Explaining Right-wing Populism
A Macroeconomic Perspective
Abstract In the past decade, there has been a major setback for globalization. Populist sentiments and increasing nationalism threaten the future of global trade. This thesis discusses which domestic macroeconomic conditions determine the rise of right-wing populist political parties in Europe. Previous research has looked at the microeconomic level, or at individual voter characteristics, I will add a macroeconomic perspective to the field. I will analyze the rise of populism by using the economic security or economic inequality perspective, and develop an econometric model that looks at the relationship between voting for right-wing populist parties and the volume of international trade, unemployment, and economic growth. My main findings are that low GDP per capita growth, decreasing unemployment, and increasing international trade determine the rising support for right-wing populist political parties in Europe.
Keywords right-wing populism, macroeconomic variables, elections, economic insecurity
JEL classification D72, F62, F69
Master Thesis
2
1 Introduction
In the past decades, rapid globalization has brought net total wealth gains all over the globe. Many emerging markets have been able to raise living standards, helped by declining protectionism. Increased integration of national economies has led to the creation of global production chains. All these developments are perceived to be positive by most. However, in the past decade, there has been a major setback for globalization. Populist sentiments and increasing nationalism threaten the future of global trade. Especially in developed countries, such anti-globalization sentiments have come to the surface. Increasingly, people feel that global production chains are taking away jobs and economic security. Standard economic theory, starting from Ricardo, argues that there are only winners from trade. However, recently people notice that there are also losers from trade and many Europeans feel that they are losing. This has led to increasing populist sentiments in Europe and the electorate has started to lend their support to populist parties. A prominent example is the election of Donald Trump as president of the United States (US). In addition, the majority of the population of the United Kingdom (UK) voted Leave in the referendum on 23 June 2016, when asked whether the UK should stay in the European Union (EU). Others have voiced the concern among many Europeans, who fear the rise of populism. Many politicians of the established political parties in Europe also reacted negatively to the increasing support for populist parties.
Broadly, there are two different types of populism, left-wing and right-wing populism. In the next section this distinction will be analyzed in more detail, using the argument of Dani Rodrik (2017). However, it is important to note here that right-wing populism mainly targets foreigners and minorities, they are to blame for the loss of national identity and this viewpoint especially gains attention in times of economic downturn. Support for right-wing populism is especially large in Europe, whereas left-wing populism is prominent in Latin America. Left-wing populism mainly focuses on landowners and large corporations, as they are argued to be accumulating all the benefits of capitalism.
3 populism opposes capitalism and thereby rejects the presence of large multinational corporations, but the main implication is increasing anti-globalization sentiment.
The topic of this thesis is to look into what economic elements give rise to the increasing support for populism in Europe. Especially in the past 15 years this subject gained interest among academic scholars. Some researchers tried to explain the rise of populism by looking at the characteristics of voters (Norris and Inglehart, 2016, Lubbers et al, 2002, Arzheimer, 2012, and Rooduijn, 2014), whereas other researchers have looked at economic and trade motives (Rodrik, 2017, Swank and Betz, 2003, Autor et al., 2017, and Colantone and Stanig, 2017).
Here, the focus will be on right-wing populism, as it is more prominent in Europe, mainly targeting foreigners and immigrants. It will be interesting to see whether the domestic macroeconomic conditions and influence of globalization also interact with the support for right-wing populism. Furthermore, there has been an increase in the support for right-wing populist parties since 1980, shown in Figure 1, taken from Johansson Heinö (2017). He presented a similar figure for left-wing populist parties, which shows that there has been a significant decrease in electoral support until 2006, before showing a comeback in the past decade.
Figure 1: Average electoral support right wing populist parties, source: Johansson Heinö (2017). Note: The graph shows average voter support for right-wing populist parties for 1980-2017.
4 parties, as recent elections in the Netherlands, Germany, and France did not lead to the expected outcome for these parties. However, these do not mean that the right-wing populist parties lose influence in parliament or dominance in the media. Since right-wing populism is especially prominent in Europe, it is interesting and relevant to choose this continent for my research.
I will analyze the rise of populism by using the economic security or economic inequality perspective, prominently brought forward by Norris and Inglehart (2016). According to these authors; “This perspective emphasizes the consequences for electoral behavior arising from profound changes transforming the workforce and society in post-industrial economies. (…) According to this view, rising economic insecurity and social deprivation among the left-behinds has fueled popular resentment of the political classes.” Several academic authors (e.g. Norris and Inglehart, 2016, and Swank and Betz, 2003) looked at economic conditions that give rise to the support for populism by looking more at a micro level. This thesis will test this premise from a different angle compared to what most others have done before. I will assess the consequence of changes in domestic macroeconomic conditions and the influence of globalization, being the volume of international trade, unemployment, and economic growth. This is important, because globalization has led to increased migration, which is perceived to have led to job loss in advanced economies. As argued, rising anti-globalization sentiments lead to an increase in support for populist parties, therefore, it is sensible to analyze the relationship between populism and macroeconomic factors. Assessing several macroeconomic indicators and comparing the impact these have on the rise of populism, will provide a good indication of what macroeconomic indicator might best predict an increase in votes for populist parties in Europe.
Bringing all elements of this master thesis together, the main research question is:
What domestic macroeconomic indicators are most important in determining the rising support for right-wing populist political parties in Europe?
5
2 Defining populism
The main element to explain first is how to define ‘populism’, and especially right-wing populism, since that is central here. By discussing this, I will lay the foundation for the remainder of the thesis.
When asking the public about a definition of populism, this will probably result in a wide variety of answers, which is not surprising, since even academic authors do not agree on a particular definition. Gidron and Bonikowski (2013) claim that the term ‘populism’ is both widely used and widely contested. Here, some of the leading definitions will be outlined and the particular definition I follow in this thesis – a definition based on Van Kessel (2015) – will be explained.
A leading author in the field of political economics and an expert in right-wing populism is Cas Mudde. Mudde (2004) explains that there are two different interpretations of the term populism, in the first one, populism refers to the highly emotional and simplistic discourse that is directed at the ‘gut feelings’ of the people. By this, he means that the people react positively to populist politicians, because they know what topics the people think are important and they use simple terminology. The second interpretation somewhat follows from this, and describes populism as opportunistic policies by policymakers with the aim of pleasing the people or voters. Thereby they ‘buy’ support from the voters for the next elections and ignore a possible ‘better’ solution. One typical example is lowering taxes just before the elections.
Mudde (2004) defines populism as an ideology that considers society to be ultimately separated into two homogeneous and antagonistic groups, ‘the pure people’ versus ‘the corrupt elite’, which argues that politics should be an expression of the volonté générale (general will) of the people.
6 be left-wing oriented. In contrast, political populism looks at the feelings of the people, the sentiments surrounding political decision-making.
Guiso et al. (2017) find the definition of populism in the Encyclopedia Brittanica very useful. This entails that populists claim to promote the interest of common citizens against the elites; they pander to people’s fears and enthusiasm; and they promote policies without regard to the long-term consequences. From this definition they highlight three important concepts: 1) the claim to be on the side of the people against the elite, they label this the “supply rhetoric”; 2) the “fears or enthusiasms” of people against the elite, the demand conditions to which populists pander; and 3) the disregard of longer-term consequences. Rodrik (2017) also claims that the first two concepts are very important, as there are both the demand and supply side that play an important role in the rise of support for populist parties. There is the demand side: the distributional and other economic fault lines created or deepened by globalization that generate potential public support. In addition, there is the supply side: populist movements supply the narratives required for political mobilization around common concerns. It is important to note this distinction and in this respect, my analysis will look at the demand side. I want to look at the demand side, because it is interesting to analyze how people in their voting behavior react to macroeconomic conditions or changes.
Acemoglu et al. (2013) explain that the label ‘populist’ is often used to emphasize that politicians use a rhetoric that aggressively defend the interests of the common man against the privileged elite. This supposed conflict between the common man and the privileged elite is mentioned in all the definitions we have seen so far and seems to be a key part of the definition.
7 To summarize, there are many differing views on how exactly to define the term populism, however, they all share some aspects. The definition of Van Kessel (2015) is very clear and considers all of these aspects and it is a definition of what a populist party is, making it especially useful in this thesis. Therefore, his definition of populism will be followed here. Van Kessel (2015) defines a party as populist if it a) portrays “the people” as virtuous and essentially homogeneous; b) advocates popular sovereignty, as opposed to elitist rule; c) defines itself as against the political establishment, which is alleged to act against the interest of the people (Guiso et al., 2017).
Right-wing populism
As mentioned in the introduction, there is a clear distinction between right-wing and left-wing populism. In this thesis I look at right-wing populism, so I will discuss it further detail here.
The distinction between both strands of populism is emphasized by Rodrik (2017); right-wing populism targets foreigners or minorities, while left-wing populism targets the wealthy and large corporations. Rodrik (2017) finds that in Europe the support for populist parties has increased sharply over the past three decades, with right-wing populism being the only driver of this increase. The opposite case holds for Latin America, in which support for left-wing populism has been staying at a high level. The focus in this thesis will be on the macroeconomic domestic variables that explain the rise of right-wing populism in Europe. Therefore, it is useful to explain what right-wing populism is, and why it is different from left-wing populism.
8 which leans more directly to economic populism, while right-wing populism is more about politics, without downplaying the role of economics.
Mudde and Kaltwasser (2012) indicate, there are very few parties that identify themselves as openly right-wing. Instead, some of these parties would claim they are neither left, nor right. The best distinction, which Mudde and Kaltwasser (2012) have found in their literature review, is by Bobbio (1996). He defines left and right parties on the basis of the key feature of (the propensity to) egalitarianism. The left considers inequalities between people to be artificial, something that can be overcome by active state involvement. The right believes that these inequalities are natural and outside the purview of the state.
9
3 Literature Review
The goal of this section is to provide an overview of existing literature on the topic of this thesis, both empirically and theoretically. This section will start with the theory literature, in order to provide the right theoretical background. Thereafter, the empirical work related to this topic will be discussed, this also forms the basis for the hypotheses that I will set out here.
As explained in the introduction, this thesis will analyze the rise of populism by using economic indicators. In this context, the economic security or economic inequality perspective, prominently brought forward by Norris and Inglehart (2016), is of importance. Their main theoretical point of view is that rising economic insecurity and social deprivation create increasing support for populist parties. Particularly, changes in society with regard to economic welfare and transformation of the workforce are seen to push people towards populist parties.
This view is discussed in a recent paper by Rodrik (2017), which tries to describe the economic and historic reasons that cause a rise of populism, in the presence of high levels of globalization. Rodrik (2017) refers to the Stolper-Samuelson theorem (S-S), which entails a model with two goods, two factors of production, and full labor mobility. An increase in the price of a good will cause an increase in the price of the factor used intensively in that industry and a decrease in the price of the other factor. In this case, Rodrik (2017) refers to the S-S because he wants to point out that opening up to international trade makes at least one factor of production worse off (under competitive conditions). Rodrik (2017) indicates that there are always losers from trade, as it produces both winners and losers. Therefore, the process of globalization gives rise to populism, since in Europe and the United States (US) there is a growing belief that people are losing from globalization. It is often argued that the participation in Global Value Chains (GVC) transfers jobs from Europe and the US, to emerging markets (Gereffi, 2014). For example, in the US the right-wing populists blame China for the decrease in employment in their home country.
10 advanced economies. However, Rodrik (2017) argues that globalization and trade are blamed because it is easy to pinpoint the enemy, like China. Technological change or progress is harder to blame.
Additionally, Timmer et al. (2013) argue that there is not necessarily a net loss of jobs in Europe due to production fragmentation. Instead, in their research of 27 European countries between 1995 and 2008, they find that in most countries manufactures GVC job increase in services is even higher than job loss in manufacturing. Also, most European countries were able to keep the wage level for manufactures jobs at roughly the same level or show a small increase. The authors also analyze the skill structure of GVC workers and find a shift away from low-skilled towards high-skilled workers. The main resistance against global production fragmentation might therefore well come from low-skilled workers.
Rodrik (2017) argues that European populism is different from US populism as it is not against trade. This is likely due to the better developed welfare states in Europe. Consequently, there is a better redistribution of the benefits of trade and globalization in Europe. This is also what Swank and Betz (2003) imply in their research, a better developed welfare state is better able to redistribute the gains of trade among winners and losers. Therefore, in these states, there is less support for right-wing populist parties.
In the past decade there has been increasing interest among academic researchers for the rise of populism. Various authors have looked into the causes of this rise from different perspectives, here the empirical literature on the determinants of populism is presented.
The first line of empirical evidence relates to what Rodrik (2017) has argued theoretically, the volume of international trade might affect the support for right-wing populist parties. Advanced economies have increasingly started to import both goods and services, and hence there has been an increase in the volume of international trade. Unfortunately, an increase in international trade is associated with job losses, for example there are Asian low-wage countries that compete away domestic industries in Europe, simply because they can produce more cheaply. This is also found or hinted at by various authors.
11 or the decline of manufacturing employment, increases the support for radical right-wing parties. This relates to the argument made before, global integration of production and production fragmentation lead to job losses among low-skilled workers in advanced economies. As a consequence, support for right-wing populist parties is increasing, as they promise to secure those jobs. Secondly, these authors find that general economic performance does not have a significant effect on the votes for these parties. I will return to this discussion later in this section. Then, there are two political reasons to explain the electoral success of radical right-wing parties, these parties benefit from a system with proportional representation, as opposed to majoritarian electoral systems. And, the success of these parties is less when there are electorally successful established right parties, who are more consensus-oriented and have embraced centrist policies.
Autor et al. (2017) argue that the electoral consequences of trade are distinct from those of generalized changes in labor demand. What distinguishes trade in terms of impact on electoral and political outcomes is that its disruptive effects are so concentrated demographically and geographically. There has been an enormous loss in manufacturing jobs in the US due to import competition from China. Rising import penetration has a disproportional effect on local labor markets specialized in labor-intensive manufacturing. Colantone and Stanig (2017) carry out a similar analysis, finding that a stronger import shock leads to 1) an increase in support for nationalist parties; 2) a general shift to the right in the electorate; and 3) an increase in support for radical right parties. Authors like Jensen et al. (2017) reach similar conclusions. These studies also look into macroeconomic indicators, however there is hardly any evidence on this perspective with regard to Europe, where we see a continuous increase in right-wing populism. Therefore, I have set up the first hypothesis:
Hypothesis 1: An increasing volume of international trade will increase the support for right-wing populist parties.
12 reaction, especially among the older part of the electorate, who want the value change in European societies to stop. For economic inequality, the indicators used by the authors are skilled and unskilled manual workers, unemployment, urbanization, and living on social benefits. For cultural values, some of the variables Norris and Inglehart (2016) used are authoritarian values, anti-immigration and mistrust in national governance. The control variables are age, sex, education, religiosity, and ethnic minority. They find more evidence for the cultural backlash thesis than for the economic security perspective. One important finding is that unemployed people were more inclined to support the Brexit than employed people. It is important to note here that the authors are looking at populism in general and take both right-wing and left-right-wing populism into account.
Norris and Inglehart (2016) base their research partly on the Brexit, and various other authors similarly looked at reasons for the Brexit, or Vote Leave in the United Kingdom (UK). Becker et al. (2016) explain that their paper and the Brexit vote it studies is not solely relevant in a European Union (EU) perspective, but is also related to populism. Meaning that the referendum and its outcome are an example of populism-driven sentiments. They find that older people, the lower educated, and areas with low pay and high unemployment were more inclined to Vote Leave. Furthermore, they find strong evidence that immigration was linked to the Brexit Vote. Additionally, fiscal cuts also increased support for leaving the EU. The authors argue that reliance on central government grants are a proxy for deprivation. Areas that were subject to more severe fiscal cuts tend to be those with relatively more deprivation. The authors therefore conclude that there is a correlation between fiscal cuts and the determinants of the Vote Leave share. The less well-off in society are hurt by fiscal cuts and therefore were more inclined to vote Leave.
13 There are even some authors that found that unemployment has a negative relationship with voting for right-wing populist parties. When using a model estimated using the ordinary least squares estimator, Knigge (1998) concludes that unemployment does not have the expected effect on voter support for right-wing parties. Her results suggest that higher levels of unemployment do actually lead to decreasing support for right-wing extremism. Knigge (1998) argues that when unemployment becomes a real problem in the economy, people tend to rely on mainstream political parties. Extreme right-wing parties lose appeal when the economic problems, like unemployment, are pronounced and salient, the electorate then rather trust the mainstream political parties with a longer track record of solving these issues. Stoyanova (2016) reaches a comparable conclusion, that the coefficient’s sign for unemployment is opposite to what was expected. This indicates that the right-wing political parties vote share decreases, when unemployment increases.
This contradictory evidence on unemployment and voting for right-wing populist parties makes it interesting to test for my sample. An increase in unemployment typically relates to the economic security perspective. People value having a job and are dependent into a large extent on their job for receiving an income. Based on this and the majority of the evidence from the aforementioned empirical work, the second hypothesis is:
Hypothesis 2: An increase in unemployment will increase the support for right-wing populist parties.
The last empirical relationship I want to discuss here is the one between GDP per capita and populism. Dornbusch and Edwards (1991) argue that populist policymakers, and the population at large, are deeply dissatisfied with the economy’s performance, they feel that the situation should and can be improved. This is an initial condition for growing support for right-wing populism. Typically, very moderate growth, stagnation, or outright depression increases the support for the right-wing populist parties.
14 increase in support of populist parties after the globalization shock and the 2008-2011 financial and sovereign debt crisis. They prove that populist parties are likely to gain support when mainstream parties and national institutions fail to manage the shocks faced by their economies. Institutional constraints increase frustration among voters and result in a turn towards populist parties, this especially holds for Eurozone countries.
Funke et al. (2016) even find that in general financial crises are often followed by partisan conflict. They base this on election data from 1870 to 2014 in 20 developed economies. As a consequence of a financial crisis, uncertainty and polarization tend to increase. Voters seem to move towards right-wing populist parties that attribute blame to minorities and foreigners. Other authors find a decreasing support for incumbent parties after the Great Recession (Hernandez and Kreisi, 2016; Dustman et al., 2017; Foster and Frieden, 2017). Based on this evidence I have set up the third and final hypothesis:
Hypothesis 3: A decrease in growth of GDP per capita will increase the support for right-wing populist parties.
This thesis will empirically test these three different hypotheses, which are based on what is discussed in the literature review. The relevant theory is the economic security or economic inequality perspective and all hypotheses are largely based on this. Also, the hypotheses can be directly linked to an increase in globalization, and this is ultimately the goal of this thesis, to explain the link between globalization and increasing support for right-wing populist parties.
To do this, I also need to include relevant control variables, which I will also base on existing literature. These are not of an economic nature, as groups in society can also be categorized on the basis of various characteristics. These characteristics are for example; sex, education, age, religion. Several authors have looked into who are more inclined to support right-wing populist parties.
15 Norris and Inglehart (2016) find in relation to the Brexit vote. But also Arzheimer (2012) reaches the conclusion that young(ish) people are more likely to support the extreme right, as well as males, people concerned about immigration and with moderate educational achievement. Somewhat similarly, Rooduijn (2014) argues there is no consistent proof that the voter bases of populist parties consist of individuals who have lower incomes, come from lower classes, or hold a lower education. Moreover, Euroscepticism, political distrust and a negative attitude towards democracy, are also no reasons for supporting populist parties. This is highly surprising and contrary to what other authors find, however it should be noted that this author looks at both left-wing and right-wing populist parties. His analysis suggests that ‘the’ populist voter does not exist.
Then, there are two more different models used in order to explain voting for right-wing populist parties. Firstly, Ivarsflaten (2008) analyzes from the grievance mobilization models. Grievances over immigration, economic changes, and elitist and corrupt politics were successfully mobilized by political parties in Western Europe according to Ivarsflaten and are explanations for why people are supporting these parties.
Secondly, there is the protest vote theory. According to this model, protest voters are rational individuals who use their vote to support this parties that are an outcast in the political arena in order to express their discontent with the traditional political parties. It is a vote against the main political actors, not a vote that is motivated by policy-driven factors (Serani, 2016). This theory is debated by Van der Brug et al. (2000), who argue that the protest vote does not exist. In their research this is based on anti-immigrant parties who do not attract more protest votes than other parties do.
16
4 Methodology
This section will explain the empirical strategy that is used in this research. The main element here is the model specification which sets out which variables are used. I will not explain here what data sources were used, this will be discussed in the ‘Data’ section.
To analyze whether my hypotheses hold, this research will analyze the outcomes of 89 parliamentary elections in 24 European countries in the period 2000 until 2013. By looking at parliamentary elections, I follow Swank and Betz (2003), who argue correctly that these are relatively consistent in structure and political importance, both across nations and time. The following model specification will test the effects the three macroeconomic variables have on the popular vote for right-wing populist parties (RWPP) during parliamentary elections.
RWPP(t) = α + β1(TRADE(t-1)) + β2(UNEMP(t-1)) + β3(ΔGDP_p(t-2, t-1,t)) + ΣβiXi + ε (1)
Where the dependent variable RWPP is the popular vote for right-wing populist parties. TRADE is the volume of international trade, and UNEMP is the unemployment rate. ΔGDP_p is the growth rate of per capita GDP, and Xi’s are a set of control variables.
I use the following control variables; ASYLUM, AGE, CRISIS, and EDUCATION. The variable
ASYLUM is the number of asylum seekers in the year before the election. Several authors have
17 population. Various researchers have shown that the level of education influences support for (right-wing) populist parties (Lubbers et al., 2002, Arzheimer, 2012).
18
5 Data
This section will explain which data sources were used, and will also provide summary statistics on the most important data. By doing this, the reader is able to get a feel for the data.
The first issue is the identification of political parties. In section 2, I explained that the definition of populist political parties of Van Kessel (2015) will be followed. Van Kessel defines a party as populist if it a) portrays “the people” as virtuous and essentially homogeneous; b) advocates popular sovereignty, as opposed to elitist rule; c) defines itself as against the political establishment, which is alleged to act against the interest of the people. This is a recent and elaborate study which identified all parties in Europe that received enough popular votes to enter parliament. Since Van Kessel looked at the period 2000 to 2013, I will do the same here. I choose to do this, because otherwise there might have been changes in political parties, and new right-wing populist parties will not be included. It should be noted, however, that it is striking that Germany, for example, is not included in my sample. Van Kessel did not identify a right-wing populist party in Germany in the sample period.
Methodologically, it is important to note here that the variable RWPP measures the total percentage of votes for all right-wing populist parties. Hence, if there are three right-wing populist parties in a country, then all their popular support is added up. I choose to take this step, because I am interested in the popular support for right-wing political thinking. Furthermore, I included all countries in which Van Kessel (2013) identified one or more right-wing populist parties. There have been years in which there was no right-right-wing populist party participating in elections in these countries, but I have included those elections in the sample and set the RWPP in that year to 0. There might be important reasons as to why there were no RWPP participating in the elections in those instances, I included those years to prevent a selection bias.
19 ones from 2000 to 2013. The included countries, the respective right-wing populist parties, and the election years are shown in Table 1.
Table 1 Countries and Populist Parties in the sample
Country Right-wing populist parties Election years
1. Austria FPÖ, Team Stronach, 2002, 2006, 2008, 2013
Alliance for the Future of Austria
2. Belgium Vlaams Blok, Front National, List Dedecker 2003, 2007, 2010 3. Bulgaria Coalition Ataka, RZS, GERB 2001, 2005, 2009, 2013
4. Croatia HSP-AS 2000, 2003, 2007, 2011
5. Czech Republic ANO, Public Affairs, Usvit 2002, 2006, 2010
6. Denmark Dansk Folkeparti 2001, 2005, 2007, 2011
7. Finland True Finns 2003, 2007, 2011
8. France Front National 2002, 2007, 2012
9. Great Britain UKIP 2001, 2005, 2010
10. Greece ANEL 2000, 2004, 2007, 2009, 2012
11. Hungary FYD-HDF, MIEP, FIDESZ-MPSZ, 2002, 2006, 2010 Movement for a Better Hungary
12. Iceland Citizen's Movement 2003, 2007, 2009, 2013
13. Italy Forza Italia, Lega Nord, PdL 2001, 2006, 2008, 2013
14. Latvia LNNK, All for Latvia 2002, 2006, 2010, 2011
15. Lithuania Party "Order and Justice" 2000, 2004, 2008, 2012 16. Luxemburg Alternative Democratic Reform Party 2004, 2009, 2013
17. Netherlands List Pim Fortuyn, Liveable Netherlands, PVV 2002, 2003, 2006, 2010, 2012 18. Norway Progress Party, Democrats 2001, 2005, 2009, 2013
19. Poland SRP, PiS 2001, 2005, 2007, 2011
20. Romania Greater Romania Party 2000, 2004, 2008, 2012
20
22. Slovenia SNS 2000, 2004, 2008, 2011
23. Sweden Sweden Democrats 2002, 2006, 2010
24. Switzerland Swiss People's Party, Swiss Democrats, 2003, 2007, 2011
Lega dei Ticinesi
Here, I turn to a description of both the explanatory variables and the control variables. The first explanatory variable is international trade (TRADE), which is measured in three different ways. I set up three different regressions, with imports growth (ΔIMPORTS), foreign content in exports (EXPORTS_fc) and imports share in GDP (IMPORTS_sh), all these are measured at period t-1. It is important to look at period t-1, because the elections might take place early in a year. Therefore, it might be ineffective to look at period t, since this has no effect on the outcome of the elections and the share of people voting for right-wing populist parties. By taking period t-1, there is certainly the possible effect included of international trade on RWPP.
ΔIMPORTS is included, because if the number of inflowing goods and services into a country
is growing, this indicates that there is more international trade and more competition for national firms and workers. The growth of imports is taken at the year before the elections, and the data on this variable are from the World Bank. So, it is the growth from period t-2 to period t-1. EXPORTS_fc is the share of imported inputs in the overall exports of a country. This share has increased over the past decade, everywhere in Europe. It is therefore a good indicator of how production has become more specialized and how this led to an increase in international trade. Production fragmentation has put pressure on manufacturing jobs. The data for this measure are created by the OECD and is taken from the year before the each election.
The data for the last measure, IMPORTS_sh, is taken from the World Bank. It measures all imports of goods and services as a share of total GDP.
21 The third and final explanatory variable is ΔGDP_p, which is a three-year average of the growth rate of per capita GDP, the two years prior to the election and the election year. I have chosen to take this methodological step, since economic voting theory suggests that voters take into account economic growth when voting. Also, they evaluate incumbent parties based on their economic performance (Wilkin et al. 1997). Since typically government are in place for about four years, I choose to include more than one year to base ΔGDP_p on, the evaluation of incumbent parties takes place based over a longer period. I do not expect that the full period matters, therefore I included the two years prior to the election and in this case also the election year. The data are also from Eurostat.
For three out of four control variables (ASYLUM, AGE, and EDUCATION) the data are derived from Eurostat. ASYLUM looks at the number of asylum applicants in the year before the elections. This is the number of asylum applicants over the total population (which was divided by 1000), which I calculated using the data by Eurostat. Thereafter, I have taken logs of that number, all these steps are to ensure that there are not really small number. Unfortunately, the data are not complete, some observations were missing.
As mentioned before, AGE is the median age of the population at the year of the elections, and
EDUCATION measures the number of medium or highly educated people of the population, so
it is a percentage.
The control variable CRISIS is a dummy variable for whether the country was affected by a financial or banking crisis and based on the work of the International Monetary Fund (IMF). In the Appendix, Table A1 gives a quick overview of the data, the description and the source.
Table 2 provides the descriptive statistics with respect to all the variables in this research. All statistics are related to the actual election years, it is important to keep this in mind when analyzing the data. The variable CRISIS is excluded from Table 2, since it is a dummy variable and the statistics will not tell us anything as such.
Table 2 Descriptive statistics
Variables N mean min max sd variance
RWPP 89 14.41 0 72.10 14.15 200.29
22 EXPORTS_fc (t-1) 89 29.44 16.16 58.92 9.66 93.33 IMPORTS_sh (t-1) 89 49.63 24.76 156.45 23.01 529.49 UNEMP (t-1) 89 8.55 2.40 20.10 4.37 19.14 ΔGDP_p (t-2, t-1, t) 89 2.26 -7.05 11.43 3.11 9.65 ASYLUM (t-1) 82 0.89 0.01 3.92 0.96 0.93 AGE 89 39.23 33.50 44.40 2.17 4.70 EDUCATION 87 20.23 7.00 35.20 7.12 50.74
The dependent variable (RWPP) shows an average of 14.41 percent of votes to RWPP in the sample. The minimum is 0, since there are some elections in which the RWPP did not receive any votes or did not participate. The maximum of 72.1 percent is relatively high, this is the result of the 2010 elections in Hungary. The FIDESZ-MPSZ party received about half of the votes, and there were also smaller parties that received a significant percentage of popular support. This is definitely an outlier, however there are multiple elections in which the RWPP received about half of the votes.
RWPP mostly received large popular support in Eastern European countries. Besides Hungary, these are countries like Bulgaria, Poland and Slovakia. Also, Italian RWPP were supported quite extensively by their population, with Forza Italia and Il Popolo della Liberta (PdL) as the largest parties. Another large right-wing populist party in Western Europe is the Swiss People’s Party in Switzerland, which received on average about 28 percent of the Swiss votes in the election years in the sample. Some countries in which the RWPP did not receive many votes are Great Britain, Croatia, Iceland, Sweden, and Slovenia. So, there is not a clear region to be identified in that respect.
The number of elections per country differ, there are several countries with three elections in the sample period, while there are some countries in which five elections took place. This leads to an unbalanced panel, but this allows me to include more observations.
23 of foreign inputs for their exports. The statistics for IMPORTS_sh entail that some countries import 1.5 times the amount of their GDP, the maximum is reached by Luxembourg.
The average level of unemployment is 8.55 percent and again there are quite some differences between countries. The Western European countries tend to perform better here.
For the growth rate of per capita GDP, there are also differing growth rates, and there are even countries that experienced a declining gross domestic product per capita. This happened typically during the Great Recession in which many European countries were hit.
24
6 Results
In this section, I will discuss the outcomes of the general model specifications, in order to conclude whether my hypotheses hold.
Before introducing the results of the general model, I will discuss the Hausman test, which is necessary to determine whether I need to use fixed or random effects models. The results of the Hausman test are depicted in Table 3, the p-value of 0.8462 means that I fail to reject the null hypothesis. This conclusion means that the difference between the fixed and random effects is not systematic and there is no strong case for using a fixed effects model. I will use the random effects model, because it has smaller standard errors.
Table 3 Hausman test
Coefficients
(b) (B) (b-B) sqrt(diag(V_b-V_B))
fixed random difference S.E.
ΔGDP_p -1.40 -1.18 -0.22 0.23 UNEMP -0.98 -0.67 -0.31 0.21 ΔIMPORTS 0.25 0.18 0.06 0.05 ASYLUM -0.63 -0.90 0.27 1.10 AGE -0.45 0.26 -0.71 0.59 CRISIS 3.66 3.17 0.49 0.86 EDUCATION -0.27 -0.48 0.21 0.20 b = consistent under Ho and Ha; obtained from xtreg
B = inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic
chi2(7) = (b-B)'[(V_b-V_B)^(-1)](b-B) = 3.40
Prob>chi2 = 0.8462
25 Table 4 depicts the results from the general model specifications. The main variables of interest are ΔGDP_p, UNEMP, ΔIMPORTS, EXPORTS_fc, and IMPORTS_sh. The other variables are the control variables. The first hypothesis relates the volume of international trade to the support for right-wing populist parties, which I measure by looking at three different variables, being
ΔIMPORTS, EXPORTS_fc, and IMPORTS_sh. According to all three estimations, an increase
in the volume of international trade leads to an increase in the number of votes for right-wing populist parties in Europe, as shown by the positive coefficients. For the measure EXPORTS_fc, this relation holds at the 5% significance level and the coefficient is larger, meaning that this is a stronger determinant for the number of votes. An intensification of intermediate trade leads to increasing job losses in Europe, as argued by Timmer et al. (2013), especially among low-skilled workers. When countries increasingly rely on foreign intermediate products for their own exports, more low-skilled jobs disappear, and their electorate is more inclined to vote for right-wing populist parties. For the variables ΔIMPORTS and IMPORTS_sh, the effect on the
number of votes for right-wing populist parties in Europe holds at the 5% significance level. The effect of EXPORTS_fc on the support for right-wing populist parties, is relatively larger than for ΔIMPORTS and IMPORTS_sh, which might be explained by the increased importance
of global value chains all over the world (Baldwin, 2012). All in all, these findings suggest that the first hypothesis holds and is in line with what leading and recent academic research has found (Autor et al., 2017; Colantone and Stanig, 2017; Jensen et al., 2017).
26 longer track record of solving these issues. Stoyanova (2016) does not give any theoretical reason for this finding. All in all, this is still a very striking finding, since the majority of empirical evidence points in the contrary direction (Swank and Betz, 2003; Norris and Inglehart, 2016; Algan et al. 2017). In line with the argument by Knigge (1998), my results might well be explained by reliance on mainstream political parties in difficult times. In Europe there is a long tradition of political parties, the vast majority of right-wing populist parties is relatively new.
The third hypothesis relates GDP per capita growth to the support for right-wing populist parties and the results in Table 3 are in line with the expectations beforehand. A decrease in GDP per capita, or low growth, has a positive impact on the support for right-wing populist parties during elections in Europe. For the first and third model, the results for ΔGDP_p are significant at the
5% level. The result for the second model is significant at the 10% level. This suggests that the third hypothesis hold, and this is in line with leading empirical evidence in this field (Guiso et al., 2017; Funke et al., 2016). The models suggest that the influence of GDP per capita growth is the most important determinant of voting for right-wing populist parties, compared to international trade and unemployment.
Table 4 General model results
27 Constant 22.17 7.873 17.26 (33.93) (31.08) (31.80) Observations 81 81 81 Number of countries R-squared 23 0.176 23 0.110 23 0.218 Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Table 4 shows some other important results with regard to the control variables. Only the variable EDUCATION has a significant effect on the number of votes for right-wing populist parties in Europe. This means that based on all three models, a higher education level means less votes for the right-wing populist parties in Europe, as shown by the negative sign. All other three control variables do not have a significant effect on the dependent variable RWPP. In the case of AGE, I also tried to re-estimate the model replacing the median age by the share of the young population in the total population. They young population is between age 20 and 30. This, however, makes no difference to the results and therefore the results are not
28
7 Robustness checks
As a robustness check, I also ran the regression excluding Eastern European countries from the model, which left me 14 countries in the sample. The results are provided in Table 5. The results are quite comparable to the general model, as shown in Table 4. Again, the variable
UNEMP has a negative relationship with RWPP, at the 10% significance level, in all three
estimations. The coefficient has become larger than before, which means that unemployment is a more important factor in determining the support for right-wing populist parties in Western Europe. Relating this to Knigge (1998), the reason might be that Western European countries have a longer history with traditional political parties, since they tend to be democratic for a longer period than Eastern European countries.
Similar to the general model, ΔIMPORTS has a positive relationship with RWPP, and is now
even significant at the 5% level. The coefficient has become somewhat larger, meaning that the level of imports growth is more important in Western Europe, pushing people more towards right-wing populist parties. The other variables that relate international trade with voting for right-wing populist parties, EXPORTS_fc and IMPORTS_sh, are no longer significant. In Western Europe, apparently these do not have a significant influence on RWPP, the coefficients also turned negative. A possible explanation is that Western European citizens are better supported by welfare systems. If they lose their job, in this case due to outsourcing as a consequence of international production fragmentation, or in general increasing imports, they know that they have access to sufficient welfare support.
Another finding that is different for this model with only Western European countries is that the variable ΔGDP_p is not significant. The signs are the same as in the general model. The
sample with only Western European countries shows that the differences in GDP per capita growth are less varying than in the general model. The positive outliers in the general model all are in Eastern European countries. In this model there is less variation, which means that the effect of ΔGDP_p on RWPP is not significant anymore.
29 more value to traditional values and what to keep those in place. This is similar to what right-wing populist parties aim for, they are concerned with national identity (Mudde and Kaltwasser, 2012). Therefore, age is an important determinant for voting for right-wing populist parties in Western Europe.
Table 5 Robustness check
(1) (2) (3) VARIABLES RWPP RWPP RWPP ΔGDP_p -1.543 -0.423 -0.396 (0.953) (0.800) (0.811) UNEMP -1.061* -1.139* -1.200* (0.630) (0.652) (0.659) ΔIMPORTS 0.381** (0.185) EXPORTS_fc -0.204 (0.188) IMPORTS_sh -0.0217 (0.0619) ASYLUM 0.293 -0.0649 -0.125 (1.382) (1.420) (1.460) AGE 1.757** 2.362*** 2.328*** (0.857) (0.834) (0.848) CRISIS 3.441 2.230 2.230 (2.329) (2.450) (2.497) EDUCATION -0.681*** -0.634*** -0.657*** (0.231) (0.235) (0.241) Constant -33.92 -51.88 -54.24* (33.14) (32.49) (32.97) Observations 42 42 42 Number of countries R-squared 13 0.296 13 0.204 13 0.202 Standard errors in parentheses
30
8 Conclusion
Over the past two decades, right-wing populist sentiments have come to surface in the Europe and North America, with the election of Donald Trump as president of the United States and the Brexit as prominent recent examples. Additionally, all over Europe, right-wing populist political parties have attracted attention among the population, and they have received more votes in elections. With the increasing support for these parties, there has also been an increasing interest among academic scholars to find out what drives these sentiments. In explaining the rise of populism, some have focused on the individual characteristics of voters (Norris and Inglehart, 2016, Lubbers et al, 2002, Arzheimer, 2012, and Rooduijn, 2014), while other researchers have looked at economic and trade motives (Rodrik, 2017, Swank and Betz, 2003, Autor et al., 2017, and Colantone and Stanig, 2017).
This thesis has analyzed the rise of populism by using the economic security or economic inequality perspective, prominently brought forward by Norris and Inglehart (2016). What distinguishes my research from other authors (e.g. Norris and Inglehart, 2016; Swank and Betz, 2003) is that I have assessed the consequence of changes in domestic macroeconomic conditions and the influence of globalization, being the volume of international trade, unemployment, and economic growth. So far, researchers have focused at the micro level.
The main research question was: What domestic macroeconomic indicators are most important
in determining the rising support for right-wing populist political parties in Europe?
The sample period was from 2000 to 2013, and the right-wing populist parties were identified by Van Kessel (2015).
To discuss the main research question, I have tested three hypotheses, being: (1) An increasing volume of international trade will increase the support for right-wing populist parties, (2) An increase in unemployment will increase the support for right-wing populist parties, and (3) A decrease in growth of GDP per capita will increase the support for right-wing populist parties. In order to test these hypotheses I created an econometric model, in which the dependent variable was the share of votes received by right-wing populist parties. The explanatory variables for international trade were ΔIMPORTS, EXPORTS_fc, and IMPORTS_sh. For the
31 Based on the results of both the general model and the model excluding Eastern European economies, hypotheses (1) and (3) do hold, as an increasing volume of international trade will increase the support for right-wing populist parties, and this is based on three different measurements, being imports growth (ΔIMPORTS), foreign content in exports (EXPORTS_fc) and imports share in GDP (IMPORTS_sh). Especially the finding for foreign content in exports is interesting, since it supports the view that increasing international fragmentation of production leads to increasing insecurity, and consequently, more support for right-wing populist parties. Furthermore, growth of GDP per capita has a negative relation with the support for right-wing populist parties, as expected beforehand.
However, hypothesis (2) does not hold, as my research actually points at the opposite, namely that an increase in unemployment will decrease the support for right-wing populist parties. This can be explained, when unemployment becomes a real problem in the economy, people tend to rely on mainstream political parties.
To answer the main research question, low GDP per capita growth and increasing international trade are most important in determining the rising support for right-wing populist political parties in Europe. According to my research, a decrease in unemployment could also explain the rise of right-wing populist parties.
When looking at the control variables, I have shown that age and education matter in determining the support for right-wing populist parties. Older voters and lower educated people are more inclined to vote for right-wing populist parties. I believe that the elderly want to preserve national values and traditions, as do right-wing populist parties. The lower educated citizens tend to be losing their jobs due to increasing international fragmentation of production and are therefore lending their support to right-wing populist parties. The presence of an international economic or financial crisis, and the share of asylum seekers, are not found to be important in determining popular support for right-wing populist parties.
There are some limitations to my research. Firstly, for two control variables, namely
EDUCATION and ASYLUM, I have some missing observations. For EDUCATION there are
32 Secondly, as already mentioned in the ‘Data’ section, there are some European countries that are not in the sample of Van Kessel (2015). Germany is not included in this research, even though there are number of small right-wing populist parties, as well as a large right-wing populist party, Alternative for Germany, however, this has been founded only in 2013. Spain is also not included in Van Kessel’s sample, which is another large country not in my research, but in contrast to Germany, in Spain there are no right-wing populist parties that gained electoral support at the national level. I have followed Van Kessel’s definition and selection of right-wing populist parties, but there are other samples, as identified for example by Inglehart and Norris (2016), and Johansson Heinö (2017). Using their samples might yield different results, and this might be interesting to look at in the future.
Further research could be directed into the question as to why my research shows that increasing unemployment leads to a decreasing support for right-wing populist parties. This is contrary to what other researchers have found (Norris and Inglehart, 2016; Algan et al., 2017), although Rooduijn (2014) has found no role for unemployment in explaining the rising support for right-wing populist parties. Also, I reached a conclusion that is similar to what Knigge (1998) and Stoyanova (2016) found. The contrasting views indicate that this issue is still debated, and it would be interesting to keep doing research in this area.
Additionally, looking at income inequality as a determinant for right-wing populist voting, might be interesting. Due to missing observations, I have not included this here. However, an important future direction for research could be to explore the role of inequality.
Another interesting research could be developed when looking at left-wing populism. A similar research set-up could be used to determine which macroeconomic indicators are most important in explaining left-wing populism, however this might be more problematic to gather data on, as left-wing populism is mostly popular in Latin America.
33
Bibliography
Acemoglu, Daron, Georgy Egorov, and Konstantin Sonin (2013), “A Political Theory of Populism”, Quarterly Journal of Economics 128, no.2, pp. 771-805.
Algan, Yann, Sergei Guriev, Elias Papaioannou, and Evgenia Passari (2017), “The European Trust Crisis and the Rise of Populism”, Brookings Papers on Economic Activity.
Arzheimer, Kai (2012), “Electoral Sociology: Who Votes for the Extreme Right and Why – and When?”, in: Uwe Backes, and Patrick Moreau (Eds), “The Extreme Right in Europe. Current Trends and Perspective”, Göttingen: Vendenhoeck & Ruprecht, pp. 35-50.
Autor, David, David Dorn, Gordon Hanson, and Kaveh Majlesi (2017), “Importing Political Polarization? The Electoral Consequences of Rising Trade Exposure”, NBER Working Paper
no. 22637.
Baldwin, Richard (2012), “Global Supply Chains: Why They Emerged, Why They Matter, and Where They Are Going”, CEPR Discussion Paper no. DP9103.
Becker, Sascha O., Thiemo Fetzer, and Dennis Novy (2016), “Who Voted for Brexit? A Comprehensive District-Level Analysis”, Warwick Working Paper Series.
Betz, Hans-Georg (2002), “Conditions favouring the success and failure of radical right-wing populist parties in contemporary democracies?”, in: Yves Mény, and Yves Surel (Eds), “Democracies and the Populist Challenge”, Springer, Berlin, pp. 197-213.
Bobbio, Norberto (1996), “Left and Right: The Significance of a Political Distinction”, Chicago
University Press.
34 Colantone, Italo and Piero Stanig (2017), ”The Trade Origins of Economic Nationalism: Import Competition and Voting Behavior in Western Europe”, Bocconi University Working Paper.
Dornbusch, Rudiger, and Sebastian Edwards (1991), “The Macroeconomics of Populism”, in: The Macroeconomics of Populism in Latin America, University of Chicago Press, pp. 7-13.
Dustmann, Christian, Barry Eichengreen, Sebastian Otten, André Sapir, Guido Tabellini, and Gylfi Zoega (2017), “Europe’s Trust Deficit: Causes and Remedies”, CEPR Press.
Foster, Chase, and Jeffry Frieden (2017), “Crisis of Trust: Socio-economic determinants of Europeans’ confidence in government”, Mimeo Harvard University.
Funke, M., M. Schularick, and C. Trebesch (2016), “Going to the extremes: Politics after financial crises, 1870-2014”, European Economic Review 88, pp. 227– 260.
Gereffi, Gary (2014), “Global value chains in a post-Washington Consensus world”, Review of
International Political Economy 21, no. 1, pp. 9-37.
Gidron, Noam, and Bart Bonikowski (2013), “Varieties of Populism: Literature Review and Research Agenda”, Weatherhead Center for International Affairs Working Paper Series no.
13-0004.
Golder, Matt (2003), “Explaining Variation in the Success of Extreme Right Parties in Western Europe”, Comparative Political Studies 36, no. 4, pp. 432-466.
Guiso, L., H. Herrera, M. Morelli, and T. Sonno (2017), “Populism: Demand and Supply”.
Guiso, L., H. Herrera, M. Morelli, and T. Sonno (2017), “Global Crises and Populism: the Role of Eurozone Institutions”.
35 Inglehart, Ronald F., and Pippa Norris (2016), “Trump, Brexit, and the Rise of Populism: Economic Have-Nots and Cultural Backlash”, Working paper.
Ivarsflaten, Elisabeth (2008), “What unites right-wing populists in Western Europe? Re-examining grievance mobilization models in seven successful cases”, Comparative Political
Studies 41, no. 1, pp. 3–23.
Jensen, J. Bradford, Dennis P. Quinn, and Stephen Weymouth (2016), ”Winners and Losers in International Trade: The Effects on U.S. Presidential Voting”, NBER Working Paper No.
21899.
Johansson Heinö, Andreas (2017), “Timbro Authoritarian Populism Index 2017”.
Kessel, Stijn van (2015), “Populist Parties in Europe – Agents of Discontent?”, Palgrave
Macmillan.
Knigge, Pia (1998), “The ecological correlates of right-wing extremism in Western Europe”,
European Journal of Political Research 34, pp. 249-279.
Lubbers, Marcel, Mérove Gijsberts, and Peer Scheepers (2002), “Extreme right-wing voting in Western Europe”, European Journal of Political Research 41, pp. 345-378.
Mudde, Cas (2004), “The Populist Zeitgeist”, Government and Opposition 39, no.4, pp. 541– 63.
Mudde, Cas (2007), “Populist Radical Right Parties in Europe”, Cambridge University Press.
Mudde, Cas, and Cristóbal Rovira Kaltwasser (2011), “Voices of the Peoples: Populism in Europe and Latin America Compared”, Kellogg Institute Working Paper 378.
Rodrik, Dani (2017), “Populism and the Economics of Globalization”.
36 Serani, Danilo (2016), “Explaining vote for populist parties: the impact of the political trust, the economic and the political context”, Working paper.
Stoyanova, Anastasia (2016), “Unemployment and European Right Wing Populism”,
Economics and Nationalism Seminar.
Swank, Duane, and Hans-Georg Betz (2003), “Globalization, the welfare state and right-wing populism in Western Europe”, Socio-Economic Review 1, pp. 215-245.
Timmer, Marcel P., Bart Los, Robert Stehrer, and Gaaitzen J. de Vries (2013), “Fragmentation, incomes, and jobs: an analysis of European competitiveness”, Economic Policy October 2013, pp. 613-661.
37
Appendix
Table A1 Data
Variable Description Source
RWPP
The number of votes received by right-wing populist parties in Europe, see also 'Data' section.
Van Kessel (2015),
European University Institute
ΔIMPORTS Measures the growth of imports from period t-2 to t-1. World Bank
EXPORTS_fc
Share of imported inputs in the overall exports of a
country OECD
IMPORTS_sh Imports of goods and services as a share of total GDP World Bank
UNEMP
Level of unemployment as a percentage of the total
workforce, at t-1. Eurostat
ΔGDP_p
Average growth of the Gross Domestic Product, of the
periods t-2, t-1, and t. Eurostat
ASYLUM
Number of asylum applicants over total population (which is divided by 1000) in the year before the
elections. Eurostat
AGE
Median age of the population in the year of the
elections. Eurostat
EDUCATION
Number of medium or highly educated people, as a
percentage of the total population. Eurostat
CRISIS
38 Table A2 Correlation table 1
Correlation
matrix ΔGDP_p UNEMP ΔIMPORTS ASYLUM AGE CRISIS EDUCATION constant
ΔGDP_p 1 UNEMP 0.2835 1 ΔIMPORTS -0.4002 -0.0800 1 ASYLUM 0.2312 0.2086 -0.0879 1 AGE 0.3330 0.1649 -0.3547 0.1897 1 CRISIS 0.2933 0.0032 0.1004 0.1513 -0.1941 1 EDUCATION 0.1770 0.2121 0.0861 -0.0245 -0.2989 -0.0641 1 constant -0.4293 -0.3155 0.3485 -0.2643 -0.4683 0.1609 0.1030 1
Table A3 Correlation table 2
Correlation matrix ΔGDP_p UNEMP EXPORTS_fc ASYLUM AGE CRISIS EDUCATION constant
39 Table A4 Correlation table 3
Correlation
matrix ΔGDP_p UNEMP IMPORTS_sh ASYLUM AGE CRISIS EDUCATION constant
