Dualization = Polarization? Policy Preferences of Labor Market Outsiders

Transcription

1 Dualization = Polarization? Policy Preferences of Labor Market Outsiders Svenja Hense, WWU Münster (svenja.hense@uni-muenster.de) Paper prepared for the DVPW Kongress 2018, Universität Frankfurt (Main), Work in progress. Please do not cite or circulate. This paper for the first time examines policy preferences of labor market insiders and outsiders in various policy fields and over 25 years. The findings show a consistent gap in preferences between those who are at high risk and those who belong to more secure occupational groups. The mean differences are greatest in the fields of European integration, economy and finance, and social policies, hence exceeding issues that are directly related to labor market risks. Comparing differences over time shows that divergence on cultural issues grows while economic questions became less disputed in recent years. This lends empirical evidence to assumptions in the literature on political mobilization and dualization proclaiming that cultural issues become of greater importance and that labor market dualization fuels political controversy and polarization. (1) Introduction As working conditions in Western countries became more unstable and unpredictable from the 1980s onwards (at least for parts of the workforce), social scientists began to wonder if this growing insecurity causes changes in people s preferences and political behavior. From the perspective of dualization theory, it is especially labor market outsiders as they are the ones to be unhappy with dualization that may change their preferences and demand more redistribution or access to secure jobs from the government. Various empirical studies have shown that this is indeed the case: People who are at economic risk demand more government intervention than those who are more secure. However, there is reason to believe that differences in policy preferences between insiders and outsiders exceed the field of labor market policies. One reason for this belief is the growing importance of cultural issues in determining election outcomes and shaping public debates. It may be that this development just coincides with the rise of dualization but is no consequence of it. Another possibility is that people experience disadvantage in the working sphere (be it personally or in their social environment) and that their disappointment manifests itself in rejecting the liberal society and its values. If this was the case, outsiders should not only opt for more redistribution or measures that help them to become more secure but also against 1

2 European integration, immigration or other policies that have shaped politics and public discourse in recent years. To date, evidence on outsiders preferences remains partial. Studies usually focus on one or very few items at one certain point in time. What is more, most restrict themselves to issues of social security or labor market policies. The contribution of this paper is to extend existing evidence on how labor market risk shapes preferences to policy issues other than those closely related to labor and social security. Using a database containing more than 1,200 survey questions asked in Germany between 1985 and 2016, I contrast positions on a variety of issues for groups of low and high risk. The findings corroborate the expectations drawn from the literature: Outsiders opt for more state intervention and are more particularistic than insiders are. A comparison of three decades shows that preference differences on cultural issues have grown while differences on economic issues were greatest in the late 1990s and early 2000s. The mean differences over all questions exceed 10 percentage points, making labor market risk a predictor comparable to education in its explanatory power. There is indeed reason to believe that dualization is related to a stronger polarization of the political discourse. The paper proceeds as follows: The next section reviews the existing literature on the connection of dualization and policy preferences. Section three introduces the database and methods, followed by the results in section four. I first present findings on single issues that were repeatedly surveyed and, in a second step, look at differences in the aggregate. Section five concludes. (2) Literature Review Recent years have seen a growth of literature on how changes on the labor market may lead to a new political conflict between workers at the core and those at the margins (cf. Schwander Forthcoming). One of the authors expecting a mobilization of outsiders is Guy Standing (2011). So far, only few movements like the ones he has in mind have emerged. If they showed up at all, it has mainly been in Southern countries and it is questionable if they will persist. Kriesi et al. (2008) see a new class cleavage emerging, based not on the antagonism of labor and capital anymore but on winners and losers of globalization. At the same time, they expect economic issues to become of less importance and the influence of cultural issues on people s vote choice to grow. The authors obviously deduce the description as winner or loser from people s position on the labor market but it remains unclear if this concept is congruent with insiders and outsiders in the literature on dualization. 2

3 Others are more skeptical. Esping-Andersen (1999), for instance, rightly points to the fact that many people move in and out of disadvantageous labor market positions and that the latter are therefore unlikely to exacerbate a lasting influence on people s political preferences and behavior. The critique of the insider/outsider divide being too fluent can be overcome by taking factors into account that have a long-term influence. Scholars studying the amount and consequences of dualization have more and more turned away from simply differentiating between the unemployed and those who have a job (as done e.g. by Rueda 2007; Emmenegger 2009). Instead, occupational risks to experience unemployment, atypical employment or automation are used more frequently (Rehm 2009; Schwander and Häusermann 2013; Häusermann, Kurer, and Schwander 2015; Thewissen and Rueda Forthcoming; Helgason and Mérola 2017; Kevins Forthcoming). As few people change their occupation during their work life, these occupation group-based measures should be able to better catch risk-exposure in a long-term perspective. This in turn is more likely than short-term changes in their labor market situation to shape people s preferences and values. Studies on electoral behavior have also shown sociotropic evaluations to be more important than egotropic evaluations (Mughan and Lacy 2002; Braun and Tausenpfund 2018). Hence, I argue that people s preferences are not only influenced by their personal experiences but also by observing their social surroundings. As the workplace is a Site of Political Preference Formation (Kitschelt and Rehm 2014), experiencing a lot of fluctuation and insecurity among colleagues is likely to alter the things voters want from politics. Previous empirical studies indeed show a correlation of occupation-based risks with policy preferences. Even though, overall, support for welfare state spending erodes (Ebbinghaus and Naumann 2018), Häusermann, Kurer, and Schwander (2015) show that being an outsider correlates positively with demand for welfare spending, especially among high-skilled outsiders and for policies that enhance redistribution and activation. As high-skilled people are more likely to bring forward their political preferences (Brady, H. E., Verba, and Schlozman 1995), Häusermann and her co-authors see the potential of a cross-class coalition where high-skilled outsiders act in the interest of low-skilled outsiders (see also Thewissen and Rueda Forthcoming; Rehm, Hacker, and Schlesinger 2012). The findings on welfare preferences are backed by other studies: People facing income risks frequently opt for more redistribution (Rehm 2009; Rehm, Hacker, and Schlesinger 2012; Emmenegger, Marx, and Schraff 2015; Thewissen and Rueda Forthcoming) or governmental assistance (Burgoon and Dekker 2010). Living together with a partner who is an insider mitigates the effects of outsiderness on demands for redistribution, but only for women (Häusermann, Kurer, and Schwander 2016; 3

4 Marx and Picot 2013). In addition, most outsiders do not live in households with insiders (Marx and Picot 2013), so it is still plausible to expect to see differences between insiders and outsiders in the aggregate. Surely, welfare state preferences are not the only policy field that shapes voting behavior. Cultural issues become of ever greater importance, increasing the vote share of both green and radical-right parties (Häusermann and Kriesi 2015). However, cultural preferences of outsiders have so far attracted little scholarly attention. Previous studies mostly focus on classes and occupational groups, respectively, instead of risk exposure or vulnerability as independent variable (Häusermann 2010, Ch. 4; Dolezal and Hutter 2012; Oesch 2013; Rydgren 2013). Nevertheless, existing studies provide some hints where to expect outsiders preferences in a two-dimensional policy space on the cultural axis. 1 It has been found that non-voters hold more conservative (or particularistic) preferences on the cultural axis (Dolezal and Hutter 2012) and that immigration and EU-integration are opposed by low-skilled people (Dancygier and Walter 2015). The latter finding is backed by Park and Kim (Forthcoming) who study pro-brexit voting. For the case of Germany, Lux (Forthcoming) shows that so called modernization losers (lowskilled people, workers, and the unemployed, respectively ) are more likely to state a voting intention for the AfD. According to the (somewhat dated) findings by van der Brug, Fennema, and Tillie (2000), we can assume that affinity to the AfD is based on ideology rather than on the intention to cast a protest vote. As the party takes clearly particularistic positions, its voters and hence Lux losers of modernization - can be located at this end of the cultural axis, too. Oesch's (2013) findings point in a similar direction: While the working class is located at the particularistic end of the cultural axis, socio-cultural professionals occupy its other end. Bringing these findings together, outsiders, defined as people facing high occupational risks, should favor state intervention on the economic axis and particularism on the cultural axis. As dualization and inequality have risen over time in Germany both in numbers and in salience (Palier and Thelen 2010; Palier 2012; Thelen 2014; Brady, D. and Biegert 2017; but see also Fevre 2007), we should see a growing divergence or polarization between opinions of insiders and outsiders over time. 1 There are various ways in the literature to name the poles of the two axes but all concepts agree on a horizontal axis symbolizing the old conflict of (re-)distribution and a vertical axis representing cultural issues. Here, I draw on Häusermann and Kriesi (2015) and designate the axes as state-market and universalism-particularism, respectively. 4

5 (3) Data and Methods To access information on policy preferences, I rely on an original dataset from another research project. 2 It includes information on public opinion and respective political decisions for more than 1,600 policy proposals. The questions were selected from two German representative surveys polled for different media: Politbarometer and DeutschlandTrend. The former covers the period from 1980 to 2016 and the latter from 1998 to Here, I restrict myself to survey questions from 1985 to The reason is twofold: First, there are very few questions from the early years that are structured the way I need them to be. Second, I merge information from the German Socio-Economic Panel (SOEP) to the dataset and the former is only available since Questions in the database usually deal with political decisions that were high on the political agenda at the time or that are of general public interest and ask about the respondents agreement with a specific policy proposal. Issues range from the minimum wage or cuts in social insurance benefits to proposed changes in abortion rights or same-sex marriage. We divided them into seven policy fields. 3 Additionally, we included information on whether the proposed policy change is located on the economic or the cultural axis and its direction (more state/more market; universalistic/particularistic). 4 I only kept questions that had two answer options (yes/no; agree/do not agree; prefer policy 1/ prefer policy 2), leaving me with an N of 1,211. For each question in the dataset, I calculated the degree of support within groups of different risk levels. My operationalization of occupational risk level draws on the proceeding of Rehm (2011), Rehm, Hacker, and Schlesinger (2012) and Schwander and Häusermann (2013). However, due to limitations of the ResPOG data, my occupational groups are somewhat cruder. Neither the Politbarometer nor the DeutschlandTrend contains detailed information on respondents occupation (e.g. measured on an ISCO scale) but just differentiates between unskilled blue-collar workers, skilled workers, low- and high-skilled employees, civil servants and the self-employed. I exclude self-employed respondents from my analyses because the risks they face differ severely from those of other employees. They self-employed do not depend on one single employer, so being laid-off does not have the same severe implications for them. As Schwander and Häusermann (2013) convincingly argue that dualization unfolds differently for 2 The database is called Responsiveness and Public Opinion in Germany (ResPOG) and was set up together with Lea Elsässer and Armin Schäfer as part of a project analyzing responsiveness towards different income and occupation groups (see Elsässer, Hense, and Schäfer 2018). 3 These are: 1. International Politics, 2. Economy & Finance, 3. Work & Social Policies, 4. Environmental Policies (also containing questions on energy supply), 5. Migration, 6. Societal Policies (e.g. questions of equal rights, a mandatory service to the armed forces etc.), 7. European Integration. 4 The coding always refers to the first answer option. 5

6 male and female, older and younger employees, I further divide occupational groups between gender and age groups (18 to 44 and 45 to 65, respectively). 5 This leaves me with 20 subgroups. Risk is measured as the difference between the unemployment rate of the particular group and the mean unemployment rate over all groups. Information on unemployment was derived from the SOEP that contains detailed information on a variety of labor-related variables for a representative sample of the German population. Unemployment rates were calculated for each year separately to account for macroeconomic changes. I grouped SOEP respondents the same way I did with respondents in the ResPOG data and calculated the difference between the particular group and the intergroup mean. Results were z-transformed to ensure comparability even though the mean and the variance changes from year to year. Finally, the year-specific standard deviation was merged to the ResPOG dataset. For each question, I calculated the predicted probability of preferring a given reform proposal at a mean risk level of unemployment for the particular year as well as for a risk that lies one standard deviation below and one above the mean. The next section first presents the results for a few selected questions that were repeatedly asked over the years. In the next step, the aggregated findings are presented by policy field and by dimensions of policies. (4) Results a. Repeatedly asked questions There are only few questions in my dataset that have been asked numerous times. This is because both Politbarometer and DeutschlandTrend are produced foremost for media and not for scientific purposes. Hence, their questions focus on policy proposals that are highly salient and controversial at the time the question is asked. Once the parliament or government made a decision on the topic or it becomes clear that the topic is off the agenda for now, the surveys will not ask about the issue again. That makes it difficult to track changes in public opinion over time but allows researchers to track which topics were when salient in public discourse and the way they were framed. Even if questions were asked repeatedly, their framing sometimes varies over time, making it difficult to disentangle effects of framing from other effects causing different amounts of support. However, this should not pose a major concern for my endeavor, as I am not so much interested in the absolute amount of support but in the question if support differs between insiders and outsiders. To see if there is a sufficient difference between the two and if it 5 The cut-off points were chosen due to the age categories in the original data. 6

7 exceeds policies that are immediately related to their different labor market position, I picked four questions that have been asked repeatedly and that cover both the economic and the cultural dimension of policies. Figure 1 Agreement to cut unemployment benefits SD +1SD Source: own calculations. Figure 2 Agreement to wealth tax Source: own calculations. -1SD +1SD Figures 1 and 2 show the amount of support for two redistributional policies. The question if respondents agreed to cutting unemployment benefits was asked multiple times in a rather short period. This is due to the Hartz reforms that were developed and debated in the early 2000s. Changes in social policies were highly salient at that time but quickly lost salience after the parliament had agreed to implement them. The rise and fall of support rates in figure 1 tells an interesting story about the effects of framing in surveys: The proposal received most support when the survey question emphasized that 7

8 cutting the benefits will save around 3 Billion Euros or that it will incentivize the long-term unemployed to get a job. Support was much lower when respondents were simply asked if unemployment benefits should be cut. However, throughout all questions, respondents belonging to a group with higher risk of becoming unemployed were less in favor of cutting benefits than their more secure counterparts were. On average, the difference was 10 percentage points. This was not the case for all questions touching on redistributional policies, as figure 2 shows. In 1996, the German government decided to suspend a special tax for the wealthy. When the decision went into effect, the Politbarometer asked its respondents if they agreed to that decision. During the next years, the question re-appeared again, asking if the tax should be turned into effect again (as frequently proposed by Members of Parliament). Whenever the question was asked, a majority of both insiders and outsiders supported taxing the wealth with no group being clearly more in favor than the other. Mean support rates deviate less than one percentage point. So far, the findings are in line with previous research: Insiders and outsiders show different preferences concerning social policies, especially if they are closely related to their labor market status. One possible explanation for this finding is pocketbook reasoning: Outsiders face a higher risk of becoming unemployed; hence they are less in favor of cutting unemployment benefits. Insiders, on the other hand, face a low risk of having to rely on benefits, hence they prefer saving tax money over securing an income for the unemployed. The two groups also hold different positions on questions of European integration, even though, again, the picture is mixed. As figure 3 shows, people in secure positions are more in favor of giving Turkey the chance to become a member of the EU one day. This question only received support from a minority of respondents in most years and support dropped sharply in 2016, with only every 10 th respondent still agreeing. The mean support rate over all eight times the question was asked was 39.6% for insiders and 33.9% for outsiders. Even though this is a noticeable difference, it shows that both groups agree that Turkey should not become an EU country. 8

9 Figure 3 Agreement that Turkey should become an EU country one day SD +1SD Source: own calculations. Figure 4 Agreement that introduction of the Euro will be (until 1998)/ was (since 1999) a good thing SD +1SD Source: own calculations. This agreement changes if we turn to another question related to European integration that was asked repeatedly: The question if respondents think that introducing a common currency, the Euro, was a good thing. Figure 4 plots the support rates from 1991 to As the Euro was only introduced (as a virtual currency) in 1999, the question asked until 1998 was if people think it is a good thing that the D-Mark will be replaced by the Euro (or ECU, as it was initially called) in the future. Support among both groups was the highest right after the introduction of the currency and again in the years after the Euro crisis when it turned out that Germany was one of the few European countries that went smoothly through the crisis and experienced a constant economic growth afterwards. However, while on average 54.4% of insiders believe that the Euro is a good thing, this belief is only shared by 30.4% of outsiders. 9

10 This section yields three preliminary findings: First, insiders and outsiders diverged on the willingness to cut unemployment benefits, corroborating previous findings that different labor market situations correlate with different preferences for redistribution. Second, the question on taxing the wealthy showed that this is not necessarily the case for all questions related to redistribution. Insiders do not always support market liberal measures. Hence, there may be policies that benefit outsiders that find also support from insiders (minimum wages being one example). Third, preference gaps between the two groups exceed economic questions. The question evaluating the introduction of the common European currency was the only one of the four questions where both groups stood on different sides of the 50% cut-off point. This gives us reason to believe that Kriesi's (2008) assumption of a growing importance of cultural issues may hold for insiders and outsiders as well. These findings are very tentative, of course. However, they show that the level of risk a person experiences in her social environment comes with different policy preferences. The next section thus includes all questions and looks at the size of preference gaps on different dimensions more systematically. b. Findings from aggregated questions The previous section showed that there are significant differences to what degree insiders and outsiders support certain policies. To get an idea if these were just coincidental findings, this section takes larger numbers of survey questions into account. Figure 5 plots the mean differences in support rates between low and high-risk groups by policy field. 10

11 Figure 5 Mean Differences between low-risk and high-risk groups across different policy fields Societal Policies Environmental Policies International Politics Migration Work & Social Policies Economy & Finance European Integration Differences in percentage points Source: own calculations. It shows that preferences of the two groups diverge more than 10 percentage points in all policy field but societal policies. Differences are greatest in the field of European integration, followed by economy and finance and work and social policies. Studies by other authors examining the nexus of dualization and policy preferences focused on social policies and as figure 5 shows, this is indeed a field that is contested between insiders and outsiders. But it also shows that questions of migration are hardly less contested and that European integration is even more. Differences between insiders and outsiders thus seem to not be restricted to economic questions. The figure also shows that my operationalization of insiders and outsiders works as it is able to capture differences between these groups. The mean differences are comparable in size to differences between respondents with low and high levels of education (cf. Elsässer, Hense, and Schäfer 2017). In fact, labor market risk and education level are highly correlated but not exchangeable (Häusermann, Kurer, and Schwander 2015). What is more, the insideroutsider measurement adds information a variable merely measuring education would not contain. Hence, the differences we see here should not be simply a matter of low- vs. higheducated people. Rather, it shows that risk exposure marks a significant difference that becomes visible in diverging policy preferences. There are several reasons to believe that political polarization in Germany has grown over the last decades. Scholars of dualization theory put their focus on growing inequality through the decline of trade union power, the transformation from an industry to a service economy, and 11

12 political decisions that protected insiders but allowed the numbers of outsiders to grow. In Germany, the latter development is at least in the public s eye - closely connected to the Hartz reforms. In addition, election outcomes in recent years showed that the old cleavage of labor vs. capital lost its predictive power. The descent of Social Democracy is the most evident consequence but many conservative or Christian Democratic catch-all parties also lost their influential position in European party systems. At the same time, the presidential elections in Austria and France in 2017, where candidates of the (former) great parties did not survive the first round, fueled the belief that issues of the second axis become more important. The fact that the Grand Coalition in Germany is not so grand anymore due to the success of the AfD might be another example. If proponents of a cultural cleavage are right, we should be able to see a growing polarization on the cultural axis over time. Figure 6 differs from the last figure as it is not restricted to differences between groups with exceptionally low and high levels of risk but displays how the two groups are located on the two axes, i.e. economic and cultural policies. It contains all questions from the database that were coded as belonging to one of the two dimensions and which pointed in a clear direction on the axes. The figure shows that insiders are more liberal on both dimensions while outsiders prefer more interventional and more particularistic politics. Figure 6 Mean positions of insiders and outsiders on two political axes.6.55 Cultural Axis Risk: -1 SD Risk: +1 SD Economic Axis Source: own calculations. 12

13 Figure 7 Mean positions of insiders and outsiders on two political axes in three periods Source: own calculations. Figure 7 shows the different dimensions over time. There were very few questions from the early years available, so the first graph covers twelve survey years while the other two only cover ten years. Comparing the three periods shows that the two groups changed their positions over time. In the late Eighties and early Nineties, they only differed in the extent to which they wanted the state to intervene in the market but hardly on their cultural positions. As one would expect, people with a risk one SD above the mean were stronger supporters of state intervention. This, again, is in line with previous findings on redistribution preferences. Both groups moved to the left during the next decade with the size of the gap between them slightly growing. Even though both groups preferred left positions of state intervention over market-friendly measures, those at high risk did so clearly more than respondents who were more secure did. Differences on cultural issues grew, too, with those at low risk being in favor of universalistic policies while those at high risk became more particularistic. The last decade shows smaller differences on economic issues but a greater differentiation on the cultural axis. However, one should keep in mind that the changes we see here might be an effect of salience. During the social reforms of the 2000s, many questions from our database 13

14 revolved around issues located on the economic axis. Higher salience of an issue might also lead to higher polarization among the respondents. On contrast, questions from recent years mostly dealt with immigration or European integration. Thus, these kinds of issues may have been of higher salience and hence shaping the picture. However, setting the cut-off points of the periods differently does not change the picture significantly. Respondents with low levels of risk moved to the left and became more universalistic over time. Respondents that are more vulnerable kept their leftist orientation and became slightly more particularistic over time. Putting these findings in a greater perspective, figure 7 corroborates Kriesi's (2008) assumption that cultural issues become more important compared to economic issues. 5) Conclusions This paper s aim was to extend existing evidence on policy preferences of labor market insiders and outsiders. This was done both on the dimensions of issues and time. Outsiderness was measured as occupational risk of unemployment, further divided by age and gender. I analyzed support rates of low- and high-risk groups for 1,211 policy questions that were asked in German surveys between 1985 and Results showed that persistent preference gaps exist between the two groups, comparable in size to the differences between low and highly educated respondents in the same surveys. Additionally, the findings showed that differences are not restricted to policies that are closely connected to outsiders disadvantageous position on the labor market. Rather, European integration turned out to be the policy field where mean differences are greatest. For example, while a majority of insiders think that the introduction of the Euro was a good thing this thought is shared by less than a third of outsiders. Locating policy preferences in a two-dimensional space affirmed the hypothesis derived from the literature that outsiders are stronger in favor of redistribution and governmental interventions and that they prefer particularistic over universalistic policies. Breaking down these findings to three decades showed a growing divergence of insiders and outsiders on cultural questions. Positions on economic issues differed the most from the mid-1990s until the mid-2000s but approached each other again since However, two caveats should be mentioned. First, the differences on policy dimensions may partly appear due to the nature of the underlying data. The two surveys analyzed here focus on highly salient and highly controversial topics. Polarization might thus appear greater than it actually is over all possible policies. What is more, certain issues are surveyed very frequently in some years but loose salience quickly afterwards and are never asked again. Economic issues, especially those concerning social security, unemployment benefits etc., were of high salience in 14

15 the second period analyzed in figure 7 and at that time, many questions were asked that dealt with the proposed welfare retrenchment of the Hartz reforms. In contrast, such questions were hardly asked in 2015 and 2016 when, due to the economic boom and very high employment rates, cutting back the welfare state was not an issue. Instead, questions revolve around the consequences of the high numbers of refugees and other sorts of immigrants coming to Germany in these years. The polarization we saw in figure 7 might thus partly be a spurious finding due to the varying numbers of questions asked on both axes in different periods. Another shortcoming of the data is the broad definition of occupational groups. The variable provided by the Politbarometer and DeutschlandTrend comprises very heterogeneous occupations and socio-economic backgrounds. A more fine-grained measure such as the ISCOscale would be desirable as a robustness check and to ensure comparability with other studies using occupational risk as an independent variable. On the other hand, the fact that we saw persistent differences between the two groups despite the data s shortcomings brings more evidence to the assumption that the consequences of labor market vulnerability run deeper than making people simply demanding more active labor market policies or governmental support for the unemployed. Rather, it seems that the insideroutsider divide has the potential to become a permanent conflict and to increase political polarization. 15

16 References Brady, David, and Thomas Biegert The Rise of Precarious Employment in Germany. SOEPpapers on Multidisciplinary Panel Data Research 936. Brady, Henry E., Sidney Verba, and Kay Lehman Schlozman Beyond SES: A Resource Model of Political Participation. The American Political Science Review 89 (2): Braun, Daniela, and Markus Tausenpfund Europawahlen 2014: Die Wahl euroskeptischer Parteien im Schatten der Wirtschaftskrise. In Parteien und die Politisierung der Europäischen Union, edited by Lisa H. Anders, Henrik Scheller, and Thomas Tuntschew, Wiesbaden: Springer VS. Burgoon, Brian, and Fabian Dekker Flexible employment, economic insecurity and social policy preferences in Europe. Journal of European Social Policy 20 (2): Dancygier, Rafaela, and Stefanie Walter Globalization, Labor Market Risks, and Class Cleavages. In The Politics of Advanced Capitalism, edited by Pablo Beramendi, Silja Häusermann, Herbert Kitschelt, and Hans-Peter Kriesi, Cambridge: Cambridge University Press. Dolezal, Martin, and Swen Hutter Participation and party choice: comparing the demand side of the new cleavages across arenas. In Political Conflict in Western Europe, edited by Hans-Peter Kriesi, Edgar Grande, Martin Dolezal, Marc Helbling, Dominic Höglinger, Swen Hutter, and Bruno Wüest, Cambridge u.a. Cambridge University Press. Ebbinghaus, Bernhard, and Elias Naumann The Popularity of Pension and Unemployment Policies Revisited: The Erosion of Public Support in Britain and Germany. In Welfare State Reforms Seen from Below. Comparing Public Attitudes and Organized Interests in Britain and Germany, edited by Bernhard Ebbinghaus and Elias Naumann, Basingstoke: Palgrave Macmillan. Elsässer, Lea, Svenja Hense, and Armin Schäfer 'Dem Deutschen Volke'? Die ungleiche Responsivität des Bundestags. Zeitschrift für Politikwissenschaft (ZPol) 27 (2): Elsässer, Lea, Svenja Hense, and Armin Schäfer Government of the People, by the Elite, for the Rich: Unequal Responsiveness in an Unlikely Case. MPIfG Discussion Paper 18/5. Emmenegger, Patrick Barriers to entry: insider/ outsider politics and the political determinants of job security regulations. Journal of European Social Policy 19 (2): Emmenegger, Patrick, Paul Marx, and Dominik Schraff Labour market disadvantage, political orientations and voting: how adverse labour market experiences translate into electoral behaviour. Socio-Economic Review 13 (2): Esping-Andersen, Gösta Politics Without Class? Postindustrial Cleavages in Europe and America. In Continuity and Change in Contemporary Capitalism, edited by Herbert Kitschelt, Peter Lange, Gary Marks, and John D. Stephens, Cambridge: Cambridge University Press. Fevre, Ralph Employment insecurity and social theory: the power of nightmares. Work, employment and society 21 (3): Häusermann, Silja The Politics of Welfare State Reform in Continental Europe: Modernization in Hard Times. Cambridge: Cambridge University Press. 16

17 Häusermann, Silja, and Hans-Peter Kriesi What Do Voters Want? Dimensions and Configurations in Individual-Level-Preferences and Party Choice. In The Politics of Advanced Capitalism, edited by Pablo Beramendi, Silja Häusermann, Herbert Kitschelt, and Hans-Peter Kriesi, Cambridge: Cambridge University Press. Häusermann, Silja, Thomas Kurer, and Hanna Schwander High-skilled outsiders? Labor market vulnerability, education and welfare state preferences. Socio-Economic Review 13 (2): Häusermann, Silja, Thomas Kurer, and Hanna Schwander Sharing the Risk? Households, Labor Market Vulnerability, and Social Policy Preferences in Western Europe. The Journal of Politics 78 (4): Helgason, Agnar Freyr, and Vittorio Mérola Employment Insecurity, Incumbent Partisanship, and Voting Behavior in Comparative Perspective. Comparative Political Studies 50 (11): Kevins, Anthony. Dualized trust: risk, social trust and the welfare state. Socio-Economic Review (forthcoming). Kitschelt, Herbert, and Philipp Rehm Occupations as a Site of Political Preference Formation. Comparative Political Studies 47 (12): Kriesi, Hans-Peter Contexts of Party Mobilization. In West European Politics in the Age of Globalization, edited by Hans-Peter Kriesi, Edgar Grande, Romain Lachat, Martin Dolezal, Simon Bornschier, and Timotheos Frey, Cambridge: Cambridge University Press. Kriesi, Hans-Peter, Edgar Grande, Romain Lachat, Martin Dolezal, Simon Bornschier, and Timotheos Frey Globalization and its impact on national spaces of competition. In West European Politics in the Age of Globalization, edited by Hans-Peter Kriesi, Edgar Grande, Romain Lachat, Martin Dolezal, Simon Bornschier, and Timotheos Frey, Cambridge: Cambridge University Press. Lux, Thomas. Die AfD und die unteren Statuslagen. Eine Forschungsnotiz zu Holger Lengfelds Studie Die Alternative für Deutschland : eine Partei für Modernisierungsverlierer? Kölner Zeitschrift für Soziologie und Sozialpsychologie (forthcoming). Marx, Paul, and Georg Picot The party preferences of atypical workers in Germany. Journal of European Social Policy 23 (2): Mughan, Anthony, and Dean Lacy Economic Performance, Job Insecurity and Electoral Choice. British Journal of Political Science 32 (3): Oesch, Daniel The class basis of the cleavage between the New Left and the radical right. An analysis for Austria, Denmark, Norway, and Switzerland. In Class Politics and the Radical Right, edited by Jens Rydgren, London / New York: Routledge. Palier, Bruno Turning Vice Into Vice: How Bismarckian Welfare States have Gone from Unsustainability to Dualization. In The Politics of the New Welfare State, edited by Giuliano Bonoli and David Natali, Oxford: Oxford University Press. Palier, Bruno, and Kathleen Thelen Institutionalizing Dualism: Complementarities and Change in France and Germany. Politics & Society 38 (1): Park, Youngduek, and Yongmin Kim. Explaining the Brexit Referendum: The Role of Worker Skill Level in Voter Decisions. The Political Quarterly (forthcoming). 17

18 Rehm, Philipp Risks and Redistribution. An Individual-Level Analysis. Comparative Political Studies 42 (7): Rehm, Philipp Risk Inequality and the Polarized American Electorate. British Journal of Political Science 41 (2): Rehm, Philipp, Jacob S. Hacker, and Mark Schlesinger Insecure Alliances: Risk, Inequality, and Support for the Welfare State. American Political Science Review 106 (2): Rueda, David Social Democracy Inside Out: Partisanship and Labor Market Policy in Industrialized Democracies. Oxford: Oxford University Press. Rydgren, Jens Introduction: Class politics and the radical right. In Class Politics and the Radical Right, edited by Jens Rydgren. London / New York: Routledge. Schwander, Hanna. Labor Market Dualization and Insider Outsider Divides: Why This New Conflict Matters. Political Studies Review (forthcoming). Schwander, Hanna, and Silja Häusermann Who is in and who is out? A risk-based conceptualization of insiders and outsiders. Journal of European Social Policy 23 (3): Standing, Guy The precariat: The new dangerous class. London: Bloomsbury Publishing. Thelen, Kathleen Varities of Liberalization and the New Politics of Social Solidarity. New York: Cambridge University Press. Thewissen, Stefan, and David Rueda. Automation and the Welfare State: Technological Change as a Determinant of Redistribution Preferences. Comparative Political Studies (forthcoming). van der Brug, Wouter, Meindert Fennema, and Jean Tillie Anti-immigrant parties in Europe: Ideological or protest vote? European Journal of Political Research 37 (77-102). 18