All data for these variables will be retrieved from the European Social Survey (2018b).
23 Table 3. Classification of authoritarian populist parties based in Europe.
Country Authoritarian populist parties Austria Freedom Party of Austria (FPÖ) Belgium Flemish Block (VB)
National Front (FN)
Czechia Freedom and Direct Democracy (SPD)* Denmark Danish People’s Party (DF)
Estonia Conservative People’s Party of Estonia (EKRE)* Finland Finnish Party – True Finns (SP-P)
France National Rally (RN, earlier called National Front) Germany Alternative for Germany (AfD)
Greece
People’s Association – Golden Dawn (XA) Independent Greeks (AE)
Popular Orthodox Rally (LAOS)
Hungary Fidesz – Hungarian Civic Union (Fi-MPSz)
Iceland N/A
Ireland N/A
Italy Northern League (LN)
Brothers of Italy – National Centre-right (FdI-CN) Latvia For Latvia from the Heart (NsL) Latvian Association of Regions (LRa)
Luxembourg Action Comm. Pensions | Alt. Demo Ref (AR|ADR) The Netherlands
Pim Fortuyn List (LPF) Party for Freedom (PVV) Political Reformed Party (SGP) Forum for Democracy (FVD)* Norway Progress Party (Fr)
Poland Law and Justice (PiS)
Real Politics Union – Congress of the New Right (UPR|KNP)
Portugal N/A
Slovakia
Ordinary People and Independent (OLaNO) Christian Democratic Movement (KDH) Slovak National Party (SNS)
Slovenia Slovenian Democratic Party (SDS)
New Slovenia – Christian People’s Party (NSI)
Spain N/A
Sweden Sweden Democrats (SD) Switzerland
Swiss People’s Party (SVP-UDC)
Federal Democratic Union of Switzerland (EDU-UDF) Ticino League (LdT)
Turkey National Action Party (MHP) United Kingdom UK Independence Party (UKIP)
British National Party (BNP)
Notes. Based on Norris and Inglehart (2019). *Party that Norris and Inglehart (2019) do not define as authoritarian populist. Explanations for the decisions that were made can be found in the main text. On a micro-level, the variable support for authoritarian populist parties can have two values: closest to an authoritarian populist party (1) or not (0). The same variable has been created for all individuals in the
24 longitudinal analysis, and the percentage of the population that feels closest to an authoritarian populist party is calculated for each country in each year.
INDEPENDENT VARIABLES
As mentioned in the conceptualization paragraph, the economic situation is conceptualized in objective and subjective ways. This paragraph will discuss all the economic variables. The objective micro-level operationalization of the economic situation will focus on the respondent’s personal situation by using data from the European Social Survey (2018b). The survey asked its respondents if they are employed or not. Besides employment, the household’s net income is used as a variable. The ESS ranks these incomes in deciles from 1 to 10 (European Social Survey, 2018b).
The most common way to operationalize the objective state of a country’s economy is the growth of the Gross Domestic Product (GDP). GDP is one of the variables in the macro-level analysis, as well as unemployment. As discussed, welfare spending and budget deficits can also be essential factors in explaining opposition against immigration. Citizens worry that their governments spent their money on immigrants and that this increases the budgetary problems of their country (Hatton, 2016). Studies on welfare state retrenchment have used different conceptualizations of welfare spending (Green-Pedersen, 2004). The used definitions are often too small, but broader definitions are unworkable. This study will use the total social spending in the percentage of GDP. It might be more valid only to use cash benefits, or to make a distinction between different social welfare programs, but this would lead to a significant loss of observations, as these data were not available for 2018, 2017, and for some countries even 2016. All objective macro-economic data come from OECD (2019) datasets.
The study uses data from the European Social Survey (2018b) for the subjective economic situation. Respondents were asked how satisfied they are with the present state of the national economy on a scale from zero to ten. The average value in each country for this variable has been computed for all years in which the ESS took place to be able to use the variable in the macro-level analyses. The feeling about the current income of the household will be added in the individual analysis, as this might differ from the satisfaction with the national economy.
25 Table 4. Operationalization of the independent variables.
Note. *Data from European Social Survey **Data from OECD.
CONTROL VARIABLES
The conceptualization mentioned all control variables of this study. Unfortunately, it is not possible to use all control variables in both the micro-level and the macro-level analyses. This paragraph will discuss all the control variables.
As discussed, the cultural backlash is conceptualized by generational replacement and opposition against the post-materialist values of the so-called silent revolution. On a micro-level, the respondent’s age is the operationalization of the generational replacement. On a macro-level, this study will not look at the average age, but at the share of the population that is above 50 years old. The opposition against freedom of the LGBT community is the variable used to operationalize the opposition against post-materialist values. LGBT rights are one of the characteristics of the silent revolution mentioned by Norris and Inglehart (2019). It would have been ideal to use more characteristics, but this would have caused a decrease in observations as the ESS did not ask all questions in every edition of the European Social Survey (2018b). As this study focuses on strong opposition against these post-materialist values, the analysis will make a distinction between respondents who strongly disagree with freedom of the LGBT community (1) and those who gave a different answer (0). For the macro-level analysis, the variable is operationalized by a computation of the share of the respondents that strongly oppose LGBT rights for every country in each year.
Variables Micro-level (cross-sectional) Macro-level (longitudinal & cross-sectional) 3. Objective economic
situation
3.1 GDP N/A GDP growth (%)**
3.2 Unemployment Currently unemployed (0-1)* Unemployment (% of population)**
3.3 Welfare spending N/A Welfare spending (% of GDP)**
3.4 Government debt N/A Government debt (% of GDP)**
3.5 Income household Household’s total net income (1-10th decile)* N/A
4. Subjective economic situation
4.1 Satisfaction with present state of the economy (0-10)*
Satisfaction with present state of the economy (0-10, average)* 4.2 Feeling about household’s
income (Very difficult 1 – Comfortable 4)*
26 The conceptualization already explained that data on individuals’ estimations of the number of immigrants are not sufficiently available. The European Social Survey (2018b) asked respondents to estimate this in 2014 for the last time. As a result of this, this study decides to leave this control variable out of the micro-level analysis. The macro-micro-level analysis consists of data from the OECD (2019) to determine the objective number of immigrants coming into countries and the share of foreign-born people in countries for each year.
To be able to compare the educational level of different countries, the European Social Survey (2018b) uses the International Standard Classification of Education. The levels vary from 1 (less than lower secondary education) to 7 (higher tertiary education). As there is no classification for all countries for each year, it is not possible to use the educational level as a variable in the macro-level analysis. Gender, religiosity, parenthood, marriage, and the country of birth of the respondent are also control variables in the individual analysis. Of these variables, only religiosity is part of the macro-level analysis as the other variables are considered to be irrelevant for the macro-level analysis or because they would have caused a substantial decrease of observations. Finally, the percentage of people living in an urbanized area in each year and each country will be used in the macro-level analysis. Unfortunately, the European Social Survey (2018b) has no variable available about whether the respondent lives in an urbanized area, so this variable cannot be used in the individual analysis.
27 Table 5. Operationalization of the control variables.
Variables Individual (cross-sectional) National (longitudinal & cross-sectional) 5. Cultural backlash
5.1 Generation Age* Percentage older than 50**
5.2 Opposition
post-materialist values Strong opposition toward LGBT-rights (No 0 – Yes 1)* Strong opposition toward LGBT-rights (% Respondents)* 6. Immigration
6.1 Inflow N/A Inflow immigrants (% Population)**
6.2 Foreign born N/A Foreign born (% Population)**
7. Educational level Highest level of education
individual (ES-ISCED, 1-7)* N/A 8. Gender Individual (Male 1 – Female 2)* N/A
9. Religiosity How religious is respondent (Not 0 – Very 10)* How religious is respondent (0-10, average)* 10. Parenthood Ever given birth to/ fathered a child (No 0 – Yes 1)* N/A
11. Marriage Are or ever been married (No 0 – Yes 1)* N/A 12. Respondent born in foreign
country
Born in foreign country (No 0 –
Yes 1)* N/A
13. Urbanization N/A % living in urbanized area**
Notes. In the analysis on a national level, dummy variables will be used for countries and years. In the analysis on an individual level, dummy variables will be used for countries. * Data from European Social Survey **Data from OECD.
