Trade Union Trade-Offs: Unions, Voters, and the Rise of Right-Wing Populism

Trade union membership in European Union member states has been in decline for decades, which has many concerned about the future of workers’ rights. While existing work examines the reasons for this decline, my research shifts the focus from union density to the functions unions serve and how these functions affect and are affected by changing electoral behavior. I examine the rise of right-wing populist movements in Europe and how these movements and the challenges today’s labor unions face can be traced to the same underlying forces. I argue that, as the relevance of trade unions declines for blue-collar workers, support for right-wing movements increases. I test this claim with the help of statistical analyses of European Values Study data. I find preliminary support for my hypothesis, and suggest how this can inform our responses to the rise of populism in Europe.

In recent years, the rise of radical right-wing populist movements throughout the European Union has been observed by a wide variety of social scientists and journalists.While this rise has been significant since the 1990s in some areas, the success of the Leave campaign in the United Kingdom (UK) and the election of Donald Trump in the United States (US) have brought these movements into the forefront of the greater political dialogue.A variety of explanations have been fielded for the increasing popularity of these movements, but none adequately explain the electoral behavior of the working class.In this paper, I present an alternative explanation: as trade unions increasingly fail to engage a changing working class, low-wage and blue collar workers turn to populist movements touting anti-establishment and anti-globalization rhetoric.
I examine the relationship between the declining strength of labor unions and the rise of radical right-wing populist movements, and use that analysis to understand how the same underlying forces cause both.I then present suggestions for how the European Union can use this understanding to inform the responses to illiberal, Eurosceptic movements while simultaneously addressing the legitimate concerns of the working class.

The Rise of Populism
Founded in 1972, the French National Front party was nothing more than a fringe movement with virtually zero electoral success for the first decade of its existence.Under Jean-Marie Le Pen, however, these right-wing radicals gained 9.6 percent of the vote in 1988 and more than 14 percent in French presidential elections.Similarly, the Swedish radical right-wing Autopartei (literally "Automobile Party") went from two seats in parliament to 10 in just four years between 1987 and 1991.Similar radical right populist parties like Vlaams Blok and Ny Demokrati emerged from virtual unrecognition to relative electoral success across Europe in the 1980s and 1990s (Betz, 1994).While populist movements vary greatly by geography, they usually share common ultra-nationalist, Eurosceptic, and anti-globalization themes.Jens Rydgren (2005) theorizes that extreme right-wing populist parties are part of a family of ideologies, supported by and learning from one another.However, these radical right-wing parties often don't match traditional right-wing beliefs in every way.
Consider the Italian Five Star Movement (M5S).M5S is one of the most distinctive populist movements currently facing Europe because it strives to remain above either the right or the left, instead seeking an anti-party agenda.While many movements have found a wider audience for their ideas through the internet, M5S considers it its genesis.For M5S, the internet is "a medium remorselessly eliminating mediation of all kinds that will eventually destroy parties and make possible a form of direct democracy if the people control the government through constant voting over the web (The Economist, 2016)."Some have called this strategy unrealistic, even Messianic in its idealism, but M5S can give us great guidance in understanding how these populist movements grow into disruptive electoral factions.By repeatedly telling the story of hyper-partisanship ruining the function of government and the establishment using these functions to their own gain, an aggressively anti-system rhetoric can be largely void of tangible policy and still attract a high volume of voters.
In June of 2016, the United Kingdom shocked the world when a referendum to leave the European Union passed despite the advice of former prime ministers, business http://scholarship.claremont.edu/urceu/vol2017/iss1/8 and financial leaders, and despite polls predicting the opposite.Markets, media, and even pro-Brexit voters were shocked by the move, the first major populist upset of the year.The primary question of the aftermath was how this movement, which had been dismissed as far-fetched and foolish in the 1990s, gained so much support so quickly.The Financial Times suggests four ingredients which caused the change: dissatisfaction with the elite, the ubiquitous urban/rural divide, immigration conflicts, and a perceived loss of control.The Vote Leave campaign harnessed voters with these concerns by peddling an anti-immigration message focused on the inability of the political elites, like former Prime Minister David Cameron, to lead change.By advertising Brexit as the move that would finally set the UK free from the elites and the bureaucracy, the Leave campaign came out the other side stronger, louder, and smarter than the Remain campaign.
As Michael Broenig (2017) suggests, populist movements have gained strength in Europe partially due to the convergence of the mainstream center-left and center-right parties, which has left those outside the center without mainstream leaders or representation.Disenfranchised, often working class voters turn to the rising populist movements who promise a refuge from the mainstream 'establishment' or 'elite.'This also manifests in antidemocratic sentiments, as noted by scholars such as Robert Foa and Yascha Mounk (2016) who examined a World Values Survey wave published in 2016 and found that support for authoritarian forms of government is on the rise, while support for democratic institutions is declining, especially among the younger generations.Social and electoral support for radical right-wing populist movements across the western world is on the rise.
While it may be tempting to assume this trend will dwindle in favor of the traditional centrist parties and institutions that have defined the western world for the last century, there is a danger to this: the election of Donald Trump shows that even the most stable democracies can be susceptible to these anti-establishment movements.If we follow this trend to its logical end, there is a very real risk that the same liberal institutions we rely on for protection of human rights, basic environmental policy, and even conflict resolution will become obsolete to the detriment of people and societies worldwide.Trade, human migration, and international cooperation all risk major setbacks from which it could take decades to recover.

The Decline of Trade Unions
Trade union membership has been on the decline across western countries for the last three decades, in many areas falling to record lows, as reported by the International Labour Organization (2015).Many have long assumed that labor unions are inevitably going to become extinct; unless trade unions substantially shift their focus towards new and emerging demographics within the labor force and reevaluate their relationships with government entities, they may indeed.Union leaders have oft conceded these challenges.As the president of the American Federation of Labor and Congress of Industrial Organizations (AFL-CIO) said in 2013; "[O]ur still-struggling economy, weak laws and political as well as ideological assaults have taken a toll on union membership, and in the process, have also imperiled economic security and good, middle-class jobs" (Trumka, 2013).
European trade union leaders are responding with more optimism to the similar situation they face.Brendan Barber, General Secretary of the Trades Union Congress, said: Trade Union Trade-Offs Today's relatively small fall in the number of union members is actually a union success story, given the continuing decline in traditionally unionized jobs in sectors such as manufacturing.Indeed, after falling strongly through the 1980s and early 1990s, union membership has roughly stabilized since 1997 … economic changes mean that unions have to run hard just to keep still.(Waddington, 2005) Increased trade and rapid globalization has left the unskilled working class feeling victimized by lower wages and the uncertain futures of their jobs in manufacturing and industry, and the labor movement has failed to keep up with their changing concerns.Furthermore, trade unions have yet to design a successful method of organizing low-wage service workers, the still-growing largest sector of low-wage labor.Existing research focuses on three central explanations for this trend: macroeconomic conditions (sometimes called the 'business cycle' explanation), the changing composition of the labor force, and increasing legislative barriers.To understand the relationship between the decline of trade unions and the rise of populism, I will consider each of these explanations briefly, leaving further debate regarding the reasons for continuing trade union decline to future research.
Using macroeconomic trends to understand changes in union density is an overall rather successful approach.This model shows that, in the short run, union density is related positively to changes in employment and price inflation and negatively to unemployment and wage inflation.Furthermore, in the long run, union density is negatively related to real wage growth (Disney, 1990).Union membership is also closely tied to the number of employees in a business (measured as average firm size), such that union density is strongest with large firm sizes and low unemployment-signs of a strong economy-and weakest when unemployment is high and firm sizes are small.This model therefore suggests that we cannot expect to see union density increase substantially until real wages and employment are also increasing.
The most popular and most widely accepted and cited explanation for the decline of labor unions is the changing composition of the labor force both from demanders and suppliers of labor.In virtually every developed country, we can observe the economic shift from manufacturing and industry to service economies and white-collar labor.For labor unions seeking new members, workers in a service economy (what we might traditionally call 'white-collar') can be prohibitively difficult to organize.Less homogenous preferences and the lack of a historically organized community among these workers are factors, as is the transient relationship between workers and employers, which is caused by the growing trend of flexibility and individualism in the workplace.Indeed, a paper from the European Trade Union Institute for Research, Education and Health and Safety claims: Employment is shifting from industry, where unionization rates tend to be relatively high, to private sector services, where unionization rates are lower.Associated with this shift are the growth of atypical forms of employment, employment at small workplaces and new forms of employment relationship.Trade union movements have yet to adjust to these developments with the consequence that members are being lost from industry at a higher rate than they are being recruited in private sector services.(Schnabel, 2012) http://scholarship.claremont.edu/urceu/vol2017/iss1/8 As workers become more likely to work short-term jobs and change jobs frequently, workplace culture has shifted toward growing emphasis on the individual relationships between employees and employers, which draws workers away from the sense of community that is so important to union recruitment techniques.The lower unionization rates of this new workforce have contributed substantially to decreasing density.
Public opinion has changed as well, and unions are seen more and more negatively in the western world.In a collection of surveys conducted by Gallup in the United States, workers point to concerns about corruption and greed within labor unions.Overall, public opinion of labor unions has decreased significantly since its high point in 1955.Currently, only 56% of people in the US have a favorable view of labor unions (Gallup, 2016).The same polls also suggest that more people think labor unions are unnecessary thanks to the gains already made in labor rights.
Structural changes in domestic industrial relations have made unions increasingly less relevant to the average worker.The difference in wages between unionized and nonunionized workers has shrunk substantially, leading the costs of union membership to outweigh the benefits.Employment law has also increasingly favored the militantly anti-union stance of employers recently, and right-to-work laws, which have inhibited union access to the workplace, make it increasingly difficult to recruit new members.Using a simple linear regression analysis to explore the long-term effects of changes in legal regulations, Freeman and Pelletier (1990) find that more recent legislation has been a major driver of decline in union membership.These authors also find a major difference in membership in workplaces where unions are recognized versus those where unions have not gained recognition, showing that the attitudes of management toward trade unions is also a major determinant of the strength and density of unions (Freeman & Pelletier, 1990).
Similarly, states that follow the Ghent system, in which unemployment insurance is distributed by trade unions, see significantly higher rates of union density as compared to non-Ghent states.But even where unions play a vital legal function, membership is still affected by the other variables discussed here.However, even these systems are eroding in favor of more centralized welfare systems (Böckerman, 2005).

Where Trade Unions and Populism Intersect
The rise of globalization has substantially changed the interest of the working class.Regions rich in capital continue to see significant decreases in jobs and slowing growth in the manufacturing and industry sectors due to the significantly cheaper labor markets elsewhere and the automation of many unskilled jobs.These displaced workers, in need of secure employment, must either enter further occupational training and education programs or must accept significantly lower wages and competition from migrant workers, automation, and off-shoring for their jobs.Thus, it is reasonable for the unskilled working class to move to the right, towards protectionist trade policy, anti-immigration stances, and a general distrust of democratic institutions, or 'the establishment,' which are benefiting from the cheaper labor markets available outside of capital rich regions.Unfortunately for trade unions and other democratic institutions, their commitment to the liberal ideals attractive to skilled workers, skilled laborers, or working professionals are no longer attractive to the unskilled, low-wage sectors of the working class.
As labor unions increasingly represent these groups who aren't as threatened by Trade Union Trade-Offs factors like globalization and immigration-groups which may in fact benefit from them in the long run-they are accordingly maintaining left-leaning policies.Unions throughout North America and Europe have released statements on the importance of assisting refugees in workplace integration and on the benefits of globalization and increased trade.As trade union density decreases over time and membership shifts toward professionals and skilled workers, the low-wage earning working class, which is more likely to perceive a threat from increased trade and immigration, does not have the traditional advocate they had during the early decades of the labor movement.They therefore seek representation in emerging populist movements which tout antiestablishment, ultranationalist ideals.The rise of radical right populist movements in North America and Europe provides an opportunity for massive swaths of the working class who feel their interests are not represented by trade unions to find leaders who propose increasing barriers to human migration, heavily penalizing outsourcing of jobs and businesses through high tariffs, and a return to domestic manufacturing and industry as solutions to the specter of a highly-educated workforce and service based economy.What emerges is a struggle between the liberal trade unions and the radical right wing populists for control of the working class voting bloc.
If this trend can be demonstrated, it means there is a clear shortfall in our current institutions: the ideals of liberal democracy do not fully represent a working class which is becoming less and less satisfied with the organizations meant to guarantee their interests.

Research Design
To test my theory, I analyze the relationships between public opinion related to labor unions and political parties, including union membership and other variables.I restrict my analysis to data from the European Values Study, because it is sufficiently uniform and reliable.The European Values Study (hereafter called "EVS") data comes from waves 1 thru 4, between 1981 and 2008.These results include nearly 165,000 unique observations in 48 countries 1 .Questions are translated and altered for best fit in each geographic region (e.g.confidence in the European Union is not measured in the United States), and all waves use a universe which includes the adult population of a country, ages 18 and older 2 .This survey used representative multi-stage or stratified random sampling throughout and was conducted through face-to-face interviews.The relevant survey questions are: whether the respondent belongs to a trade union, the level of confidence in trade unions, and where the respondent places her-or himself on the political scale from left to right, with 1 being the furthest left and 10 being the furthest right (see Table 1).I use IBM's SPSS Statistics 24 for my analysis.
http://scholarship.claremont.edu/urceu/vol2017/iss1/8If the null hypothesis is correct, there will be no difference between union members and non-union members.To display the difference which I believe exists, I create contingency tables for each EVS wave using variables a067 and e033, and variables e069_05 and e033.If my hypothesis is correct, I should find that union members are increasingly more likely to place themselves on the right side of the political scale over time, though they should still be much less likely than non-union members to identify with the far right.Analysis should also show that those with high confidence in labor unions (variable e069_05) are more likely to position themselves to the left of the political spectrum, while those with low confidence are more likely to position themselves to the right.These differences should remain constant over time; however, the full sample should show a shift to the right.
To further test the relationships between these variables, I attempt an ordinal regression analysis.I use the same variables from the EVS Longitudinal dataset, but recode the data to be conducive to an ordinal regression analysis by creating a series of dummy variables for each of the two independent variables, such that: confidence_1 is E069_05=Missing, Unknown; confidence_2 is E069_05=No answer; confidence_3 is E069_05=Don´t know; confidence_4 is E069_05=A great deal; confidence_5 is E069_05=Quite a lot; confi-dence_6 is E069_05=Not very much; confidence_7 is E069_05=None at all; and so forth, using the same logic for the membership variable.

Findings and Analysis
My analysis of data gathered by the European Values Study between 1981 and 2008 shows preliminary support for my theory.A comparison of contingency tables from each EVS wave between 1981 and 2008 showed that union members were less likely than non-union members to place themselves on the right end of the political spectrum until 2008 (see Table 2).However, the proportion of union members who placed themselves on the right end of the political spectrum increased much more quickly than that of nonunion members.Indeed, the number nearly doubled between the third and fourth waves (see Figure 1).  2 shows the percentage of respondents who position themselves 9-10 on the political scale from left to right, with 1 being the furthest left and 10 being the furthest right.Adapted from EVS (2015): European Values Study Longitudinal Data File 1981-2008(EVS 1981-2008).GESIS Data Archive, Cologne.ZA4804 Data file Version 3.0.0,doi:10.4232/1.12253.These data suggest that labor union members are increasingly identifying with the far right of the political spectrum, contrary to historical norms.Furthermore, the spike in union members placing themselves at 9 or 10 on the spectrum between 1999 and 2008 may be indicative of the growing popularity of right-wing populist movements.Analysis of EVS2017, which will begin in August of 2017, may provide additional evidence of this trend.Additional research into the rates of membership decline and the rates of increasing right-wing identity could also help better describe these changes.
Analysis of respondents stratified by their level of confidence in labor unions shows that those with high confidence are more likely to position themselves to the left on the political spectrum (1 or 2), while those with low confidence are more likely to position themselves to the right (9 or 10).These differences remain constant over time and display a measurable shift toward the far-left and -right ends of the spectrum, supporting the idea that voters overall are more likely to identify with radical political movements than with the moderate or centrist ruling parties.As displayed in Figure 2, those with very high confidence in labor unions are the most likely to identify as either a 1 or 2, or far-left, on the political spectrum.Likewise, those with extremely low confidence are the most likely to identify as either a 9 or 10, or the far-right (see Figure 3).This finding suggests that as confidence in labor unions declines over time, total right-wing identity increases over time, supporting the theory that individuals losing confidence in labor unions may instead be placing that confidence in rightwing populist movements.
Based on the figures displayed above and the parameter estimates observed in Table 3a, we can assume there is a relationship between both membership in a labor union and level of confidence in labor unions and political identity.Using a logistic regression means we are testing the probabilities associated with observing certain responses.Thus, our model is: where the coefficients in our regression output tell us how much the logit, or probability of a specific value in the dependent variable, changes based on the values of our independent variables.The sign of the coefficient tells us if each independent variable correlates with a higher or lower dependent variable value.
http://scholarship.claremont.edu/urceu/vol2017/iss1/83b below, we can reject the null hypothesis that a model without our independent variables is as good as a model with the independent variables.However, this doesn't mean we can assume the model is a perfect fit.Observing Table 3c, we can see that the chi-square values are exceptionally large and our significance remains at the .000output.This suggests that our model does not fit the data in question.A model that fits properly should show a small chi-square value and a high significance level.Goodness-of-Fit Output Link function: Logit.
While there are certain transformations of data and variations of models that could provide a better fit for our data, outlined below, the output we can observe suggests that decreasing confidence in labor unions increases the probability of selecting a higher value on the political identity scale.In other words, the lower one's confidence in labor unions, the more likely one is to identify with right-wing political ideologies.Furthermore, having "a great deal" of confidence in labor unions (confidence_4) indicates a higher likelihood of placing oneself on the left-wing of the political spectrum.Not belonging to a labor union also increases the likelihood of a respondent selecting a higher value, or further right position, on the political spectrum.

Limitations
While the data show some support for my hypothesis, there are significant caveats to be considered in our data.First, the analysis does not control for use of the Ghent system.Combining union density statistics in both Ghent and non-Ghent states means our membership numbers are not fully reliable or consistent, and we instead see two distinct sets of membership rates: one with membership in the 80% range (Ghent states), and one http://scholarship.claremont.edu/urceu/vol2017/iss1/8 with membership clustered around 13%. Moving the analysis from the full dataset to a much smaller scale, such as a state-level analysis, would help us to gain a better idea of how different industrial relationships affect our hypothesis and would better control for the variations in the roles of labor unions, the unique flavors of each nationalist or radical right-wing movement, and differing population size, resources, economies, etc.
We must also consider the model fit.Recall that the output shown in Table 3c suggested a straightforward ordinal regression was not the best fit model.To correct this, future research should attempt a multivariate analysis to control for time.The dataset used here is longitudinal, covering four distinct EVS waves from 1981 to 2008.Using a generalized linear model with fixed effects could solve this problem and provide a much more reliable analysis of the relationships in question, including an analysis of these rates of change.The EVS 2017 wave will be introduced to the field in the fall of 2017, and will provide even more relevant information.Each of the trends I have observed has also increased its rate of change substantially in recent years, so adding the EVS2017 data to the mix will provide a greater range of data in which to observe these trends.
Furthermore, if we are trying to show support for the hypothesis that these two trends share similar underlying causes, we should test those variables as well.The theoretical causes discussed in my literature review section are the convergence of mainstream political movements, globalization manifested in increased trade and migration, and the growing distrust of the political and economic elite.To test the degree to which these causes may explain variation in both trends, we could run a similar analysis using variables which quantify these changes.For example, using attitudes toward trade and immigration, we could create a "protectionist index" which could be used to determine if those holding more protectionist ideologies are more or less likely to distrust labor unions.We could use the same index to evaluate political identity.We could then assume that similar coefficients and significance suggests the hypothetical "protectionist index" is indeed a factor causing both trends and therefore the basis of the existing relationship for which we found evidence.We could also use such measurements as occurrence of foreign direct investments or multilateral trade agreements as measures of globalization.The same process could be followed using indexes representative of all three theoretical causes to better understand their roles in these trends.

Conclusion
Ultimately, more research is needed to confirm my findings, but this preliminary analysis shows a worrying trend.As union density continues to fall, we can expect support for right-wing populist movements to grow.It is possible that falling confidence and membership can be reversed by increasing organization efforts among white collar workers, advocating for greater interaction between government and labor unions, and allowing labor unions greater access to the workplace.By re-engaging the working class, and therefore providing them better political and social representation, labor unions can regain the strength of the labor movement in decades past and may be able to slow the rising distrust in democratic institutions that has closely followed radical right-wing populist movements.The economic concerns of blue collar workers are neither illegitimate nor inexorably tied to the rising intolerances characteristic of radical right-wing populist movements.Rather, it is the failure of liberal institutions like labor unions to adequately represent the economic Trade Union Trade-Offs concerns of the working class that is continuing to encourage these shifts.Correcting these failures doesn't necessitate a move to the right for labor unions.Instead, focusing on job placement for unemployed manufacturers and occupational education for those left behind by the shift to a service economy could bring this disenfranchised population back into the workforce.
The shared liberal values enshrined in democratic international regimes are essential to keeping peace and human security around the world; rising support for radical right wing populist parties across North America and Europe pose an unparalleled domestic threat to those regimes.But these trends are not inexorable.Trade unions and other liberal institutions can-and must-reconnect with and re-engage the working class and provide legitimate political and social representation. http://scholarship.claremont.edu/urceu/vol2017/iss1/8

Figure 2 .
Figure 2. Left-Wing Identity by Union Confidence.Displays changes in the proportion of respondents who identify as a 1 or 2 on the political spectrum, stratified by level of confidence in labor unions.Adapted from EVS (2015): European Values Study Longitudinal Data File1981-2008  (EVS 1981-2008)).GESIS Data Archive, Cologne.ZA4804 Data file Version 3.0.0,doi:10.4232/1.1225

Figure 3 .
Figure 3. Right-Wing Identity by Union Confidence.Displays changes in the proportion of respondents who identify as a 9 or 10 on the political spectrum, stratified by level of confidence in labor unions.Adapted from EVS (2015): European Values Study Longitudinal Data File 1981-2008 (EVS 1981-2008).GESIS Data Archive, Cologne.ZA4804 Data file Version 3.0.0,doi:10.4232/1.12253

Table 3a PLUM
Ordinal Regression