The impact of Chinese import penetration on the support for populist parties in Europe

The impact of Chinese import penetration on the

support for populist parties in Europe

19th June 2018

Master Thesis

Faculty of Economics and Business

Double-Degree Master in International Economics

Master in International Economics, University of Göttingen

Master in International Economics and Business, University of Groningen

Laura Salemink

l.salemink@student.rug.nl

Student number University of Groningen: S3192253 Student number University of Göttingen: 21620968

Supervisors:

Dr. Robbert K.J. Maseland, University of Groningen

Abstract

This paper analyzes the effect of increasing imports originating from China on the voting support for populist parties in 25 member states of the European Union. To test this impact, a Chinese import shock variable is calculated based on the ex-ante industry specialization of each country. We expect the vote shares for populist parties to increase with growing Chinese imports due to the associated increasing economic and cultural insecurity experienced by the society. The general findings contradict the hypotheses. Since the results are not in line with previous research and especially the studies for the United States, several explanations like differences in trade openness, culture, as well as party and welfare systems are put forward to understand the outcomes better.

Table of contents

6.1 IMPORTS FROM 10 “NEW” MEMBER STATES OF THE EU ... 26

7 DISCUSSION ... 29

8 LIMITATIONS AND FUTURE RESEARCH ... 32

9 CONCLUSION ... 34

REFERENCES ... 36

1 Introduction

Recent years have been characterized by the gain in strength of populist movements in several developed countries. The election of Donald Trump, Brexit or the increasing voting support for the Alternative für Deutschland in Germany, the Front National in France and, a couple of weeks ago, for the Lega and the Five Star Movement in Italy are only some examples. These illustrations indicate that the anti-establishment philosophy proclaimed by populist politicians increasingly seems to provide answers to the challenges faced by the general public. Inglehart and Norris (2016) show that the average share of votes for populist parties has more than doubled since the 1960s. This mainly occurred at the expense of more centralist and long-established parties. Additionally, they observe that these parties need fewer representatives in parliament in order to exert pressure on the political agenda. The role of the UK Independence Party in pushing the Brexit campaign clearly confirms this phenomenon (Goodwin & Heath, 2016).

This paper focuses on 25 member states of the European Union. Considering the elections of the last 10 years, one can notice that several countries in Southern Europe faced a left-shift of their political environment. Especially the financial crisis in 2008, the following bailouts of governments and the associated austerity measures induced this shift in Greece and Spain (Stavrakakis & Katsambekis, 2014). Similar to Latin America after the implementation of the structural adjustment programs by the International Monetary Fund and the World Bank (Rodrik, 2017), these countries blame the European Union and the European Central Bank for the ongoing bad economic and social situation. Thus, the populist movements mainly target the income cleavage between rich and poor European nations. In Western Europe, however, the story is a different one. Right-wing populism seems to prevail with the Flemish Block in Belgium and the Alternative für Deutschland in Germany (Bröning, 2016). In these states, the populist parties principally focus on the impact foreigners and migrants have on the domestic economy (Rodrik, 2017). Hence, these examples indicate that European states face different problems, which are addressed by populist parties on either side of the political spectrum.

increasing import penetration from low-income countries such as China changed the voting behavior in many developed countries (Autor et al., 2016b; Dippel et al., 2016; Colantone & Stanig, 2017; Mughan et al., 2003). This change can mainly be associated with the increasing income inequality in these states, even though globalization can only partly explain this phenomenon (Alvaredo et al., 2018). Overall, low- and middle-skilled workers disproportionately seem to lose from the worldwide integration process. This favors the support for non-establishment parties. Hence, the relationship between expanding imports from China and the voting support for populist parties will be analyzed in this paper. Therefore, I investigate the following research question: How does the increase of imports originating from China affect the support for populist parties in 25 member states of the European Union?

The results obtained performing this analysis contradict the general findings of the literature claiming that growing globalization increases the support for more radical parties. However, the paper provides some insights into why the outcomes differ from previous research that was mainly focusing on the United States. First, Europe is composed of countries with different political, legal and cultural backgrounds. Hence, the results might not be as uniform as for the United States. Second, the welfare and the party system in Europe and the US are fundamentally different. Third, the trade and migration policy in the EU cannot be decided at the national level. Fourth, several countries faced a general left or right shift of their government lately. This indicates that parties, in general, have changed their positions towards certain issues. Therefore, the political landscape might have changed even though an increase of populist politicians in parliament has not been observed.

2 Theoretical background

Lately, there has been a vast literature analyzing the consequences of growing Chinese import penetration on labor market characteristics (Autor et al., 2013), wages (Autor et al., 2016a) and voting preferences (Autor et al., 2016b) in the United States. The main outcome of these studies is that imports from China decreased employment in the US industrial sector, as expected. However, it did not lead to an increase in employment in other sectors, as predicted by trade theory (Autor et al., 2016a). Moreover, citizens living in trade-exposed local labor markets, in other words in labor markets with a high share of manufacturing employment, decreased the support for moderate politicians and increased it for more radical ones on either side of the political spectrum (Autor et al., 2016b). Based on these studies, some research has been done regarding the same effects in the European market. Nevertheless, these investigations have been primarily limited to Western Europe and radical right-wing parties (Colatone & Stanig, 2017; Swank & Betz, 2003) or to the analysis of a specific country (Dippel et al., 2016). In order to contribute to the ongoing research, this paper investigates the impact of increased imports from China on the voting support for left- and right-wing populist parties in 25 member states of the European Union. Since the European Union is composed of different countries with various cultural, political and economic backgrounds, it will be interesting to see whether the effects of growing globalization are as clear as in the studies for the United States.

order and oppose to immigration and rights for minorities. Cosmopolitan liberalists, on the other side, defend the post-materialist values present in the society nowadays (Inglehart & Norris, 2016).

Each value scale ranges from 0 to 100. A populist party is characterized as a party that scores higher than 80 on the cultural value scale. In that sense, populism can be defined as an anti-establishment orientation (Rodrik, 2017). It claims to speak for the ordinary people and against the corrupt elites; it puts national interest above cooperation across borders; it prefers protectionist policies regulating the movement of labor, goods and capital over free trade; and it supports xenophobia over tolerance of multiculturalism to name only some aspects of populism (Inglehart & Norris, 2016). Evidently, the group of populist parties can then easily be split into left- and right-wing populist parties on the basis of the economic standpoint of the party. The former focus mainly on the uneven income distribution between wealthy and deprived groups; the latter emphasize the cleavage between national, ethnic, religious or cultural identities (Rodrik, 2017).

Figure 1: Theoretical model of party competition

Source: Heuristic model of party competition in Western societies (Inglehart & Norris, 2016).

2.1 Economic perspective

globalization faced by the workforce and the society in post-industrial economies. There is not only a trend for greater income and wealth inequality in developed countries but also rising economic insecurity and social deprivation among the losers of globalization (Inglehart & Norris, 2016).

A big strand of the literature has focused on the impact of globalization on the voting behavior of citizens from developed countries. Globalization is here defined as the opening of nationalistic mindsets to a broader vision of the interdependent world with free transfer of capital, goods, and services across national frontiers (Buelens, 1999). A major component of globalization is international trade. Generally speaking, international trade is beneficial for the country as a whole (Feenstra & Taylor, 2014). Nevertheless, it creates winners and losers in the society as already claimed by the Stopler-Samuelson theorem (Feenstra & Taylor, 2014). According to this theory, the gains from the winners could compensate the losses from the losers to make the entire economy win. However, compensation is costly and winners need to be induced to stick to the redistribution plan even after the trade agreement has been signed (Rodrik, 2017). Thus, international trade can be considered as an important cause of increasing income and wealth inequality (Williamson, 2002). Therefore, it only seems to be sustainable if appropriate redistribution policies are set up in order to compensate the losers.

Figure 2: Share of manufacturing value added in global manufacturing value added

Comparing the shares of world manufacturing value added among the three main regions of interest, Figure 2 clearly indicates that China has constantly increased its share over the last 15 years whereas the Western countries’ shares gradually declined. This illustrates that China gained a leading position in manufacturing over the past years. This is principally due to the fact that China has a revealed comparative advantage in manufacturing as a result of the large supply of cheap low-skilled labor compared to the rest of the world (Autor et al., 2016a). Following the theory of comparative advantage, China will specialize in manufacturing and the Western countries in different industries (Feenstra & Taylor, 2014). This, in turn, means that manufacturing workers and thus, a larger share of the lower middle-class in developed countries disproportionately lose from trade liberalization with China (Swank & Betz, 2003). Clearly, the decreasing demand for manufacturing workers, the associated deteriorating wages in this sector and the resulting rising income inequality are not only due to trade, however, it is the number one reason blamed by politicians. It is easier to point at “them” than to blame “us” (Rodrik, 2017). Thus, politicians tend to accuse international trade while neglecting the role that other factors such as technological change or varying consumer behaviors play.

Due to the rising inequality, the lower and the middle class from Western societies are expected to participate in the political backlash associated with globalization. This backlash favors populist parties, which I defined as following a non-establishment philosophy. The thirst for radical politicians was clearly observable in the last presidential election in the United States (Cohen, 2018). Both, Donald Trump on the right and Bernie Sanders on the left, received support from people that felt stuck and perceived the system to run against them. Thus, since Europe can be considered as similar to the United States according to the development status, my first hypothesis can be stated as follows:

Hypothesis 1: Increasing imports from China increases the share of votes for populist parties in Europe.

economic left-wing parties would gain more vote shares, since they campaign, amongst other things, for social equality, fairness and egalitarianism (Graham et al., 2012). This implies that the state has to take care of groups that lose out by retraining programs and unemployment insurances. The second effect would give rise to economic right-wing parties, as people want to protect themselves from losing their job to competitors from other countries. Right-wing parties support the ingroup-thinking related to the opposition of “them” versus “us” (Graham et al., 2012).

One crucial aspect through which increasing import penetration affects for the support of populist movements is that more competition from other countries impacts the perceived employment security. Personal job insecurity is therefore a very important determinant of whether a person votes for a non-establishment party or not. If the worker feels insecure, he may choose to support the right-wing populist parties arguing in favor of protectionism or endorse left-wing movements fighting for a more extensive social welfare system (Colantone & Stanig, 2017). The latter argument follows the literature about embedded liberalism (Ruggie, 2003). This economic theory has mainly two objectives. First, opposing the right-wing populist view, it wants to promote free trade. Second, it wants to allow the government to intervene in order to provide generous welfare systems and full employment. According to this view, the economic problem provoked by globalization is addressed by an economic solution, namely the compensation of the losers. Thus, my second hypothesis can be stated as follows:

Hypothesis 2: Growing imports from China increases the share of votes for left-wing populist parties in Europe more than the share for right-wing populist parties due to the growing desire for a more generous welfare system.

emphasize the domestic control of the economy. Often, this requires imposing restrictions on the movement of goods, capital and labor. Economic nationalism can thus be seen as an opposing view to globalization. This line of reasoning clearly focuses on the “them” versus “us” aspect of populism (Rodrik, 2017).

In Western societies, this rise of economic nationalism is favored by groups that currently face a lot of economic insecurity such as low-skilled manufacturing workers (Oesch, 2008). Mughan et al. (2003) observed that Australian workers who worry about the provision of good jobs in the future are more likely to endorse the limitation of imports. As indicated by the study, higher employment insecurity increases the demand for protectionism. Hence, these citizens disproportionately support the radical right-wing parties. Even if they support free markets and liberalization at the domestic level, these parties heavily criticize international openness (Swank & Betz, 2003).

Nevertheless, the economic reason is not the only reason explaining the support especially for right-wing populist parties in Europe. The fear of identity loss plays an important role (Inglehart & Norris, 2016). The study performed by Oesch (2008) confirms that in Western Europe the want for cultural protectionism, which basically is the defense of the national identity against outsiders, clearly outcompetes the economic reasons to vote for right-wing populist parties. Hence, we will address this aspect of protectionism in the next subsection.

2.2 Cultural perspective

As Inglehart and Norris (2016) noticed, the surge for anti-establishment parties is not only an economic phenomenon. The increasing number of non-economic issues in the party manifestos clearly indicates this. Therefore, the authors decided to add another dimension to the classic economic left-right spectrum, namely the cultural dimension.

well as a diversity of lifestyles. Moreover, the post-materialists tend to emerge from the more secure and college-educated strata of Western societies and are therefore relatively favorable towards progressive social change and humanistic values (Piketty, 2018). On the other side of the scale, we have the populists. They feel threatened by the erosion of their former predominant values (Oesch, 2008). Moreover, the experience of a less secure welfare system during their lifetime led to lower levels of existential security. Hence, they put the well-being of the nation at first and argue for protectionism and stability (Braithwaite et al., 1996). Thus, according to this theory, xenophobia is only part of a broader cultural backlash among the older generation rejecting many other liberal and cosmopolitan values characterizing the post-industrial societies nowadays (Inglehart & Norris, 2016). As a deduction, the cultural dimension confirms the claim of the economic perspective that high levels of existential security are conductive to support for more post-materialist values and to vote for non-populist parties.

Following this line of reasoning, migration provides a good example to make the distinction between economic and cultural insecurity clearer. As already shown, the fear of losing one’s job increases the support for non-establishment parties on either side of the political spectrum. Generally speaking, immigrants compete directly for jobs with lower-skilled workers in developed countries (Finseraas et al., 2016). But, these workers already face economic insecurity due to the ongoing globalization process. Hence, with increasing numbers of immigrants, they are even more inclined to support parties fighting for protectionism or a more generous welfare system for nationals (Oesch, 2008). From this point of view, immigrants will decrease the perceived job security of the lower middle class even more. Nevertheless, immigration is not only a problem from the economic standpoint, but it also changes the cultural landscape of a country. Each immigrant brings his own cultural background and thus the cultures become more and more interconnected (Inglehart & Norris, 2016). This leads to growing support for populist parties since citizens fear to lose their national identity. Clearly, this illustration shows that migration might increase the support for populist parties from the economic and the cultural perspective indicating that it is an important aspect that needs to be taken into account. This leads us to my third hypothesis:

3 Data

As I will analyze the effect of increasing imports from China on the vote shares gained by populist parties in several European countries, I will now explain the construction of my dependent and independent variables. The focus lies on countries that joined the European Union the latest in 2004 (European Union, 2018). This allows including data for Eastern European countries as well, which has often been missing from previous research (Colantone & Stanig, 2017; Swank & Betz, 2003). Thus, in total, the sample consists of 25 countries. The period of analysis ranges from 1996 to 2017. Since China joined the World Trade Organization in 2001 (wto.org, 2018), this period allows analyzing the growing importance of this particular emerging country in international trade.

3.1 Election data

First of all, I will focus on the election data. The dependent variable varies according to the analysis I am conducting. To test Hypothesis 1, the total vote share for populist parties in the 25 European countries of consideration is used. For Hypothesis 2 and Hypothesis 3, the dependent variable is respectively the vote share for left- and right-wing populist parties.

In a second step, I match the information about the populist parties with the data provided by the Comparative Manifesto Project (CMP) database. This dataset provides data on the national elections of each country of interest (Volkens et al., 2017). Since there is no election data for Malta between 2000 and 2017, I drop the observations for this country. Hence, I obtain a final sample of 24 countries and 99 elections between 1996 and 2017. The CMP data gives us information about the percentage vote share each party obtained in a particular election. The change in these vote shares for the populist parties defined between two elections gives us my dependent variables.

3.2 Import data

In the next step, my independent variable needs to be defined. The main problem with the exploration of the causal effect of imports on domestic market characteristics is that changes in trade policies are often related to changes in the behavior of the trade partners. Thus, there might be an endogeneity problem. However, the benefit of analyzing the effect of growing Chinese exports is that the shock can be considered as exogenous (Autor et al., 2016a). Not only the unexpected Chinese export growth, the degree of isolation under Mao and the associated catch-up potential, but also the distinct comparative advantage in manufacturing allow to overcome the challenges in identifying the causal effects of trade shocks according to the authors.

Here, c is an index for each country of the sample, j represents each specific industry and e is an index for Europe, which would be equal to the total of the 24 European countries considered in the final sample of this paper. As the formula indicates, we consider the share of workers in one industry Lcj(pre-sample) relative to the overall working population in that country Lc(pre-sample). Hence,

the resulting weights capture the relative importance of each industry in a given country. The larger this term in one of the manufacturing industries in which China has a comparative advantage, the more vulnerable a country is to the increase in imports from China. ΔIMPChinaejt

is the change in imports from China over the past n years in Europe e in industry j. This is normalized by the number of workers in Europe e in that industry j at the beginning of the sample period, Lej(pre-sample). The equation indicates that differences in import shocks across countries

stem entirely from the variation in the local industry employment structure in the pre-sample period.

All the relevant data is provided by the World Input-Output Database (WIOD) (Timmer et al., 2015). Taking the release from 2016, I have data for each country in 56 different industries from 2000 to 2014. The Socioeconomic Account data provides information about the labor force and the World Input-Output Tables about imports from China. In this paper, I use the number of persons engaged (EMP) in the year 2000 from the Socioeconomic Account to measure a country’s exposure to imports from China. The choice of the year 2000 as the pre-sample period can be explained by the fact that China entered the WTO in 2001 (wto.org, 2018). Thus, 2000 is the last data available before this change.

As the equation indicates, I also need the difference between imports of two different moments in time in order to calculate the import shock. Therefore, I first calculate the total European imports stemming from China for each election year in each country. Then, I take the difference between European imports between two consecutive national elections. The data is given in US dollars and I adjust for inflation using the consumer price index (Data.worldbank.org, 2018). Combining the import data with the labor data, I can finally calculate the import shock. The resulting shock is given in 1000 US$ per person engaged.

3.2.1 Endogeneity

voting behavior, imports and the economic performance of the country may simultaneously be affected by these unobserved shocks. Due to this possible endogeneity, Autor et al. (2013) introduced the following instrument. They decided to instrument the imports from China to the United States by imports to other high-income countries. Hence, in this paper, I follow the same approach and instrument the imports from China to the European countries by the imports from China to the US. This instrument is meant to capture the variation in Chinese exports due to exogenous changes in the supply conditions in China, rather than to domestic factors that could be correlated with electoral outcomes (Autor et al., 2013; Colantone & Stanig, 2017). Pursuing this approach, I obtain the following import shock variable:

𝐼𝑚𝑝𝑜𝑟𝑡 𝑆ℎ𝑜𝑐𝑘!"# = 𝐿!" !"#!!"#$%& 𝐿_{! !"#!!"#$%&} ! ∗∆𝐼𝑀𝑃𝐶ℎ𝑖𝑛𝑎_𝐿 !"#$ !" !"#!!"#$%&

The only difference with the import shock calculated previously is that we use the change in imports to the US instead of the one to Europe. The instrument used seems to be reasonable, since it is exogenous (Autor et al., 2013) and highly correlated (0.922) with the actual imports to Europe. This can also be seen looking at the imports from China to the United States and to Europe. Figure 3 clearly indicates that total imports evolve similarly, even though the growth rate of European imports is larger.

Figure 3: Imports from China (2000-2014) in million US$

3.3 Controls

4 Methodology

In the previous section, I defined my dependent, independent and control variables. Hence, I need to introduce the methodology used to analyze the provided data. In order to catch the effect of the Chinese import shock on the voting support for populist parties, I will use the following regression model:

∆𝑃𝑒𝑟𝑣𝑜𝑡𝑒_!" = 𝛼_!"+ 𝛽 𝐼𝑚𝑝𝑜𝑟𝑡𝑠ℎ𝑜𝑐𝑘_!"#+ 𝛾 𝑍_!+ 𝜀_!"

Here, ΔPervotect is the change in vote shares for populist parties between two consecutive

elections in country c in election year t. This variable will be adapted throughout the analysis in order to test my three hypotheses. In the first step, it includes the sum of the vote shares for all national populist parties in a particular election. In the second and third step, I will distinguish between the change in vote shares for left- and right-wing populist parties. Importshockect is, of

course, the independent variable defined in the preceding section. In order to eliminate the endogeneity problem mentioned, I will perform an instrumental variable approach instrumenting imports to Europe by imports to the United States. αct is the intercept and includes the country

5 Statistics

As mentioned, China has a comparative advantage in manufacturing (Autor et al., 2016a). This mainly stems from the abundance of cheap labor. Clearly, the data in Figure 2 shows that China became one of the most important global players in manufacturing and the growth in manufacturing value added does not seem to diminish over time (Data.worldbank.org, 2018). Indeed, if China increasingly produces manufacturing output, the demand for manufacturing workers in the rest of the world decreases. According to trade theory, developed countries will decrease their labor share in manufacturing and specialize in other industries if they start to trade with China (Feenstra & Taylor, 2014). Considering European countries as developed countries, Figure 4 clearly illustrates this point. The share of workers engaged in manufacturing sectors has distinctively decreased throughout Europe (Timmer et al., 2015). Only in Poland, there has been a modest increase. Nevertheless, trade is clearly not the only reason explaining this decrease; technological change plays an important role in it as well (Rodrik, 2017).

Figure 4: Share of labor in manufacturing in 2000 and 2014

Source: WIOD Manufacturing Labor/Total labor, Number of persons engaged (EMP) (Timmer et al., 2015). 0 0.05 0.1 0.15 0.2 0.25 0.3 Czech R ep ub lic

Slovenia Slovakia Malta Hu

In addition, following the line of reasoning introduced by Autor et al. (2013), we can expect countries with a higher initial share of manufacturing labor to suffer more from China’s export surge because the perceived economic insecurity will increase for a larger share of the population. Figure 5 depicts the export exposure of each European country of consideration based on the pre-sample period defined previously. The darker the color in the map, the higher the share of manufacturing employment relative to the other countries. Hence, on the one hand, countries like Slovenia, Hungary or the Czech Republic are supposed to suffer more from Chinese imports, since they compete in the same industry segments as China. While, on the other hand, countries like Greece, Austria or France should be much less affected by Chinese growing exports.

Figure 5: Share of labor in manufacturing relative to other countries

Source: WIOD, Socioeconomic Account, Number of persons engaged (EMP) (Timmer et al., 2015).

However, the share of votes for left-wing populist parties has constantly increased and reached on average 12% of votes in the last period. This huge increase in votes for left-wing populist parties can clearly be associated with the financial crisis and the resulting bailouts. Especially in Greece, these parties gained vote shares up to 40% (Volkens et al., 2017). As Figure 6 indicates, the right-wing populist parties did not experience the same growth in votes. Between 1996 and 2017, the share only doubled whereas, for left-wing populist parties, the share was multiplied by 6 over the same time period (Volkens et al., 2017; Inglehart & Norris, 2016). Overall, Figure 6 clearly indicates that the share of votes for populist parties has increased in the last 20 years. This is in line with the findings of Inglehart & Norris (2016).

Figure 6: Average vote shares for populist parties over time

Source: Manifesto Project Database and CHES (Volkens et al., 2017; Inglehart & Norris, 2016).

already indicated that the average vote share for right-wing populist parties did not experience the same growth as did the one for left-wing populist parties.

One possible reason for this lower growth rate might be the fact that people prefer to fight for a more generous welfare system instead of advocating against free trade since trade has benefits such as cheaper goods’ prices for them as well. On the one hand, right-wing parties propose protectionist policies in terms of labor, capital, goods and services (Swank & Betz, 2003). But, trade- and migration-related topics are decided by the European Union (Ec.europa.eu, 2018a) and not by the national parliaments of each member state. Table XXIX and XXX in the appendix indicate that higher vote shares for right-wing populist parties increase the trade openness and immigration in certain regions. This does not mean that the populist parties do not favor protectionism, but it indicates that this topic is located outside of the national decision-making zone. Moreover, this also means that populist right-wing politicians cannot implement all election promises at the national level. In order to make the decision-makers in Brussels aware of their opinions, these politicians often decide to intensively use the media and to advocate for leaving the EU. On the other hand, left-wing populists fight for a more generous welfare system. This issue is regulated at the national level (EULEX, 2018) and, hence, requires much less international media coverage, since the politicians can directly implement the desired changes. This argumentation would make sense considering the high media coverage received by Boris Johnson and the Brexit campaign (The Independent, 2016) in contrast to the relatively small international campaign of Greek populists after the bailout (Matsaganis, 2011). These explanations could justify why we expect populist, especially right-wing populists, to have gained a lot more votes than they actually did.

definitely was the primary cause (Stavrakakis & Katsambekis, 2014). Overall, Figure 7 in combination with Figure A of the appendix confirm that the vote shares for populist parties in several countries have increased over the last years.

Figure 7: Share of votes for populist parties in 2012

Source: Manifesto Project Database and CHES (Volkens et al., 2017; Inglehart & Norris, 2016).

Figure 8: Votes for populist parties by European region (1996-2017)

Source: Percentage vote share for populist parties between 1996 and 2017 (Volkens et al., 2017; Inglehart & Norris, 2016).

6 Results

In this section, we will analyze the results of the regression analyses performed to test the three derived hypotheses. Table 1 displays the results of four different regression models. In the first model, we simply regress the main independent variable, namely the import shock, on the change in vote shares for populist parties. The second model extends the first one and adds besides the import shock also the immigration variable. This is done because the theory section indicates that immigration might impact the vote shares through both channels of interest as well. The third model includes the interaction term between those two variables in order to test for the simultaneous effect of both variables. Finally, the last model incorporates GDP, trade openness and social protection benefits as control variables. As the table indicates, we account for year and regional fixed effects as well as for countries experiencing a debt crisis. The data is clustered by country. Furthermore, each model is estimated twice. First, I perform a simple OLS regression using the European import shock from China. In the second step, I use the imports to the United States as an instrument as discussed in the data section.

Table 1: Votes for populist parties

Model 1 Model 2 Model 3 Model 4

OLS IV OLS IV OLS IV OLS IV

Import shock -4.431 -372.4 -3.578 -25.69 -3.437 75.87 -3.134 68.09 (-0.89) (-0.07) (-0.73) (-0.80) (-0.71) (-0.41) (-0.67) (-0.39) Immigration -158.3 -625.1 -117.2 5573.4 207.7 4392.6 (-0.61) (-0.82) (-0.35) (-0.39) (-0.64) (-0.39) IS * IMM -131.3 -13983.3 79.02 -11630.7 (-0.27) (-0.38) (-0.17) (-0.35) Log GDP -61.77* 2.399 (-2.29) (-0.01) Trade openness 0.0832 -0.603 (-0.66) (-0.33) Social benefits -0.00003 -0.027 (-0.01) (-0.38)

Year fixed effects Yes Yes Yes Yes Yes Yes Yes Yes

Region fixed effects Yes Yes Yes Yes Yes Yes Yes Yes

Debt crisis Yes Yes Yes Yes Yes Yes Yes Yes

Observations 102 84 82 82 82 82 82 82

Adjusted R-squared 0.005 0.027 0.013 0.135

t statistics in parentheses

Since the results are insignificant at any conventional significance level and display mainly the opposite sign of what was predicted, I reject Hypothesis 1. According to this hypothesis, I expected the import shock originating from China to have a positive effect on the percentage share of votes for populist parties.

If we consider the effect each dependent and control variable has on the overall vote share for populist parties by itself, we can notice that the results are also very limited. Table 2 shows the outcomes of such individual regressions. Here, I distinguish the effect each variable has on the support for populist parties in each region of Europe. For instance, I regress the change in social protection benefits between two consecutive elections on the change in votes for populist parties in a particular country. As the results indicate, the bigger the increase in social benefits, the lower the support for populist parties in Eastern European countries. Even though this effect is small, it shows that people are less likely to support parties fighting for an even better welfare system in a radical way if improvements are made. Considering the main independent variable of interest, we can see that in Eastern Europe the effect is negative and is confirmed by the IV import shock variable. This contradicts Hypothesis 1 because the outcome claims that the higher the imports from China, the lower the share of votes for populist movements.

Table 2: Effect of variables on the vote share for populist parties

North South East West

Import shock 0.977 19.280 -25.256 -9.469 (1.791) (11.390) (3.248)*** (11.601) Immigration 62.434 -46.214 -902.916 492.371 (95.108) (299.230) (2670.304) (941.361) Log GDP -0.883 -7.589 -248.646 40.715 (4.066) (23.850) (54.091)** (130.734) Social benefits -0.000 -0.003 -0.052 0.004 (0.002) (0.005) (0.011)** (0.004) Trade openness -0.013 -0.119 0.755 -0.033 (0.044) (0.114) (0.415) (0.349) Unemployment 0.117 0.668 1.699 0.111 (0.116) (1.007) (3.322) (2.478) IV Import shock -3.053 -2.950 -73.789 -19.179 (15.920) (13.957) (2.359)*** (26.159) * p<0.1; ** p<0.05; *** p<0.01

distinction between countries with a communist past and others. The results are mainly insignificant as well. Thus, it seems that the outcomes are consistent. Moreover, this table confirms the negative outcome, which we observed in Table 2 for the import shock in Eastern Europe. Another interesting aspect is that unemployment has a positive effect on the vote share of populist parties overall as Table III indicates. This is a reasonable outcome since we expect more trade with China to increase the economic insecurity and unemployment can be related to this aspect. All in all, the variables chosen do not seem to influence the vote shares for populist parties by themselves and if they do, the effect often has the opposite sign than expected. However, these results can also be associated with the fact that I have very few observations for each region. Therefore, the outcomes might be limited in that sense. Moreover, the vote shares for populist movements increased only a little between two consecutive elections, especially for right-wing populist parties. Hence, it might be that the change provoked by Chinese imports is only minimal regarding these small increases.

I will now turn to the subsamples in order to test hypotheses 2 and 3 and consider the effects of Model 1, 2, 3 and 4 on the vote shares for left- and right-wing populist parties. The theoretical analysis stated that both groups might benefit from growing globalization for different reasons. Hence, I will check for that using my regression results. The setup of Table 3 and Table 4 is consistent with the one from Table 1.

Table 3: Votes for left-wing populist parties

Model 1 Model 2 Model 3 Model 4

OLS IV OLS IV OLS IV OLS IV

Import shock -39.3* -194.7 -85.9* -94.74 -61.55 136 -57.87*** -57.87*** (-2.70) (-0.71) (-2.66) (-1.88) (-0.82) (-0.48) (-1.09e+13) (-7.25) Immigration -6233* -6624* -2634 28681 -3459*** -3459*** (-2.57) (-2.32) (-0.25) (-0.71) (-1.63e+13) (-9.06) IS * IMM -5224 -52011 . . (-0.31) (-0.89) . . Log GDP -69.38*** -69.38*** (-4.27e+12) (-3.56) Trade openness 0.668*** 0.668*** (-5.60e+12) (-3.86) Social benefits 0.009*** 0.009*** (-5.23e+12) (-3.67)

Year fixed effects Yes Yes Yes Yes Yes Yes Yes Yes

Region fixed effects Yes Yes Yes Yes Yes Yes Yes Yes

Debt crisis Yes Yes Yes Yes Yes Yes Yes Yes

Observations 30 24 23 23 23 23 23 23

Adjusted R-squared 0.76 0.934 0.929 1

t statistics in parentheses

* p<0.05, ** p<0.01, *** p<0.001

Second, I need to look at the share of votes for right-wing populist parties in order to test Hypothesis 3. This hypothesis states that right-wing populists should gain more votes since they argue for protectionist policies. Looking at the results for right-wing parties in Table 4, we can observe that, similar to Table 1, the table displays mainly insignificant outcomes. Again, this can be related to the fact that the vote shares for right-wing populist parties increased only slowly. Nevertheless, the results for my main independent variable display the expected sign. Furthermore, the outcomes indicate that migration has a small negative effect on the vote shares for right-wing populist parties. This goes against the expectation derived from the theoretical analysis that migration has a positive effect through the economic and the cultural perspective. Nevertheless, it is in line with the findings for left-wing populist parties in Table 3.

import shock and immigration starts having a negative effect on the vote shares. Splitting the sample into before and after the financial crisis indicates that the import shock and immigration have a negative effect on the vote shares for both left- and right-wing populist parties, whereas the interaction term of these two variables has a positive effect (Table IX, Table X). This outcome indicates that in the most recent years, the Chinese import shock has the expected effect on the support for left- and right-wing populist parties when it is interacted with the migration variable. Moreover, it confirms the negative effect of the Chinese shock observed in Table 2 and Table 3. Additionally, I also test whether lagging the independent variables for one period, meaning taking the change that occurred between the two previous elections as independent variables, impacts the main results. I do so in order to determine whether higher imports from China affect the support for populist parties with a delay. The Tables XVI, XVII and XVIII of the appendix show that this does not affect the main results either.

Table 4: Votes for right-wing populist parties

Model 1 Model 2 Model 3 Model 4

OLS IV OLS IV OLS IV OLS IV

Import shock 3.731 -75.19 9.535 -45.42 17.51 17.94 17.16 -5.629 (-0.5) (-0.4) (-1.55) (-0.28) (-1.27) (-1.07) (-1.73) (-0.08) Immigration -1592.7* -1214.6 -121.9 -8.163 -204.7 779 (-2.36) (-0.48) (-0.06) (-0.00) (-0.15) (-0.19) IS * IMM -6348.6 -6833.8 -1425.1 -15398.9 (-0.92) (-1.05) (-0.34) (-0.42) Log GDP -125.9 73.4 (-1.58) (-0.14) Trade openness -0.308 0.605 (-0.48) (-0.21) Social benefits -0.00393 -0.00633 (-0.27) (-0.25)

Year fixed effects Yes Yes Yes Yes Yes Yes Yes Yes

Region fixed effects Yes Yes Yes Yes Yes Yes Yes Yes

Debt crisis Yes Yes Yes Yes Yes Yes Yes Yes

Observations 40 34 34 34 34 34 34 34

Adjusted R-squared 0.443 0.652 0.712 0.755

t statistics in parentheses

* p<0.05, ** p<0.01, *** p<0.001

6.1 Imports from 10 “new” member states of the EU

the European Union (Ec.europa.eu, 2018). Nevertheless, it could be that other trade partners have a bigger impact on the European market than does China. Therefore, it seems reasonable that the effects in Europe are not as pronounced as in the United States (Autor et al., 2016b). Furthermore, Europe is characterized by a lot of intra-European trade. Therefore, it might be the case that the trade shock originating from China did not affect the European countries as much as it affected the United States.

Figure C of the appendix clearly indicates that the 10 “new” member states export more to the 15 “old” member states than China exports to whole Europe over the same time period (Timmer et al., 2015). Thus, the impact of the 10 “new” members that entered in 2004 into the European Union might have been much bigger than the effect stemming from China (Marin, 2017). Since both of those shocks occurred almost during the same time period, it is difficult to distinguish the effects and it might be that the political backlash related to China cannot be clearly seen in the data. In order to test whether imports from the 10 “new” member states have an effect on the voting support for populist parties in the 15 “old” member states, I perform the same regressions as before and substitute imports from China by imports from the countries that joined the European Union in 2004. We suspect imports from these mainly Eastern European countries to have the same effect on the voting support for populist parties as imports from China, since they are lower income countries than the 15 “old” member states (imf.org, 2018) and are characterized, in general, by higher manufacturing labor shares (Timmer et al., 2015).

Table 5 illustrates the outcomes obtained for right-wing populist parties. The results indicate that the growing imports from lower-income European states have the expected positive effect on the voting support for right-wing populist parties in the 15 “old” member states as shown by Model 2. This confirms the outcomes found by Colantone and Stanig (2017) as well as the ones obtained by Swank and Betz (2003) for the support of radical right-wing parties in Western Europe. Moreover, the vote share for left-wing populist parties is positively affected as well, whereas the results for populist parties in general are insignificant as the Tables XIV and XV in the appendix respectively show.

to the growing desire for a more generous welfare state. I assume these outcomes to be robust since the previous results obtained for imports from China seem to be robust as well. However, testing the robustness of these regression results is beyond the scope of this paper.

Table 5: Votes for right-wing populist parties using imports from 10 new member states

Model 1 Model 2 Model 3 Model 4

OLS IV OLS IV OLS IV OLS IV

Import shock EE 18.62 13.86 23.06* 69.59 -35.43 -290.4 -107 -600.1 (-1.55) (-0.35) (-2.91) (-0.53) (-0.53) (-0.18) (-0.91) (-0.37) Immigration -1382 -2818.8 -5872.2 -22127.7 -8518.5 -57798 (-0.69) (-0.61) (-0.88) (-0.22) (-0.50) (-0.35) IS * IMM 20750 100271.3 36108.9 261500.9 (-0.88) (-0.2) (-0.49) (-0.35) Log GDP -161.9 802.4 (-0.25) (-0.24) Trade openness 0.182 1.168 (-0.23) (-0.26) Social benefits 0.0109 0.132 (-0.18) (-0.32)

Year fixed effects Yes Yes Yes Yes Yes Yes Yes Yes

Region fixed effects Yes Yes Yes Yes Yes Yes Yes Yes

Debt crisis Yes Yes Yes Yes Yes Yes Yes Yes

Observations 28 28 28 28 28 28 28 28

Adjusted R-squared 0.783 0.802 0.831 0.921

t statistics in parentheses

7 Discussion

According to the recent literature about the relationship between voting and globalization, I expected the vote shares for populist parties to increase due to the growing economic and cultural insecurity that can be associated with the ongoing global integration process. To contribute to the research about the adverse effects of trade integration with low-income countries, and especially with China, this paper investigated the impact of growing Chinese import penetration on the voting support for populist parties in 24 European countries. The regression results contradict the general claims of the literature that growing import penetration from China leads to a political backlash in developed countries increasing the support for anti-establishment movements (Autor et al., 2016b; Colantone & Stanig, 2017). Taking imports from Eastern European countries, however, indicates that international trade seems to be an important aspect influencing the support of populist parties at least in Western Europe. Therefore, the results require some further discussion.

seems that Chinese import penetration only has the predicted effect in combination with immigration. This can also be related to the fact that China’s export growth accelerated quickly after the financial crisis (Lardy, 2011).

Since the main purpose of this paper was to investigate the impact of Chinese imports on the voting support for populist parties, we now need to discuss why the results differ from the results obtained in previous research and especially from the literature about the United States.

Overall, one can notice that I only have significant results for left-wing populist parties. However, these results reject my hypothesis. The higher the imports from China, the lower the support for left-wing populist parties. I expected the vote share for those parties to increase because I supposed workers to fight for more extensive welfare systems. This assumption was based on the results found for the United States (Autor et al., 2016a). Nevertheless, one can argue that the welfare system in Europe is much better than the one in the US (Alesina & Glaeser, 2006). Since a good system of social insurance and redistribution is already in place, people might be less inclined to vote for populist parties fighting for an even more generous welfare system. Moreover, the welfare system is one reason explaining why there is more inequality in the United States than in Europe (Piketty, 2018). Since inequality is an important channel through which the growing imports from low-income countries affect the voting support for more radical politicians, this might explain why in Europe the effect is not as pronounced.

A second important aspect is that Europe is much more open to trade than the US (Data.worldbank.org, 2018). The US can be considered as a relatively closed economy. Since the presidency of Donald Trump, the United States wants to withdraw from or renegotiate several international trade agreements what could even accentuate this difference (MccGwire, 2017). Overall, the European countries were too small to rely on themselves even before the European Union was created (Braudel, 1994). Therefore, trade has existed in Europe for a long time. This familiarity with trade integration might indicate that Europe is able to adapt its economy quicker to more competition (Marin, 2017). Moreover, some authors noticed that the Chinese export surge improved innovation in European countries. Hence, growing imports had a positive effect on the overall productivity of the European countries (Bloom et al., 2011) what has not been observed in the United States.

Another important aspect is the difference in the political landscape of the US and Europe. European countries are mainly characterized by multi-party systems whereas the United States has a two-party system. Hence, the studies based on the US do not really consider higher support for populist parties; they just look at how the representatives in both parties get more radical (Autor et al., 2016b).

Recently, experts often speak about left or right shifts of governments in Europe (Colantone & Stanig, 2017). This does not necessarily mean that the support for populist parties increases. It simply implies that parties change their attitudes towards certain issues to either side of the political spectrum leading to more political polarization (The Economist, 2018). Thus, considering only the change in vote shares for populist parties, this study does not allow us to determine whether parties moved more to the right or to the left on the political spectrum. The results of this research only show that with growing import penetration from China the share of votes for left-wing populist parties declines. Generally speaking, populist parties have very extreme points of view. Hence, if long-established parties promise to implement the same policies in less drastic ways, people might prefer to support these movements instead.

8 Limitations and future research

Based on the preceding discussion, I will now outline some limitations of the study as well as some future research areas. First, the study indicates that there are definitely some shortcomings with the data and the methodology. Performing tests on the data used show that the main dependent and independent variables are not perfectly normally distributed. This violates an important assumption of the ordinary least squares regressions (Carter Hill et al., 2011). Thus, the standard statistical tests calculated might be biased. Using more advanced econometric procedures could solve this problem in future papers. Not having done this in this paper is related to the fact that the standard OLS and IV regressions have been used in the baseline papers as well (Autor et al., 2013; Colantone & Stanig, 2017).

The CMP database provides only data for national elections and for parties that actually managed to get some seats in the parliament (Volkens et al., 2017). Hence, some data will be missing. For instance, in Germany, a party has to reach a 5% level of votes in order to be represented in the national parliament (Deutscher Bundestag, 2018). This example indicates that the provided data might be biased in some countries based on the legal background for elections. Furthermore, I only used national data. At the regional level, populist politicians might gain more vote shares based on the region they are nested in and based on the fact that it is easier to reach political positions at the regional level. In addition, the basis of this paper is the classification made by Inglehart and Norris (2016). It could be that different parties could be identified as populist using some other definitions.

Looking at the import data, there might be some bias as well. Even though I corrected for inflation, there might be other unobserved issues affecting the data. An exchange rate shock could affect imports and exports even if the shock did not occur between any currency and the Euro or the Yuan. Moreover, the period of the financial crisis was associated with a lot of uncertainty that might have affected the data. In addition to those problems, data over a longer period of time would have increased the robustness of the results as well. Having import data for only 14 years limits the research since I can only observe the outcomes of three or four national elections. Using the WIOD, however, was a choice made based on the fact that it provides information on imports and labor market characteristics in 56 industries and the data has been computed in a uniform way for each country.

Chinese and the Eastern European import shock do not have the same impact on the voting support over the period of analysis. It is important to understand the effect globalization has on the voting behavior of citizens because this aspect is expected to influence domestic politics even more in the future. Moreover, this subject seems to be more relevant than ever with the growing voice of populist parties all over Europe. Even though in this study I do not observe a clear link between international trade with China and increasing vote shares for populist parties, it seems clear that other trade partners have the expected positive effect leading to a more solid voter base for populist parties. Hence, it might be a good start to not only look at China but also at other trade partners and compare the different impacts.

Furthermore, it might be reasonable not to look at populist parties per se, but to consider whether growing international trade has an effect on the political polarization of each party like Autor et al. (2016b) did for American politicians and Colantone and Stanig (2017) for radical right-wing parties in Western Europe. This might be the better option in determining whether trade has an effect on the more radical thinking of the population since populist parties might often have too extremist points of view. These restrain citizens to support these parties.

9 Conclusion

To sum up this paper, it is clear that the main results obtained using imports from China contradict the general outcomes of the literature in this field of study. Nevertheless, some further explanations have been put forward in order to explain these results in the European context. The methodology used is based on the most recent literature on the subject and has been implemented by several authors so far. Even though the general statistics indicate that the share of labor in manufacturing decreased in almost all European nations and the average vote shares for populist parties increased since 1996, the regression analysis does not confirm that these effects are related to growing trade with China. Clearly, globalization seems to have an effect on the voting outcomes in the 25 European countries of interest. Nevertheless, the results contradict the expectation that more imports from China lead to an increase in the vote shares for populist parties. Although the general results are insignificant, I find a clear negative relationship between the Chinese import shock and the share of votes for left-wing populist parties. The robustness checks support these basic findings. Since this outcome contradicts my hypotheses, some further tests were performed in order to determine whether growing imports from other trade partners would lead to the expected effect. The results from this analysis confirm the theoretical assumptions that growing imports from low-income European countries increase the support for populist parties and especially for left-wing populist parties.

One important aspect mentioned throughout the paper is that the EU is composed of states with different cultural, political and legal backgrounds. This might impact the results obtained, since not every country faces the same problems and has the same want for populist parties. In addition, the time period chosen might have impacted the growing support for populist parties more than the actual globalization process. The paper clearly shows that the average European vote share of populist parties increased over the last 20 years. Two important events favoring populism clearly were the financial crisis and the refugee crisis that hit Europe in recent years. The former led to large populist movements especially in countries of Southern Europe experiencing a debt crisis, the latter to growing support for right-wing populist claims in Western Europe. Nevertheless, looking at the countries that have not been affected by the European debt crisis, the share of votes increased only by a little.

as well as the history might explain why the results vary between those two regions. Another point is that I only consider populist parties. Hence, if European governments experience left and right shifts like various experts noticed, this does not necessarily indicate that the vote shares for populist parties increase. If other parties implement the desired policies, citizens might refrain from supporting populist parties because of their, often too extreme, standpoints.

References

Alesina, A. & Glaeser, A. A. (2006). Why are welfare states in the US and Europe so different? Horizons Strategiques n° 2(2), 51–61.

Alvaredo, F., Chancel, L., Piketty, T., Saez, E., & Zucman, G. (2018). World inequality report 2018. Retrieved 24 Mai 2018, from http://wir2018.wid.world/.

Ashcroft, M. (2018). Election 2015: Voting tactically has become more fraught in the new. The

Independent. Retrieved 12 March 2018, from

http://www.independent.co.uk/voices/comment/election-2015-voting-tactically-has-become-more-fraught-in-new-political-order-10227034.html.

Autor, D. H., Dorn, D., & Hanson, G. H. (2013). The China Syndrome: Local labor market impacts of import competition in the United States. American Economic Review, 103(6), 2121–2168.

Autor, D. H., Dorn, D., & Hanson, G. H. (2016a). The China Shock: Learning from Labor-Market Adjustment to Large Changes in Trade. Annual Review of Economics, 8(1), 205– 240.

Autor, D., Dorn, D., Hanson, G., & Majlesi, K. (2016b). Importing Political Polarization? The Electoral Consequences of Rising Trade Exposure. MIT Working Paper.

Bakker, R., Edwards, E., Hooghe, L., Jolly, S., Marks, G., Polk, J., Rovny J., Steenbergen, M., & Vachudova, M. (2015). 2014 Chapel Hill Expert Survey, Version 2015. Chapel Hill, NC, University of North Carolina, Chapel Hill.

Balsvik, R., Jensen, S., & Salvanes, K. G. (2014). Made in China, sold in Norway: Local labor market effects of an import shock. Journal of Public Economics, 127, 137–144.

Beker, V. A. (2014). The European debt crisis: causes and consequences. Economics, Finance and Public Policy, 3(2), 9.

Berger, S. (2000). Globalization and politics. Annual Review Political Science, 3, 43-62.

Bernard, A. B., Smeets, V., & Warzynski, F. (2017). Rethinking deindustrialization. Economic Policy, 32(89), 5–38.

Braithwaite, V., Makkai, T., & Pittelkow, Y. (1996). Inglehart’s materialism-postmaterialism concept: Clarifying the dimensionality debate through Rokeach’s model of social values. Journal of Applied Social Psychology, 26(17), 1536–1555.

Braudel, F. (1994). A History of Civilizations. University of Virginia: The Penguin Press.

Bröning, M. (2018). The Rise of Populism in Europe. Foreign Affairs. Retrieved 21 April 2018, from https://www.foreignaffairs.com/articles/europe/2016-06-03/rise-populism-europe. Budge I. (2013). The standard right–left scale. Manuscript. University of Essex.

Buelens, F. (1999). Globalisation and the Nation-State. Cheltenham, United Kingdom: Edward Elgar Publishing.

Carter Hill, R., Griffiths, W. E., & Lim, G. C. (2011). Principles of Econometrics: Fourth Edition. John Wiley & Sons, Inc.

Census.gov. (2018). Foreign Trade - U.S. Top Trading Partners. Retrieved 15 May 2018, from https://www.census.gov/foreign-trade/statistics/highlights/top/top1312yr.html.

Cohen, R. (2018). The Politics of Backlash. Nytimes.com. Retrieved 10 March 2018, from https://www.nytimes.com/2016/04/05/opinion/the-politics-of-backlash.html.

Colantone, I., & Stanig, P. (2017). The Trade Origins of Nationalist Protectionism: Import Competition and Voting Behavior in Western Europe. Working Paper, 1–27.

Data.worldbank.org. (2018). World Bank Open Data | Data. Retrieved 20 April 2018, from https://data.worldbank.org/.

Dauth, W., Findeisen, S., & Suedekum, J. (2014). The rise of the east and the far east: German labor markets and trade integration. Journal of the European Economic Association, 12(6), 1643–1675.

Deutscher Bundestag (2018). Fünf-Prozent-Hürde. Retrieved 20 May 2018, from http://www.bundestag.de/service/glossar/glossar/F/fuenf_prozent_huerde/247296.

Dippel, C., Gold, R., Heblich, S., Dippel, C., & Gold, R. (2016). Globalization and Its ( Dis- ) Content : Trade Shocks and Voting Behavior Globalization and Its ( Dis- ) Content : Trade Shocks and Voting Behavior.

Ec.europa.eu. (2018). Database - Eurostat. Retrieved 20 April 2018, from http://ec.europa.eu/eurostat/data/database.

EULEX. (2018). EUR-Lex: Article 184-185. Retrieved 20 May 2018, from http://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=CELEX:12012E/TXT&from=EN.

European Union. (2018). Countries - European Union - European Commission. Retrieved 19 April 2018, from https://europa.eu/european-union/about-eu/countries_en#tab-0-1.

Feenstra, R., & Taylor, A. (2014). International Trade. Worth Publishers MacMillan.

Finseraas, H., Roed, M., & Schone, P. (2016). Labor Market Competition with Immigrants and Party Choice : Evidence From A Skill Cell Approach. Working Paper.

Glatzer, M., & Rueschemeyer, D. (2005). Globalization and the Future of the Welfare State. Pittsburgh, Pa: University of Pittsburgh Press.

Goodwin, M. J., & Heath, O. (2016). The 2016 Referendum, Brexit and the Left Behind: An Aggregate-level Analysis of the Result. Political Quarterly, 87(3), 323–332.

Graham, J., Nosek, B. A., & Haidt, J. (2012). The Moral Stereotypes of Liberals and Conservatives: Exaggeration of Differences across the Political Spectrum. PLoS ONE, 7(12).

Imf.org. (2018). Report for Selected Countries and Subjects. Retrieved 10 June 2018, from http://www.imf.org/external/pubs/ft/weo/2018/01/weodata/weorept.aspx?pr.x=64&pr.y=15 &sy=2016&ey=2022&sort=country&ds=.&br=1&c=914%2C946%2C137%2C962%2C911 %2C122%2C912%2C181%2C913%2C124%2C921%2C943%2C963%2C918%2C138%2C 142%2C964%2C182%2C960%2C423%2C935%2C968%2C128%2C922%2C942%2C939 %2C936%2C172%2C961%2C132%2C915%2C184%2C134%2C174%2C144%2C146%2C 944%2C176%2C178%2C136%2C186%2C916%2C926%2C112%2C967%2C941&s=PPPP C&grp=0&a=.

Inglehart, R. (1977). The Silent Revolution: Changing Values and Political Styles Among Western Publics. Princeton University Press.

Inglehart, R., & Norris, P. (2016). Trump, Brexit, and the Rise of Populism: Economic Have-Nots and Cultural Backlash. SSRN Electronic Journal.

Ladner, A. (2008). The Polarization of the European Party Systems – New Data , New Approach, New Results, ECPR General Conference in Glasgow, 1–19.

Lardy, N. R.. (2011). Sustaining China’s Economic Growth After the Global Financial Crisis. Peterson Institute for International Economics.

Marin, D. (2017). The China Shock: Why Germany is different | VOX, CEPR’s Policy Portal. Retrieved 19 May 2018, from https://voxeu.org/article/china-shock-why-germany-different. Matsaganis, M. (2011). The welfare state and the crisis: The case of Greece. Journal of European

Social Policy, 21(5), 501–512.

MccGwire, Jake. (2017). A game theory analysis of Donald Trump’s proposed tariff on Chinese exports. The Student Economic Review Vol. XXXI, 69-77.

Mughan, A., Bean, C., & McAllister, I. (2003). Economic globalization, job insecurity and the populist reaction. Electoral Studies, 22(4), 617–633.

Oesch, D. (2008). Explaining workers’ support for right-wing populist parties in Western Europe: Evidence from Austria, Belgium, France, Norway, and Switzerland. International Political Science Review, 29(3), 349–373.

Piketty, T. (2018). Brahmin Left vs Merchant Right: Rising Inequality and the Changing Structure of Political Conflict, Evidence from France & the US, 1948-2017. WID.world Working Paper Series N°2018/7.

Pryke, S. (2012). Economic Nationalism: Theory, History and Prospects. Global Policy, 3(3), 281–291.

Rodrik, D. (2017). Populism and the Economics of Globalization. Journal of International Business Policy, 1-22.

Ruggie, J. G. (2003). Taking Embedded Liberalism Global : The Corporate Connection. Taming Globalization: Frontiers of Governance, (May 2002), 1–29.

Stavrakakis, Y., & Katsambekis, G. (2014). Left-wing populism in the European periphery: The case of SYRIZA. Journal of Political Ideologies, 19(2), 119–142.

Swank, D. & Betz, H.-G. (2003). Globalization, the welfare state and right-wing populism in Western Europe. Socio-Economic Review, 1(2), 215–245.

The Economist. (2018). European political parties shift right. Retrieved 19 May 2018, from https://www.economist.com/europe/2018/04/26/european-political-parties-shift-right.

The Independent. (2016). Boris Johnson confirms he will campaign for UK to leave EU. Retrieved 14 June 2018 from https://www.independent.co.uk/news/uk/politics/boris-johnson-confirms-he-will-campaign-for-uk-to-leave-eu-referendum-a6887596.html.

Unstats.un.org. (2018). UNSD — Methodology. Retrieved 20 April 2018, from https://unstats.un.org/unsd/methodology/m49/.

Volkens, A., Lehman, P., Matthiess, T., Merz, N., Regel, S. & Wessels, B. (2017). The Manifesto Data Collection. Manifesto Project (MRG/ CMP/ MARPOR). Version 2017b. Berlin: Wissenschaftszentrum Berlin für Sozialforschung (WZB).

Williamson, J. (2002). Winners and Losers Over Two Centuries of Globalization. NBER Working Paper Series, Working Paper 9161.

WorldAtlas. (2018). What Was the Eastern Bloc?. WorldAtlas. Retrieved 20 April 2018, from https://www.worldatlas.com/articles/what-was-the-eastern-bloc.html.

Appendix

Figure A: Share of votes for populist parties in 2004

Figure B: Populism in Europe

Source: Inglehart and Norris (2016).

Figure C: Imports from China and the 10 new member states of the EU