Disappointed Expectations:Downward Mobility and Electoral

Change

Thomas Kurer† Briitta van Staalduinen∗

August 10, 2020

Abstract

Postindustrial occupational change has brought an end to an era of unprecedentedupward mobility. We consider whether this immense structural shift is relatedto ongoing political disruptions across Western democracies. To do so, we in-troduce the concept of status discordance, which we operationalize as the dif-ference between expectations formed during childhood based on parental statusand outcomes realized in adulthood. Our innovative empirical approach lever-ages German household panel data and out-of-sample predictive modelling toprovide empirical estimates of respondents’ status expectations. The analysis re-veals that political dissatisfaction is widespread among voters who fall short ofintergenerational status expectations. We show that such dissatisfaction reducessupport for mainstream parties and increases abstention and support for radicalparties. We also identify di�erent types of status discordant voters by their edu-cational and occupational background, which we show in�uences whether theychoose to abstain from voting or vote for radical left or right parties. Our �nd-ings highlight how expectations about opportunities underpin generational vot-ing patterns, shedding light on the ongoing breakdown of the postwar politicalconsensus.

†University of Zurich. kurer@ipz.uzh.ch∗Harvard University. vanstaalduinen@g.harvard.edu.

A previous version of this paper has been presented at ECPR 2019 and SVPW 2020. We thank the par-ticipants as well as Tarik Abou-Chadi, Per Engzell, Lukas Ha�ert, Sophie Hill, Arno Van Hootegem,Peter Hall, Alex Mierke-Zatwarnicki, Denise Traber, David Weisstanner, and Delia Zollinger forvery helpful comments.

1

1 Introduction

The rise of populism and decline of mainstream parties across Western democracieshas renewed scholarly interest in the factors driving political dissatisfaction and sup-port for radical parties. A burgeoning literature studies the structural roots of thesedevelopments (Inglehart and Norris 2017; Rodrik 2018; Colantone and Stanig 2018;Burgoon et al. 2018; Ballard-Rosa et al. 2018; Im et al. 2019; Margalit 2019; Anelli et al.2019; Kurer 2020). Recent research in political science emphasizes the role of statusloss, particularly so among white men without college degrees often viewed as ‘leftbehind’ by an emerging global knowledge economy and increasingly disillusioned bythe mainstream parties that promote it. Radical parties have attracted these voters byattributing their status loss to the advancement of other groups such as immigrants,minorities, or women, or to the mismanagement of globalization by an unchecked,political-�nancial elite (Hobolt 2016; Gidron and Hall 2017; Gest et al. 2017; Steenvo-orden and Harteveld 2018; Hooghe and Marks 2018).

We build on these explanations by recasting status loss in terms of occupational changeand intergenerational mobility. Postwar economic expansion spawned an era of un-precedented upward mobility. As Lipset (1959: 11) put it, “in every industrial country,a large proportion of the population have had to �nd occupations considerably dif-ferent from those of their parents", a phenomenon rooted in “a period of prosperitythat followed on the heels of great dislocations" (1959: 33). As a result, occupationalreorientation most often meant occupational upgrading. Leading accounts at the timepresupposed that economies would expand inde�nitely, sustaining upward mobilityfor future generations (e.g. Treiman 1970).

However, the rate of upward mobility began to decrease towards the end of the 20thcentury and continues today (e.g. Chetty et al. 2017; Breen and Müller 2020). Scholarsposit that this is largely due to postindustrial occupational change. Deindustrialization,job polarization, and the recent slowdown in the growth of relative demand for high-skilled workers (see, e.g., Goos et al. 2009; Autor and Dorn 2013; Oesch 2013; Autor et al.2020) have impaired the social escalator, eliminating decent, middle-skill jobs whileat the same time gradually reducing the value of a college degree. Recent empiricalstudies report a striking contrast between the fortunes of those born in the 50s and 60s

2

and those born in the 80s and early 90s, a pattern that holds both in the US (e.g. Chettyet al. 2017; Hout 2018) and across many advanced economies in Western Europe (e.g.Chauvel and Schröder 2014; Hertel 2017; Bukodi et al. 2020).

We contend that a rise in disappointed expectations has followed this large-scale struc-tural shift, and that these disappointments shape voter behavior. The upward mobilityof postwar cohorts gave way to a societal and political consensus of progress. Mostpeople could reasonably expect to at least meet, if not exceed, status standards set bytheir parents. We expect that the growing share of citizens who fall short of theseexpectations and who instead experience what we term "status discordance", attributesome of the blame to mainstream parties.

Relying on German household panel data and machine-learning methods, we proposean innovative empirical approach to overcome a key obstacle to studying status dis-cordance: measuring expectations. We start from the assumption that status standardsset by parents have a lasting impact on how individuals understand their own statusachievements. We create out-of-sample predictions for respondents’ expected occupa-tional status based on key socio-economic characteristics of their fathers1 (education,job, social status, citizenship, etc.) and various pre-treatment characteristics of therespondents themselves (gender, age, region in childhood, moving pattern, etc.). Weargue that such a prediction provides a reasonable proxy for the status position respon-dents expect to achieve in adulthood. We then use panel data to calculate the di�erencebetween expected status (prediction) and realized outcome (observed) in adulthood.This di�erence is what we term status discordance. It captures respondents’ statusachievements relative to salient socio-economic reference points in childhood.

1Although contemporary research in social mobility should take much more seriously the past and fu-ture role of mothers (Beller 2009), we follow the majority of existing work and use only fathers’ occu-pation. The main reason is data availability: missing values on occupation are much more prevalentamong mothers than fathers, most likely a result of higher shares of atypical employment or unpaidwork, which are only inadequately captured by standard occupational classi�cations. However, ourresults are robust to the inclusion of mother’s occupational status in the out-of-sample predictions.The correlation between a measure of status discordance predicted on the basis of father’s back-ground only versus father’s and mother’s background is high (p=0.885, see Appendix Figure A1)and we are able to reproduce our key results within the reduced sample of respondents for whomwe have information on mothers’ occupational status. Still, we hope that our analysis serves as astarting point for future analyses, including a consideration of how the e�ects of status discordancedi�er in homes where the mother is the primary breadwinner or in single-mother homes.

3

Our analysis provides strong and consistent evidence that negative status discordanceis systematically related to abstention and support for radical parties. We further showthat the political e�ects of status discordance depend on a given voter’s educational andoccupational background: college-educated, high-skilled voters who experience statusdiscordance relative to highly educated parents are more likely to vote for the radicalleft, while low-skilled workers with high-school degrees from middle- or working-class families are more likely to vote for the radical right. In a �nal step, we con�rmour reasoning with evidence of the attitudinal mechanisms linking status discordanceto either radical left or radical right support.

These �ndings shed light on the forces driving recent electoral outcomes in post in-dustrial societies. We hope to contribute to research examining these outcomes bydrawing attention to the role of expectations in shaping generational voting patterns.Due to occupational change and fading prospects for mobility, growing shares of vot-ers are having to adjust expectations that have been long rooted in postwar projec-tions of progress. Our �ndings suggest that, although for now still outnumbered bythe upwardly mobile, these disappointed voters have and will continue to change theelectoral landscape of advanced democracies.

2 The Politics of Expectations

Interest in the factors driving support for radical parties expanded substantially fol-lowing a string of major radical right victories in the �rst decade of the 21st century.Building on earlier inquiries into why and when radical parties �nd electoral success,many recent contributions point forcefully towards the importance of postindustrialoccupational change. These explanations were particularly popular in the aftermath ofBrexit and the triumph of Donald Trump, events viewed as being propelled by supportfrom low-skilled, high-school educated voters who have been hit especially hard bydeindustrialization, automation, and globalization.

Evidence from an in�uential body of work indeed reveals that certain voters su�eredincome losses and increased unemployment risk following the shift to a postindus-trial economy, which in many cases was overlooked by mainstream political elites

4

(Inglehart and Norris 2017; Owen and Johnston 2017; Rodrik 2018; Eichengreen 2018;Ballard-Rosa et al. 2018; Colantone and Stanig 2018; Burgoon et al. 2018; Anelli et al.2019; Kurer 2020; Broz et al. 2020). Recent contributions to this literature have fo-cused on developing innovative ways to measure status loss. Gidron and Hall (2017)show how the disappearance of manufacturing jobs and the concentration of knowl-edge economies in urban centers encourages support for the radical right among whitemen without a college degree. Evidence from interviews similarly reveals that whitemen often view their loss in status relative to gains in status among women and mi-norities, in�uenced also by the purported replacement of low-skilled native workerswith immigrant workers (Hochschild 2016; Gest 2016).

While this literature provides an important point of departure, we argue that exist-ing work has neglected long-term changes in economic opportunity in general, andintergenerational occupational mobility in particular. Similar to existing research, weaim to capture status loss but propose to do so with a focus on disappointed expec-tations resulting from declining occupational mobility across generations. We arguethat occupational standards set by parents are a particularly salient reference point forunderstanding status (Cohen 1987; Hill and Duncan 1987). As the �rst example of whatwork is and why people work, parental occupations orient individual understandingsof status. In some cases, parental occupations serve as a standard that children can rea-sonably expect to meet, and in other cases as a standard that children should expect tosurpass. While this can vary, especially across class, race, and gender (Kao and Tienda1998; Lareau 2011; Polavieja and Platt 2014), parental occupations establish enduringassociations between work, income, and standard of living. Children use their parentsas a reference point when forming ideas about the �nancial and social prestige attachedto certain occupations, ideas that later play an important role in career choices, even-tual earnings, and evaluations about their own achievements. These understandingsmay change in response to new experiences or expanded awareness about meritocracyor equality, but standards set by parents nonetheless carry lifelong weight, anchoringindividual status expectations (Johnson 2002; Dustmann 2004; Checchi 2006).

Evidence from a longstanding literature on relative deprivation shows that discrepan-cies between expectations and reality have political consequences (e.g. Geschwender1964; Gurr 1970; Kurer et al. 2018; Burgoon et al. 2018; Mitrea et al. 2020). Another

5

stream of research has long attempted to determine whether socially mobile votersmaintain the political preferences of the class into which they were born, or adoptthose of the class into which they move later in life (Lipset 1960; Abramson and Books1971; Benabou and Ok 2001; Lahtinen et al. 2017; Ares 2019). Building on these insights,we propose that the discrepancies resulting from declining occupational mobility a�ectpolitical behavior because those experiencing status discordance feel they have beendenied the opportunity to secure the status they had come to expect for themselvesbased on standards set in childhood—expectations also encouraged by the postwar po-litical consensus. Mainstream political parties have seemed to assume that, at least inthe long term, economic advancement would sustain upward mobility to some degree,regardless of political interventions. While such assumptions have suited many well,they have served to alienate the increasing share of downwardly mobile voters.

We therefore expect that status discordant voters are more likely to feel politicallyalienated and dissatis�ed with mainstream politics. Given evidence that abstention isa common reaction to political alienation (Levi and Stoker 2000; Blais 2006), we expectthat negative status discordance goes hand in hand with lower rates of political par-ticipation. At the same time, the growing literature on status loss suggests that otherstatus discordant voters will choose to voice their discontent in the political arena,especially when there are compelling anti-establishment party options. We thus alsoexpect that negative status discordance increases the likelihood of supporting partiesoutside the political mainstream.

Beyond these general patterns, we hypothesize that there are di�erent types of statusdiscordant voters. That is, just as political alternatives to mainstream parties comein di�erent ideological shades, status discordant voters come from a diverge range ofsocio-economic backgrounds, which we argue shapes their vote choice. More speci�-cally, we argue that whether status discordant voters support the radical left or rightdepends on their fathers’ occupation and their own status position. We expect that sta-tus discordant voters at lower positions in the occupational hierarchy are more likelyto vote for radical right parties. These parties tend to hearken back to earlier timesmarked by economies dependent on routine work, clearly demarcated social and cul-tural identities, and traditional workplace hierarchies—times when low-skilled work-ers earned a respectable standard of living and even had some prospects for career

6

mobility (Steenvoorden and Harteveld 2018; Gidron and Hall 2017). For the parents oflow-skilled status-discordant voters, postwar labor markets provided ample economicopportunity. This is no longer the case; prospects for those without college degreeshave clearly worsened. Radical right parties’ promise to undo recent societal changeis thus an e�ective way of attracting low-skilled status discordant voters. Of course,the anti-immigrant and EU-skeptic views of radical right parties are also important,but we consider such positions as part of their broader platform calling for a return toearlier— pre-immigration, pre-EU—times.

In contrast, status discordant voters at higher positions in the occupational hierarchyare more likely to vote for radical left parties. It may be less obvious why high-skilledvoters—who are objectively better o� than low-skilled voters—would feel neglected bymainstream parties in the �rst place. However, while it is true that those with collegedegrees are still less likely to experience downward mobility, it is important to empha-sis that high-skilled voters do experience status downgrades relative to their parents.Consider, for example, someone who grew up in a �nancially-stable home headed bylawyer or doctor—someone who seamlessly attended and graduated from college. Oc-cupational trajectories after college, however, are becoming increasingly complicatedby increasing competition due to the ’massi�cation’ of higher education (Ansell andGingrich 2018), the declining college wage premium (Autor et al. 2020), and the fact thatsome college degrees land graduates in public or arts sectors where the work may beintellectually or socially rewarding but poorly remunerated and with limited prospectsfor career mobility. We hypothesize that these highly educated yet status discordantvoters are more likely to support radical left parties. Radical left parties appeal to high-skilled voters not by calling for a return to earlier times, but instead by calling for aradically di�erent economic system—a system designed to combat the inequality thatpostwar capitalism has over time come to promote, and to economically reward thosewith goals other than pro�t growth, be they environmental, socio-cultural, or human-itarian (Rooduijn et al. 2017). High-skilled status discordant voters, then, see neitherthe current nor the postwar political consensus as su�cient to reform the economy ina way that would recognize and reward their skills and aspirations.

In other words, the distinction between these di�erent types of status discordant votersis whether the economic and political context that shaped the occupational trajectory

7

of their parents would present the kinds of opportunities that they feel are lackingtoday. For low-skilled voters, such a context would indeed be ideal; for high-skilledvoters, such a context would likely be a regression.

3 Declining Occupational Mobility and Status

Discordance

Why do disappointed status expectations result in such remarkable political disrup-tion now but not so much before? While such disappointments have always been afeature of any society, they were less frequent and thus less politically relevant. In theprosperous and highly mobile post-war period, most people experienced status up-grades relative to their parents. In contrast, the trend of declining upward mobility inpost-industrial democracies suggests that a growing number of people are falling shortof expectations based on parental standards. Declining mobility, then, not only mat-ters for politics because it is becoming more pervasive, but also because expectationsare still undergoing adjustment to a postindustrial context in the shadow of postwarprosperity.

Figure 1 provides a graphical contextualization of our argument. The x-axis shows anindividual’s intergenerational reference point, in other words an individual’s expectedoccupational status based on parental background. The y-axis shows the actual occu-pational outcome realized by that same individual. If the realized status perfectly coin-cides with the expected status, this individual observation would lie on the 45-degreeline. If the realized occupational status exceeds (falls below) the intergenerational ref-erence point, we would instead observe positive (negative) status discordance.

Now, any given society consists of a large number of such comparisons of individualexpectations and outcomes, resulting in a distribution of observations as the one sim-ulated in Figure 1. If all observations aligned on the 45-degree line, this would be asociety with zero mobility; parental background would be a perfect predictor of in-dividual status. In reality, there is a fair amount of mobility and the pattern in most

8

postindustrial societies more closely resembles the depicted distribution (dashed redline, see Figure A3 in the Appendix for the empirical distribution in Germany).

The indicated downward shift of the red line illustrates the background condition forour argument, namely the decline in occupational mobility. Almost all those born inthe postwar era managed to exceed expectations based on the economic circumstancesof their parents, which in this �gure would result in a population consisting almost ex-clusively of individuals experiencing positive status discordance, and thus lying abovethe 45-degree line (dotted red line). However, due to declining occupational mobility,the respective shares of positive and negative status discordance are becoming increas-ingly similar. That is, the downward shift of the line directly increases the number ofpeople facing negative status discordance. As such, the pool of disappointed citizensis steadily growing, which we argue helps explain why we see political disruption atthis point in time but not before.

declining absolute

mobility

0

10

20

30

40

50

60

70

80

90

100

0 10 20 30 40 50 60 70 80 90 100

Intergenerational Reference Point (Father)

Rea

lized

Sta

tus

(Chi

ld)

Status Discordance

positive

negative

Figure 1: The increasing importance of status discordance when absolute mobility declines (simu-lated data)

9

Finally, we want to again emphasize that, as the �gure illustrates, status discordancecan occur across the occupational hierarchy, at all status positions. Our focus is strictlyon any deviation from an expectation based on parental background.2 Appreciatingthe reach of occupational change is important to understanding its e�ects. How dis-appointed expectations shape voting behaviour—that is, how disappointed voters ul-timately express their discontent with mainstream parties—depends heavily on thesocio-economic context in which status expectations were formed. As theorized in theprevious section, there are di�erent types of status discordant voters across the occu-pational hierarchy who may well respond di�erently to the experience of disappointedexpectations. We examine this possibility in more detail in the empirical analysis.

4 Status Discordance: Concept and Measure

For the purpose of our analysis, we need an empirical measure that captures devia-tions in realized socio-economic status from expected socio-economic status based onparental background. As we have argued, this intergenerational reference point is crit-ical because it shapes individual status expectations. Two individuals with the sameabsolute socio-economic status are likely to think very di�erently about their achieve-ments depending on the socio-economic context in which they were raised and thestatus expectations this generated.

The empirical challenge thus lies in acquiring an estimate of expected status outcomesas proxied by parental background. We propose an innovative approach to arrive atan empirical estimate of such status expectations. We draw on rich household paneldata from the German Socio Economic Panel (SOEP) and recent advances in machinelearning to create out-of-sample predictions of socio-economic status given parentalbackground. More speci�cally, we use earlier waves of SOEP to train a random forestalgorithm (Breiman 2001) to predict an individual’s occupational status on the basisof an array of variables capturing his or her father’s socio-economic status as well asvarious time-invariant characteristics of the respondent his/herself. Tree-based models

2In fact, due to ceiling e�ects, negative status discordance may just as likely occur among people withfavorable parental backgrounds as people with less fortunate backgrounds.

10

like random forests are a valuable tool for prediction that make few assumptions on thedata-generating process and very �exibly handle non-linearities, interactions betweencovariates, or the inclusion of irrelevant covariates (Montgomery and Olivella 2018).We have evaluated the performance of di�erent learning methods (regression trees,bagging, random forest) on our data set and decided on the basis of the mean squarederror (MSE), i.e. the most commonly-used measure to assess model accuracy (Jameset al. 2013), to proceed with a random forest algorithm.

After training the algorithm, we create individual predictions of occupational statusfor all respondents in the latest available SOEP wave from 2018. This latest release con-tains information on respondents’ voting behavior in the 2017 German federal election,which allows us to study the relationship between mobility perceptions and politicalbehavior.3 The di�erence between the realized occupational status and status expec-tations proxied by predictions based on parental background is our measure of inter-generational status discordance (ISD).

ISDi = yi - yi,

where yi is the realized individual socio-economic status, captured by the Interna-tional Socio-Economic Index (ISEI), a widely used measure of occupational status. Incontrast to categorical indices of occupations, such social class schemes, the ISEI is acontinuous, uni-dimensional measure, created by optimally scaling occupation as anintervening variable between education and income (Ganzeboom et al. 1992). Moreintuitively, the ISEI can be understood as a weighted sum of mean education and meanincome for each occupational group, adjusted for age to account for life-cycle e�ects.A detailed comparison of the available indices of class, occupational prestige and occu-pational status shows that the ISEI accounts well for what drives the intergenerationaloccupational mobility process (Ganzeboom et al. 1992: p. 7).

We estimate status expectations, yi = f (Xj , X i), based on a set of variables of parentalbackground, Xj , and various potentially relevant characteristics the respondent, X i.

3Note that while we exploit the longitudinal nature of the household panel data set covering multiplegenerations for the creation of our key explanatory variable, the main analysis of political repercus-sions is cross-sectional.

11

Xj includes father’s occupational status, father’s education level, father’s age, andfather’s nationality (German/not German). X i is a vector of variables including basicdemographic and childhood circumstances of the respondent him/herself, all of whichcan be considered pre-treatment covariates (age, gender, nationality, region of schooldegree, having moved since childhood, urban/rural childhood location). Figure A2in the appendix displays variable importance plots of the underlying random forestmodel.

Panel A in Figure 2 displays the distribution of the resulting measure. As expected,most respondents achieve a socio-economic status that corresponds closely to predic-tions based on parental background, resulting in most ISD values clustering aroundzero. Deviations from the predictions occur to a similar extent in both directions,meaning that we have a comparable number of negative status discordance (downwardmobility) and positive status discordance (upward mobility) in the total sample. PanelB shows that this distribution is relatively similar between male and female respon-dents, whereas Panel C shows the expected pattern with respect to education: negativestatus discordance is much more common among those without a college degree. Asindicated previously, although highly educated voters do experience downward mobil-ity, completing a college degree increases the chances of higher socio-economic statusand thus the probability to achieve a position in society that matches or exceeds ex-pectations based on parental background. Panel D, �nally, reveals important spatialvariation. Even today, respondents who grew up in East Germany, on average, faceless favorable mobility prospects compared to those who lived in the West in 1989.Both the roots and persistence of this pattern have been described in existing work,perhaps most impressively in Mau’s (2019) recent sociological account on German re-uni�cation from an East German perspective.

12

0.00

0.01

0.02

0.03

-60 -30 0 30

Intergenerational Status Discordance

Positive ← → Negative

De

nsity

(a) Total

0.00

0.01

0.02

0.03

-60 -30 0 30

Intergenerational Status Discordance

Positive ← → Negative

De

nsity

malefemale

(b) By gender

0.00

0.01

0.02

0.03

-60 -30 0 30

Intergenerational Status Discordance

Positive ← → Negative

De

nsity

(c) By education

0.00

0.01

0.02

0.03

-60 -30 0 30

Intergenerational Status Discordance

Positive ← → Negative

De

nsity

WestEast

(d) By region at the time of re-uni�cation(1989)

Figure 2: Distribution of Intergenerational Status Discordance (Original Scale)

For all analyses below, we transform the raw ISD values, i.e. the actual di�erencebetween prediction and status, into a percentile ranking variable in order not to over-interpret small di�erences in the estimated numerical values created in a relativelynoisy procedure. The upward/downward mobility threshold corresponds to percentile53, which makes sense given the relatively symmetric distribution of the raw measurepresented in Figure 2. Note that our results do not depend on this transformation.All results can be recovered using the raw values of status discordance instead of theranking variable (see Appendix Table A6).

13

5 Findings

5.1 Status Discordance and Political Alienation

As a �rst step in our analysis of the political e�ects of status discordance, we showsimple bivariate correlations between the ranked percentile-version of ISD and dif-ferent indicators of political alienation from mainstream parties. Figure 3 shows therelationship between status discordance and (a) abstention and (b) anti-establishmentvoting, or support for non-mainstream parties in the 2017 German federal election.Non-mainstream includes radical parties from both left ("Die Linke") and right ("Alter-native für Deutschland").

0.0

0.1

0.2

0.3

0 25 50 75 100

Intergenerational Status Discordance

Pr(

Abste

ntion)

(a) Abstention

0.0

0.1

0.2

0.3

0 25 50 75 100

Intergenerational Status Discordance

Vote

Pr(

Anti−

Esta

blis

hm

ent)

(b) Anti-Establishment Voting

Figure 3: Intergenerational Status Discordance and Voting Behavior

The Figure reveals strong positive correlations for both indicators of political alien-ation. Those in the lowest percentiles of the ranked ISD variable, i.e. individualswhose socio-economic status is higher than expected based on parental background,have a close-to-zero probability of abstention, and show very low levels of supportfor non-mainstream parties. In contrast, the level of support strongly and steadilyincreases with percentile ranks including higher numbers of citizens who have experi-enced negative status discordance. The positive correlation persists when only looking

14

at support for the radical-right (Alternative für Deutschland), is robust to more �exi-ble non-linear speci�cations, and also holds when using a decile instead of a percentileversion of the explanatory variable (see Appendix Figures A4 - A5).

This clear and consistent pattern, in line with our expectations, demonstrates the facevalidity of our measure. We consider these positive correlations as the �rst piece of evi-dence in support of our core hypothesis that the disappointment of status expectationsincreases the likelihood of dissatisfaction with mainstream politics.

As a next step, we examine the relationship between status discordance and politicalalienation in a more robust multivariate setting. The main explanatory variable re-mains the ISD rank. We include socio-demographic (age, gender) and socio-economic(education, income) characteristics of individuals in the models, as well as a variablecapturing the East/West location of a respondent in 1989. In addition, Bundesland-�xed e�ects take care of regional variation in the contemporary electoral landscapeof Germany. Following our theoretical expectations, the �rst models presented in Ta-ble 1 focus on abstention (column 1), general political alienation, proxied by whetherrespondents feel close to any party in the party system (column 2), and support fornon-mainstream parties in the federal election 2017 (column 3). We supplement the�rst set of models with an analysis of support for mainstream parties, which shouldyield the empirical mirror image of the results on support for non-mainstream parties.We show results from a broad de�nition that includes the by now fairly establishedGreen party (column 4). The coe�cients for ISD thus report the change in the proba-bility to support a given party family (or the probability to identify with a party) fora one-percentile increase in negative status discordance net of current socio-economicstatus. That is, we are not interested in whether more educated or higher-income re-spondents are more likely to support a given party, but rather in the e�ect of statusdiscordance compared to otherwise similar respondents.

The results in Table 1 demonstrate that the strong correlations presented before holdwhen controlling for socio-demographic and socio-economic characteristics of respon-dents. A one standard deviation increase in ISD (sd=28.86) is associated with a 2-3percentage point increase in abstention and/or support for radical left or right parties.The magnitude of the e�ect is notable given relatively low baseline probabilities, par-

15

ticularly when it comes to voting for radical parties (about 14 percent). The e�ect onidenti�cation with political parties goes in the same direction and is even slightly morepronounced. The last column of Table 1 shows the mirror image of this e�ect, i.e. thepronounced negative correlation between negative status discordance and support formainstream parties.4

This �rst set of evidence thus con�rms our hypothesis that the disappointed expecta-tions following declining mobility prospects have political e�ects. Net of other vari-ables that capture absolute socio-economic conditions, downward mobility—at leastwhen conceptualized as intergenerational status discordance—serves as a signi�cantdriver of political alienation and dissatisfaction with mainstream parties.

5.2 Status Discordance and Party Choice

As a next step, we extend our focus beyond non-mainstream voting and examine sup-port for particular parties. In further disaggregating voting behavior, our aim is tobegin exploring the possibility of di�erent types of status discordant voters. Figure4 summarizes results from various models assessing the relationship between statusdiscordance and (a) vote choice and (b) party identi�cation. We include the same setof individual-level controls as in the previous analysis to keep current socio-economicconditions of respondents constant.

Intergenerational status discordance indeed carries considerable explanatory powerwhen it comes to party choice. In line with our expectations, negative status discor-dance is consistently related to higher values of support for radical parties. Radical-right parties in particular attract the downwardly mobile—the coe�cients approxi-mately double in size and are estimated much more precisely in comparison to theradical left. Much in contrast, mainstream parties are not appealing to voters whoexperience negative status discordance. It is worth noting that this pattern is robustwhen looking at former West and Eastern German regions separately.

4The inclusion or exclusion of the Green party in this de�nition hardly changes the result, re�ectingthe relatively established nature of the German Greens, epitomized not least by their participation inthe coalition government between 1998 and 2005, long before the time window under considerationhere.

16

Table 1: Intergenerational Status Discordance and Political Alienation

Abstain No Party ID Vote Radical Vote MainstreamStatus Discordance (Pctiles) 0.074∗∗∗ 0.084∗∗∗ 0.055∗∗∗ −0.128∗∗∗

(0.012) (0.018) (0.014) (0.017)Female (1=yes) −0.858 7.307∗∗∗ −3.258∗∗∗ 4.920∗∗∗

(0.633) (0.984) (0.744) (0.918)Age −0.360∗∗∗ −0.487∗∗∗ 0.051 0.403∗∗∗

(0.033) (0.052) (0.039) (0.048)Educ: Lower Elementary Ref. Ref. Ref. Ref.– Upper Elementary −4.170 5.287 −0.253 4.806

(3.746) (5.743) (4.398) (5.430)– Lower Second. −13.238∗∗∗ −3.227 −4.653 17.712∗∗∗

(3.711) (5.690) (4.357) (5.379)– Secondary −20.804∗∗∗ −11.611∗ −5.889 24.681∗∗∗

(3.842) (5.898) (4.511) (5.568)– University Prep. −23.423∗∗∗ −17.678∗∗ −7.682 28.879∗∗∗

(3.777) (5.797) (4.435) (5.475)– Tertiary I −20.925∗∗∗ −14.010∗ −4.673 24.856∗∗∗

(3.794) (5.826) (4.455) (5.500)– Tertiary II −21.378∗∗∗ −20.259∗∗∗ −7.008 28.521∗∗∗

(3.758) (5.752) (4.413) (5.447)Income (log) −2.235∗∗∗ 0.088 −1.610∗∗ 4.450∗∗∗

(0.458) (0.710) (0.537) (0.663)In 1989: East Ref. Ref. Ref. Ref.– West −2.914∗∗ −10.327∗∗∗ −9.899∗∗∗ 13.555∗∗∗

(1.067) (1.656) (1.253) (1.547)– Abroad 15.413∗∗∗ 15.476∗∗∗ 0.047 −12.021∗∗∗

(1.901) (2.676) (2.232) (2.755)– Born after 1989 −1.135 −1.743 −5.144∗∗ 6.020∗∗

(1.469) (2.286) (1.724) (2.129)Intercept 58.444∗∗∗ 77.448∗∗∗ 32.567∗∗∗ −2.075

(5.604) (8.681) (6.581) (8.124)R2 0.106 0.095 0.060 0.142Adj. R2 0.104 0.093 0.058 0.140Num. obs. 10758 12020 10758 10758∗∗∗p < 0.001; ∗∗p < 0.01; ∗p < 0.05. All models include regional (Bundesland) �xed e�ects.

17

While there is no clear relationship with mainstream left parties, the association withmainstream right parties is clearly negative. This means that support for CDU, CSU andFDP is signi�cantly higher among voters who experience positive status discordance.Support for the mainstream right among those on the winning side of mobility trendsaccords with existing �ndings and prevailing theoretical expectations (Abramson andBooks 1971; Benabou and Ok 2001; Alesina et al. 2018).

Figure 4: Intergenerational Status Discordance and Party Support

●

●

●

●

ISD

−0.1 0.0

Linke

SPD

Green

Mainstream Left

Mainstream Right

CDU/CSU

AfD

Radical Right

Outer bars are 0.95 CIs

(a) Vote Choice (Bundestagswahl 2017)

●

●

●

●

●

ISD

−0.1 0.0 0.1

Linke

SPD

Green

Mainstream Left

Mainstream Right

CDU/CSU

AfD

Radical Right

Outer bars are 0.95 CIs

(b) Party Identi�cation

In terms of magnitude, the coe�cient of ISD is comparable to that of gender and hasa larger e�ect than current socio-economic status, measured as log income, as well asage of the respondent (Appendix Figure A7 displays standardized coe�cients). Thisfavorable comparison to important and well-known predictors of radical right votingunderscores the importance of expectations in shaping voter behavior. At the sametime, level of education is clearly a more powerful factor driving support for the Al-ternative für Deutschland.

5.3 Socio-Economic Mechanisms Linking Status Discordanceand Voting

The results thus far suggest that radical right and left parties share a common themethat attracts voters who experience disappointment with regard to status expectations.

18

However, given stark di�erences in the political platforms of radical left and right par-ties, it is unlikely that downwardly mobile voters more or less arbitrarily opt for oneor the other. Rather, we suggest that di�erent types of status discordant voters seekdi�erent political solutions to address their fading prospects for upward mobility. Aswe have emphasized, our measure captures deviations from status expectations basedon parental background and is thus agnostic about current occupational status or in-come level. An individual from a lower-middle class background who faces povertymay have the same ISD rank as someone from an upper-class background who endsup in the middle class. However, while they share ISD ranks and an increased likeli-hood to support parties outside the mainstream, we expect that individual and parentalsocio-economic status shapes which anti-establishment party status discordant cometo support.

Table 2 shows e�ects of status discordance conditional on whether an individual hasa college degree or not. The results indeed con�rm that there are di�erent types ofdownwardly mobile voters. A college degree strongly decreases the likelihood of sup-porting a radical right party among respondents experiencing negative status discor-dance, while it reinforces, albeit relatively weakly, the likelihood to vote for radical leftparties.

Figure 6 provides a di�erent perspective on the same narrative by showing e�ects ofstatus discordance conditional on father socio-economic status, i.e. a standard pre-treatment times treatment interaction. Again, the analysis con�rms the existence ofdi�erent types of status discordant voters. Respondents who experience downwardmobility relative to fathers who did not have college degrees and who worked low- andmiddle-skill jobs are more likely to support radical right parties while those who ex-perience downward mobility relative to college-educated fathers and fathers at higherpositions in the occupational hierarchy are more likely to support radical left parties.The e�ect on support for the radical left, then, is con�ned to voters who have achieveda certain socio-economic status, proxied by education, but who consider this status asfalling short of expectations formed in childhood.

19

Table2:

Stat

usD

isco

rdan

cean

dRa

dica

lPar

tySu

ppor

t,by

Educ

atio

n

rrvo

terr

idaf

dvo

teaf

did

linke

vote

linke

idIS

D(P

ctile

s)0.

088∗∗

∗0.

106∗∗

∗0.

082∗∗

∗0.

092∗∗

∗−

0.00

20.

029

(0.0

12)

(0.0

17)

(0.0

12)

(0.0

17)

(0.0

12)

(0.0

18)

Colle

geD

egre

e−

3.00

6∗∗−

3.48

8∗∗−

2.96

3∗∗−

3.65

2∗∗1.

788

0.59

9(1

.032

)(1

.346

)(1

.023

)(1

.327

)(1

.012

)(1

.396

)IS

Dx

Colle

geD

egre

e−

0.04

2∗−

0.03

8−

0.03

8−

0.02

70.

037

0.04

6(0

.021

)(0

.027

)(0

.021

)(0

.027

)(0

.020

)(0

.028

)In

terc

ept

6.14

311

.886

∗7.

500∗

12.0

10∗

14.8

45∗∗

∗13

.777

∗∗

(3.4

68)

(4.7

42)

(3.4

40)

(4.6

73)

(3.4

02)

(4.9

17)

Cont

rols

Bund

esla

ndFE

sR2

0.04

30.

060

0.04

20.

055

0.03

00.

081

Adj

.R2

0.04

10.

056

0.04

00.

051

0.02

90.

077

Num

.obs

.10

828

5396

1082

853

9610

828

5396

RMSE

25.8

3924

.983

25.6

3324

.615

25.3

4825

.902

∗∗∗p

<0.

001,

∗∗p

<0.

01,∗

p<

0.05

.Con

trol

sasi

nTa

ble

1ex

cept

the

educ

atio

nva

riabl

e(r

epla

ced

with

colle

gedu

mm

y).

20

As we proposed in section 3, one can imagine di�erent versions of status discordantvoters from higher socio-economic strata. One example includes high-skilled voterswho cannot replicate the economic prosperity of their parents despite having a univer-sity degree. In contrast to their parents’ generation, they su�er from a growing supplyof college graduates and increasingly globalized markets, which results in sti�er com-petition and mounting wage pressure even in highly pro�table sectors of the economy.Another example includes high-skilled voters who seek satisfaction in areas that tran-scend traditional ideas of career and material success, e.g. many workers in the non-pro�t or arts sectors, and who voluntarily give up economic ambitions for such work.Both types of highly educated downwardly mobile voters are more likely to supportradical left parties, even if for slightly di�erent reasons.

Figure 6: Status Discordance and Radical Party Support, by Father Status

●

●

●

−0.05

0.00

0.05

0.10

0.15

low mid highFather Occupational Status

Con

ditio

nal M

argi

nal E

ffect

of S

tatu

s D

isco

rdan

ce

VoteChoice

● afd

linke

In contrast, radical right parties successfully mobilize those whose achievements arenot only lower relative to expectations based on parental background, but also lowerin the occupational hierarchy. This group is also victim to postindustrial occupationalchange, though the e�ects are of a di�erent type and degree than those a�ecting high-skilled voters, who as we mentioned before are still less likely to experience downwardmobility. While postwar labor markets created ample economic opportunity for work-ers without college degrees, job prospects for the children of such workers have un-

21

doubtedly worsened. Common examples are semiskilled routine work in factories andback o�ces vulnerable to displacement due to automation and technological change.The once-promising occupational trajectory of relatively well-paid and well-protectedroutine employment has all but disappeared for current labor market entrants withouta college degree, forcing them instead into low-skilled and often poorly remuneratedjobs in the service sector. Radical right parties are attractive for these voters becausethey advocate for a return to precisely the times in which low-skilled work still hadintegrity and still posed opportunities for mobility.

5.4 Attitudinal Mechanisms Linking Status Discordance andVoting

As a last step, and in the hopes of pointing to fruitful avenues for future research, weexplore mechanisms linking status discordance to voting behavior and party choice byexamining some attitudinal implications of downward mobility. First, Table 3 showsdirect and conditional (on education) e�ects of ISD on life satisfaction (Model 1 and2). As expected, respondents who experience negative status discordance are generallyless satis�ed, which again validates our measure and con�rms one of the basic premisesof our theoretical reasoning. General dissatisfaction with life is likely to translate intolower levels of political satisfaction, which, unfortunately, we are not able to measureas it did not feature in the SOEP questionnaire. The negative interaction term in thesecond column of Table 3 indicates that, if anything, downward mobility seems es-pecially di�cult to bear among those whose expectations were quite high in the �rstplace. This result implies that the group of the high-skilled downwardly mobile doesnot primarily consist of respondents who "voluntarily" give up economic ambition.

A second indication of political dissatisfaction is provided by the negative correlationbetween status discordance and political interest in the third column. This e�ect is sig-ni�cantly more pronounced among low-skilled compared to high-skilled downwardlymobile voters. The latter are only slightly less interested in politics than the rest ofthe population with more favorable social mobility experiences. We interpret this asevidence that the strong abstention e�ects shown previously are a political responsethat is particularly prevalent among less educated voters, which aligns with existing

22

evidence (Levi and Stoker 2000). College-educated status discordant voters may feelpolitically dissatis�ed to the same degree but may be more likely to nonetheless viewthe political arena as an appropriate place to voice this dissatisfaction.

The remaining columns in Table 3 show results on two selected items of a batteryasking respondents about the extent to which certain issues worry them (on a 1-3 scale;"big concern", "some concern", or "no concern"). We show results with respect to twotypical issues representing the two dimensions of the political space: immigration andpersonal economic circumstances. Higher levels of status discordance are associatedwith higher levels of concern with regard to both issues (Models 5 and 7), con�rmingour expectation that downward mobility generates political dissatisfaction insofar asmainstream parties are unable to address, or unaware of, core concerns.5

Finally, models 6 and 8 provide further evidence that di�erent types of status discordantvoters have di�erent preferences when it comes to the societal changes they envision,thus motivating the choice between a left- or a right-wing alternative to the politi-cal establishment. Concerns about immigration are slightly more pronounced amongstatus discordant voters without a college degree, while concerns about the economyare more pronounced among voters with a degree. The partisan pattern among down-wardly mobile voters shown previously—that the low-skilled tend to turn right, andhigh-skilled tend to turn left—thus seems to support our hypothesis outlined in sec-tion 3. The tendency of radical right parties to hearken back to a time when indus-trial work still carried integrity and opportunity—often portrayed by these parties as atime before immigrants replaced native labor—is particularly appealing to low-skilleddownwardly mobile voters. High-skilled voters, meanwhile, likely support radical leftparties for their emphasis on addressing income inequality and reforming global �-nancial institutions.

5In line with research showing that these concerns are often accompanied by a general sense of socialpessimism shown to drive support for radical parties (e.g. Steenvoorden and Harteveld 2018), wesimilarly �nd weakly positive correlations between status discordance and a general concern aboutwhere society is heading (see Table A8 in the Appendix).

23

Table3:

Stat

usD

isco

rdan

cean

dA

ttitu

des,

Dire

ctan

dby

Educ

atio

n

Satis

fact

ion

Polit

ical

Conc

ern:

Conc

ern:

with

Life

Inte

rest

Mig

ratio

nEc

onom

y(p

erso

nal)

12

34

56

78

ISD

(Pct

iles)

−0.

003∗∗

∗−

0.00

2∗∗∗−

0.00

3∗∗∗−

0.00

3∗∗∗

0.00

3∗∗∗

0.00

3∗∗∗

0.00

3∗∗∗

0.00

2∗∗∗

(0.0

01)

(0.0

01)

(0.0

00)

(0.0

00)

(0.0

00)

(0.0

00)

(0.0

00)

(0.0

00)

Colle

geD

egre

e0.

075∗

0.16

0∗∗0.

334∗∗

∗0.

265∗∗

∗−

0.34

4∗∗∗−

0.26

3∗∗∗−

0.08

4∗∗∗−

0.15

3∗∗∗

(0.0

32)

(0.0

56)

(0.0

16)

(0.0

27)

(0.0

16)

(0.0

27)

(0.0

13)

(0.0

23)

ISD

xCo

llege

Deg

ree

−0.

002

0.00

2∗∗−

0.00

2∗∗∗

0.00

2∗∗∗

(0.0

01)

(0.0

01)

(0.0

01)

(0.0

00)

Inte

rcep

t6.

934∗∗

∗6.

908∗∗

∗1.

714∗∗

∗1.

736∗∗

∗1.

873∗∗

∗1.

848∗∗

∗2.

648∗∗

∗2.

670∗∗

∗

(0.1

86)

(0.1

87)

(0.0

90)

(0.0

90)

(0.0

90)

(0.0

90)

(0.0

76)

(0.0

76)

Cont

rols

Bund

esla

ndFE

sR2

0.01

90.

020

0.16

90.

170

0.10

90.

110

0.08

80.

089

Adj

.R2

0.01

80.

018

0.16

80.

168

0.10

70.

108

0.08

70.

088

Num

.obs

.12

120

1212

012

111

1211

112

097

1209

712

115

1211

5RM

SE1.

473

1.47

30.

712

0.71

20.

708

0.70

80.

603

0.60

3∗∗

∗p

<0.

001,

∗∗p

<0.

01,∗

p<

0.05

.Con

trol

sasi

nTa

ble

1ex

cept

the

educ

atio

nva

riabl

e(r

epla

ced

with

colle

gedu

mm

y).

24

6 Conclusion

We have argued that the disappointed expectations following fading prospects for mo-bility have political consequences, and that these consequences shed light on the polit-ical disruptions ongoing across Western democracies today. To support our argument,we introduced the innovative concept of status discordance, which we capture empiri-cally as the di�erence between status expectations and status outcomes. Using parentalsocio-economic background as the reference point for status expectations, we demon-strate that the experience of disappointing these expectations—an experience becom-ing increasingly frequent—is an important factor underlying political alienation anddissatisfaction with mainstream parties. Our evidence, based on German householdpanel data, reveals that the downwardly mobile are more likely to abstain from votingand to vote for radical parties.

While these results hold regardless of a voter’s socio-economic background, we alsodemonstrate that there are di�erent types of downwardly mobile voters. More speci�-cally, low-skilled downwardly mobile voters tend to support radical right parties, whilethose at higher positions in the occupational hierarchy gravitate towards the radicalleft. The attitudinal measures in our dataset provide further support for this mecha-nism, the primacy of concerns about immigration relative to concerns about the econ-omy being one factor distinguishing these voters. Notwithstanding this �nding, theunderlying mechanisms linking negative status discordance to party choice remain afruitful pursuit for future research.

We believe our results carry implications beyond the German context. Downward mo-bility is increasing across postindustrial democracies, as are electoral victories amongradical parties—trends that our results indicate may be linked. While the degree ofdownward mobility and party systems vary across Western democracies, our �ndingssuggest that even in two-party systems such as the US, the growing share of down-wardly mobile voters across all socio-economic backgrounds may help to explain therecent rise of candidates such as Donald Trump and Bernie Sanders, whose politi-cal platforms diverged markedly from the usual centrist tendencies of American poli-tics.

25

More generally, we hope our analysis draws attention to the importance of expec-tations when it comes to political behavior, and to the link between social mobilityand political (in)stability. Upward mobility was an undervalued cornerstone of post-war politics. As increasing shares of citizens face declining opportunities for mobility,disappointed expectations are on the rise, a�ecting voter behavior and electoral out-comes. Importantly, this is distinct from the usual dissatisfaction that is part of anyhealthy democracy in that it cannot be addressed through short term measures or re-placement of party leaders. It instead re�ects the sense among disappointed votersthat the political solutions prevailing in recent decades are out of touch with the en-croaching reality of downward mobility across postindustrial societies. While changeis inherent to democratic politics, the degree of political disruption taking place inadvanced democracies is unprecedented. Given that ongoing occupational and edu-cational trends seem likely to continue, as the task of social and economic inclusionlooms, party politics in advanced democracies may never be the same.

26

References

Abramson, P. R. and Books, J. W. (1971). Social mobility and political attitudes: Astudy of intergenerational mobility among young british men. Comparative Politics,3(3):403–428.

Alesina, A., Stantcheva, S., and Teso, E. (2018). Intergenerational Mobility and Prefer-ences for Redistribution. American Economic Review, 108(2):521–554.

Anelli, M., Colantone, I., and Stanig, P. (2019). We Were The Robots: Automation andVoting Behavior in Western Europe. IZA Discussion Paper Series, No. 12485.

Ansell, B. and Gingrich, J. (2018). Welfare Democracies and Party Politics. ExplainingElectoral Dynamics in Times of Changing Welfare Capitalism, chapter Skills in de-mand? Higher education and social investment in Europe, pages 225–256. OxfordUniversity Press.

Ares, M. (2019). Changing classes, changing preferences: how social class mobilitya�ects economic preferences. West European Politics, 0(0):1–27.

Autor, D. and Dorn, D. (2013). The Growth of Low-Skill Service Jobs and the Polariza-tion of the US Labor Market. American Economic Review, 103(5):1553–1597.

Autor, D., Goldin, C., and Katz, L. (2020). Extending the Race between Education andTechnology. Technical Report w26705, National Bureau of Economic Research, Cam-bridge, MA.

Ballard-Rosa, C., Jensen, A., and Scheve, K. (2018). Economic Decline, Social Identity,and Authoritarian Values in the United States. Manuscript.

Beller, E. (2009). Bringing intergenerational social mobility research into the twenty-�rst century: Why mothers matter. American Sociological Review, 74:507–528.

Benabou, R. and Ok, E. A. (2001). Social mobility and the demand for redistribution:The poum hypothesis. The Quarterly Journal of Economics, 116(2):447–487.

Blais, A. (2006). What a�ects voter turnout? Annual Review of Political Science, 9:111–125.

27

Breen, R. and Müller, W. (2020). Education and intergenerational social mobility in Eu-rope and the United States. Studies in social inequality. Stanford University Press,Stanford California.

Breiman, L. (2001). Random Forests. Machine Learning, 45(1):5–32.

Broz, J. L., Frieden, J., and Weymouth, S. (2020). Populism in Place: The EconomicGeography of the Globalization Backlash. Manuscript.

Bukodi, E., Paskov, M., and Nolan, B. (2020). Intergenerational Class Mobility in Eu-rope: A New Account. Social Forces, 98(3):941–972.

Burgoon, B., van Noort, S., Rooduijn, M., and Underhill, G. (2018). Positional Depriva-tion and Support for Radical Right and Radical Left Parties. Economic Policy, online�rst.

Chauvel, L. and Schröder, M. (2014). Generational inequalities and welfare regimes.Social Forces, 92(4):1259–1283.

Checchi, D. (2006). The economics of education: Human capital, family background andinequality. Cambridge University Press.

Chetty, R., Grusky, D., Hell, M., Hendren, N., Manduca, R., and Narang, J. (2017). Thefading American dream: Trends in absolute income mobility since 1940. Science,356(6336):398–406.

Cohen, J. (1987). Parents as educational models and de�ners. Journal of Marriage andFamily, 49(2):339–351.

Colantone, I. and Stanig, P. (2018). Global competition and Brexit. American PoliticalScience Review, 112(2):201–218.

Dustmann, C. (2004). Parental background, secondary school track choice, and wages.Oxford Economic Papers, 56(2):209–230.

Eichengreen, B. (2018). The Populist Temptation: Economic Grievance and Political Re-action in the Modern Era. Oxford University Press.

Ganzeboom, H., Graaf, P. D., and Treiman, D. (1992). A standard international socio-economic index of occupational status. Social Science Research, 21:1–56.

28

Geschwender, J. A. (1964). Social Structure and the Negro Revolt: An Examination ofSome Hypotheses. Social Forces, 43(2):248–256.

Gest, J. (2016). The NewMinority: White Working Class Politics in an Age of Immigrationand Inequality. Oxford University Press.

Gest, J., Reny, T., and Mayer, J. (2017). Roots of the Radical Right: Nostalgic Deprivationin the United States and Britain. Comparative Political Studies, 51(13):1694–1719.

Gidron, N. and Hall, P. A. (2017). The politics of social status: economic and culturalroots of the populist right. The British Journal of Sociology, 68:S57–S84.

Goos, M., Manning, A., and Salomons, A. (2009). Job polarization in Europe. TheAmerican Economic Review, pages 58–63.

Gurr, T. R. (1970). Why men rebel. Princeton University Press, Princeton, New Jersey.

Hertel, F. R. (2017). Absolute Mobility in Germany over the 20th century. In Social Mo-bility in the 20th Century, pages 185–213. Springer Fachmedien Wiesbaden, Wies-baden.

Hill, M. and Duncan, G. (1987). Parental family income and the socioeconomic attain-ment of children. Social Science Research, 16:39–73.

Hobolt, S. B. (2016). The brexit vote: a divided nation, a divided continent. Journal ofEuropean Public Policy, 23(9):1259–1277.

Hochschild, A. R. (2016). Strangers in Their Own Land: Anger and Mourning on theAmerican Right. New Press, The.

Hooghe, L. and Marks, G. (2018). Cleavage theory meets europe’s crises: Lipset, rokkan,and the transnational cleavage. Journal of European Public Policy, 25(1):109–135.

Hout, M. (2018). Americans’ occupational status re�ects the status of both of theirparents. Proceedings of the National Academy of Sciences, 115(38):9527–9532.

Im, Z. J., Mayer, N., Palier, B., and Rovny, J. (2019). The “losers of automation”: Areservoir of votes for the radical right? Research & Politics, 6(1).

29

Inglehart, R. and Norris, P. (2017). Trump and the Populist Authoritarian Parties: TheSilent Revolution in Reverse. Perspectives on Politics, 15(2):443–454.

James, G., Witten, D., Hastie, T., and Tibshirani, R. (2013). An Introduction to StatisticalLearning, volume 103 of Springer Texts in Statistics. Springer New York, New York,NY.

Johnson, M. K. (2002). Social origins, adolescent experiences, and work value trajecto-ries during the transition to adulthood. Social forces, 80(4):1307–1340.

Kao, G. and Tienda, M. (1998). Educational Aspirations of Minority Youth. AmericanJournal of Education, 106(3):349–384. Publisher: The University of Chicago Press.

Kurer, T. (2020). The Declining Middle: Occupational Change, Social Status, and thePopulist Right. Comparative Political Studies, 53(10-11):1798–1835.

Kurer, T., Häusermann, S., Wüest, B., and Enggist, M. (2018). Economic Grievancesand Political Protest. European Journal of Political Research.

Lahtinen, H., Wass, H., and Hiilamo, H. (2017). Gradient constraint in voting: Thee�ect of intra-generational social class and income mobility on turnout. ElectoralStudies, 45:14–23.

Lareau, A. (2011). Unequal Childhoods: Class, race, and family life. Berkeley: Universityof California Press.

Levi, M. and Stoker, L. (2000). Political trust and trustworthiness. Annual review ofpolitical science, 3(1):475–507.

Lipset, S. M. (1959). Some Social Requisites of Democracy: Economic Developmentand Political Legitimacy. The American Political Science Review, 53(1):69–105.

Lipset, S. M. (1960). Political Man. Doubleday, New York.

Margalit, Y. (2019). Economic Insecurity and the Causes of Populism, Reconsidered.Journal of Economic Perspectives, 33(4):152–170.

Mau, S. (2019). Lütten Klein: Leben in der ostdeutschen Transformationsgesellschaft.Suhrkamp, Berlin.

30

Mitrea, E. C., Mühlböck, M., and Warmuth, J. (2020). Extreme Pessimists? ExpectedSocioeconomic Downward Mobility and the Political Attitudes of Young Adults. Po-litical Behavior.

Montgomery, J. M. and Olivella, S. (2018). Tree-Based Models for Political Science Data.American Journal of Political Science, 62(3):729–744.

Oesch, D. (2013). Occupational change in Europe: how technology and education trans-form the job structure. Oxford Univ. Press, Oxford.

Owen, E. and Johnston, N. P. (2017). Occupation and the Political Economy of Trade:Job Routineness, O�shorability, and Protectionist Sentiment. International Organi-zation, 71(4):665–699.

Polavieja, J. G. and Platt, L. (2014). Nurse or mechanic? the role of parental socializationand children’s personality in the formation of sex-typed occupational aspirations.Social Forces, 93(1):31–61.

Rodrik, D. (2018). Populism and the economics of globalization. Journal of InternationalBusiness Policy, 1(1-2):12–33.

Rooduijn, M., Burgoon, B., Van Elsas, E. J., and Van de Werfhorst, H. G. (2017). Radicaldistinction: Support for radical left and radical right parties in europe. EuropeanUnion Politics, 18(4):536–559.

Steenvoorden, E. and Harteveld, E. (2018). The appeal of nostalgia: the in�uence of so-cietal pessimism on support for populist radical right parties. West European Politics,41(1):28–52.

Treiman, D. J. (1970). Industrialization and Social Strati�cation*. Sociological Inquiry,40(2):207–234.

31

7 Appendix

Appendix A

●

●

●●●

● ●

●●

●

●

●●

●

●

●

●●

●

●

●

●●

●●●●

●

●●●

●

●

●

●

●

●●

●

●

●●

●●●●

●

●

●●●

●

●

●

●

●

●

●●●

●

●●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●●●●

●

●

●●

●

●

●

●

●

●●●●●

●

●

●

●

●●●●●●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●●

●●●

●

●●

●

●

●

●●

●

●●●●

●●

●

●

●●

●●●

●

●

●

●

●●●

●

●●

●

●

●●

●

●

●

●

●

●

●

●●●

● ●

●●●

●

●●

●●●●●●●

●

●

●●

●●

●●

●●●●●●

●

●●●●●

●

●

●

●

●●

●

●

●

●

●●●●

●●●

●

●

●

●

●●●●●

●

●

●●

●●●●●

●●●●

●●

●

●

●●●

●●●

●

●

●

●●

●

●

●

●●●

●

●

●

●

●

●●

●

●

●

●

●●

●

●●

●●

●

●

●●●

●

●

●

●

●

●

●

●●●●

●

●

●

●

●

●

●

●

●

●

●●●●●

●

●●●●

●

●

●●

●●

●

●

●

●●

●●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●●●●●

●●●●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●●●

●

●●●●

●

●

●●●●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●●

●●●

●

●●●●

●●●

●

●

●

●

●

●

●●●●●●●

●

●

●

●

●

●

●

●

●

●●

●●●

●

●

●

●

●

● ●●●●

●

●●●

●

●

●

●

●●●●●

●

●

●●●●●

●

●

●●

●

●●●●

●

●

●

●

● ●●

●●

●●

●●

●

●

●●

●

●

●

●●

●●●●●●

●

●

●

●●

●

●

●

●●

●●●●

●●

●

●● ●

●

●●●

●

●

●

●

●

●●

●●●

●

●●●

●●●

●

●●●

●

●

●●

●

●

●

●

●

●●

●●●●

●

●

●

●

●●●

●

●●●●

●

●●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●●

●●

●●●●

●

●

●●●●●●

●

●

●●

●

●

●

●

●

●

●●

●

●

●

●●●

●●

●●

●●● ●

●

●

●●●●

●

●

●●

●●

●

●●

●●

●

●●

●

●

●

●

●●●

●

●

●

●

●●

●

●

●

●●

●

●

●●●

●

●●●

●

●● ●

●

●●

●

●

●

●

●

●

●●

●

●●●●

●

●●●

●

●

●●●●

●

●

●

●

●●

●●●●

●

●

●●

●●●

●

●

●

●

●●

●

●

●

●

●

●●●●●●●●●●●●●

●

●●●●

●

●

●

●●●

●●

●

●

●

●

●●

●

●●●●

●

●

●

●

●●●●●●●

●

●

●●

●

●

●

●

●

●

●

●

●●

●●

● ●●●

●

●

●

●

●

●●●●●

●

●

●

●

●●●●

●

●

●●●

●

●

●●

●●

●

●

●

●

●●●

●

●

●

●

●

●

●●●●●●●

●●

●

●●●●

●

●

●

●●

●●

●●●

●

●

●●

●

●●●

●

●●●

●●●

●●●

●

●●

●

●

●

● ●

●●

●

●●●●●●●

●

●

●

●

●●●●

●●●●

●●

●

●●

●●

●●●●

●

●

●●●●●

●

●

●

●

●

●

●●●●

●

●

●

●

●

●●

●

●

●●●●●

●

●

●●

●●

●

●

●

●●

●

●●●

●●●●●

●

●●

●

●●●●

●

●●●

●

●

●

●

●

●

●

●

●●●●

●●●

●

●●●●

●

●

●

●●●●

●

●

●●●

●●

●

●

●

●

●●●

●●

●

●

●●●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●●

●●

●●

●

●

●●●●

●●●

●

●

●●

●

●

●

●●

●●

●

●

●

●

●

●

●●

●

●●●

●

●

●●●●

●

●●●

●●

●

●

●

●

●

●

●

●

●●●

●

●

●

●●

●

●

●

●

●

●

●

●

●●●

●

●

●

●

●

●

●

●

●

●

●

●

●●●●

●●●●

●

●●

●

●

●

●●●

●

●

●

●

●

●

●

●

●

●

●

●●●

●●

●

●

●

●

●

●

●●●●

●

●

●●

●

●●

●

●●●

●

●●

●●

●

●

●

●

●

●●

●

●●●●●

●

●

●

●

●

●●

●

●

●●●

●●

●

●

●●

●

●

●●●●●●●●●

● ●

●

●●

●

●

●

●

●

●

●●●

●

●●

●

●

●

●

●

●

●

●●●

●

●

●

●

●

●

●

●●

●

●●●

●

●●

● ●

●

●●●●●●

●

●●●

●

●

●

●●

●

●

●

●

●

●

●

●●

●●●

●

●●●

●

●●

●

●

●

●

●●

●

●

●

●●

●

●

●

●

●

●

●

●

● ●

●●

●

●●

●●

●

●

●

●

●●

●

●

●

●

●

●

●●●

●

●●●

●

●

●●●

●

●

●●

●

●

●

●

●

●

●

●●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●

● ●

●

●

●●

●

●

●

●

●

●

●

●

●●

●●●●

●

●

●

●●●

●

●

●

●

●

●

●

●●●●●

●●

●

●

●●

●

●●

●●●

●

●

●

●●●

●

●●

●●

●

●●

●

●

●●●

●●

●

●

●

●●

●

●

●

●

●●

●

●

●

●●

●●

●●●●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●●

●●

●

●

●●

●●●●

●

●●●

●

●●

●●

●●●

●

●

●●●●

●

●

●●

●

●

●

●

●

●●

●●

●

●

●

●

●●

●

●●●

●

●

●

●

●

●

●

●

●

●●●

●

●●●●

●

●

●

●

●

●

●

●

●

●

●●

●●●

●

●

●●

●

●●

●

● ●

●●●

●

●

●

●

●●

●

●

●●

●

●

●●

●

●

●

●●●

●●●

●

●●

●●

●

●●

●

●●●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●●●

●

●

●

●●

●

●●●●●●●

●

●

●

●●●●

●

●●●

●

●

●

●●●●

●

●●

●

●

●●

●

●

●

●●

●

●

●●●

●●

●

●●●

●

●

●●●●●●●

●

●

● ●

●

●

●●

●●

●

●●●

●●●●

●●

●●

●

●●

●

●

●

●●●

●

●

●

●●

●

●

●●●

●

●●

●

●●

●●

●

●●

●

●●●

●

●

●

●

●

●

●

●●●

●

●●

●

●

●●

●

●

●

●

●

●

●

●●

●

● ●

●

●

●

●●

●

●

●

●

●

●●●●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●●

●

●

●●

●●●

●●●

●

●

●

●●

●

●●●

●

●●

●

●

●●

●

●●●●●

●

●●

●

● ●●●●

●●

●

●● ●

●

●●

●

●

●●

●

● ●●●

●●

●

●

●

●

●●

●●●

●

●●●●●●

●

●●

●●

●

●

●

●

●●●●

●

●

●

●

●

●

●●

●

●●●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●●●●●●●●

●

●

●●●

●

●

●

●●●●

●●

●

●

●●

●●●

●

●

●

●

●●●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●●●●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●●

●●●

●

●●●●●

●●●

●

●●●

●●

●

●

●

●

●

●

●

●

●●●

●

●●●●

●

●●

●

●

●

●

●●●

●

●

●●●●

●●

●

●●●

●

●

●

●

●●

● ●

●

●

●

●

●

●●

●●

●

●

●

●

●

●

●

●●

●

●

●

●●

●

●●

●●●●●●

●

●

●●●●

●●

●

●

●

●

●●●●

●

●

●

●●●●

●

●

●

●●

●

●

●●●●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

● ●●

●●

●

●

●●

●

●

●

●

●●

●

●

●

●●

●

● ●

●

●

●

●●

●

●

●

●

●

●●

●

●

●

●●

●

●

●●

●●●

●

●

●

● ●

●●●

●

●

●

●

●

●●

●

●

●

●

●

●

●●●

●

●

●●

●

●

●

●

●

●●

●

●

●●

●

●

●

●

●

●

●●

●●

●

●●●●●●

●

●● ●

●

●

●●

●●

●●●

●

●●●

●

●

●

●●

●

●

●

●

●

●●●●●●

●

●●

●

●

●

●●●●

●

●●●

●●●●

●●

●

●

●●●

●●

●

●

●●

●

●

●

●

●●

●●●

●●●●●●●

●

●●

●

●

●

●

●●●●

●

●

●

●

●

●●●●

●

●

●●●●

●

●●

●

●

●

●●●●

●

●

●

●

●●●

●●

●

●

●

●

●●●●

●

●

●

●

●

●

●

●●

●

●●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●●●●●

●

●

●

●

●

● ●

●

●

●●

●

●

●

●

●

●

●

●●

●

●

●●●

●

●

●

●

●

●

●

●●●●

●●

●

●

●

●●

●

●

●

●●●●●

●

●

●●

●●●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●●●

●●

●

●

●●● ●

●●

●

●

●

●●

●

●

●●

●

●

●●

●●

●

●

●

●

●

●

●

●

●●●●

●

●●

●●●●

●●

●

●

●

●●

●

●

●●

●

●

●

●

●

●●

●●●

●

●

●●●

●●

●●●

●

●

●

●●

●●

●●

●●●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●●

●●

●

●●●●●● ●

●

●

●●●●●●●●

●

●●

●

●

●

●

●

●●●●

●

●

●

●

●

●●

●

●

●●●

●

●●●

●

●

●

●

●

●●●

●

●

●

●

●

●

●

●

●

●●

●

● ●●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

● ●

●

●

●

●

●

●

●●

●●●●

●

●●

●

●

●

●

●

●

●

●●

●

●

●

●●

●●●

●

●

●

●

●

●

●

●

●

●●● ●

●●

●

●

●

●

●●●

●●●

●

●

●

●●

●●●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●●

●

●

●●●

●

●

●

●

●●●●

●

●

●●

●

●

●●

●

●●

●●●

●●

●

●

●

●

●●●●●

●

●

●●

●

●

●

●

●

●

●●

●

●

●

●

●●

●

●

●

●

●

●

● ●

●●●●●

●

●

●

●

●

●

●

●

●

●

●●

●

●●

●

●

●

●●

●

●

●●

●

●

●

●

●

●

●●●

●

●

●●

●●

●

●●

●

●●●●

●●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●●●

●●●●●

●

●●●●

●

●

●●●

● ●

●

●

●

●●●

● ●

●●●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●●●

●

●

●

●

●

●

●

●

●

●

●

●●

●●

●

●

●

●

●

●

●

●

●●●●

●

●●

●

●

●

●●●

●

●

●●●●

●

●●

●

●

●●

●

●

●

●

●

●●●●

●

●

●●●●

●

●●●●

●

●●

●●

●●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●●●●

●

●

●●

●

●●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●●●●● ●●

●

●

●

●

●

●

●

●●

●

●●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●●

●

●

●

●

●

● ●●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●● ●

●

●

●

●

●

●

●

●

●

●

●

●

●

●● ●●

●

●●●●

●

●●

●

●

●

●

●

●

●

●●●

●●●

●●

●

●

●

●●●●●●

●

●

●●●●

●

●

●

●●●●

●●●

●

●

●

●●●

●●

●

●

●

●

●●●

●

●

●

●

●

●●●●●

●

●

●

●●●●

●●

●

●●

●

●

● ●

●●

●

●

●

●

●●

●●

●

●

●

●●

●

●

●

●

●

●●●

●

●

●●●

●

●

●

●

●

●●

●

●●●●●●

●

●●●

●

●

●

●●● ●

●

●

●

●●

●

●

●●●●●●

●

●

●

●

●●●●●

●

●

●●●●

●

●

●

●●

●●●●●●

●●●●●

●●

●

●●●●●●

●●

●

●

●●

●

●

●

●

●

●

●●

●

●●

●

●

●

●

●

●

●

●

●●●●●

●

●●●

●

●

●●

●

●

●●●

●

●

●●

●●

●

●●●●

●●● ●

●

●

●

●●●

●

●

●

●●

●

●

●

●

●

●●

●●

●●●●

●

●

●●

●●

●

●

●

●

●

●●●●

●

●●

●●

●●●

●●●

●

●

●●●

●

●

●●

●

●

●

●●●●

●

●●●

●●●

●

●

●

●●

●

●

●

●

● ●

●●

●

●

●●●●

●●

●

●

●●●●●●●

●

●

●

●

●

●●●●

●●●

●●●●

●●

●●

●

●●●

●

●

●●

●

●●

●●●●●

●●●

●

●●

●

●

●

●●●

●

●

●●●●

●

●●●

●

●●

●

●●●

●

●●

● ●

●●

●

●

●

●

●

●

●

●●

●

●

●

●

● ●●

●

●●●

●

●

●

●

●

●

●

●

●

●

●

●

●●●

●

●

●●

●●

●

●●

●

●

●

●

●

●●

●

●

●●●

●

●

●●

●

●

●

●

●

●

●

● ●

●

●●●●●

●

●●●●

●●

●

●

●

●

●

●●

●

●

●

●

●

●●●

●

●

●●● ●

●

●●●●●●

●

●

●

●

●●●

●●●●

●

●●

●

●

●

●●● ●

●

●

●

●●

●●

●

●●

●

●

●

●

●●

●

●●

●

●

●●●●

●●

●

●●●●●

●●

●

●●●

●

●●

●

●

●●

●●

●

●●

●●

●

●●

●

●

●●●●●

●●●

●

●

●●●

●

●

●●

●

●

●

●●

●

●

● ●

●

●

●

●●

●

●

●

●

●

●

●●

●

●

●

●

●

●●●

●

●

●

●

●

●

●●

●

●●●●●●●

●

●

●

●

●●

●●●●

●●

●

●

●

●

●●

●

●

●

●

●●●

●

●

●

● ●

●

●●

●

●●

●

●●●

●●●

●

●

●

●

●

●

●

●

●

●●●

●

●

●●●

●

●

●

●

●

●●●

●

●

●

●

●

●

●

●●●

●

●

●

●●

●

●●

●

●

●

●

●

●

●

● ●

●●●

●

●

●

●

●●●

●

●

●

●

●

●

●

●

●

●

●●●●●

●

●

●

●●●

●

●

●

●

●

●

●

●

●●

●

●●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●●●

●

●●

●●●●

●

●

●

●●

●

●

●

●

●

●

●●

●

●●

●●●●

●

●

●●●

●

●●

●

●

●●●●●

●

●

●

●●●●

●

●

● ●

●

●● ●

●●

●●

●

●

●●

●

●

●

●●

●

●

●●

●

●

●

●

●●●

●

●

●

●

●●

●

●

●

●

●

●●●●

●

●

●

●●

●●

●●●● ●

●●●

●

●

●

●

●●

●●●

●

●

●

●

●

●

●●

●

●●

●

●●

●

●

●

●

●

●

●●●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

● ●

●

●

●

●

●

●

●

●

●

●

●●

●●●

●

●●

●●●●●●

●

●

●●

●

●

●

●

●

●

●

●●

●

●

●

●

●●

●●

●

●●●●

●

●

●

●

●

●●

●

●

●●●

●

●

●

●

●●

●●●●

●

●

●

●

●

●●●

●

●

●●●

●

●

●●●●

●●●●

●●

●

●

●

●

●●●

●●●

●

●

●●

●

●●

●

●

●●

●

●●

●

●

●●

●

●

●

●

●●

●

●

●

●●●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●●●

●●

●

●●

●

●

●

●

●

●

●

●

●

●●●

●

●

●

●

●

●

●

●

●●●

●

●

●

●

●●●

●

●●

●

●●●

●

●●

●

●

●

●

●

●

●

●●

●

●●

●

●

●●

●

●

●●●●●

●

●●

●

●●

●

●

●

●

●

●●

●

●

●

●

●●

●

●●●

●

●●●●

●●●●

●

●

●●●

●

●

●

●

●●●● ●●

●●●●●●●●●●

●●●●

●●●

●

●

●

●

●●●

●

●●

●●

● ●●●●

●●●

●

●●●●●

●

●●

●●●●

●

●●●

●●●

●

●●●

●

●

●●

●

●●●

●●

●

●●

●●●●

●●●●

●●

●

●

●

●●●

● ●

●●● ●●

●

●

●

●

●●

● ●

●●●

●

●

●

●

● ●●

●

●

●●●

●

●●●

●

●

●●●●

●

●

●●●●

●

●●

●●

●●●●

●●●●

●

●

●●●

●

●

●

●

●

●

●

●

●

●●●●●

●●●●

●

●●●

●●●

●

●●●

●

●

●

●

●

●

●●●

●

●●

●

●

●●●

●●●

●

●

●

●

●

●●

●

●

●

●

●

●

●●

●●●●

●

●●●

●

●●●●

●

●

●●●●●

●

●●●

●

●

●

●●●

●

●●

●

●

●

●●●

●●

●

●●●

●

●●●

●●●

●

●

●●●●●●●

●●

●

●●●●●●●●●●

●

●●

●

●●●●●●●●●●●

●●●●

●

●●

●

●●

●●

●

●

●

●●

●●●

●●●

●

●●●

●●● ●

●

●

●●●

●

●●

●●●●●

●

●●

●

●

●

●●●●

●●●●●

●●

●●

●

●●

●●

●

●●●●

● ●●●●●●

●●●●●

●● ●●●

●

●

●●●●

●●

●●

●

●●

●

●

●●●

●

●

●●●

●

●

●

●

●

●

●

●●●

●●●

●●●

●●

●

●

●

●

●●●

●

●●●

●

●

●●●

●

●

●

●●

●

●

●

●

●●

● ●

●

●

●

●

●

●

●●

●

●●

●

●

●

●●

●

●●●

●

●●●

●

●●

●●

●

●

●●●

●

●

●

●●●

●●●●●

●

●

●●

●

●●●

●●●

●

●

●●

●●●●●

●●

●●●●●

●

●

●

●●

●

●●●

●

●●●

●

●

●

●●

●●●

●●

●

●

●

●●●●

●●●●

●

● ●

●

●

●●●

●●

●●●●●●

● ●

●●●

●●●● ●

●●

●

●●●●

●●

●

●

●

●

●●●

●●●●

●

●●●

●●

●●●

●

●

●

●●●

●

●

●

●●●

●

●

●

●●●●●

●●●●

●

●●

●

●

●●

●●●

●●●

●

●

●●●

●

●●

●

●●●

●

●●

●

●

●●●●

●

●

●

●

●●●●●●

●

●●●

●

●●●

●

●

● ●

●●●●

●

●

●●●

●

●●●●

●●●●

●

●

●

●

●

●

●

●

●●

●●●

●

●●●●

●

●

●●●

●

●●●●●●

●●●

●●●●

●

●●●●

●

●

●●●

●●●●

●

●

●

●●

●

●

●●●

●

●●●●

●

●●●●●

●●●●

●

●

●

●●●

●

●●

●●

●

●

●●●●

●

●

●●●●●●

●

●

●

●

●

●

●

●●●●

●

●●

●

●

●●

●

●

●

●

●

●●●

●

●●

●

●

●●●

●●●●

●●

●

●●●

●

●

●●●

●●●

●

●●●●

●

●

●

●●

●

●●

●●

●

●●

●

●

●

●●

●●●●

●●●

●

●

●

●

● ●

●

●●●

●

●

●

●●

●

●

●

●●●

●

●

●

●

●●●

●

●●

●●●●

●

●

●

●●

●●

●●

●●

●

●

●

●●

●● ●●

●

●●

●●

●

●

●●

●

●●

●

●

●●●●

●●

●●●●

●

●

●●●

●

●

●●

●●

●

●

●●●

●

●

●

●

●

●

●

●●●

●●

● ●

●

●●●●●●●●

●

●●●

●

●

●●●

●●●●

●

●

●

●

● ●●

●

●

●

●

●●●●●

●

●●

●

●

●●

●

●

●

●

●●●●●

●●

●

●●

●

●

●

●●

●

●

●●●●

●

●

●●

●

●●

●

●●●

●●●

●

●

●●

●●

●

●

●●●

●

●

●●●

●

●●

●●●

●

●●

●

●●●●

●

●●●

●●●●

●●●●●

●

●●●●

●

●●

●●●●

●●

●

●●

●

●●

●●●

●●●●

●

●

●

●●●

●●●●

●

●

●●●

●

●

●●●●

●●●●

●●●●

●●●●●

●

●●●

●

●●●

●

●

●●●●●●

●●●●

●

●●●●

●

●●●●

●●

●●

●

●

●

●

●

●●●

●

●

●●

●

●

●●● ●●●●

●

●●●●

●

●

●

●

●●●●●●

●

●

●●

●

●

●●

●

●

●

●●●●

●●

●

●

●●●

●

●

●

●

●

●

●●

●●

●

●

●

●

●

●●

●

●

●

●●

●

●

●

●

●●

●

●

●●●●●

●●

●●●

●

●

●●

●●●●●

●

●

●

●●

●

●●●●

●●

●

●

●

●

●●●●●

●

●

●

●●

●

●

●●●

●

●●

●

●

●

●●

●

●

●●●

●

●

●●

●●

●●●●●

●

●●●

●

●

●●●

●●

●●

●

●●●●

●

●●

●●●

●

●●●●●●●

●●●●

●

●●●●

●

●

●●● ●

●

●●●●●

●●

●

●

●●●

●

●●●

●

●

●●●●

●

●●●

●

●●●●

●

●●

●

●●

●

●

●●●

●

●

●●

●

●●●●

●●

●

●

●

●

●

●

●

●●●

●●●

●

●●

●

●●

●

●●●●

●

●●

●●●●●

●

●

●●

●

●

●●●

●

●

●

●

●

●

●●

●

●●●●

●●

●

●●●

●

●

●●●●●

●

●●

●

●

●●●

● ●

●

●●●●●

●

●●

●

●

●

●

●●●

●●

●

●

●●

●

●

●

●●●●●

●●●●

●

●

●

●●●

●

●●●●

●●●●

● ●

●

●

●

●●●●

●●●●

●

●

●

●

●

●●●●●●●

●

●

●

●

●

●

●

●●●●●●

●●●●● ●●

●●

●●●●●●

●

●

●

●

●

●●●

●●

●●

●

●●●

●

●

●●

●

●●●●

●

●●●

●

●

●

●●

●

●

●

●●●●

●

●●

●●●●

●

●

●

●●

●●

●●●

●

●

●

●●●

●

●

●

●

●

●

●●●

●●●

●●

●●

●

●

●

●

●●

●

●

●

●

●

●

●●

●

●

●●

●●

●

●●

●

●

●

●●●

●●●

● ●

●●

●

●

●

●●

●●

●●

●

●

●

●●●

●●

●●

●

●

●

●

●

●

●

●●

●●

●

●●●

●●

●

●●

●

●●

●

●●

●●

●●

●

●

●●●● ●

●

●

●●●

●

●

●

●

●

●●●

●●

●

●

●

●●

●

●●●

●

●

●●

● ●

●●

●●

●●●

● ●●●

●●●

●

●●

●●

●

●

●● ●●

●

●●

●●

●

●●●

●

●●●

●●

●

●●

●

●●

●●

●

●●

●●

●●●

●●●●●

●

●

●

●●

●

●

●●●

●●

●

●

●●

●

●●●●●●

●

●

●

●

●

●●●●●

●

●

●

●

●●

●

●

●

●●

●●●●

●

●●

●

●

●

●●

●

●●

●

●●

●

●

●

●●

●

●●●

●

●●

●

●

●●

●

●

●

●

●

●

●●●

●●●

●●●

●

●●●

●

●

●●●

●

●●●●

● ●●

●●●

●●

●●●●●●

●

●●●●

●

●●●●●

●●●●

●

●

●●●

●●●●●

●

●●

●●●

●

●●●●●

●

●

●●●

●

●●●●

●●●●●●

●

●●●

●●

●

●

●●●

●●●

●

●

●●●●●●●

●●●●

●

●

●

●

●

●●●●●●

●●

●

●●●

●●●●

●

●●

●

●

●●●

●

●●●●●

●

●●●●●

●●

●

●

●●

●●

●

●●●

●

●●

●●

●

●●

●●●

●●●●●

●

●

●●●●●

●●●

●●●●●●

●●●●●

●

●●●●●

●

●

●●●●●●

●

●●●●●●

●

●●●

●●

●

●

●●●

●

●

●

●

●

●

●

●

●

●

●●

●

●●●●

●

●●

●

●

●●

●

●

●●

●

●●

●

●

●●

●●●

●

●●●

●

●●●

●

●●●●

●●

●

●

●

●

●●●

●●

●

●●●

●●●●●

●

●●●●

●●●

●

●●

●●●●●●

●●●●

●

●●

●

●

●

●

●

●●●

●

●●●

●

●

●

●

●

●

●

●●

●

●●●● ●

●●

●●●●

●

●●●

●

●

●●●

●

●●●●●

● ●

●●●

●●

●●●

●

●●●

●

●● ●

●●

●

●

●

●●●

●

●●●●

●●●●

●

●

●

●

●

●

●

●

●

●●●

●●

●

●●

●●●

●

●

●●

●

●●●

●

●●●

●

●●●

●

●●●

●

●

●

●●●

●●●

●

●

●●●●●

●

●●●●

●

●●

●●●●

●●●●●

●●●●●●

●●●

●●

●●●●

●

●

●●●●●

●●●●

●

●

●●●●

●

●

●

●●

●●●●●

●●

●●●

●●●●●●

●

●

●●●●●●

●

●

●●●●

●

●●●

●

●

●

●●●

●●

●●●●

●

●●●

●

●●●●●

●

●● ●

●●●●

●●

●

●

●●

●●●●

●

●●●

●

●●●

●

●●●

●

●

●●

●●●

●

●●●

●

●●●

●

●●

●●

●●

●

●

●

●●

●

●

●●

●●● ●

●

●

●

●

●●

●

●

●●●●

●

●●●●●

●●●

●●●●●●

●

●

●

●●●●

●●●●

●

●●●●●

●

●

●

●●●

●●

●●

●

●●

●

●

●●●●

●●●●

●●

●

●●●●●

●●●

●●●●

●

●●●●●

●●

●●●

●

●

●●●● ●

●

●●

●●●●

●

●●●●

●

●

●

●●

●●●●●

●●●●

●

●

●●●●

●

●

●●

●●●

●●●●●

●

●●

● ●

●●●●

●

●

●

●

●

●●●

●

●

●

●

●

●

●

●

●●●●●

●●

●●

●

●●

●●●

●

●

●

●●●●●

●

●

●●

●

●●

●●

●●●

●

●●

●●●●●

●

●

●●

●

●●●● ●

●●●●

●●

●

●

●●●

●

●●●

●

●●●●

●

●●●●

●●●

●

● ●●●

●

●

●

●●

●

●●

●

●

●●●●

●

●

●●

●

●

● ●●

●●

●●

●

●

●

●

●

●●

●●●●●

●●

●

●●●

●

●

●●●●

●●●●●●●

●

●

●

●●

●

●

●●

●●

●

●●●

●

●●

●

●●●

●●

●●●

●

●●

●

●●●●

●

●●●

●●●

●

●●●

●●

●●

●

●

●●●●●●●

●●●

●●●

●●●●

●●

●

●

●●●

● ●●

●

●

●●●●●

●

●●●●●

●

●

●●●●●

●

●●●●

●●●●

●●●●●

●●●

●●●●

●●●

●●●●

●

●●●●

●

●

●●●●

●●●

●

●●●

●●●●

●●●●●

●●●●

●

●

●

●●●●

●

●

●

●●●

●

●●●●●

●

●●

●●

●

●●●●

●●●●

●

●●●

●

●●●●

●●●●●●●

●●

●●●

●

●

●●●●

●

●●● ●

●

●●●

●●●

●

●●●●

●

●●●

●●●

●●●

●

●●●

●

●

●●●

●

●●●

●●●● ●

●●

●●●●●

●●●

●●●●●

●

●●●●●●●●●

●●●●●

●●●●

●

●●●●

●

●

●●●

●●●

●

●

●●●

●●●●●●●

●●●

●●●

●

●●●

●●●

●●●●

●

●

●

●●

●●●

●

●●●

●●●●

●

●●

●

●

●●

● ●●●

●●●●

●

●

●●●●●

●

●●●

●●●

●●

●

●●●

●

●●●

●● ●●●

●●

●

●

●●●●

●

●●

●●●●

●

●●●

●

●●●●

●●

●

●●●

●

●

●●●

●

●●●●

●

●

●●●●

●

●

●●●●

●●●●●

●

●

●●●

● ●●●●

●

●●●

●

●●●●●●●

●●●●

●

●●

●

●●●

●●●

●●

●

●●●

●

●●●●

●●●

●●●●●●

●

●●●

●

●●●

●

●●●

●

●

●●●●●

●●●●●

●●●

●

●

●

●●●●

●

●●●

●●●

●●●

●

●

●

●●●●

●●●

●

●

●●●

●

●●●●●

●●●●● ●●●

●

●

●●●●

●

●

●

●

●●

●●●

●

●

●

●●●

●

●

●●●● ●●●●●

●

●●●

●●

●●●●●

●●●●●

●

●●●●●

●

●●

●●●●

●

●●●

●

●●●●●

●●●●

●

●●●●

●●●●

●

●●●

●●

●

●●●●●●

●

●

●●

●●●●

●

●●●●●●

●

●●●

●●●

●

●●●

●

●

●

●●●

●

●

●●●●●

●

●●●

●

●●●●

●●●●

●●●

●

●●●●

●

●●●

●

●●●

●

●●●

●

●●

●●●●

●

●●●

●●●●

●

●●●●

●●●●

●

●●●●

●●●●

●●●●●

●

●

●●●●

●

●●●

●●●

●

●●●

●●●●

●

●●●●

●

●

●●●●●●●

●●

●

●●●●

●

●●●

●

●●●

●

●

●●●

●●●

●●●

●

●●●

●●●

●●

●

●●●●●●

●

●

●●

●●●●

●

●

●●●

●●●

●●●

●●

●●●●

●

●●●

●●●

●

●●●

●

●●

●

●●●

●

●

●●●●●

●

●

●

●●●

●●

●●●●

●●●

●

●

●

●●●

●

●●●

●

●●●

●

●●●

●●●●

●

●

●●●

●●●

●●●

●

●●●

●

●●●●

●

●●●

●●

●●●●

●

●●●

● ●●

●

●

●

●●●

●●●●

●

●●●

●●●

●●●

●

●●●●●

●

●●●

●●

●

●●

●

●●●

●

●●

●

●●●●●●●●●

●

●●●

●●

●

●

●●●

●●●●●

●

●●

●

●●●

●●●

●

●

●●●

●●●●

●●●

●

●●●

●●●

●●

●

●●●

●

●●●

●●●

●

●

●●●

●●

●

●●

●

●

●●●

●●●●

●●●●

●●●

●

●

●●●

●●●●

●

●

●

●

●●

●●●●

●●●●

●●●●

●

●

●●●●●

●●

●●●●●

●●●

●

●●

●

●

●●●●

●●●●

●●●●● ●

●●●●

●

●●●●

●

●●●●

●●●● ●●●

●

●●●

●

●●●●

●●●●

●

● ●

●

●●●●●●●

●

●●●●

●

●●●●

●

●

●

●●●

●●●

●

●●●

●

●●●

●●●●

●●● ●

●

●●●

●

●●●

●

●

●●

●●● ●

●●●●

●

●●

●●●

●●●

●

●

●●●●

●

●

●●●●

●●●

●

●●●●

●

●●●

●

●

●●●●

●●●

●

●

●

●●●

●

●

●

●●●

●

●●●●●

●●●

●

●

●●

●●●

●●

●●●

●

●●●

●

●

●●●

●

●

●●

●●●

●●●

●●●

●

●●●

●●●●

●

●

●●●

●

●

●●●●●

●

●●●

●

●●●●

●

●●

●●●

●●●

●

●●●●

●●●

●

●●

●

●●●

●●●●

●

●●●

●●

●

●

●●●●

●●

●

●

●●●

●

●●

●●●●

●

●

●

●

●●●●●

●●●

●

●

●

●●

●

●

●●●

●●●●●

●●●●

●

●●●

●

●

●

●●●●●●●

●●

●

●

●●●

●

●●●●●●

●●●

●

●

●

●●●

●

●●●

●●●●

●●

●

●●

●

●

●●●

●●

●●●

●●●●●

●

●●●●●●●

●

●●●

●●●●

●●●●

●

●

●●●

●●●

●

●

●●●

●●●

●

●

●

●●●

●

●●●●

●●●●

●●●

●●●●●●●

●●●

●●●

●

●●●

●●●●●●●

●●●

●

●●●●●●

●

●●

●

●●●

●

●●●

●

●●●●

●●

●

●●

●●●●

●●●

●

●●

●●●●

●

●

●●●●

●

●

●

●●

●

●

●

●●●●●

●●●

●

●●●

●

●

●●

●●●●

●●●

●●●

●

●●●●

●

●●●

●●●

●●

●●●

●●●

●

●●

●

●●

●

●●●●●

●●●

●●●

●

●

●●●

●

●●

●

●●●

●●

●●●●

●

●●●

●

●●●

●

●●●

●●●●●

●●●●

●●

●

●●●

●

●●●

●●

●●●●

●

●●●

●●●●

●●●

●

●

●●●

●●

●●●

●

●●●●

●●●

●

●●●

●●●

●

●

●●●●●

●

●●●●

●●●

●●●

●

●

●

●

●●●

●●●●

●

●●●

●●●

●●●●●

●●●

●

●●●

●

●●●●●

●●●●

●

●

●●●

●

●●●

●

●●●

●●

●●●

●●

●●

●●

●●

●●●●

●●●

●

●

●●●●

●

●●●●

●●

●●●

●

●●●

●

●●●●●

●

●●●

●●●●

●

●●●

●●●

●●

●●●●

●●●●●

●●●

●

●●●●

●●●●

●

●●●

●●●●●

●●●

●

●

●●●●

●

●

●

●●●●●● ●

●

●

●●●●

●●●

●

●

●●●●

●

●●●●●●

●

●●●

●●●●

●

●●

●

●●●

●

●●●

●●●

●

●●●

●

●●●●●

●●●●

●

●●

●●●

●

●●●●●

●

●

●●

●

●●●●●●●

●●●

●

●●

●●

●●●

●●

●

●

●●●●

●

●●●

●

●●●

●

●●

●●●

●

●●●

●

●●●

●●●

●

●

●

●

●

●●●

●

●

●●●●

●●●

●

●●●●●●●●

●●

●

●

●

●

●

●

●●●

●

●

●●

●

●

●

●●

●

●●

●

●●

●●

●●●

● ●●

● ●●

●

●●●

●

●●●

●●

●●●

●

●●●

●●

●

●

●

●

●

●●●

●●

●

●

●

●

●

●●●●●

●

●

●

●

●●

●●●

●

●

●●

●

●●

●

●●

●●●●

●●

●

●

●

●●●

●

●

●●

●

●

●

●

●

●

●●●

●

●

●

●●●●

●

● ●

●●●●

●●●●●

●

●●

●

●●

●●

●

●

●

●●

●●●

●

●

●

●

●

●

●

●●●●

●

●

●

●

●

●

●

●●●●

●

●●●●●

●●

●

●

●

●●●

●

●●●

●

●●●

●

●

●

●

●●

●

●

●

●●

●●

●

●

●●

●

●●

●●

●

●●●●

●●

●

●

●●

●●●

●

●●

●

●●●●●●●●●

●

●

●●●

●

●●

●

●

●

●●

●

●

●

●

●

●

●●●

●

●

●

●

●●●

●

●

●

●

●

●

●●

●●●●

●●●

●●●●

●

●●●●

●

●

●

●●●●

●

●

●

●

●●●

●

●●

●

●

●●

●

●●

●●●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●●●

●●

●●

●●●●

●●●

●●●●●

●

●●●●

●●●●

●

●

●●●

●

●

●

●●●

●

●

●●

●

●●●

●

●

●●●●

●

●

●●●

●

●

●●

●●●

●

●

●

●

●●●●

●●

●

●

●●●

●●

●

●●

●

●●

●

●●●

●

●

●

●

●

●●

●

●

●

●

●●●

●●●●

●

●●

●●●

●●

●

● ●●●

●

●

●

●

●●●●●●

●

●

●

●●

●

●●●●●●

●●●

●

●

●

●

●

●●

●

●●

●●●●●

●

●●

●●

●

●●

●

●●●●●

●

●●●

●

●●●●●

●

●●●●

●●●●

●

●

●●●

●●

●●●

●

●

●

●●

●

●

●

●

●

●

●●

●

●●●●

●

●●●

●

●●●

●

●

●

●●●●

●

●●●●

●

●●

●

●

●●

●

●

●●●

●●

●

●●●●

●

●●

●

●

●●

●

●

●

●

●

●

●

●●●●●

●

●

●●●●

●

●●

●●●

●

●

●

●

●

●●

●

●

●

●●

●

●

●

●

●

●●

●

●●

●●

●

●

●

●

●●●●●

●

●

●

●

●●●

●●●

●

●

●

●●●●●

●

●

●●●

●

●

●

●●

●

●

●●

●

●●●

●●

●●

●

●

●●

●●●

●

●●

●

●

●●

●

●

●

●

●●

●●

●●●

●

●●●

●●

●

●

●

●●

●

●●●

●●●●

●

●●

●●

●

●

●

●

●●

●

●

●

●●

●

●

●

●●●● ●●●

●

●

●●●●

●

●

●●●●

●●●

●

●●

●

●

●

●●

●

●

●●●●

●

●●●●

●

●●

●

●

●●●

●●●

●●

●●

●

●

●

●

●

●

●●●

●●●

●●●●

●●●

●

●●●

●●

●●●●

●●

●

●●●●

●●

●●●

●

●●

●●●●

●

●●

●

●

●●●

●●

●●●

●

●●●●

●

●●●●

●

●●

●●

●●●

●

●

●

●●

●

●

●

●

●

● ●

●

●

●● ●

●●●

●●

●●●●

●●●●●

●●

●●●

●

●

●●

●●

●

●●●

●

●

●

●

●●

●

●

●

●●

●

●●●●●

●

●●●●

●

●●●

●

●●

●

●●●●●

●

●

●

●

●●●

●●

●

●

●●●●

●●●

●

●●●

●●●

●

●

●

●

●●● ●

●●

●●●

●●

●

●●

●

●●●

●●

●●●

●●●

●

●

●

●●

●●

●

●●

●●●●

●●

●

●

●

●

●●

●

●

●

●●●

●

●

●●●

●

●●●

●●

●

●

●

●●●●

●

●

●

●

●

●

●●●

●

●

●

●

●

●

●●

●●

●

●

●

●

●

●

●

●●●●

●

●●

●●

●

●

●●

●

●

●●

●●

●

●

●

●

●

●●

●●

●●

●

●

●

●

●●●

●

●

●

●●

●

●

●●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●●

●●

●●

●●●

●

●●

●●

●●●

●

●●●

●

●●

●

●●

●

●

●●

●●●

●

●

●●●●●●

●●

●●●

●● ●

●

●●●

●

●●

●

●

●

●

●

●

●

●

●

●●●●

●●●

●●

●

●●

●

●●●

●

●●

●

●

●●●

●●

●

●●

●●●

●

●

●

●●

●

●

●

●

●●●

●

●

●●

●●●

●

●●●

●

●

●

●

●

●

●

●

●

●

●●●●●

●

●

●●

●

●●

●

●

●●

●●●

●

●●●●●●

●

●

●

●●

●

●

●

●●●

●

●●●

●●●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●●●

●

●

●

●●

●●●●●

●●●●

●

●●

●●●●

●●●

●

●●●

●●

●

●

●

●

●

●●

●

●

●

●

●

●●●

●

●

●

● ●

●●●●

●

●●

●

●●●

●

●

●●●

●●

●●

●●

●●

●

●

●

●

●●

●●●

●

●●

●

●

●●●

●●

●

●

●

●●

●

●

●

●

●

●

●

●●●●

●●

●●●

●●

●

● ●●

●●

●●

●●●

●

●●

●●

●

●

●

●

●

●

●

●●

●

●

●

●

●●

●

●

●

●●

●

●

●

●●

●

●● ●●

●●●●

●

●●●

●●●

●●

●●●●

●

●●

●

●●●●

●

●●●

●

●

●●

●

●●●

●

●

●

●

●●

●●●●

●● ●

● ●

●●●

●

●●●

●

●

●

●

●

●●

●●●

●

●

●

●

●●●●

● ●

●●●

●

●●●●

●

●

●

● ●

●

●

●

●

●●●

●

●

●●

●

● ●

●

●●

●

●●●

●

●●●●

●

●●

●

●

●

●

●

●

●

●●●

●

●●●

●

●

●

●●●●

●●●●

●●

●

●●●●●

●●●

●●●●

●●

●

●●●

●

●●

●

●●●

●

●●●

●●

●

●

●

●●

●

●●●

●

●

●

●●●●

●●●●●

●

●

●

●

●

●

●

●●●

●

●

●

●

●

●●

●

●●

●

●●●●●

●

●●

●●●●

●

●●

●●●

●● ●●

●

●●●●

●

●

●

●

●

●●

●●

●

●●●●

●

●

●

●●

●

●●

●●

●

●

●

●

●●

●

●●●●

●●

●

●

●

●●

●

●●

●●

●

●

●●

●

●

●

●

●●

●

●

●

●

●●

●●●

●

●●

●

●

●●

●●

●● ●

●●

●

●

●

●●

●

●

●

●

●

●

●●●

●

●

●●●

●

●

●●

●

●●

●●●●

●

●

●●

●

●

●

●

●

●●

●

●

●

●●

●

●

●

●●●

●

●

●●●

●

● ●

●●●

●

●

●●●

●

●

●●●●

●●

●

●

●

●

●

●

●

●

●●

●

●

●

●●

●

●

●

●

●

●

●

●●

●●●

●●● ●

●

●●●

●

●

●●

●

●

●

●

●●

●

●

●

●

●

●●●●

●

●

●

●

●

●

●●●

●

●

●

●

●●

●

●

●●●

●●

●●

●

●

●

●

●●

●

●

●●●

●

●

●

●

●

●●

●

●

●

●

●●

●●

●

●●

●●●

●●●

●●

●

●●●●●●●

●

●

●

●

●

●●●●●

●

●

●

●

●

●

●

●

●●

●

●

●●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●●●

●

●

●

●●

●●●●

●

●

●●

●

●●●●●

●

●●

●

●

●

●●●

●●

●●●●●

●●●●●●●●●

●

●

●●

●●

●●●●●

●●

●

●

●●●

●

●● ●

●

●

●

● ●

●

●

● ●

●

●

●

●●●

●

●●●

●

●

●●

●●●

●

●

●

●●●●

●

●

●●●●●

●●●●

●

●●●

●●

●

●

●●●●●

●

●

●

●●

●●

●

●●●●

●●●

●●●

●

●

●●

●●

●

●

●

●

●

●

●

●

●●●

●

●●

●

●●

●

●

●

●

●●●

●●

●

●

●

●

●

●●●●

●●

●

●●

●

●●

●

●

●

●

●

●●

●

●

●

●●

●

●●

●

●

●

●

●●●●

●

●●

●

●

●

●

●●●●

●

●

●

●●

●●●●

●

●

●

●

●●

●●

●●

●●●●

●

●

●●

●●

●

●

●

●

●

●

●●●

●●

●●

●

●●

●●

●

●

●

●

●

●

●

●

●

●

●●●

●●●● ●●●●

●

●

●●●

●●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●●

●● ●

●●●

●

●●●

●

●●

●●

●

●

●●

●●

●●

●●

●●●

●●●

●

●●

●

●

●

●●●●●

●

●

●

●

● ●

●

●

●

●

●●●

●

●

●

●

●

●

●

●

●●

●●●●●

●

●

●

● ●

●

●●●

●

●

●

●●

●●

●

●

●

●

●●

●

●●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●●

●●

●

●

●

●●

●

●●

● ●

●●●●

●●

●●

●●

●●●

●

●

●

●

●

●

●●

●

●●●●

●

●

●

●

●

●

●●●●●

●

●

●

●

●

●

●●●

●

●

●

●

●●

●

●

●●●●

●

●●

●

●

●●

●

●●●

●

● ●

●

●

●

●●●

●

●

●

●

●

●●

●

●

●

●●

●

●●

●●

●

●

●

●●●●●

●

●

●

●

●

●

●

●

●●●

●

●●●●●

●●●●

●

●

●●

●

●●●

●

●●

●

●

●

●

●

●

●

●●

●

●●●

●●●●

●●

●

●●

●

●●

●

●

●

●

●

●● ●●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●●●

●

●

●

●

●

●

●

●

●

●●●

●

●●●

●●

●●●

●●

●●

●

●●

●

●

●●

●

●

●

●●●●●●

●

●

●

●●●●

● ●

●

●●●

●

●

●

●

●

●

●●

●

●

●●●●

●●

●

●

●

●

●

●

●

●●

●●

●

●

●

●

●● ●

●

●

●

●●

●

●●●●

●

●

●

●●

●●

●

●

●●

●

●

●

●

●●●

●

●

●●

●

●●●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●●● ●●●●●

●

●

●●●●●

●●●●

●●●●●

●●●●●●●●

●●

●

●

●

●●

●

●●●

●

●●●

●

●

●

●●

●

●●

●●

●

●

●●●●

●

●●

●

●

●

●●●●●

●

●● ●

●

●

●

●

●

●●

●

●●

●●

●

●

●●

●●

●

●●●●●

●●●

●

●●

●

●

●

●

●

●●

●

●

●●●●

●●

●

●

●●

●●

●●●

●●

●●

●

●

●

●

●

● ●●

●

●

●●

●

●

●

●●

●●

●

●

●

●

●●

●●

●

●

●●

●

●●●

●

●●● ●

●

●

●

●

●●●●●●●

●

●

●

●

●●●●

●

●●

●

●

●●●

●

●●●●

●

●

●●

●●

●

●

●

●●

●

●●

●

●

●

●●

●

●

●

●

●

●

●

●●

●

●●

●

●

●●●●●●

●

●

●

●●●

●●

●

●

●

●●

●●

●

●●

●●

●●●●

●●

●

●●

●

●●●●

●

●●●●

●

●●

●

●

●●

●●

●

●●

●●

●

●●

●●

●●

●●

●

●

●

●

●●●

●

●

●

●

●

● ●

●

●●

●

●●

●●

●

●

●●●●

●

●

●

● ●

●●

●

●●●●●

●●●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●●●

●

●

●●●

●

●

●

●

●

●●

●

●

●

●

●

●

●●

●●●

●

●●●●

●●

●

●●●●●

●

●

●

●

●●

●●●

●

●●

●

●

●

●

●●●

●

●

●

●

●

●

●

●

●

●

●

●

●●●

●

●●●●

●

●

●●

●

●

● ●

●

●

●

●

●●

●

●

●

●●●●●●

●●●●

●●●●

●●●

●

●●●

●

●

●

●

●

●●●

●

●●●

●

●●

●

●

●●

●

●

●

●

●

●●

●●●

●

●●

●●●●

●

●●●

●

●

●

●●●

●●

●●●●

●

●

●●●

●

●●

●●

●●

●

●●

●

●●●

●●●

●●

●●

●●●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●●

●●

●

●●

●

●

●

●

●

●

●●

●

●●

●

●●

●

●

●●

● ●

●

●

●

●●

●

●

●●●●● ●●

●

●

●●●

●●●●

●

●●●

●●

●

●

●

●

●●●●●

●

●

●

●

●● ●

●

●

●

●

●

●

● ●

●

●●

●●●●

●

●

●

●

●

●●●●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●●

●●

●

●

●

●●

●●

●●●●

●

●

●●● ●●●●

●

●

●

●

●●●●

●

●

●●●

●

●

●

●

●

●

●

●●

●●●●

●

●

●●

●●

●●

●

●●●●

●

●●

●●

●

●

●

●●

●

●●

●

●

●

●

● ●

●

●●

●●●●

●

●

●●●

●●●

●

●

●

●●

●●

●

●

●

●

●

●●●

●

●

●

●

●

●

●

●

●

●

●●

●

●●

●

●

●

●

●

●

●

●

●●●

●

●

●●●●

●

●●

●

●

●

●●

●

●

●

●

●●

●

●

●

●

●

●●

●

●

●●

●

●

●

●●

●

●

●

●●

●

● ●●

●

●

●

●●

●●●

●●

●

●●●●●●

●

●●

●

●●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●●

●

●

●●

●

●

●

●

●

●●

●

●●●●●●

●●●

●

●

●●

●

●

●

●

●

●

●

●

●●

●

●●●

●●●●

●

●●

●●●

●

●

●

●

●

●●

●

●

●●●●●

●

●

●●●●

●

●●●

●

●

●

●

● ●

●●

●●●●●

●●

●●

●

●

●

●

●

●

●●●●

●

●

●

●●●●

●

●●●●●

●●

●

●

●

●

●●●

●

●

●

●

●

●●

●●

●●●

●●

●●●●

●

●

●

●

●

●

●

●

●●

●

●●

●

●

●

●●●

●

●●

●

●

●

●

●●

●● ●

●

●

●●

●

●

●

●●

●

●

●

●

●

●

●●

●

●

●●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●●●●●●●●

●

●

●

●

●

●

●

●

●

●●

●●●●

●

●●●●

●

●

●

●●●●●

●

●

●●

●

●●●●

●

●●

●

●

●

●

●

●●

●●

●●

●●

●

●

●

●

●

●●

●●

●

●

●

●

●●●

●

●

●

●

●

●

●

●

●

●

●●●

●

●

●

●

●

●●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●●●

●

●

●

●

●

●

●

●

●●

●

●

●●

●

●

●

●

●

●●

●

●

●

●●●

●

●

●

●●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●●●●

●●●

●●●

●

●

●

●

●

●

●●●

●

●

●

●●

●●

●

●

●

●

●

●

●

●

●

●

●

●

● ●

●

●

●

●

●●

●●●

●

●

●

●●●

●

●

●●●

●●●●●

●

●●●

●

●

●

●●

●

● ●

●●

●●

●●●●

●●●

●

●

●

●●

● ●

●

●

●

●

●●

●

●

●

●●

●

●

●

●●●●●

●

●●

●●

●●

●

●

●●●

●●

●●●

●●●●

●

●●

●●

●

●●●

●

●

●

●

●

●●

●

●

●

●●●●

●

●

●● ●

●

●

●

●

●●

●●

●

●

●

●●●

●●

●

●●●

●●

●

●

●

●

●

●

●●

●●

●

●

●

●

●●●

●

●●

●

●

●●

●●

●

●

●

●●

●

●●●

●

●

●

●●

● ●

●

●

●

●

●

●

●

●●

●

●

●

●

●●

●

●●

●

●●

●●●

●

●●●

●

●

●●

●

●

●

●

●

●

●

●●●

●

●

●

●●

●

●

●

●

●

●●

●

●●

●●

●

●

●●

●

●

●

●

●

●

●

●

●●

●

●●●●●

●

●

●●●

●

●●

●

●

●●●

●●●●

●

●

●

●

●●●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●●

●

●●

●

●●

●

●

●●

●

●●

●

●●

●●●

●

●

●

●

●●

●

●

●

●

●●●●●

●

●●

●

●●●

●

●

●

●

●

●●

●

●●●

●

●

●●

●●●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

● ●

●

●●

●

●

●

●

●

●

●

●●●

●●●●

●●●●●

●●●●

●●●●●

●

●

●

●●

●

●

●

●

●

●●

●

●

●

●●

●

●

●●

●

●

●

●

●

●

●

●

●

● ●

●

●●●●●

●●

●

●

●

●●●●

●●●

●

●●

●●

●

●

●

●

●●

●

●●

●

●

●●●

●

●

●

●

●●●

●

●

●

●

●

●●●●

●●●●

●

●●

●

●

●

●

●

●

●●

●

●●

●

●

●●●

●

●

●

●

●

●

●●●●

●●●

●

●●●

●

●●

●

●●

●

●●●●

●

●

●

●●

●●

●

●

●

●

●●

●

●

●

● ●●

●

●

●

●

● ●●●

●●●● ●

●

●

●●●

●

●

●●

●

●

●

●

●●●●●

●●●

●

● ●

●

●

●

●

●

●

●

●

●

●●●●

●

●

●

●

●●●●

●

●

●

●

●

●

●

●

●

●

●

●

●

● ●

●●●

●

●●

●●●

●

●

●

●

●

●

●

●●●

●

●●

●

●●

●●●●

●

●

●●●

●

●●

●●●

●

●●●

●

●

●

●

●

●

●

●

●

●

●●●

●

●

●●

●●

● ●

●

●

●●

●

●

●

●●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●●●●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●●●●

●

●

●

●

●

●

●●

●

●

●

●

●●

●●

●

●

●

●●●●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●●

●●●●●

●

●

●●●

●

●

●●

●●●

●●

●

●●

●●●●●●

●

●●●

●

●

●

●●●●

●

●●

●

●

●●●

●

●

●

●

●

●

●

●

●

●

●

● ●

●

●●●

●●●

●

●●●●

●

●●

●

●

●

●●●

●

●●

●●

●

●●●

●

●

●●

●●

●

●●●

●

●

●

●

●

●●●

●

●●●

●

●

●●

●

●

●

●

●

●●●●

●●

●

●

●

●

●●

●

●●●

●

●

●

●●

●●●

●

●●

●●

●●●

●●●●

●●●

●

●

●●

●

●

●

●

●●●●●

●●

●

●

● ●

●

●●●

●

●

●●●●●●●●●●●

●

●

●●●

●

●●

●

●

●●●●

●

●

●

●●

●●●●

●

●●

●●●

●●

●●

●

●●●●●

●

●

●

●

●●

●●

●●

●●●●

●●

●

●●

●

●

●●●

●

●●●

●●

●●●

●

●

●● ●

●

●●

●●

●●

●

●●●

●●

●

●

●

●●

●

●●●

●●

●

●

●●

●

●

●

●●

●

●

●

●

●●●

●

●

●

●

●●

●●●●

●

●●●

●

●

●●●●

●

●

●

●●

●●●

●

●

●

●

●

●●●

●●

●

●

●

●●

●●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●●

●●●●

●

●●

●

30 40 50 60 70

3040

5060

70

Predicted status based on Father's socio−economic background

Pre

dict

ed s

tatu

s ba

sed

on F

athe

r's a

nd M

othe

r's b

ackg

roun

d Correlation: 0.885

Figure A1: Predictions based on both parents’ vs. fathers’ information

Appendix B

32

childloc2

childloc3

german

female

sameloc1

childloc1

fgerman

sameloc2

fybirth

fedu

ybirth

fisei

0 20 40 60 80

%IncMSE

german

sameloc2

fgerman

childloc2

childloc1

childloc3

sameloc1

female

fedu

fybirth

ybirth

fisei

0 50000 150000

IncNodePurity

Random forest variable importance plot

Figure A2: Random Forest Variable Importance

1

2

3

4

5

6

7

8

9

10

1 2 3 4 5 6 7 8 9 10Intergenerational Reference Point (Father)

Rea

lized

Sta

tus

(Chi

ld)

(a) Heatplot Status Discordance (Decile)

0

10

20

30

40

50

60

70

80

90

100

0 10 20 30 40 50 60 70 80 90 100Intergenerational Reference Point (Father)

Rea

lized

Sta

tus

(Chi

ld)

(b) Jitterplot Status Discordance (Per-centile)

Figure A3: Empirical Distribution of Status Discordance in Germany

33

Appendix C

0.0

0.1

0.2

0.3

0 25 50 75 100

Intergenerational Status Discordance (Percentiles)

Pr(

Abste

ntion)

(a) Abstention, LOESS

0.0

0.1

0.2

0.3

0 25 50 75 100

Intergenerational Status Discordance (Percentiles)

Vote

Pr(

Anti−

Esta

blis

hm

ent)

(b) Anti-Establishment Voting, LOESS

Figure A4: Intergenerational Status Discordance (Percentiles) and Political Alienation. Loesssmoothed.

0.0

0.1

0.2

0.3

2.5 5.0 7.5 10.0

Intergenerational Status Discordance (Percentiles)

Pr(

Abste

ntion)

(a) Absention, ISD Deciles

0.0

0.1

0.2

0.3

1 2 3 4 5 6 7 8 9 10

Intergenerational Status Discordance (Deciles)

Vote

Pr(

Anti−

Esta

blis

hm

ent)

(b) Anti-Establishment Voting, ISDDeciles

Figure A5: Intergenerational Status Discordance (Deciles) and Political Alienation.

34

0.0

0.1

0.2

0.3

0 25 50 75 100

Intergenerational Status Discordance

Vote

Pr(

Anti−

Esta

blis

hm

ent)

(a) Voted Radical Party in 2017

0.0

0.1

0.2

0.3

0 25 50 75 100

Intergenerational Status Discordance

Pr(

Abste

ntion)

(b) Abstention

0.0

0.1

0.2

0.3

0 25 50 75 100

Intergenerational Status Discordance

Vote

Pr(

AfD

)

(c) Voted AfD in 2017

Figure A6: Intergenerational Status Discordance and Political Alienation.

35

Appendix D

Figure A7: Standardized Coe�cients Radical Right Vote

Education

Gender: Female

Age

Income

Status Discordance

−10 0

Standardized Coefficient: Vote for AfD

36

TableA1:

Inte

rgen

erat

iona

lSta

tusD

isco

rdan

cean

dPo

litic

alA

liena

tion,

East

Ger

man

y

Abs

tain

Vote

Radi

cal

No

Part

yID

Vote

Mai

nstr

eam

1Vo

teM

ains

trea

m2

ISD

(Pct

iles)

0.07

5∗0.

099∗

0.11

7∗∗−

0.13

0∗∗−

0.14

5∗∗∗

(0.0

31)

(0.0

42)

(0.0

40)

(0.0

45)

(0.0

44)

Fem

ale

(1=y

es)

3.78

5∗−

2.86

26.

769∗∗

∗0.

804

1.61

7(1

.523

)(2

.063

)(2

.000

)(2

.226

)(2

.179

)A

ge−

0.44

2∗∗∗

−0.

065

−0.

365∗∗

∗0.

713∗∗

∗0.

614∗∗

∗

(0.0

71)

(0.0

97)

(0.0

93)

(0.1

04)

(0.1

02)

Educ

:Upp

erEl

emen

tary

18.0

795.

069

8.56

3−

24.8

11−

24.0

72(1

0.91

5)(1

4.78

4)(1

5.01

6)(1

5.94

6)(1

5.61

0)Ed

uc:L

ower

Seco

nd.

−2.

854

9.38

92.

769

−12

.862

−10

.592

(10.

621)

(14.

386)

(14.

643)

(15.

517)

(15.

189)

Educ

:Sec

onda

ry−

8.42

3−

2.56

8−

6.76

1−

3.90

68.

723

(11.

285)

(15.

285)

(15.

520)

(16.

486)

(16.

138)

Educ

:Uni

vers

ityPr

ep.

−17

.311

−0.

541

−13

.890

0.20

09.

803

(10.

764)

(14.

580)

(14.

820)

(15.

726)

(15.

394)

Educ

:Ter

tiary

I−

12.1

978.

067

−6.

068

−11

.535

0.00

7(1

0.89

1)(1

4.75

1)(1

4.98

2)(1

5.91

0)(1

5.57

5)Ed

uc:T

ertia

ryII

−14

.657

5.46

7−

17.9

34−

10.2

128.

450

(10.

840)

(14.

683)

(14.

913)

(15.

836)

(15.

502)

Inco

me

(log)

−3.

800∗∗

−3.

294

−2.

081

8.61

1∗∗∗

8.75

1∗∗∗

(1.2

76)

(1.7

29)

(1.6

62)

(1.8

65)

(1.8

25)

Inte

rcep

t66

.712

∗∗∗

43.5

20∗

96.6

75∗∗

∗−

36.1

14−

33.2

43(1

4.86

6)(2

0.13

5)(1

9.97

8)(2

1.71

8)(2

1.25

9)R2

0.12

90.

025

0.07

40.

073

0.12

1A

dj.R

20.

123

0.01

90.

068

0.06

70.

115

Num

.obs

.20

5920

5923

0120

5920

59∗∗

∗p

<0.

001;

∗∗p

<0.

01;∗

p<

0.05

.All

mod

elsi

nclu

dere

gion

al(B

unde

sland

)�xe

de�

ects

.

37

TableA2:

Inte

rgen

erat

iona

lSta

tusD

isco

rdan

cean

dPa

rty

Choi

ce,E

astG

erm

any

rraf

dcd

ucsu

msr

msl

gree

nsp

dlin

keIS

D(P

ctile

s)0.

089∗∗

0.07

6∗−

0.12

4∗∗−

0.11

3∗∗−

0.02

6−

0.01

5−

0.01

10.

013

(0.0

31)

(0.0

31)

(0.0

39)

(0.0

42)

(0.0

37)

(0.0

24)

(0.0

31)

(0.0

32)

Fem

ale

(1=y

es)

−2.

656−

2.91

22.

265

0.29

60.

531

0.81

3−

0.28

3−

0.11

7(1

.544

)(1

.533

)(1

.949

)(2

.098

)(1

.830

)(1

.183

)(1

.550

)(1

.614

)A

ge−

0.14

5∗−

0.12

40.

433∗∗

∗0.

443∗∗

∗0.

168

−0.

099

0.26

7∗∗∗

0.07

1(0

.072

)(0

.072

)(0

.091

)(0

.098

)(0

.086

)(0

.055

)(0

.073

)(0

.076

)Ed

uc:U

pper

Elem

enta

ry5.

932

5.23

07.

048

9.60

3−

33.5

05∗

0.73

9−

34.2

44∗∗

−0.

843

(11.

066)

(10.

985)

(13.

964)

(15.

034)

(13.

114)

(8.4

74)

(11.

103)

(11.

561)

Educ

:Low

erSe

cond

.−

1.41

7−

2.09

119

.174

22.8

29−

32.9

55∗∗

2.27

0−

35.2

25∗∗

11.0

05(1

0.76

8)(1

0.68

9)(1

3.58

8)(1

4.62

9)(1

2.76

0)(8

.246

)(1

0.80

4)(1

1.25

0)Ed

uc:S

econ

dary

−9.

530−

9.61

833

.044

∗33

.159

∗−

24.1

6812

.629

−36

.797

∗∗7.

036

(11.

440)

(11.

357)

(14.

437)

(15.

542)

(13.

557)

(8.7

61)

(11.

479)

(11.

952)

Educ

:Uni

vers

ityPr

ep.

−13

.378−

13.5

3323

.687

36.4

93∗

−24

.330

9.60

3−

33.9

32∗∗

12.8

83(1

0.91

3)(1

0.83

3)(1

3.77

1)(1

4.82

5)(1

2.93

2)(8

.357

)(1

0.94

9)(1

1.40

1)Ed

uc:T

ertia

ryI

−12

.810−

13.1

9022

.476

25.3

79−

25.0

5011

.543

−36

.593

∗∗∗

21.0

20(1

1.04

1)(1

0.96

0)(1

3.93

3)(1

5.00

0)(1

3.08

4)(8

.455

)(1

1.07

8)(1

1.53

5)Ed

uc:T

ertia

ryII

−13

.965−

14.4

6717

.485

22.4

82−

13.7

0518

.662

∗−

32.3

67∗∗

19.5

91(1

0.99

0)(1

0.91

0)(1

3.86

8)(1

4.93

0)(1

3.02

3)(8

.416

)(1

1.02

6)(1

1.48

1)In

com

e(lo

g)0.

566

0.44

04.

831∗∗

7.12

9∗∗∗

1.83

20.

140

1.69

1−

3.86

8∗∗

(1.2

94)

(1.2

85)

(1.6

33)

(1.7

58)

(1.5

33)

(0.9

91)

(1.2

98)

(1.3

52)

Inte

rcep

t16

.487

17.8

01−

50.9

76∗∗−

66.2

69∗∗

32.3

662.

870

29.4

9627

.174

(15.

071)

(14.

961)

(19.

018)

(20.

475)

(17.

860)

(11.

542)

(15.

121)

(15.

745)

R20.

067

0.06

40.

048

0.05

80.

061

0.07

10.

038

0.02

5A

dj.R

20.

061

0.05

80.

041

0.05

20.

055

0.06

50.

031

0.01

9N

um.o

bs.

2059

2059

2059

2059

2059

2059

2059

2059

∗∗∗p

<0.

001;

∗∗p

<0.

01;∗

p<

0.05

.All

mod

elsi

nclu

dere

gion

al(B

unde

sland

)�xe

de�

ects

.

38

TableA3:

Inte

rgen

erat

iona

lSta

tusD

isco

rdan

cean

dPo

litic

alA

liena

tion,

Wes

tGer

man

y

Abs

tain

Vote

Radi

cal

No

Part

yID

Vote

Mai

nstr

eam

1Vo

teM

ains

trea

m2

ISD

(Pct

iles)

0.08

1∗∗∗

0.05

1∗∗∗

0.09

0∗∗∗

−0.

110∗∗

∗−

0.13

6∗∗∗

(0.0

13)

(0.0

14)

(0.0

20)

(0.0

21)

(0.0

18)

Fem

ale

(1=y

es)

−1.

678∗

−3.

282∗∗

∗7.

537∗∗

∗1.

490

5.22

7∗∗∗

(0.6

91)

(0.7

65)

(1.1

18)

(1.1

73)

(1.0

06)

Age

−0.

390∗∗

∗−

0.00

9−

0.70

1∗∗∗

0.39

9∗∗∗

0.50

6∗∗∗

(0.0

29)

(0.0

32)

(0.0

47)

(0.0

49)

(0.0

42)

Educ

:Upp

erEl

emen

tary

−11

.922

∗∗−

4.70

1−

3.52

220

.090

∗∗16

.517

∗∗

(3.9

48)

(4.3

69)

(6.2

19)

(6.7

00)

(5.7

42)

Educ

:Low

erSe

cond

.−

19.5

15∗∗

∗−

9.54

5∗−

12.2

19∗

28.1

35∗∗

∗29

.019

∗∗∗

(3.9

24)

(4.3

42)

(6.1

82)

(6.6

60)

(5.7

08)

Educ

:Sec

onda

ry−

27.8

44∗∗

∗−

9.72

6∗−

21.0

86∗∗

∗25

.409

∗∗∗

35.1

41∗∗

∗

(4.0

44)

(4.4

75)

(6.3

83)

(6.8

64)

(5.8

83)

Educ

:Uni

vers

ityPr

ep.

−28

.695

∗∗∗−

10.5

56∗

−25

.836

∗∗∗

25.8

67∗∗

∗37

.478

∗∗∗

(3.9

90)

(4.4

15)

(6.2

93)

(6.7

71)

(5.8

04)

Educ

:Ter

tiary

I−

27.0

18∗∗

∗−

9.21

7∗−

23.6

22∗∗

∗23

.363

∗∗∗

35.5

42∗∗

∗

(4.0

11)

(4.4

38)

(6.3

33)

(6.8

07)

(5.8

34)

Educ

:Ter

tiary

II−

26.1

67∗∗

∗−

10.9

70∗

−26

.661

∗∗∗

18.4

72∗∗

36.9

12∗∗

∗

(3.9

77)

(4.4

01)

(6.2

65)

(6.7

50)

(5.7

85)

Inco

me

(log)

−2.

283∗∗

∗−

1.56

7∗∗−

0.09

44.

115∗∗

∗4.

222∗∗

∗

(0.4

79)

(0.5

30)

(0.7

75)

(0.8

14)

(0.6

97)

Inte

rcep

t64

.237

∗∗∗

31.1

57∗∗

∗88

.957

∗∗∗

−8.

999

−4.

169

(5.5

73)

(6.1

67)

(8.9

14)

(9.4

58)

(8.1

07)

R20.

089

0.01

90.

069

0.03

80.

091

Adj

.R2

0.08

70.

017

0.06

70.

036

0.08

9N

um.o

bs.

8784

8784

9818

8784

8784

∗∗∗p

<0.

001;

∗∗p

<0.

01;∗

p<

0.05

.All

mod

elsi

nclu

dere

gion

al(B

unde

sland

)�xe

de�

ects

.

39

TableA4:

Inte

rgen

erat

iona

lSta

tusD

isco

rdan

cean

dPa

rty

Choi

ce,W

estG

erm

any

rraf

dcd

ucsu

msr

msl

gree

nsp

dlin

keIS

D(P

ctile

s)0.

050∗∗

∗0.

048∗∗

∗−

0.07

9∗∗∗−

0.13

3∗∗∗−

0.00

5−

0.02

60.

020

0.00

1(0

.010

)(0

.010

)(0

.020

)(0

.021

)(0

.021

)(0

.015

)(0

.018

)(0

.010

)Fe

mal

e(1

=yes

)−

2.73

4∗∗∗−

2.70

1∗∗∗

4.01

4∗∗∗

1.75

84.

098∗∗

∗3.

738∗∗

∗0.

072

−0.

548

(0.5

68)

(0.5

66)

(1.1

16)

(1.1

73)

(1.1

50)

(0.8

50)

(0.9

78)

(0.5

43)

Age

−0.

029

−0.

025

0.23

8∗∗∗

0.14

2∗∗0.

339∗∗

∗0.

107∗∗

0.24

4∗∗∗

0.02

0(0

.024

)(0

.024

)(0

.047

)(0

.049

)(0

.048

)(0

.036

)(0

.041

)(0

.023

)Ed

uc:U

pper

Elem

enta

ry−

8.79

2∗∗−

8.98

3∗∗6.

567

9.85

16.

893

−3.

573

10.4

444.

091

(3.2

43)

(3.2

30)

(6.3

73)

(6.7

01)

(6.5

66)

(4.8

52)

(5.5

87)

(3.0

99)

Educ

:Low

erSe

cond

.−

14.3

19∗∗

∗−

14.4

00∗∗

∗11

.845

17.5

41∗∗

11.5

750.

884

10.7

154.

774

(3.2

23)

(3.2

11)

(6.3

34)

(6.6

61)

(6.5

27)

(4.8

23)

(5.5

53)

(3.0

80)

Educ

:Sec

onda

ry−

14.5

66∗∗

∗−

14.5

49∗∗

∗13

.551

∗14

.912

∗20

.019

∗∗9.

733

10.3

794.

841

(3.3

22)

(3.3

09)

(6.5

29)

(6.8

65)

(6.7

27)

(4.9

70)

(5.7

24)

(3.1

75)

Educ

:Uni

vers

ityPr

ep.

−18

.024

∗∗∗−

18.0

38∗∗

∗11

.740

17.4

97∗∗

21.2

23∗∗

11.6

11∗

9.13

87.

468∗

(3.2

77)

(3.2

64)

(6.4

40)

(6.7

72)

(6.6

36)

(4.9

03)

(5.6

46)

(3.1

32)

Educ

:Ter

tiary

I−

17.1

19∗∗

∗−

17.1

31∗∗

∗13

.420

∗18

.432

∗∗17

.914

∗∗12

.179

∗5.

587

7.90

2∗

(3.2

94)

(3.2

82)

(6.4

75)

(6.8

08)

(6.6

71)

(4.9

29)

(5.6

76)

(3.1

48)

Educ

:Ter

tiary

II−

18.1

83∗∗

∗−

18.2

01∗∗

∗4.

244

11.0

6226

.502

∗∗∗

18.4

40∗∗

∗7.

660

7.21

2∗

(3.2

67)

(3.2

54)

(6.4

20)

(6.7

51)

(6.6

15)

(4.8

88)

(5.6

29)

(3.1

22)

Inco

me

(log)

0.21

20.

154

3.18

7∗∗∗

3.53

0∗∗∗

0.99

40.

107

0.68

9−

1.78

0∗∗∗

(0.3

94)

(0.3

92)

(0.7

74)

(0.8

14)

(0.7

97)

(0.5

89)

(0.6

78)

(0.3

76)

Inte

rcep

t19

.146

∗∗∗

19.3

53∗∗

∗−

16.9

98−

9.52

13.

623

4.83

10.

153

12.0

12∗∗

(4.5

78)

(4.5

60)

(8.9

97)

(9.4

60)

(9.2

70)

(6.8

49)

(7.8

87)

(4.3

75)

R20.

038

0.03

60.

023

0.02

50.

042

0.07

40.

027

0.01

7A

dj.R

20.

035

0.03

40.

021

0.02

30.

040

0.07

10.

025

0.01

5N

um.o

bs.

8784

8784

8784

8784

8784

8784

8784

8784

∗∗∗p

<0.

001;

∗∗p

<0.

01;∗

p<

0.05

.All

mod

elsi

nclu

dere

gion

al(B

unde

sland

)�xe

de�

ects

.

40

TableA5:

Inte

rgen

erat

iona

lSta

tusD

isco

rdan

cean

dPa

rty

Iden

ti�ca

tion

rraf

dcd

ucsu

msr

msl

gree

nsp

dlin

keIS

D(P

ctile

s)0.

068∗∗

∗0.

063∗∗

∗−

0.11

0∗∗∗−

0.12

0∗∗∗

0.01

60.

018

0.00

80.

036∗

(0.0

14)

(0.0

13)

(0.0

26)

(0.0

27)

(0.0

27)

(0.0

22)

(0.0

23)

(0.0

14)

Fem

ale

(1=y

es)

−4.

047∗∗

∗−

3.96

3∗∗∗

2.15

4−

0.73

35.

711∗∗

∗5.

186∗∗

∗1.

136

−0.

435

(0.7

41)

(0.7

28)

(1.4

32)

(1.4

86)

(1.4

84)

(1.1

95)

(1.2

64)

(0.7

83)

Age

−0.

121∗∗

−0.

076∗

−0.

090

−0.

191∗

0.20

4∗∗−

0.01

10.

245∗∗

∗0.

199∗∗

∗

(0.0

38)

(0.0

38)

(0.0

74)

(0.0

77)

(0.0

77)

(0.0

62)

(0.0

66)

(0.0

41)

Educ

:Upp

erEl

emen

tary

−16

.905

∗∗−

18.5

69∗∗

∗17

.597

17.9

39−

7.69

2−

4.43

9−

3.39

64.

603

(5.5

10)

(5.4

17)

(10.

652)

(11.

054)

(11.

036)

(8.8

89)

(9.4

01)

(5.8

25)

Educ

:Low

erSe

cond

.−

25.4

68∗∗

∗−

26.1

23∗∗

∗19

.492

20.0

27−

3.86

53.

885

−7.

946

6.24

0(5

.466

)(5

.373

)(1

0.56

6)(1

0.96

5)(1

0.94

7)(8

.818

)(9

.325

)(5

.778

)Ed

uc:S

econ

dary

−30

.632

∗∗∗−

30.6

95∗∗

∗19

.131

15.6

025.

664

14.1

57−

8.54

26.

700

(5.5

70)

(5.4

76)

(10.

768)

(11.

174)

(11.

156)

(8.9

86)

(9.5

03)

(5.8

89)

Educ

:Uni

vers

ityPr

ep.

−32

.998

∗∗∗−

33.0

42∗∗

∗13

.703

15.9

914.

891

14.8

18−

10.2

329.

714

(5.5

10)

(5.4

17)

(10.

653)

(11.

055)

(11.

037)

(8.8

90)

(9.4

01)

(5.8

26)

Educ

:Ter

tiary

I−

30.5

80∗∗

∗−

30.7

32∗∗

∗12

.943

12.7

464.

440

20.3

62∗−

17.0

8710

.415

(5.5

19)

(5.4

26)

(10.

670)

(11.

073)

(11.

055)

(8.9

04)

(9.4

17)

(5.8

35)

Educ

:Ter

tiary

II−

32.6

07∗∗

∗−

32.8

08∗∗

∗2.

873

6.06

214

.618

27.0

39∗∗

−12

.586

9.27

0(5

.483

)(5

.391

)(1

0.60

1)(1

1.00

1)(1

0.98

3)(8

.847

)(9

.355

)(5

.797

)In

com

e(lo

g)0.

238

0.14

33.

566∗∗

∗3.

718∗∗

∗−

1.67

9−

2.91

0∗∗∗

1.59

7−

1.57

7∗∗

(0.5

32)

(0.5

23)

(1.0

29)

(1.0

68)

(1.0

66)

(0.8

59)

(0.9

08)

(0.5

63)

Inte

rcep

t41

.746

∗∗∗

41.1

99∗∗

∗−

11.1

981.

968

40.1

43∗∗

21.5

0914

.011

9.01

6(7

.486

)(7

.359

)(1

4.47

2)(1

5.01

8)(1

4.99

3)(1

2.07

7)(1

2.77

1)(7

.914

)R2

0.09

50.

088

0.04

10.

032

0.06

20.

099

0.04

40.

082

Adj

.R2

0.09

00.

083

0.03

60.

027

0.05

70.

095

0.03

90.

077

Num

.obs

.53

8653

8653

8653

8653

8653

8653

8653

86∗∗

∗p

<0.

001;

∗∗p

<0.

01;∗

p<

0.05

.All

mod

elsi

nclu

dere

gion

al(B

unde

sland

)�xe

de�

ects

.

41

TableA6:

Inte

rgen

erat

iona

lSta

tusD

isco

rdan

ce(N

on-T

rans

form

ed)a

ndPa

rty

Choi

ce,

rraf

dcd

ucsu

msr

msl

left

linke

ISD

(Pre

cise

orig

inal

mea

sure

)0.

089∗∗

∗0.

083∗∗

∗−

0.16

0∗∗∗−

0.22

7∗∗∗

0.00

90.

010

0.00

3(0

.018

)(0

.018

)(0

.033

)(0

.034

)(0

.033

)(0

.034

)(0

.018

)Fe

mal

e(1

=yes

)−

2.88

4∗∗∗−

2.90

3∗∗∗

3.65

1∗∗∗

1.66

43.

619∗∗

∗3.

109∗∗

−0.

377

(0.5

54)

(0.5

51)

(0.9

86)

(1.0

41)

(0.9

95)

(1.0

38)

(0.5

47)

Age

−0.

030

−0.

023

0.21

3∗∗∗

0.16

3∗∗0.

213∗∗

∗0.

290∗∗

∗0.

080∗∗

(0.0

29)

(0.0

29)

(0.0

51)

(0.0

54)

(0.0

52)

(0.0

54)

(0.0

29)

Educ

:Upp

erEl

emen

tary

−4.

390

−4.

619

7.17

19.

679

−4.

819

−0.

604

4.16

8(3

.280

)(3

.263

)(5

.830

)(6

.157

)(5

.886

)(6

.141

)(3

.238

)Ed

uc:L

ower

Seco

nd.

−9.

963∗∗

−10

.116

∗∗13

.378

∗18

.094

∗∗−

0.10

85.

241

5.29

2(3

.249

)(3

.232

)(5

.776

)(6

.099

)(5

.831

)(6

.084

)(3

.208

)Ed

uc:S

econ

dary

−11

.228

∗∗∗−

11.1

94∗∗

∗16

.224

∗∗16

.852

∗∗7.

792

13.0

77∗

5.25

7(3

.364

)(3

.346

)(5

.979

)(6

.314

)(6

.036

)(6

.298

)(3

.321

)Ed

uc:U

nive

rsity

Prep

.−

15.6

47∗∗

∗−

15.6

28∗∗

∗14

.479

∗21

.016

∗∗∗

9.44

517

.629

∗∗7.

846∗

(3.3

07)

(3.2

90)

(5.8

79)

(6.2

08)

(5.9

35)

(6.1

92)

(3.2

65)

Educ

:Ter

tiary

I−

14.6

55∗∗

∗−

14.6

73∗∗

∗14

.973

∗19

.295

∗∗6.

579

16.4

56∗∗

9.84

4∗∗

(3.3

22)

(3.3

04)

(5.9

04)

(6.2

35)

(5.9

60)

(6.2

19)

(3.2

79)

Educ

:Ter

tiary

II−

15.9

34∗∗

∗−

15.9

77∗∗

∗6.

753

12.8

55∗

16.5

33∗∗

25.5

96∗∗

∗8.

805∗∗

(3.2

90)

(3.2

73)

(5.8

49)

(6.1

76)

(5.9

04)

(6.1

61)

(3.2

48)

Inco

me

(log)

0.26

70.

184

3.16

7∗∗∗

4.02

8∗∗∗

0.67

4−

1.34

6−

1.89

6∗∗∗

(0.4

01)

(0.3

99)

(0.7

13)

(0.7

53)

(0.7

20)

(0.7

51)

(0.3

96)

Inte

rcep

t20

.990

∗∗∗

21.2

92∗∗

∗−

25.4

67∗∗−

25.0

61∗∗

14.8

8830

.657

∗∗∗

14.5

89∗∗

(4.7

71)

(4.7

46)

(8.4

81)

(8.9

56)

(8.5

62)

(8.9

34)

(4.7

10)

R20.

060

0.05

80.

029

0.03

10.

070

0.06

40.

038

Adj

.R2

0.05

70.

055

0.02

60.

028

0.06

80.

061

0.03

5N

um.o

bs.

1075

810

758

1075

810

758

1075

810

758

1075

8∗∗

∗p

<0.

001;

∗∗p

<0.

01;∗

p<

0.05

.All

mod

elsi

nclu

dere

gion

al(B

unde

sland

)�xe

de�

ects

.

42

TableA7:

Inte

rgen

erat

iona

lSta

tusD

isco

rdan

ce(D

ecile

s)an

dPa

rty

Choi

ce

rraf

dcd

ucsu

msr

msl

left

linke

ISD

(Dec

iles)

0.51

3∗∗∗

0.47

6∗∗∗

−0.

838∗∗

∗−

1.18

9∗∗∗−

0.07

5−

0.07

90.

019

(0.1

02)

(0.1

02)

(0.1

82)

(0.1

92)

(0.1

84)

(0.1

92)

(0.1

01)

Fem

ale

(1=y

es)

−2.

865∗∗

∗−

2.88

6∗∗∗

3.65

7∗∗∗

1.67

33.

557∗∗

∗3.

046∗∗

−0.

376

(0.5

55)

(0.5

52)

(0.9

86)

(1.0

41)

(0.9

95)

(1.0

39)

(0.5

48)

Age

−0.

029

−0.

023

0.20

9∗∗∗

0.15

8∗∗0.

216∗∗

∗0.

292∗∗

∗0.

080∗∗

(0.0

29)

(0.0

29)

(0.0

51)

(0.0

54)

(0.0

52)

(0.0

54)

(0.0

29)

Educ

:Upp

erEl

emen

tary

−4.

426

−4.

652

7.25

89.

804

−4.

861

−0.

647

4.16

7(3

.280

)(3

.262

)(5

.831

)(6

.158

)(5

.886

)(6

.141

)(3

.238

)Ed

uc:L

ower

Seco

nd.

−9.

929∗∗

−10

.084

∗∗13

.403

∗18

.133

∗∗−

0.24

45.

101

5.29

3(3

.249

)(3

.232

)(5

.777

)(6

.101

)(5

.831

)(6

.084

)(3

.208

)Ed

uc:S

econ

dary

−11

.170

∗∗∗−

11.1

39∗∗

∗16

.229

∗∗16

.864

∗∗7.

616

12.8

96∗

5.25

8(3

.364

)(3

.346

)(5

.980

)(6

.316

)(6

.036

)(6

.298

)(3

.321

)Ed

uc:U

nive

rsity

Prep

.−

15.5

79∗∗

∗−

15.5

63∗∗

∗14

.505

∗21

.058

∗∗∗

9.20

817

.385

∗∗7.

848∗

(3.3

08)

(3.2

90)

(5.8

80)

(6.2

11)

(5.9

35)

(6.1

93)

(3.2

66)

Educ

:Ter

tiary

I−

14.5

56∗∗

∗−

14.5

78∗∗

∗14

.992

∗19

.328

∗∗6.

265

16.1

34∗∗

9.84

6∗∗

(3.3

22)

(3.3

05)

(5.9

07)

(6.2

38)

(5.9

62)

(6.2

21)

(3.2

80)

Educ

:Ter

tiary

II−

15.8

70∗∗

∗−

15.9

15∗∗

∗6.

889

13.0

57∗

16.1

28∗∗

25.1

79∗∗

∗8.

806∗∗

(3.2

91)

(3.2

73)

(5.8

50)

(6.1

79)

(5.9

05)

(6.1

61)

(3.2

49)

Inco

me

(log)

0.27

30.

192

3.23

8∗∗∗

4.13

2∗∗∗

0.53

8−

1.48

6∗−

1.89

6∗∗∗

(0.4

01)

(0.3

99)

(0.7

12)

(0.7

52)

(0.7

19)

(0.7

50)

(0.3

96)

Inte

rcep

t17

.906

∗∗∗

18.4

26∗∗

∗−

21.1

47∗−

18.9

59∗

16.5

2232

.343

∗∗∗

14.4

80∗∗

(4.9

23)

(4.8

97)

(8.7

52)

(9.2

44)

(8.8

34)

(9.2

18)

(4.8

60)

R20.

060

0.05

80.

028

0.03

00.

070

0.06

40.

038

Adj

.R2

0.05

70.

055

0.02

60.

028

0.06

80.

061

0.03

5N

um.o

bs.

1075

810

758

1075

810

758

1075

810

758

1075

8∗∗

∗p

<0.

001;

∗∗p

<0.

01;∗

p<

0.05

.All

mod

elsi

nclu

dere

gion

al(B

unde

sland

)�xe

de�

ects

.

43

TableA8:

Stat

usD

isco

rdan

cean

dA

ttitu

des,

Dire

ctan

dby

Educ

atio

n

wor

ry:s

ocw

orry

:soc

wor

ry:e

cons

ocw

orry

:eco

nsoc

ISD

(Pct

iles)

0.00

1∗∗0.

001∗∗

∗0.

001∗∗

∗0.

001∗∗

∗

(0.0

00)

(0.0

00)

(0.0

00)

(0.0

00)

Colle

geD

egre

e−

0.01

40.

023

−0.

150∗∗

∗−

0.13

4∗∗∗

(0.0

14)

(0.0

24)

(0.0

13)

(0.0

23)

ISD

xCo

llege

Deg

ree

−0.

001

−0.

000

(0.0

00)

(0.0

00)

Inte

rcep

t2.

236∗∗

∗2.

219∗∗

∗1.

911∗∗

∗1.

904∗∗

∗

(0.0

89)

(0.0

89)

(0.0

85)

(0.0

86)

R20.

016

0.01

60.

040

0.04

0A

dj.R

20.

014

0.01

50.

038

0.03

8N

um.o

bs.

1204

912

049

1205

812

058

∗∗∗p

<0.

001;

∗∗p

<0.

01;∗

p<

0.05

.All

mod

elsi

nclu

dere

gion

al(B

unde

sland

)�xe

de�

ects

.

44