Citation
Effects of Immigrant Flows on the Vote Share for Anti-Immigrant Supranational Parties in the European Parliament

Material Information

Title:
Effects of Immigrant Flows on the Vote Share for Anti-Immigrant Supranational Parties in the European Parliament
Creator:
Flanagan, Danielle Marie
Publication Date:
Language:
English

Notes

Abstract:
The Effects of Immigrant Flows on the Vote Share for Anti-Immigrant Supranational Parties in the European Parliament ( en )
General Note:
Awarded Bachelor of Arts, cum laude, Major: Political Science and Bachelor of Arts, summa cum laude, Major: Economics on May 8, 2018.
General Note:
College or School: College of Liberal Arts and Sciences
General Note:
Advisor: Michelle Phillips. Advisor Department or School: Economics

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Danielle Marie Flanagan. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.

UFDC Membership

Aggregations:
UF Undergraduate Honors Theses

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

Effects of Immigrant Flows on the Vote Share for Anti-Immigrant Supranational Parties in the European Parliament Danielle Flanagan March 9, 2018

PAGE 2

I. INTRODUCTION Amid rising migrant numbers, increased visibility of these communities, and the emergence of xenophobic and far-right parties across Europe, immigration concerns have assumed a greater priority in European Union (EU) policy debates. The changing demographic landscape appears to have heralded a change in the European political landscape, however, the significance in which the former affects the latter has continued to incite controversy, and will thus be the topic of investigation. With the European Parliament as the only directly-elected EU body and originally conceived to represent the "citizens of Europe," it serves as the best political system held constant across the twenty-eight Member States to capture anti-immigrant and far-right political ideologies within the EU. To contextualize the functioning of this European institution, it is 1 important to note that its composition is governed according to the "principle of degressive proportionality," in which seats are allocated according to Member State size, yet in which the more populous Member States have agreed to be under-represented in purposes of allowing greater representation for the less populous countries. Each EU representative is then member 2 to a particular political group in the EU-wide party system or parties in office (see Figure 1), not necessarily affiliating with the same party group as other representatives in their national delegation. Due to the presence of national representatives in a supranational setting, the concept of divided loyalties between national and European party memberships has introduced the EU Barometer surveys poll European public sentiment on issues such as immigration, however, it only 1 consistently measures whether immigration is ranked as a concern though does not differentiate between positive and negative sentiments held across Member States, and thus the data is not included in the study. European Parliament, European Parliament: Organization and Operation, Fact Sheets on the EU, 2 www.europarl.europa.eu/ftu/pdf/en/FTU_1.3.3pdf, (accessed: October 2017), 2017 Page 1

PAGE 3

question as to whether these political groups in the EU Parliament are truly reflective of ideological preferences across the continent. These concerns, however, have been assuaged through Hogenauer's (2017) findings that national parties join supranational party groups they largely agree with, using ideological compatibility as the most important factor for membership. 3 Subsequently, the supranational parties that subscribe to anti-immigrant ideologiesEuropean Freedom and Democracy (EFD) and the Non-Inscrits (NI) which are also characterized by farright sentiments (as displayed in Figure 1), are those that will be studied. 4 Figure 1 : 5 Anna-Lena Hogenauer, "The European Parliament in Times of Crisis: Transnationalism under 3 Pressure?," Journal of Contemporary European Research, 13.2 (2017):1091-1108. The political left-right spectrum differs according to context. Within the study, the spectrum that is 4 referred to is specific to the EU Parliament and the positioning of its party groups on multiple dimensions of policy outlined in McElroy and Benoit (2011). "European Parliament 25 May 2014 Elections," Global Security, globalsecurity.org/military/world/ 5 europe/eu-parliament-2014.htm (accessed: September 29, 2017), 2014. Page 2

PAGE 4

It is to be further noted that the linkage between the presence of immigrants and voting outcomes has an extensive literature, although for the purposes of this investigation the scope is limited to the two prevailing theories: Realistic Conflict Theory and Contact Theory. According to the Realistic Conflict Theory, the presence of a significant out-group (migrants) threatens the in-group's (natives) resource pool, prompting them to vote for the far-right. An alternative to 6 this explanation is provided by Gordon Allport's Contact Theory (1954), which ascertains that instead a negative relationship exists between the presence of the out-group and far-right support. The reasoning follows that it is easier to condemn migrants when one has had little 7 contact with them. The way in which these two theories apply to the European situation will thus be explored, as the relationship between migrant flows and electoral support for anti-immigrant, far-right parties animates the paper. The study thereby posits that EU Member States with high rates of immigration will correspond to a higher vote share for anti-immigrant parties in the European Parliament. While the vote share for anti-immigrant parties is the principle focus of the study, general determinants of vote stare are still relevant and must be given note, mainly the concepts of preferential and strategic voting. Most models in voting choice literature assume that voters select parties they "prefer" or whose policy position they most closely align with, largely expected to be tightly-linked to the voter's education and income level. However, in some models, vote choice can also be strategic, such that voters subscribe to bandwagon and underdog effects, voting for second-choice parties who are perceived to have a better chance of winning or Sebastien Rojon, "Immigration and Extreme-Right Voting in France: A contextual analysis of the 2012 6 presidential elections," International Migration Institute University of Oxford, 2013. Sebastien Rojon, "Immigration and Extreme-Right Voting in France: A contextual analysis of the 2012 7 presidential elections," International Migration Institute University of Oxford, 2013. Page 3

PAGE 5

those parties who are expected to lose. Despite mention to the possibility of strategic voting in the European parliamentary elections, this variable was unable to be accounted for due to the unavailability of data. II. SAMPLE The sample used for this study consists of the twenty-eight European Union Member States across the two most recent European parliamentary election years: 2009 and 2014. It is important to note, however, that the 2013 EU enlargement saw Croatia join as the twenty-eighth Member State, and thus it is excluded from the data for the year 2009. III. DEPENDENT VARIABLE 3.1 Vote share of anti-immigrant parties in the European Parliament The dependent variable is the percentage of the combined vote share within each EU Member State for the Non-Inscrits (NI) and Europe Freedom and Democracy (EFD) supranational parties in the EU Parliament. The vote share in percentage terms of these two parties are bundled together for each Member State, as the two are both characterized by antiimmigrant and far-right sentiments. Rather than tracking the vote share of national parties within each Member State as indicators of such sentiments, the EU supranational parties provide a more uniform measure for comparative study across countries, as individual state anti-immigrant, farright national parties vary significantly along the traditional left-right political scale. The vote share percentage for the specified parties NI and EFD is calculated for both 2009 and 2014, the two corresponding years of the most recent European parliamentary elections. The data Page 4

PAGE 6

are derived from the European Parliament Election Results for the seventh and eighth parliamentary terms. 8 IV. INDEPENDENT VARIABLES 4.1 Rate of Immigration The rate of immigration accounts for the number of immigrants relative to the size of the resident population, calculated as immigrants per 1,000 persons. A five-year weighted average of immigration rates into each Member State is calculated for the years spanning 2005 to 2009 as well as 2010 to 2014, the former period corresponding to the years preceding the seventh European Parliament elected in 2009 and the latter period corresponding to the years preceding the eighth European Parliament elected in 2014. The data are obtained from Eurostat: Migration and Migration Population Statistics. However, due to missing total population data in Eurostat, 9 population statistics from the World Bank were pulled for the following countries in the specified years: Bulgaria 2005, Croatia 2005-2012, Estonia 2005, Greece 2008, Poland 2009, Portugal 2005, and Romania 2005-2011. 10 The primary purpose of the investigation is to determine whether there is a correlation between the rate of immigration and the vote share for anti-immigrant parties in the European Parliament. It is thus hypothesized that Member States with high immigration rates will have a European Parliament, Results of the 2009 and 2014 European Elections European Parliament, 8 www.europarl.europa.eu/elections2014-results/en/mobile-country-at-2009.html (accessed: October 2017), 2014. Eurostat, Migration and Migration Population Statistics, European Union, ec.europa.eu/eurostat/ 9 statistics-explained/index.php/Migration_and_migrant_population_statistics (accessed: October 2017), 2017. World Bank, "Population, Total," World Bank Data, data.worldbank.org/indicator/SP.POP.TOTL? 10 end=2012&start=2011&view=chart&year_low_desc=true (accessed: October 2017), 2017. Page 5

PAGE 7

higher vote share for the anti-immigrant parties NI and EFD in the European Parliament across the 2009 and 2014 elections. However, it is essential to note that this variable does not account for the possibility that migrants might avoid settling in countries with high antiimmigrant sentiments With the availability of data across a larger range of election years, the change in immigration rate across such years for each Member State would instead be calculated to control for this possibility in the data. 4.2 Unemployment Rate The unemployment rate, measured as the total unemployment as a percentage of the labor force, is accounted for in each Member State for the years 2009 and 2014, and obtained from Eurostat data. It should be considered that the data from this time coincided with the onset of 11 the Great Recession in Europe in 2009, when high unemployment was a common characteristic of many EU Member States. Furthermore, high unemployment has been associated with the electoral success of the far-right within political economy literature. In reference to Jackman and Volpert (1996), their findings confirm that higher rates of unemployment are linked to higher levels of electoral support for extreme right parties, as it creates favorable conditions for such political movements. In a similar vein, anti-immigrant attitudes have been observed to be more 12 pronounced with high unemployment, as migrant newcomers are perceived to flood the market for labor, thereby displacing native workers. Subsequently, it is hypothesized that higher unemployment rates in terms of percentage of the active population within EU Member States will correspond with a larger vote share for the NI and EFD in the 2009 and 2014 elections. Eurostat, Unemployment Statistics European Union, ec.europa.eu/eurostat/statistics-explained/ 11 index.php/Unemployment_statistics (accessed: October 2017), 2017. Robert W. Jackman and Karin Volpert, "Conditions Favoring Parties of the Extreme Right in Western 12 Europe," British Journal of Political Science 26.4 (1996): 501-521. Page 6

PAGE 8

4.3 Change in Income Inequality The Gini coefficient of equalized disposable income, accounting for income inequality within each Member State, can be obtained from Eurostat: EU Statistics on Income and Living Conditions Survey. The coefficient ranges from 0 to 100, 0 indicating perfect equality and 100 13 indicating perfect inequality. The change in income inequality for each respective Member State was calculated across a three-year period prior to each European parliamentary election. Thus, the years under review were 2007 to 2009 and 2012 to 2014. The change in income inequality was not measured across a larger frame of time due to the unavailability of data for two Member States Bulgaria and Romania as they joined the EU in 2007. Although income is often used as a mainstay for voter preferences, income inequality is instead selected since it more comprehensively captures intra-country wealth disparities. According to the 2015 study by Jetten, Mois, and Postmes, opposition to immigration was concluded to be higher for all income levels when inequality was growing rather than declining. Thus, within the scope of this study, 14 it is hypothesized that with an increase in the Gini coefficient within a Member State, the greater its vote share will be for the NI and EFD. 4.4 Proportion of Urban Population This independent variable is included to gauge the population distribution by degree of urbanization, and more specifically, the percentage of each Member States' population living in urban areas. The data are taken from Eurostat: Statistics on Rural Areas in the EU, for the years Eurostat, Gini Coefficient of Equivalised Disposable Income European Union, ec.europa.eu/eurostat/ 13 web/products-datasets/-/ilc_di12 (accessed: October 2017), 2017. Jolanda Jetten, Frank Mols, and Tom Postmes, "Relative Deprivation and Relative Wealth Enhances 14 Anti-Immigrant Sentiments: The V-Curve Re-Examined," PloS One, 10.10(2015): p. e0139156 Page 7

PAGE 9

2009 and 2014. Controlling for the degree of urbanization ensures that the potential 15 confounding factor of a political rural-urban divide will not influence the relationship between the rate of immigration and vote share for the NI and EFD in the EU Parliament. Substantiated by literature on voter preferences, urban areas tend to lean to the left and rural areas tend to align with the right. Thus, it is hypothesized that Member States with a smaller urban population 16 recorded in percentage terms will correspond with a higher vote share for the NI and EFD in the European Parliament. 4.5 Tertiary Education Tertiary education, defined as levels 5 to 8 according to the International Standard Classification of Education, is accounted for as a percentage of the population for each EU Member State in the years 2009 and 2014. The data are obtained from Eurostat. This variable 17 is used to account for the percentage of the population with a high education level, as it includes individuals who have pursued postsecondary education or studies beyond a high school diploma. Several studies have substantiated the relationship between education level and voting behavior. Particularly within Lubbers and Tolsma (2002), lower educated individuals were indicated to be more likely to have radical right-wing voting preferences, even including those individuals of the Eurostat, Statistics on Rural Areas in the EU European Union, ec.europa.eu/eurostat/statistics15 explained/index.php/Statistics _on_rural_areas_in_the_EU#Population_distribution_by_degree_of_ urbanisation (accessed: October 2017), 2014. Jason Roy, Andrea Perrella, and Joshua Borden, "Rural, Suburban and Urban Voters: Dissecting 16 Residence Based Voter Cleavages in Provincial Elections," Canadian Political Sciences Review 9.1 (2015): 112-127. Eurostat, Educational Attainment Statistics European Union, ec.europa.eu/eurostat/statistics-explained/ 17 index.php/Educational_attainment_statistics (accessed: October 2017), 2016. Page 8

PAGE 10

same educational category who did not agree with the ideology of the far-right. Therefore, an 18 inverse relationship is hypothesized to exist between tertiary education and vote share for the NI and EFD: the greater the percentage of the tertiary educated population of a Member State, the lower the vote share will be for the NI and EFD in the European Parliament across the two observed election years. 4.6 Omitted Independent Variable: Religiosity A significant variable in political economy literature observed to affect voter preferences is religiosity. Religiosity is defined as "the part of the population that identifies themselves as being a religious person." Within the frame of this study, the variable cannot be accounted for, 19 as there is a lack of data that accounts for the religiosity or lack thereof (the percentage of the population that is unaffiliated) that spans more than one year and is within the temporal context of the study. It is to be noted, however, that previous studies have observed that higher levels of religiosity are correlated with increased rightist voter preferences. 20 Marcel Lubbers, and Jochem Tolsma, "Education's Impact on Explanations of Radical Right-Wing 18 Voting," Conference on Migration 2011. Jan-Emmanuel De Neve, "Endogenous Preferences: The Political Consequences of Economic 19 Institutions," LSE Political Science and Political Economy Working Paper No. 04/2009, 2013. Jan-Emmanuel De Neve, "Endogenous Preferences: The Political Consequences of Economic 20 Institutions," LSE Political Science and Political Economy Working Paper No. 04/2009, 2013. Page 9

PAGE 11

V. RESULTS 5.1 Descriptive Statistics of Data (n=55) 5.2 Correlation Matrix The correlation matrix provided above addresses the potential concern of multicollinearity between the variables. As evidenced by the correlation coefficients, no two variables have a coefficient larger than 0.8, the largest being equal to a rounded 0.36. 5.3 Regressions and Results Model 1: Vote Share = 0 + 1 (Immigration Rate) + 2 (Unemployment Rate) + 3 (Income Inequality) + 4 (Urban Population) + 5 (Tertiary Education) + error 21 Variable Mean Standard Deviation Minimum Maximum Vote Share (%) 6.95 8.66 0.00 30.38 Immigration Rate (per 1,000) 8.56 7.42 0.20 38.46 Unemployment Rate (%) 9.70 4.56 4.4 26.50 Change in Income Inequality (0 to 100) -3.33 4.03 -11.70 3.30 Urban Population (%) 40.92 14.49 14.50 89.50 Tertiary Education (%) 24.46 7.07 11.20 36.60 Initially, a regression was run including the Crime Rate (per 100,000 inhabitants) and the Median Age 21 of the Population for each Member State, however, such variables were excluded, as all the results were statistically insignificant. Page 10 Imm_Rate Unemployme nt_Rate Income_Ineq uality Urban_Pop Tertiary_Edu Imm_Rate 1.00 Unemployme nt_Rate -0.23 1.00 Income_Ineq uality -0.15 0.29 1.00 Urban_Pop 0.07 -0.02 0.26 1.00 Tertiary_Edu 0.36 0.08 -0.01 0.03 1.00

PAGE 12

Rate of Immigration In examining the regression output, the immigration rate is observed to be statistically significant at the 90 percent confidence level. Thus, the null hypothesis stating that the immigration rate has no effect on the vote share for anti-immigrant partiesNI and EFDin the European Parliament can be rejected. It is especially important to consider that with the small number of observations (n=55), the statistical findings were still significant, and thus would presumably be significant at an even higher confidence level as n increases. The negative coefficient implies that with a one unit increase in the immigration rate measured per 1,000 inhabitants, the combined vote share for the two anti-immigrant parties decreases by a rounded 0.32 percentage point. This unexpected sign differs from the a priori expectation that a higher rate of immigration would correlate with a higher vote share for the NI and EFD. Interestingly, however, these findings can be explained by the Contact Theory, which claims that a negative relationship exists between the presence of migrants and far-right, antiimmigrant support. Therefore, the Contact rather than Conflict theory appears to be consistent with the results found within the context for the EU in the mentioned years. Regression Table Coefficients Standard Error t Stat P-value Intercept 7.79 6.31 1.24 0.22 Imm_Rate -0.32 0.18 -1.80 0.08 Unemployment_ Rate -0.44 0.28 -1.58 0.12 Income_Inequali ty 0.20 0.32 0.62 0.54 Urban_Pop 0.06 0.08 0.76 0.45 Tertiary_Edu 0.17 0.18 0.95 0.35 Page 11

PAGE 13

Unemployment Rate The unemployment rate variable is statistically insignificant, as indicated by the p-value of 0.12. Thus, the null hypothesis asserting that the rate of unemployment has no effect on the vote share for the anti-immigrant partiesEFD and NIin the European Parliament cannot be rejected. It is important to note, however, that this may be a result of the limited number of observations (n=55) available across the two recent election years. Change in Income Inequality The change in income inequality, measured by the Gini coefficient, is statistically insignificant with the p-value of 0.54. Subsequently, the null hypothesis, which states that a change in the level of income inequality or Gini coefficient in each Member State has no effect on the vote share for the anti-immigrant parties in the EU Parliament, cannot be rejected. Urban Population This variable capturing the percentage of the urban population within the EU Member States is statistically insignificant, evidenced by its p-value of 0.45. The null hypothesis, asserting that the percentage of the urban population within each EU Member State has no effect on the vote share of the NI and EFD in the European Parliament, cannot be rejected. Tertiary Education The final variable is also found to be statistically insignificant with a p-value of 0.35. The null hypothesis, stating that the percentage of the tertiary-educated population within each EU Member does not have an effect on the vote share for the NI and EFD in the European Parliament, cannot be rejected. Page 12

PAGE 14

5.4 Notes It is important to address that the model explains little of the variability in the data, as indicated by the low adjusted R-squared (0.01). This finding can be a potential result of the limited sample, as the data is constrained to only two elections years, thereby making it difficult to identify a consistent trend regarding effects on vote share for supranational parties in the parliament. An anomaly election year has the potential to have amplified effects in the data due 22 to the small sample size. For means of improvement, the study can be extended to encompass the elections of the ninth and tenth European Parliaments in 2019 and 2024, respectively. The effects of an increased number of observations are twofold: first, it will allow for an analysis over a larger timeframe and second, it will allow for the addition of more variables that may exert an influence on the EU parliamentary vote share. VI. CONCLUSION The findings of the investigation encourage further study, as the immigration rate is evidenced to have a statistically significant influence on the combined vote share for the EFD and the NI in the European Parliament. Specific to the context of the EU, the results saw a departure from the Conflict Theory, instead being consistent with the Contact Theory, positing Although not accounted for in the study, a potential consideration must be given to the possibility that 22 Member States more tolerant towards migrants may generally be inclined to vote against far-right supranational parties rather than voting against such parties due to the presence of migrants. Page 13 Regression Statistics Multiple R 0.32 R Squared 0.10 Adjusted R Squared 0.01 Standard Error 8.60 Observations 55

PAGE 15

that with increased exposure, people tend to become more tolerant of migrants. However, due to the lack of data across a range of years, much of this will remain speculation. With the recent development of the 2015 migrant crisis where an unprecedented influx of migrants numbering to over one million crossed EU borders, the elections of the ninth and tenth European Parliaments in 2019 and 2024 may be more indicative of the effects of the immigration rate on parliamentary vote share. A study including these years could even capture the 23 influence of a sudden surge of immigrants into each EU Member State, and whether a rapid change in the immigration rate (rather than the studied five-year weighted average) has an impact on the vote share. This expansion in the temporal frame would also increase the number of observations and thereby permit the addition of variables, potentially resulting in a higher adjusted R-squared so that the variability between immigration rates and parliamentary vote share in the model could be better explained. The contemporary nature of the crisis and polarizing political climates within Member States deem it beneficial to continue monitoring the parliamentary vote share of the far-right, anti-immigrant parties along with additional variables in order to discern whether these results are microcosms of a larger phenomenon found in migration studies literature. Within such literature, the Contact Theory has served as a guidepost in the analysis of both experimental and survey research, typifying the pattern that interaction between groups (migrant and native) reduces prejudice among them. Ultimately, the study brings the topics of migration and voter preferences to the fore, attempting to determine whether the flow of immigrants serves as a catalyst to a more politically divided or unified Europe. Jeanne Park, "Europe's Migration Crisis," CFR Backgrounders, Council on Foreign Relations, 23 cfr.org/refugees-and-the-displaced/europes-migration-crisis/p32874 (accessed: October 2017), 2015. Page 14

PAGE 16

References De Neve, Jan-Emmanuel. "Endogenous Preferences: The Political Consequences of Economic Institutions," LSE Political Science and Political Economy Working Paper No. 04/2009, 2013. Eurostat, Educational Attainment Statistics European Union, 2016. Eurostat, Gini Coefficient of Equivalised Disposable Income European Union, 2017. Eurostat, Migration and Migration Population Statistics, European Union, 2017. Eurostat, Statistics on Rural Areas in the EU European Union, 2014. Eurostat, Unemployment Statistics European Union, 2017. "European Parliament 25 May 2014 Elections." Global security.org globalsecurity.org/military/ world/europe/eu-parliament-2014.htm (accessed: October 2017). 2014. European Parliament, European Parliament: Organization and Operation, Fact Sheets on the EU, 2017. European Parliament, Results of the 2009 and 2014 European Elections European Parliament, 2014. Hogenauer, Anna-Lena. "The European Parliament in Times of Crisis: Transnationalism under Pressure?." Journal of Contemporary European Research, 13.2 (2017):1091-1108. Jackman, Robert W., and Karin Volpert, "Conditions Favoring Parties of the Extreme Right in Western Europe," British Journal of Political Science 26.4 (1996): 501-521. Jetten, Jolanda, Mols, Frank, and Tom Postmes, "Relative Deprivation and Relative Wealth Enhances Anti-Immigrant Sentiments: The V-Curve Re-Examined," PloS One, 10.10 (2015): p. e0139156 Lubbers, Marcel, and Jochem Tolsma. "Education's Impact on Explanations of Radical RightWing Voting," Conference on Migration 2011. McElroy, Gail and Kenneth Bennoit. "Policy positioning in the European Parliament," European Union Politics 13.1 (2012): 150-167. Park, Jeanne. "Europe's Migration Crisis." CFR Backgrounders Council on Foreign Relations. www.cfr.org/refugees-and-the-displaced/europes-migration-crisis/p32874 (accessed: October 2017), 2015. Rojon, Sebastien. "Immigration and Extreme-Right Voting in France: A Contextual Analysis of the 2012 Presidential Elections," International Migration Institute University of Oxford, 2013. Roy, Jason, Perrella, Andrea and Joshua Borden. "Rural, Suburban and Urban Voters: Dissecting Residence Based Voter Cleavages in Provincial Elections," Canadian Political Sciences Review 9.1 (2015): 112-127. World Bank, "Population, Total," World Bank Data, data.worldbank.org/indicator/ SP.POP.TOTL?end=2012&start=2011&view=chart&year_low_desc=true (accessed: October 2017), 2017. Page 15