How the young get more than the elderly out of society – but not out of the state

Pieter Vanhuysse, University of Southern Denmark

Accusations levelled against the baby boomer generation for hoarding too much wealth and wielding too much “grey power” in an age of “gerontocracy” have become commonplace across Europe. At the centre of this debate lies the question of what resources different generations pass on to each other – and what is fair.

But in a new research paper on intergenerational transfers, demographers Robert Gal, Lili Vargha and I argue that it’s misleading to portray older people as benefiting more from society than younger people. Weiterlesen

Veröffentlicht unter Gastbeitrag | Verschlagwortet mit , , , , , , , | Schreib einen Kommentar

There will never be a political age conflict between the young and the old.

This essay is the fifth in a series written by Achim Goerres for the project “Ageing Democracies? Political Participation and Cultural Values Among the Elderly in Europe” financed by the Open Society Foundation. The complete project report with all essays and the reports written by the other project members can be found here. In this essay Achim Goerres argues that there is no such thing as an age cleavage, and that it is highly unlikely to ever become a conflict line along to which the political system organises. Weiterlesen

Veröffentlicht unter Ageing Democracies? | Verschlagwortet mit , , , , , | Schreib einen Kommentar

Migranten auf dem Stimmzettel: Eine Vignettenanalyse zur Untersuchung der Wirkung der ethnischen Herkunft von politischen Kandidaten in Verbindung mit weiteren Kandidateneigenschaften auf die Kandidatenevaluation der Wähler

Von Erik Wenker

Dem Beitrag liegt die Bachelorarbeit des Autors zugrunde, die am Lehrstuhl für Empirische Politikwissenschaft von Dr. Sabrina Mayer und Prof. Dr. Achim Goerres betreut wurde.

Während Migranten medial, politisch und wissenschaftlich zunehmend als relevante Wählergruppe wahrgenommen werden und auf dem politischen Parkett ankommen, um beispielsweise als Parteivorsitzende zu amtieren (Street 2014: 375), sind sie in Deutschland und Europa gleichzeitig parlamentarisch unterrepräsentiert (Bird/Saalfeld/Wüst 2011; Bloemraad/Schönwälder 2013). Weiterlesen

Veröffentlicht unter Allgemein | Verschlagwortet mit , , , , , , , , | Schreib einen Kommentar

Welchen Einfluss haben persönliche Veränderungen auf die Bereitschaft nicht-institutionell politisch zu partizipieren?

von Florian Gerls.

Die Teilhabe am politischen Prozess ist zweifellos Grundpfeiler einer jeden Demokratie. Klassisch durch Wahlen realisiert fand in den 60er und 70er Jahren des letzten Jahrhunderts eine Zunahme von, bis dahin als unkonventionell bezeichneten, Formen politischer Partizipation statt. Demonstrationen, Bürgerinitiativen und Unterschriftenaktionen wurden zunehmend als Möglichkeit zur Ergänzung und später auch als Alternative zur Beteiligung an Wahlen eingeschätzt. Es lässt sich also eine Forderung nach weiteren politischen Einflussmöglichkeiten zu dieser Zeit konstatieren, welche schon von Willy Brand in seiner Regierungserklärung mit den Worten „Mehr Demokratie wagen“ aufgegriffen wurde. Heutzutage scheint die Teilnahme an Demonstrationen, das Unterschreiben von Petitionen und ein Engagement in Bürgerinitiativen, welche sich als Formen nicht-institutionalisierter Partizipation benennen lassen, beinahe selbstverständlich in der politischen Auseinandersetzung zu sein.

Auch wenn etwa bei Demonstrationen gegen Stuttgart 21 partizipierende Personen abwertend als „Wutbürger“ bezeichnet wurden, so liegen genannten Partizipationsinstrumenten wichtige demokratische Potentiale wie etwa eine Ventilfunktion zugrunde (Wagschal 2015; S.99). Zudem können durch nicht verfasste Formen politischer Partizipation bestimmte Interessen zum Ausdruck gebracht und diese anschließend in den politischen Prozess eingespeist werden (Lauth; Pickel; Pickel 2014; S.248). Daneben wird die Unzufriedenheit mit der „politischen Klasse“ besser identifiziert, was eine Stärkung der Demokratie zur Folge hat (Offe 2012; S.43). Auf Basis der Funktionen für die Demokratie als solche ist es wenig verwunderlich, dass nicht-institutionalisierte politische Partizipation schon vor einiger Zeit in den politikwissenschaftlichen Fokus gerückt ist.

Auffallend dabei ist, dass in der Fülle an Studien der Forschungsgegenstand vorwiegend im Querschnitt analysiert wird. In diesen wird untersucht, inwieweit Unterschiede bezüglich verschiedener Merkmale zwischen Menschen die Partizipationsbereitschaft selbiger beeinflussen. Doch zweifelsohne variieren viele für die Partizipation wichtige Merkmale im Lebensverlauf einer Person. Der Notwendigkeit, den Einfluss von individuellen Veränderungen auf die Bereitschaft nicht-institutionalisiert zu partizipieren zu untersuchen, wurde in der bisherigen Forschung bei weitem nicht zur Genüge Rechnung getragen, sodass eine Forschungslücke konstatiert werden muss. Doch über diese wissenschaftsinterne Notwendigkeit hinaus ergeben sich auch gesellschaftliche und politische Anforderungen, einen methodischen Perspektivwechsel vorzunehmen, um ein Gefühl dafür zu erlangen, an welchen Stellen politische Maßnahmen greifen könnten, um zu verhindern, dass Veränderungen der Lebenslagen von Menschen ihr Partizipationspotential negativ beeinflussen.

Aus der Literatur lassen sich drei Gruppen von möglichen Einflussfaktoren herausarbeiten. Die erste Gruppe umfasst soziodemografische Merkmale, wie etwa Bildung, Alter und Einkommen. Bereits in der Studie Participation in America, welche von Verba und Nie vorgelegt wurde, ließ sich ein nicht unwesentlicher Effekt postulieren (Verba; Nie 1987). Das entworfene sozioökonomische Standardmodell implizierte, dass Menschen mit höherem sozioökonomischem Status eher partizipieren als Menschen mit geringem Status. Hierauf aufbauend wurde (1995) von Verba, Schlozman und Brady das Civic-Voluntarism-Modell entwickelt. Über einen entscheidenden Einfluss von sozioökonomischen Ressourcen hinaus wurde auch ein Effekt von Motivationen und politischen Orientierungen auf die Partizipation eines Individuums suggeriert. In meiner Arbeit habe ich die Gruppe von Determinanten durch eine Auswahl von Variablen politischer Unterstützung (Easton 1965), (Norris 1999) und (Pickel; Pickel 2016) ergänzt. Die letzte theoretisch interessante Merkmalsgruppe umfasst Variablen, die die Integration eines Individuums in soziale Netzwerke darstellen. Die Wichtigkeit der sozialen Kontakte auf die Partizipationsbereitschaft fußt ebenfalls auf den Überlegungen des Civic-Voluntarism-Modells.

Um das Forschungsvorhaben zu realisieren, wurde auf einen Individualdatensatz der Bundestagswahlstudie 2002 bis 2009 zurückgegriffen. Ein aufaddierter Index von zwei Variablen, welche die Bereitschaft an einer Demonstration teilzunehmen und die Bereitschaft sich in einer Bürgerinitiative zu engagieren messen, wurde als abhängige Variable verwendet. Ein höherer Wert stellt eine höhere Partizipationsbereitschaft dar. Für die Determinanten wurde zum großen Teil auf gängige Operationalisierungen zurückgegriffen. Gegeben des Forschungsinteresses wurden verschiedene Arten von Panelregressionstechniken (Random Effects-, First Difference- und Fixed Effects Regression) verwendet. Gerade letztgenannte bieten den wesentlichen Vorteil, unbeobachtete Heterogenität zu kontrollieren. Dies bedeutet, dass Effekte von unbeobachteten Drittvariablen auf die interessierenden statistischen Beziehungen ausgekoppelt und dadurch kausale Effekte besser gesichert werden können. In einer Random Effects Regression werden auch Unterschiede zwischen Personen erfasst. Diese Technik ist also als Kompromiss eines Längsschnitt- und eines Querschnittmodells zu verstehen. Zusätzlich wurden sogenannte gepoolte Modelle berechnet, in welchen Quer- und Längsschnittdaten zusammengefasst werden und welche als OLS-Regressionsmodelle zu interpretieren sind. Dies wurde deshalb realisiert, damit reine Längsschnitteffekte mit Querschnittseffekten unmittelbar verglichen werden können. Für Leserinnen und Leser, die in dem Umgang mit den verwendeten Regressionstechniken nicht vertraut sind, befindet sich im Anhang eine Interpretationshilfe.

Abbildung 1
Abbildung 1: Verteilung der abhängigen Variable. Anmerkung: Die gelbe vertikale Linie ist der Mittelwert, die rote Linie stellt den Median dar. Quelle: Eigene Darstellung und Berechnung; Datenbasis: Bundestagswahlstudie 2002 bis 2009

Aus meiner empirischen Analyse ging eine Vielzahl interessanter Befunde hervor. Zunächst ist hervorzuheben, dass sich die Betrachtung der Längsschnittdimension in allen berechneten Modellen statistisch lohnt. Gerade das Modell mit politischen Orientierungen, Unterstützung und Interessen weist ein hohes Erklärungspotential auf. Als besonders relevant für die Erklärung einer Variation im individuellen Partizipationsverhalten zeigt sich die Wichtigkeit postmaterialistischer Werte. Ernüchternd ist jedoch, dass die Modelle zwar im Ganzen signifikant sind, sich die Effekte einzelner Variablen jedoch häufig nicht bei gängigen Signifikanzniveaus absichern lassen. Umso interessanter ist es, die signifikanten Längsschnitteffekte (Fixed Effects Modell) hervor zu heben. Querschnittsbefunde, die nahelegen, dass eine höhere Bildung das Partizipationsverhalten steigert, kann für die Längsschnittdimension nicht ohne Einschränkungen übertragen werden. Zwar haben alle Koeffizienten der Dummy Variablen, die alle eine höhere Bildung darstellen als die Referenzgruppe (noch Schüler, kein Abschluss oder Hauptschulabschluss), positive Vorzeichen, jedoch lässt sich der Effekt nur bei der Dummy Variable „Mittlere Reife“ hinreichend absichern. Der Effekt des Haushaltseinkommens hingegen ist in allen Modellen signifikant und positiv. Das heißt, eine positive Veränderung des Haushaltseinkommens hat im Schnitt auch eine positive Veränderung der Partizipationsbereitschaft zur Folge. Besonders unerwartet sind die Befunde bei der Schichtzugehörigkeit von Personen. Ein sozialer Aufstieg von der Arbeiterschicht in die Mittel- und auch in die Oberschicht vermindert tendenziell die Bereitschaft nicht-institutionell zu partizipieren.

Part.Index

(OLS)

(RE)

(FE)

Bildung (RG geringe Bildung)
   Mittlere Reife

0.53***

0.56***

0.37*

(0.09)

(0.09)

(0.21)

   Abitur

1.24***

1.23***

0.40

(0.12)

(0.12)

(0.27)

   Studium (Uni/FH)

1.42***

1.36***

0.38

(0.11)

(0.12)

(0.29)

Haushaltseinkommen

0.05***

0.05***

0.06***

(0.01)

(0.01)

(0.02)

Schicht (RG Arbeiterschicht)
   Mittelschicht

0.09

0.07

-0.29**

(0.08)

(0.08)

(0.14)

   Oberschicht

-0.30

-0.22

-0.82**

(0.25)

(0.24)

(0.40)

Alter

0.06***

0.05***

0.04

(0.01)

(0.01)

(0.04)

Alter quadriert

-0.001***

-0.001***

-0.0002

(0.0001)

(0.0001)

(0.0004)

Konfession

-0.12

-0.08

-0.02

(0=konfessionslos, 1=in Konfession

(0.08)

(0.08)

(0.21)

Konstante

4.68***

4.86***

(0.27)

(0.29)

Beobachtungen

4,806

4,806

4,806

R2

0.13

0.08

0.01

Angepasstes R2

0.13

0.08

0.01

F Statistik

79.09***

(df = 9; 4796)

41.03***

(df = 9; 4796)

2.34**

(df = 9; 1534)

Tabelle 1: Soziodemografisches Modell. Anmerkung: *p<0.1; **p<0.05; ***p<0.01; Quelle: Eigene Berechnung; Datenbasis: Bundestagswahlstudie 2002 bis 2009

Die Effekte des politischen Interesses, der Wichtigkeit postmaterialistischer Werte und der Zufriedenheit mit der Regime Performance sind hypothesenkonform und signifikant und stimmen zudem in allen Modellen überein. Während ein Steigen des politischen Interesses und der Wichtigkeit postmaterialistischer Werte im Zeitverlauf die Partizipationsbereitschaft durchschnittlich erhöht, sinkt selbige tendenziell bei Ansteigen der Demokratiezufriedenheit. Unerwarteterweise brachten andere Variablen politischer Unterstützung (Vertrauen in politische und rechtsstaatliche Institutionen, Regierungszufriedenheit und die Legitimität der politischen Gemeinschaft), ebenso wie Variablen zur Darstellung der internen und externen Efficacy und der eigenen Links-Rechts Verortung im reinen Längsschnitt (Fixed Effects) keine signifikanten Effekte hervor.

Part.Index

(OLS)

(RE)

(FE)

Politisches Interesse

0.49***

0.50***

0.33***

(0.03)

(0.04)

(0.08)

Links-Einstufung

0.18***

0.16***

0.05

(1=rechts; 11=links)

(0.02)

(0.02)

(0.03)

Interne Efficacy

0.46***

0.41***

0.09

(0.04)

(0.04)

(0.06)

Externe Efficacy

0.01

-0.0004

-0.01

(0.04)

(0.04)

(0.07)

Vertrauen in pol. Inst.

-0.02

-0.001

0.04

(0.03)

(0.03)

(0.04)

V. in rechts. Institutionen

0.45***

0.37***

0.33

(0.13)

(0.13)

(0.26)

V. in rechts. Inst. (quadriert)

-0.02**

-0.02*

-0.01

(0.01)

(0.01)

(0.02)

Demokratiezufriedenheit

-0.16***

-0.17***

-0.19***

(0.04)

(0.04)

(0.07)

Zufriedenheit mit BundesR.

0.04**

0.03*

-0.01

(0.01)

(0.01)

(0.02)

Stolz Deutscher zu sein

-0.24***

-0.23***

0.01

(0.03)

(0.03)

(0.06)

Konstante

1.35***

1.69***

(0.50)

(0.50)

Beobachtungen

4,806

4,806

4,806

R2

0.16

0.10

0.03

Angepasstes R2

0.16

0.10

0.03

F Statistik

93.06***

(df = 10; 4795)

52.36***

(df = 10; 4795)

4.44***

(df = 10; 1533)

Tabelle 2: Modell mit Orientierungen, Motivation und politischer Unterstützung. Anmerkung: *p<0.1; **p<0.05; ***p<0.01; Quelle: Eigene Berechnung; Datenbasis: Bundestagswahlstudie 2002 bis 2009

 

Part.Index

(OLS)

(RE)

(FE)

Postmaterialismus.Index

0.55***

0.52***

0.14**

(0.03)

(0.03)

(0.06)

Konstante

3.32***

3.43***

(0.12)

(0.13)

N

3,904

3,904

3,904

R2

0.11

0.08

0.01

Angepasstes R2

0.11

0.08

0.01

F Statistik

472.18***

(df = 1; 3902)

320.67***

(df = 1; 3902)

5.55**

(df = 1; 640)

Tabelle 3: Modell mit der Wichtigkeit postmaterialistischer Werte. Anmerkung: *p<0.1; **p<0.05; ***p<0.01; Quelle: Eigene Berechnung; Datenbasis: Bundestagswahlstudie 2002 bis 2009

Die Integration in Netzwerke (Mitglieds- oder Amtsträgerschaft in Parteien, Berufsvereinigungen und Gewerkschaften) erhöht tendenziell hypothesenkonform die Bereitschaft zur nicht institutionellen Partizipation (Referenzgruppe: Kein Mitglied). Es muss allerdings erwähnt sein, dass lediglich die Effekte im Querschnittsmodell signifikant sind. In einem berechneten First Difference Modell brachte nur die Gewerkschaftsvariable einen signifikanten positiven Effekt hervor.

Part.Index

(OLS)

(FD)

(FE)

Partei

0.87***

0.41

0.35

(0.14)

(0.31)

(0.30)

Berufsvereinigung

0.62***

0.21

0.20

(0.14)

(0.22)

(0.22)

Gewerkschaft

0.71***

0.39*

0.26

(0.10)

(0.21)

(0.20)

Konstante

5.94***

0.35 ***

(0.04)

(0.07)

N

4,806

1,543

4,806

R2

0.03

0.005

0.003

Angepasstes R2

0.03

0.003

0.001

F Statistik

46.52*** (df = 3; 4802)

2.32* (df = 3; 1539)

1.52 (df = 3; 1540)

Tabelle 4: Modell mit der Integration in Netzwerke. Anmerkung: *p<0.1; **p<0.05; ***p<0.01; Quelle: Eigene Berechnung; Datenbasis: Bundestagswahlstudie 2002 bis 2009

Zusammenfassend bleibt festzuhalten, dass sich die Betrachtung der Längsschnittdimension in allen berechneten Modellen rein statistisch in jedem Falle lohnt. Trotz der mangelnden Signifikanz der einzelnen Längsschnitteffekte konnten interessante Wirkungsmechanismen identifiziert werden, die teilweise von den Querschnittsbefunden abweichen. Die vermutete Wichtigkeit eines methodischen Perspektivwechsels lässt sich letztendlich bestätigen.

Technischer Anhang

Für die empirische Analyse wurden Daten des Langfristpanel 2002 bis 2009 der Bundestagswahlstudien (englisch: German Longitudinal Election Study) verwendet, welche von dem Leibniz-Institut für Sozialwissenschaften auf dessen Homepage bereitgestellt werden.

Für besseres Verständnis der Regressionsergebnisse ist untenstehend eine Interpretationshilfe dargestellt.

Variante Interpretation der Koeffizienten
Fixed Effects Wie stark verändert sich Y über die Zeit hinweg durchschnittlich pro Person, wenn X um eine Einheit steigt.
Random Effects Durchschnittlicher Effekt von X auf Y, wenn X über die Zeit hinweg und zwischen Personen um eine Einheit variiert.
First Difference Wie verändert sich Y im Schnitt, zu dem Zeitpunkt, an dem sich X um eine Einheit verändert.

Tabelle 5: Erklärung der Interpretation der Koeffizienten. Quelle: Eigene Darstellung; Angelehnt an (Torres-Reyna 2010; S.12-14) und (Giesselmann; Windzio 2012; S.61)

Um die Querschnittseffekte besser nachvollziehen zu können (OLS-Regressionsmodelle) sind selbige untenstehend visualisiert.

Abbildung 2: Querschnittsbefunde des soziodemografischen Modells. Quelle: Eigene Berechnung und Darstellung; Datenbasis: Bundestagswahlstudie 2002 bis 2009
Abbildung 2: Querschnittsbefunde des soziodemografischen Modells. Quelle: Eigene Berechnung und Darstellung; Datenbasis: Bundestagswahlstudie 2002 bis 2009

 

Abbildung 3: Querschnittsbefunde des Modells mit Orientierungen, Motivationen und politischer Unterstützung. Quelle: Eigene Berechnung und Darstellung; Datenbasis: Bundestagswahlstudie 2002 bis 2009
Abbildung 3: Querschnittsbefunde des Modells mit Orientierungen, Motivationen und politischer Unterstützung. Quelle: Eigene Berechnung und Darstellung; Datenbasis: Bundestagswahlstudie 2002 bis 2009
Abbildung 4: Querschnittsbefunde des Modells mit der Wichtigkeit postmaterialistischer Werte. Quelle: Eigene Berechnung und Darstellung; Datenbasis: Bundestagswahlstudie 2002 bis 2009
Abbildung 4: Querschnittsbefunde des Modells mit der Wichtigkeit postmaterialistischer Werte. Quelle: Eigene Berechnung und Darstellung; Datenbasis: Bundestagswahlstudie 2002 bis 2009
Abbildung 5: Querschnittsbefunde des Modells mit der Integration in Netzwerke. Quelle: Eigene Berechnung und Darstellung; Datenbasis: Bundestagswahlstudie 2002 bis 2009
Abbildung 5: Querschnittsbefunde des Modells mit der Integration in Netzwerke. Quelle: Eigene Berechnung und Darstellung; Datenbasis: Bundestagswahlstudie 2002 bis 2009

 

Verwendete Literatur:

Easton, David (1965): A systems analysis of political life. New York/London/Sydney: John Wiley & Sons.

Giesselmann, Marco; Windzio, Michael (2012): Regressionsmodelle zur Analyse von Paneldaten. Wiesbaden: VS Verlag für Sozialwissenschaften.

Lauth, Hans-Joachim; Pickel, Gert; Pickel, Susanne (2014): Vergleich politischer Systeme. Paderborn: Schöningh [u.a.] (UTB, 4000 : Politikwissenschaft).

Norris, Pippa (1999): Introduction: The Growth of Critical Citizens? In: Norris, Pippa (Hg.): Critical Citizens. Global Support for Democratic Governance. Repr. Oxford: Oxford Univ. Press, S. 1–30.

Offe, Claus (2012): Political disaffection as an outcome of institutional practices? Some post Tocquevillean speculations. In: Torcal, Mariano; Montero, José R. (Hg.): Political disaffection in contemporary democracies. Social capital, institutions and politics. Paperback ed. London: Routledge (Routledge research in comparative politics, 13), S. 23–45.

Pickel, Susanne; Pickel, Gert (2016): Politische Kultur in der Vergleichenden Politikwissenschaft. In: Lauth, Hans-Joachim; Kneuer, Marianne; Pickel, Gert (Hg): Handbuch Vergleichende Politikwissenschaft. Wiesbaden. Springer VS, S. 541-556.

Torres-Reyna, Oscar (2010): Getting Started in Fixed/Random Effects Models using R. Princeton University. Online verfügbar unter: https://www.princeton.edu/~otorres/Panel101R.pdf.

Verba, Sidney; Nie, Norman H. (1987): Participation in America. Political democracy and social equality. Repr. [der Ausg.] New York: Harper & Row, 1972, 1. [Dr.]. Chicago: Univ. of Chicago Press.

Verba, Sidney; Schlozman, Kay L.; Brady, Henry E. (1995): Voice and Equality. Civic Voluntarism in American Politics, Cambridge, Mass.: Harvard University Press.

Wagschal, Uwe (2015): Direkte Demokratie: Instrumente – Policy-Wirkungen – neue Formen der Bürgerbeteiligung. In: Wagschal, Uwe; Wenzelburger, Georg; Jäckle, Sebastian (Hg.): Einführung in die Vergleichende Politikwissenschaft. Institutionen – Akteure – Policies. 1. Aufl. Stuttgart: Kohlhammer Verlag, S. 85–102.

Veröffentlicht unter Allgemein | Verschlagwortet mit , , , , , , , , , | Schreib einen Kommentar

Social inequalities within the group of older people impede the formation of a politically uniform bloc of older people

This essay is the third in a series written by Achim Goerres for the project “Ageing Democracies? Political Participation and Cultural Values Among the Elderly in Europe” financed by the Open Society Foundation. The complete project report with all essays and the reports written by the other project members can be found here.

There are various examples of older people’s political protests in Europe. In Spain, the Iaioflautas movement is one of older people who protest on multiple political issues related to the labour market, education, health, gender and basic income. They use new modes of swift communication and conduct political actions defying stereotypes of old age, mirroring other older people’s social movements such as the Raging Nannies in Canada and the USA (Blanche-Tarragó and Fernández-Ardèvol 2014). In 2004 and 2005 in England, political protests against the Council Tax were mainly led by older people, as they were disproportionally affected by it (Goerres 2009: chap. 7). In the early 1990s, older people in Poland, Czechoslovakia and Hungary protested in favour of acceptable pension deals after the political transitions of their respective countries (Vanhuysse 2006).

But there is a common misperception in the public discourse about political protests among older people. There is an inherent portrayal of older people as a generally homogenous political group. Visuals of older people protesting seem to strengthen this conclusion, since they involve individuals who are united in their common course. Yet, these images only represent a fraction of all older people and only one particular course. To draw conclusions about older people from those images is comparable to seeing a group of top athletes long-jumping and extrapolating from this that all humans can jump that far.

Discussions about the political participation and views of older people could benefit from shifting our focus to the immense social heterogeneity that exists within the group of older people. This social heterogeneity also translates into political heterogeneity, both in terms of activism and in terms of interests. Let us consider four major social lines of stratification that are particularly relevant for politics: education, income, gender and health.

Education and income are the main social dividing lines between individuals across Europe, stratifying the social position that an individual has in a society as well as his or her political preferences and behaviour. A rich, highly educated older person is miles away from a poor, low-educated older person, both in terms of social as well as political experience. The fact that both are pensioners and therefore recipients of public pensions is not a strong bond. For the rich senior, a public pension is likely to be a smaller fraction of his or her disposable income than for the poor senior. The rich, highly educated senior is more likely to have diversified his or her pension income across a diverse set of assets and different kinds of pensions.

Let us look at the relationship between household income and support for redistribution from the rich to the poor by the state. We will use this attitude to get a sense of where people see themselves on the classic economic scale from very left to very right. Across all older people, 73% think that it is the role of the government to decrease income differences between the rich and the poor. However, when we divide older people by income groups, we get exactly the pattern that we see among younger people. Those with higher income support this notion less often than those with lower income. More concretely, among older people whose household income is in the lowest 30% of their respective country’s income distribution, the proportion that supports income redistribution by the state lies at 79%. Among the highest 30% of household incomes, that estimate lies at 62%. In other words, income divides older people as to their demand for one of the core functions of the modern state in the same way it divides younger people.

Gender is another factor which stratifies the social experience of modern life to a great extent. At old age, the accumulation of these experiential differences becomes greatest and intersects with different mortality rates. All over the world, women are on average more likely to live longer than men (Barford et al. 2006). The gap in life expectancy between men and women has been decreasing in Europe over the last two decades, yet it still varies widely (Van Oyen et al. 2010). This means that the older the age group is, the more female it is. We know of some gender-related differences in political preferences. Women tend to place more emphasis on some policy issues over others (Campbell 2004). They are also less likely to vote for right-wing parties (Norris 2005) and more likely to support social policy expenditure (Jaime-Castillo et al. 2016; Hatemi et al. 2012). Thus, the composition of political behaviour and preferences is affected by there being more women at old age.

Health is another important line of division among older people. Health discrepancies in old age are striking. As a result, pension age can typically be divided into a first period, when pensioners are still capable of many things and a second period characterised by multi-morbidity, incapacitating them in many ways. The first period is referred to as the “third age” or “young old” and the second as the “fourth age” or “old old”. The age at the time of transition from the first period to the second period varies greatly among individuals. Health is an important predictor of political participation (Mattila et al. 2013; Sund et al. 2016; Söderlund and Rapeli 2015). Yet it predicts participation in different ways. Voting can be made accessible to individuals with health problems in various ways like proxy voting (someone votes for you), postal voting and mobile voting booths (in hospitals for example), mitigating the impact of health problems. But other forms of participation, such as writing letters or demonstrating in the streets, are much more demanding in terms of cognitive and physical abilities. Health inequalities thus translate more into political inequalities among older people for those political actions that are more demanding.

We can explore this using 2014 survey data for 20 European countries. In that survey, people were asked whether they were hampered in their daily activities by any illness or disability. Among the young old (people between 60 and 74 years old), 27% said that they were to some extent hampered, while 9% said that they were hampered very much. In the group of the old-old (people aged 75 and older), 34% said that they were hampered to some extent and 18% that they were hampered a lot. This pattern is mirrored in their political activity levels.

In Figure 1, we can see two mosaic plots that classify four types of political activity and the degree of self-perceived constraint, once for the young-old and once for the old-old. Each tile of the mosaic represents the sub-group of one activity type and one type of self-assessed constraint. The four types of activities are: non-active, only voting, voting and more, and only non-institutionalised political participation.

In the previous on political participation in general, we saw that older people who only vote are the largest group followed by the very active group of voters and more, followed by the non-active and those who only use non-institutionalised forms of participation. In this plot, we can now explore how the different activity types intersect with the level of self-perceived health constraints.

Each tile is the size of the sub-group defined by the two levels of the two variables. So, in each sub-panel, the lowest tile on the left is the group that feels hampered heavily by health issues and is politically active. This share is much bigger in the group of the old-old compared to the young-old. The blue tiles are the sub-groups who do not feel hampered at all in their everyday lives. We see that the unhampered group is represented more in higher-activity groups than in the passive groups. We also see that the largest group among the young-old and the old-old are the ones that feel fine and just go to the polls. These are older people who are not constrained by their health and only partake in politics through voting. However, among the old-old, this group is smaller than among the young-old. Overall, the politically passive and those who only vote are more common among the old-old than among the young -old. The group whose members use voting in addition to other forms of political participation is smaller among the old-old than among the young-old.

In a nutshell, we see that health structures the ways in which older people participate in politics. Worse health is associated with less or no political activity. Non-institutionalised participation plays a small role among older people to begin with, and it becomes almost non-existent among the old-old.

Figure 1: Proportion of sub-groups by activity types and whether they feel hampered by health issues in their daily activities, 20 European countries in 2014

Proportion of sub-groups by activity types and whether they feel hampered by health issues in their daily activities, 20 European countries in 2014

Proportion of sub-groups by activity types and whether they feel hampered by health issues in their daily activities, 20 European countries in 2014

We have thus confirmed that older people are a divided group. They are divided by differences in attitudes and resources that relate to income, education, gender and health. These differences not only structure the social position of older people, but also what they do and want in politics. Socio-economic inequality among older people translates into political inequality among older people, a fact that is often very much neglected in public debate.

 

REFERENCES

Barford, Anna, Danny Dorling, George Davey Smith, and Mary Shaw. 2006. „Life Expectancy: Women Now on Top Everywhere.“ British Medical Journal 332 (7545):808.

Blanche-Tarragó, D., and M. Fernández-Ardèvol. 2014. „The Iaioflautas Movement in Catalonia: A Seniors’ Networked Social Movement.“ In 5th ECREA European Communication Conference. Lisbon.

Campbell, Rosie. 2004. „Gender, Ideology and Issue Preference: Is There such a Thing as a Political Women’s Interest in Britain?1.“ The British Journal of Politics & International Relations 6 (1):20-44.

Goerres, Achim. 2009. The Political Participation of Older People in Ageing Europe. Basingstoke: Palgrave.

Hatemi, Peter K., Rose McDermott, J.Michael Bailey, and Nicholas G. Martin. 2012. „The Different Effects of Gender and Sex on Vote Choice.“ Political Research Quarterly 65 (1):76-92.

Jaime-Castillo, Antonio M., Juan J. Fernández, Celia Valiente, and Damon Mayrl. 2016. „Collective religiosity and the gender gap in attitudes towards economic redistribution in 86 countries, 1990–2008.“ Social Science Research 57:17-30.

Mattila, Mikko, Peter Söderlund, Hanna Wass, and Lauri Rapeli. 2013. „Healthy voting: The effect of self-reported health on turnout in 30 countries.“ Electoral Studies 32 (4):886-91.

Norris, Pippa. 2005. Radical right: Voters and parties in the Electoral Market. Cambridge: Cambridge University Press.

Söderlund, Peter, and Lauri Rapeli. 2015. „In sickness and in health.“ Politics and the Life Sciences 34 (1):28-43.

Sund, Reijo, Hannu Lahtinen, Hanna Wass, Mikko Mattila, and Pekka Martikainen. 2016. „How voter turnout varies between different chronic conditions? A population-based register study.“ Journal of Epidemiology and Community Health:jech-2016-208314.

Van Oyen, Herman, Bianca Cox, Carol Jagger, Emmanuelle Cambois, Wilma Nusselder, Clare Gilles, and Jean-Marie Robine. 2010. „Gender Gaps in Life Expectancy and Expected Years with Activity Limitations at Age 50 in the European Union: Associations with Macro-level Structural Indicators.“ European Journal of Aging 7 (4):229-37.

Vanhuysse, P. 2006. Divide and Pacify: Strategic Social Policies and Political Protests in Post-Communist Democracies. Budapest: Central European University Press.

 

Veröffentlicht unter Ageing Democracies? | Verschlagwortet mit , , , , , | Schreib einen Kommentar

As we age, we do not grow politically more conservative. Age differences in political preferences are almost exclusively due to the ways different cohorts grew up.

This essay is the third in a series written by Achim Goerres for the project “Ageing Democracies? Political Participation and Cultural Values Among the Elderly in Europe” financed by the Open Society Foundation. The complete project report with all essays and the reports written by the other project members can be found here.

One of the great myths about ageing and older people in politics is that individuals become more conservative with age. There is the commonly known bon mot that “if you’re not a liberal when you’re 25, you have no heart. If you’re not conservative by the time you’re 35, you have no brain.” This saying, which cannot be traced unequivocally to one source but seems to have been expressed first with slightly different age groups and adjectives by John Adams in a 1799 diary entry, seems to ring a bell with many observers of our European societies (Shapiro 2011).

The simple and, to some people, very appealing idea behind this is plainly wrong. Yet there seems to be something intuitively accurate about the phrase, which might explain why it has lasted so long despite the fact that there are very concrete empirical problems with it. In modern Europe, the period between an individual’s twenties and thirties is one of many changes for many people. They settle into their jobs, maybe they start a family, they start using different services provided by the public and private sector. Thus, it seems to make sense that political preferences change, too, during this period.

To begin with, political conservatism can mean several things. It can mean a generally held belief or a set of political values that are called “conservative”. Political science differentiates between two dimensions along which political preferences are usually grouped in Europe and to which political parties and candidates respond in terms of what they offer voters. The first is the economic dimension, which is associated by most people in Europe with the idea of left and right. Individuals who are more leftist on this dimension tend to believe in a strong role of the state in regulating the economy and redistributing between various social groups, most importantly from the rich to the poor. People who are more rightist on the economic dimension tend to believe in a lean role of the state both in regulating the market and in redistribution. The second dimension is the cultural one. People who are leftist here tend to support diversity with regard to sexual orientation, religion, ethnic background, language and other defining markers, and believe that the state should provide regulations to allow such diversity. Those on the cultural right are more supportive of a dominant way of living one’s life that is typically linked to a specific and narrow set of markers, such as one ethnic origin, one type of religion and one family structure. These two dimensions are not fully independent from another. Those people who are more conservative economically tend to be more conservative culturally, the relationship between the two dimensions is, however, not very strong. This is the reason why it makes sense for parties to explore the full two-dimensional room on these dimensions. For instance, the Dutch left-liberal party D66 (Democraten 66) and the right-liberal party VVD (Volkspartij voor Vrijheid and Democraties) are both economically conservative, but they differ on the cultural dimension with the D66 being more progressive culturally than VVD.

We can analyse these dimensions of political values by looking at some public opinion data from 2010 to 2014 from the European Social Survey. To measure the economic dimension, we will use support for the idea that the government should redistribute from the rich to the poor.

Table 1 shows the two demographically oldest European societies and the two youngest alongside support for the above statement as expressed in the European Social Survey. The first column shows estimated support across all age groups, the second the level of support among the 60+ group, the third the level among those under 30 and the last column the ratio of column 2 to 3. Again, if the ratio stands at one, there is no difference in support levels between the two age groups.

In Germany, 70% of the adult population think that the state should redistribute from the rich to the poor, while 73% of older people and 70% of young people think so. So, in Germany, younger people are actually more likely to be economically conservative than older people, but only very slightly. Older people are 1.04 times more likely to be supportive of redistribution than young people. The same pattern prevails in Italy, which is demographically the most similar society to Germany in Europe. It also holds in Ireland and Slovakia, the two youngest societies in our sample.

Table 1: Support for income redistribution by age group

[table id=6 /]

Let us now look at the cultural dimension. Table 2 shows similar evidence about whether individuals support the idea that homosexuals should be able to live as they want. The numbers are levels of support for the culturally progressive position. In Germany, for example, 85% of the adult population believe that homosexuals should live their lives as they wish, while 77% of older people and 88% of young people believe the same. Older people are thus 12% less likely to support this view. All four countries show the same inter-age group pattern, namely that older people tend to be culturally more conservative than young people. The only striking difference in Table 2 is between Slovakia and the other three countries, since Slovakia has a strongly culturally conservative populace with stark age group differences and an age ratio of 0.50.

Table 2: Support for free expression of homosexual life styles in two oldest and youngest democracies

[table id=7 /]

Figure 2 gives us an overview of all European countries in terms of the differences between older and young people with respect to the economic and cultural views described above. The picture is divided by the two parity lines at the value 1. The biggest quadrant is to the top left: older people are economically less and culturally more conservative than young people. Four countries deviate from this pattern: Austria, Switzerland, Great Britain (i.e., United Kingdom without Northern Ireland) and the Netherlands. In the first three, older people are on average both economically and culturally more conservative than young people. In the Netherlands, however, older people are less conservative than young people on both dimensions, although only slightly so in the cultural dimension.

The countries are also marked by different symbols depending on their time of democratisation. Those countries that have been democratised before or slightly after World War II tend to show a smaller age ratio in the cultural dimension than the other countries. This means that the difference between older people and the young in terms of the cultural dimension is much smaller in more established democracies.

In a nutshell based on our measures, older people are on average and in most European countries economically less conservative than young people. They are also culturally more conservative, broadly, than their young country peers.

Are there any explanations for these patterns? First of all, the distribution of certain social characteristics is different among the elderly than it is among younger people. For instance, the group of older people is more female than younger groups due to gender-specific mortality rates and less well educated due to recent expansions of educational possibilities. Being female and lower educated are both associated with less economic conservatism than being male or higher-educated. At the same time, women and lower educated people are also less likely to be economically conservative, supporting the notion that the differences observed may be due to the composition of the groups.

Figure 1: Age ratios of support for redistribution (economic progressivism) and support for diversity of sexual orientation (cultural progressivism)

Figure 1Legend: Diamonds=Democracies since before 1945, rectangles=democracies since after 1945 and before 1961, circles=democracies since the 1970s, crosses= democracies since 1989. For country acronyms, see appendix.

There are further explanations for age-related differences with regard to political values. Most importantly, older people are members of a different cohort than younger people. This means that individuals who grew up during the same time, given the same historical context, share similar experiences that shapes them in late adolescence and early adulthood. Political scientists use the term “political generations” to refer to causal mechanism. These common experiences are tremendously shaped by national circumstances and political history. Being a member of a birth cohort in one country can shape an individual rather differently than being the member of the same birth cohort in another country. If these experiences were all idiosyncratic to a national context, we would not see such a common pattern across countries. Instead, there are some cohort experiences that have a similar political effect across European countries. World War II and its aftermath is one such common experience. We know that the experience of death and violence in World War II shaped the collective experience in Eastern Europe and the longing for safety in the European Union.

More importantly in the context of our discussion about conservatism, there have been broad developments in Europe that shaped the ways in which members of different cohorts relate to politics. One of these broad developments is socio-economic modernisation and democratisation (Inglehart 1997). This is a broad development at the social, economic and political level through which individuals grow more individualistic, more cosmopolitan and more accepting of diversity. This development catches cohorts differently, such that it is mostly those cohorts whose members are still young and can still be shaped by this change. When we look at a snap shot of younger and older people as we did with our data, this can explain the varying degrees of cultural conservatism among older people. Their cohorts have been less impacted by this development than cohorts of younger people. Thus, it is not a coincidence that the richer and, according to this theory, socio-economically more developed countries in Europe (Western Europe) tend to be more on the right of the x-axis in Figure 1. The further along societies are in the process of socio-economic development, the smaller the gap in cultural conservatism between younger and older people is. Lithuania, Greece, Estonia and Slovakia show culturally much more conservative older people relative to younger people in their countries because they are, according to this theory, less developed (with GDP per capita being a simple indicator of that). Iceland, the Netherlands and Belgium, in contrast, show a rather low level of difference.

This co-evolution with socio-economic development is remarkable because the social status of older people tends to decline with increasing modernisation. In pre-modern society, the social status of older men (not women!) as the heads of households was still high (Foner 1984). This status declined with increasing industrialisation and was finally removed altogether with the introduction of the modern welfare state, which allowed all individuals to seek their own material fortunes without the family having to be the main safety net.

In other words, if older people are more conservative than younger people, this is much more likely due to their cohort membership than to where they are in the life cycle. But these differences are not stable across time. For instance, analyses of the Brexit vote in the United Kingdom demonstrate that older people in 2016 were much more likely to vote for the Leave option than for the Remain option, most likely because older people belonged to cohorts that were more fond of the traditional nation-state than the supranational structure of EU governance (Goodwin and Heath 2016; Hobolt 2016).

So, is there anything left to say about older people and conservatism? There is some evidence about voters being more open to newer parties in the first elections of their lives. Later, if these parties do not make it into the establishment, they tend to shift to more established parties. As voters have had more opportunities to cast a vote, they grow increasingly disenchanted with wasting their votes on new parties. However, this effect, which one might call status-quo conservatism, is small and can only be demonstrated for countries with proportional representation systems (Goerres 2009).

Appendix

Table 3: Support for income redistribution by age group

[table id=8 /]

Table 4: Support for free expression of homosexual life styles

[table id=9 /]

 

References

Foner, Nancy. 1984. Ages in Conflict: A Cross-Cultural Perspective on Inequality between Old and Young. New York: Columbia University Press.

Goerres, Achim. 2009. The Political Participation of Older People in Europe: The Greying of Our Democracies. Basingstoke: Palgrave Macmillan.

Goodwin, Matthew J, and Oliver Heath. 2016. „The 2016 Referendum, Brexit and the Left Behind: An Aggregate‐level Analysis of the Result.“ The Political Quarterly 87 (3):323-32.

Hobolt, Sara B. 2016. „The Brexit Vote: a Divided Nation, a Divided Continent.“ Journal of European Public Policy 23 (9):1259-77.

Inglehart, Ronald. 1997. Modernization and Postmodernization. Cultural, Economic, and Political Change in 43 Societies. Princeton: Princeton University Press.

Shapiro, Fred. 2011. John Adams Said it First  [cited 18 April 2017].

Veröffentlicht unter Ageing Democracies? | Verschlagwortet mit , , , , , | Schreib einen Kommentar

69,11%

Von Jakob Kemper

69,11% betrug die Wahlbeteiligung bei der Bundestagswahl am 24. September 2017 bezogen auf die Voting Age Population (VAP), also die gesamte in Deutschland lebende Bevölkerung über 18 Jahren. Das ist im Vergleich mit anderen Staaten gar nicht so übel, wie die folgende Grafik des Pew Research Centers zeigt (Deutschland ist hier noch mit Zahlen von 2013 aufgeführt).

Note: Voting-age population (VAP) turnout is derived from estimates of each country's VAP by the International Institute for Democracy and Electoral Assistance. Registered-voter (RV) turnout derived from each country's reported registration data. Because of methodology differences, in some countries estimated VAP is lower than reported RV. Current voting-age population estimates for Iceland, Japan and Turkey unavailable. *National law makes voting compulsory. In addition, one Swiss canton has compulsory voting. Source: Pew Research Center calculations based on data from International Institute for Democracy and Electoral Assistance, European Election Database, United States Election Project, Office of the Clerk of the U.S. House of Representatives and various national election authorities.

Note: Voting-age population (VAP) turnout is derived from estimates of each country’s VAP by the International Institute for Democracy and Electoral Assistance. Registered-voter (RV) turnout derived from each country’s reported registration data. Because of methodology differences, in some countries estimated VAP is lower than reported RV. Current voting-age population estimates for Iceland, Japan and Turkey unavailable.
*National law makes voting compulsory. In addition, one Swiss canton has compulsory voting.
Source: Pew Research Center calculations based on data from International Institute for Democracy and Electoral Assistance, European Election Database, United States Election Project, Office of the Clerk of the U.S. House of Representatives and various national election authorities.

Wie man sieht, schwanken die Zahlen auch innerhalb der OECD-Mitgliedsstaaten ganz erheblich. Wieso ist etwa in manchen Staaten die Differenz zwischen der Wahlbeteiligung registrierter Wähler und der VAP so groß? Am Beispiel der USA ist ein Strang der Erklärungen gut zu erläutern: Hier muss sich, anders als in Deutschland, der Bürger um die Registrierung zur Wahl kümmern, was für ihn die Kosten zur Wahl zu gehen erhöht und zu einer geringeren Motivation wählen zu gehen führen kann. Wenn aber einmal der erste Schritt gemacht ist, dann sollten sich die (zeitlichen, organisatorischen) Kosten der Registrierung auch gelohnt haben, weshalb man eine recht hohe Wahlbeteiligung der registrierten Wähler erkennen kann. Dieser Erklärungsansatz der Rationalen Wahl ist in der Wahlforschung aber durchaus umstritten, daneben gibt es etwa noch sozioökonomische (Arbeitslose etwa wählen typischerweise seltener), politische (Bei Wahlen, die sehr knapp scheinen, wählen Bürger eher), institutionelle (Bei Wahlpflicht wählen Bürger tendenziell häufiger) und individuelle (Ältere Bürger wählen eher) Erklärungsansätze.

Literaturtipp: Das Handbuch Wahlforschung (Hrsg. Jürgen W. Falter und Harald Schoen) gibt eine gute deutschsprachige Einführung in das Gebiet der Wahlforschung.

Für einen Überblick über die Wahlbeteiligung weltweit sei der Bericht „Voter turnout trends around the World“ des International Institute for Democracy and Electoral Assistance empfohlen.

 

 

Literatur / Weblinks:

DeSilver, D. (2017, Mai 15). U.S. trails most developed countries in voter turnout. Abgerufen 6. November 2017, von http://www.pewresearch.org/fact-tank/2017/05/15/u-s-voter-turnout-trails-most-developed-countries/

Falter, J. W., & Schoen, H. (Hrsg.). (2014). Handbuch Wahlforschung. Wiesbaden: Springer Fachmedien Wiesbaden. https://doi.org/10.1007/978-3-658-05164-8

International Institute for Democracy and Electoral Assistance. (2017). Voter Turnout Database. Abgerufen 6. November 2017, von https://www.idea.int/data-tools/data/voter-turnout

Solijonov, A., & International Institute for Democracy and Electoral Assistance. (2016). Voter turnout trends around the World. Stockholm: IDEA. Abgerufen von https://www.idea.int/sites/default/files/publications/voter-turnout-trends-around-the-world.pdf

Veröffentlicht unter politische Zahlen im Fokus | Verschlagwortet mit , , , | Schreib einen Kommentar

The political participation of older people is a multidimensional phenomenon within a changing context of political participation that varies widely across Europe.

This essay is the second of a series written by Achim Goerres for the project “Ageing Democracies? Political Participation and Cultural Values Among the Elderly in Europe” financed by the Open Society Foundation. The complete project report with all essays and the reports written by the other project members can be found here.

At the onset of many liberal democracies, participation in democratic elections was the only political action that citizens would engage in. All other forms of political participation were extremely rare. This has changed dramatically since the 1970s in Europe.

In mainstream political science, political participation is an individual action intended to affect political outcomes. These outcomes can be policies, political institutions or the selection of political personnel. Purely expressive actions, such as waving a flag, do not fall under this definition. Within this definition, there are a wide variety of different political actions. Here, we will only look at four that are among the most common forms of low-intensity political participation: voting, contacting a public official or politician, signing a petition and taking part in a demonstration. They are low-intensity forms of participation because they require relatively low levels of cognitive and physical abilities. The first two are often called institutionalised political participation, since they involve the use of formalised institutions of representative democracy. The second two are often called non-institutionalised participation, since they do not require formalised forms of participation.

Political participation is changing across Europe in manifold ways that vary from country to country. Broadly, individuals are becoming less willing to subscribe to whole manifestoes of political goals, such as those offered by political parties. Instead, they are increasingly interested in single political issues, such as the protection of the environment. There is also less willingness to attach oneself to a specific type of political participation in the long run. This makes non-institutionalised forms of political participation increasingly popular. Voting and some forms of institutionalised participation, like party activism, are on the decline. This is not true for all countries and all parties, but the tendency is clearly there. Across European countries, the levels of non-institutionalised participation seem to depend on the length of democratic epoch in a country. As I will show below, the longer ago the transition to democracy was, the more common non-institutionalised forms of participation are.

How we participate in politics is linked to the patterns we observe today and what we first learned when we were more impressionable young adults. This is a typical cohort phenomenon, wherein older people are different because they belong to a certain birth cohort who shared different social and political experiences at a young age. This is not a deterministic relationship, however. There are certainly very active older people in non-institutionalised forms of participation. But generally, today’s older people use these forms less because these forms of participation were less common when they grew up. Nonetheless, this particular participation gap between older and young people is closing. In 1981, the non-institutionalised participation rate of older men in Western Europe was roughly 46% that of men under the age of 30. By 2000, this ratio had risen to 84%. Among women, the change was from 24 % to 52 % over the same time period (Goerres 2009: 129). As I will show below, there are even some European countries, such as Switzerland or Sweden, where the gap has reversed and older people are now more likely to use forms of non-institutionalised participation than young people.

That the manner of the overall participation process is changing is not only remarkable as such, but it carries an inherent message about social inequalities. High-frequency forms of political participation, such as voting, are less unequal than low-frequency forms of participation in terms of the social profile of those who resort to them. Moreover, non-institutionalised forms of participation are also higher education forms of political participation. So even though these forms are becoming more popular, they have a strong bias against people with a lower educational level than voting does. Thus, if older people are more likely to use institutionalised forms of participation such as voting, the social representation of older people is less biased than that of young people, who are more likely to use high-bias forms of participation. The effects of these differences on representation bias are fairly unexplored.

Let us examine age differences in political participation in more detail. We start by identifying four groups: (1) those who are inactive politically (Not active), meaning that they do not engage in any form of political activity; (2) those who only use non-institutionalised forms of participation (only non-institutionalised participants); (3) those who vote and do nothing else (Only Voters); and (4) those who vote and do at least one other activity, whatever that may be (Vote Plussers). For those groups, we use four different political actions that will also form the basis of the more nuanced analysis to follow: voting, contacting a public official or politician, signing a petition and demonstrating in the streets.

Table 1: Types of political participation across age groups in Europe

Not Active

Only Non-Institutionalised Participants

Only Voters

Vote Plussers

Total

Older People (60+ yrs)

13

2

57

28

100

Middle-aged (30-59 yrs)

17

6

42

36

100

Young (18-29 yrs)

28

9

36

27

100

Across all Age Groups

17

5

46

32

100

Table 1 shows the four activity groups across 27 European countries weighed by their population size. If you want to know just four numbers for the political participation of older Europeans in free European democracies, they can be found in the first row. Thirteen percent of older people in Europe are not politically active at all. Though this may seem like a lot, it is very little compared to the 28% of young people or the 17% of middle-aged people in Europe who do not participate. Meanwhile, only using non-institutionalised forms of participation is preferred by only a small fraction (2%) of older people. Among the young, however, 9% only use non-institutionalised forms of participation. Taken with the 28% of totally inactive young people, this means that nearly two-fifths of the European youth are not inside the electoral process.

In contrast, the majority of older people (57%) only participate in politics by voting, substantially more than the 42% among the middle-aged or the 36% among the young who participate in this way. Interestingly, however, 28% of older people use voting and at least one other form of political participation, a number that is very similar among the youngest group (27 %), but much lower than for the middle-aged (36%). Thus, it can be said that older people are much more firmly placed inside the electoral process than other age groups (85%), but that the middle-aged show more strength among the vote-plussers. The young are clearly the least active of the three age groups.

Figure 1: Distribution of activity types across European countries and Europe

ageing_democracies_ingles-digitalFigure 1 reveals the distribution of these four political activity groups across all 27 European countries. Lithuania has the least politically active older people whereas Iceland has the most active older people on the right. Those who only vote are the dominant activity is most European countries. Those who are active but outside of electoral politics constitute a small fraction in all European countries.

Institutionalised political participation across countries

Let us now turn to some detailed cross-country comparisons. This statistical analysis is based on all European countries that were in the European Social Survey between 2010 and 2014 and considered free democracies, a total of 27 countries. For some countries we have data from three different surveys and for others just from one.

The self-reported voting participation of older people (60+) has a minimum of 70% in Lithuania and a maximum of 96% in Denmark, with a mean of 84% (see Table 2). So while older people have high voter participation rates across Europe, there are strong differences between older people in different countries.

When we look at the ratio of elderly voting participation and young (ages 18-29) voting participation , we see that the ratio stands almost exactly at 1.0 for Italy and Belgium, meaning that in these two countries the voting rates for the two age groups are the same. This makes sense because both countries have mandatory voting with either weak enforcement (Italy) or strong enforcement (Belgium). However, for all other countries, the ratio is above 1 with an average of 1.38 and a maximum of 2.07 in Lithuania. This means that older people in Lithuania are 2.07 times more likely to vote than young people.

Overall in Europe, older people are more likely to vote than young people. The reasons for this are well-known. First, as voters have had more opportunities to vote, they get used to voting (Goerres 2007). The occasional non-voter tends to become a regular voter. This is an age effect that is likely to linger for some time to come. In addition, in many countries older people belong to cohorts that were socialised into a stronger sense of duty to vote compared to more recent cohorts (Franklin 2004). This is a historical trend that can reverse. It could be that, in the future, members of older cohorts will have a lower likelihood of voting than younger cohorts.

Table 2: Voting participation by age groups

Country Name

Country Code

Overall

Level among older people (60+)

Level among middle-aged (30-59 yrs)

Level among young (18-29 yrs)

Ratio old to young

Lithuania

LT

59

70

60

34

2.07

Britain

GB

71

85

70

44

1.92

Croatia

HR

75

83

81

43

1.91

Ireland

IE

74

88

77

48

1.83

France

FR

73

86

71

48

1.78

Switzerland

CH

66

77

65

46

1.68

Austria

AT

77

87

77

59

1.47

Cyprus

CY

83

92

84

63

1.47

Czech Republic

CZ

63

71

63

49

1.44

Slovenia

SI

71

79

72

55

1.44

Portugal

PT

72

80

70

58

1.39

Iceland

IS

89

95

91

71

1.34

Bulgaria

BG

77

82

76

62

1.34

Slovakia

SK

75

81

76

61

1.33

Estonia

EE

70

74

72

56

1.32

Finland

FI

83

88

82

69

1.28

Netherlands

NL

83

88

83

69

1.27

Greece

GR

81

86

82

69

1.24

Poland

PL

70

75

71

61

1.22

Norway

NO

88

92

89

76

1.22

Spain

ES

80

85

80

70

1.22

Hungary

HU

73

74

77

61

1.21

Germany

DE

83

88

83

73

1.20

Denmark

DK

93

96

94

81

1.19

Sweden

SE

92

95

93

84

1.13

Italy

IT

82

80

84

77

1.04

Belgium

BE

89

88

90

90

0.97

Mean

78

84

78

62

1.41

Minimum

59

70

60

34

0.97

Maximum

93

96

94

90

2.07

In Figure 2, we can see that the relationship between the ratio of voting levels is strongly correlated with the overall level of reported voting participation. The more common voting is in a country, the smaller the difference is. This is in parts a mechanical effect. If an activity is almost carried out by everyone, sizeable groups in that country cannot differ much in their activity levels. In other words, if the voting differential in participation between young and old is seen as a problem, increasing the overall turnout will tend to narrow the gap here.

Figure 2: Scatter plot overall level of reported voting and ratio of the levels of older people and young people

Fig2Legend: Diamonds=Democracies since before 1945, rectangles=democracies since after 1945 and before 1961, circles=democracies since the 1970s, crosses= democracies since 1989. For country acronyms, see Table 2.

Figure 2 also shows the timing of democratisation in each of these countries. Countries that democratised after 1989 tend to have lower levels of voter turnout overall and a stronger distortion of voting in favour of older people. In other words, more recent democracies are characterised by a voting process in which older people play a disproportionally stronger role compared to young people.

Let us turn to the second type of institutionalised political participation (Table 3, with more detailed results in the appendix). Whether people contacted a politician or public official varies a lot between countries. It is used very little in Bulgaria (5%) and very much in Iceland (28 %), with an overall average of 14%. Among older people, Croats are the least likely to use that channel (4%) compared to older Icelanders (26%).

When we look at the ratio between the participation level of older people divided by that of young people, we see much more dramatic cross-country differences than we did for voting participation. Six countries show a pattern in which older people are less likely to use this channel of participation. Croatia, for instance, has a ratio of only 0.54, meaning older Croats are 46% less likely to use this form of political action than young Croats. Meanwhile, in Belgium there is absolutely no difference. In contrast, older people were more likely to use this channel in 20 countries. The maximum is in Lithuania, where older people are 2.78 times more likely to contact their public officials than young people.

Table 3: Contacting a Public Official or Politician by Age Group

Overall

Level among older people (60+)

Level among middle-aged (30-59 yrs)

Level among young (18-29 yrs)

Ratio old to young

Mean

14

13

17

9

1.51

Minimum

5

4 (Croatia)

5

3

0.54 (Croatia)

Maximum

28

26 (Iceland)

31

19

2.78 (Lithuania)

For details, see the appendix.

Figure 3: Scatter plot of the overall reported level of contacting a public official or politician and the age ratio

Fig3

Diamonds=Democracies since before 1945, rectangles=democracies since after 1945 and before 1961, circles=democracies since the 1970s, crosses= democracies since 1989. For country acronyms, see Table 2.

Figure 3 shows that there is almost no relationship between the overall level of contacting public officials and politicians and the ratio level of elderly to young people. The fitted line is almost flat. However, we do see that older democracies (i.e. those that have existed since either before Wold War II or shortly after) cluster on the right-hand side of the graph. This means that they have higher levels of contacting public officials and are less heterogeneous in terms of the ratio of elderly to young people using this form of participation. In more recently democratised countries, the level of contacting tends to be lower and the distortion of usage varies dramatically.

Non-institutionalised participation

Signing a petition is the most common form of political participation outside of the established formal channels (Table 4). Here, the range of country averages is very large. Forty six percent of older Icelanders said they had signed a petition in the previous 12 months, compared to only 3% in Hungary and Greece. The ratio is on average tilted below 1 with a mean of 0.69, a minimum of 0.29 in Portugal and a maximum of 1.03 in Switzerland. Of the 27 countries in our sample, 25 reveal a ratio below 1, meaning that older people in Europe are less likely to use this form of political participation.

Table 4: Signing a petition by age group

Overall

Level among older people (60+)

Level among middle-aged (30-59 yrs)

Level among young (18-29 yrs)

Ratio old by young

Mean

23

17

25

24

0.67

Minimum

4

3 (Hungary, Greece)

4

3

0.29 (Portugal)

Maximum

57

46 (Iceland)

63

56

1.03 (Switzerland)

See the appendix for detailed results.

Figure 4: Scatter plot of the overall level of signing a petition and the age ratio

Fig4Diamonds=Democracies since before 1945, rectangles=democracies since after 1945 and before 1961, circles=democracies since the 1970s, crosses= democracies since 1989. For country acronyms, see Table 2.

In Figure 4 we see that, in contrast, to voting there is a distinct positive relationship between the overall level of this form of political participation and the ratio. This means that the more common signing a petition is in a country, the less difference there is between whether young and older people participate in this way. At the bottom-left of the figure, we can see a cluster of six countries (Estonia, Greece, Lithuania, Poland, Portugal and Slovenia) that are all characterised by low levels of this form of participation and a low ratio of older to young participation. Here, signing a petition is predominantly carried out by young people and not at all by older people.

Demonstrating in the streets is another form of non-institutionalised participation that is much less used and more demanding in terms of time and energy than just signing a petition. In previous studies of the 1970s, demonstrating was considered a youth thing. This finding has now been partially revised. Demonstrating in the streets is far more common in countries such as Spain or Iceland, especially compared to places like Finland or the Netherlands. This is reflected among older people, too, with the highest participation rate being that of older Icelanders (16%). Meanwhile, in the Netherlands, Finland or Slovenia, only 1% of the elderly took part in demonstrations. In most countries (26 of the 27 in our sample), demonstrating in the streets is predominantly carried out by young people and not by older people. In Iceland, older people are resorting to street demonstrations as much as young people with the ratio at exactly one. The gap between young and older people, however, varies quite substantially across countries.

Table 5: Taking part in a demonstration by age group

Overall

Level among older people (60+)

Level among middle-aged (30-59 yrs)

Level among young (18-29 yrs)

Ratio old by young

Mean

7

5

8

9

0.53

Minimum

2

1 (Finland, Netherlands, Slovenia)

2

3

0.16 (Finland)

Maximum

22

16 (Iceland)

25

28

1.00 (Iceland)

Figure 5 displays the overall level of demonstrating in a country and the participation ratio of older to young people. There is a positive association between the two, but not a very strong one. The more common demonstrating is, the less difference there is between older and young people. The less popular this form of participation is, the more it is carried forward by young people. There is no discernible relationship between the timing of democracy here: countries from all groups can be found everywhere. Two interesting countries stand out, Iceland and Spain. Both show a very high level of participation in demonstrations (18 and 22%). But in Iceland, older people and young people have the same level of participation, thus showing an age gap of zero. In contrast, Spain shows a significant age gap, with the ratio standing at 0.48. Older Spanish people are 52% less likely to demonstrate compared to the level of young Spanish people. Even though both countries show high levels of demonstration, in Spain this is carried out much more by young people than it is in Iceland where this kind of political action is equally popular among young and older people.

Figure 5: Scatter plot overall level of demonstrating and the age ratio

Fig5Diamonds=Democracies since before 1945, rectangles=democracies since after 1945 and before 1961, circles=democracies since the 1970s, crosses= democracies since 1989. For country acronyms, see Table 2.

Figure 6 puts our evidence together. Here we see the mean age ratios for institutionalised and non-institutionalised participation in one graph. Where the two lines meet, a hypothetical country would show no difference in participation levels between older and young people in any of the two dimensions. Iceland comes closest to this point. In institutionalised participation, Icelandic older people are a little bit more active than Icelandic young people. After Iceland, Sweden comes closest to this two-dimensional parity. In non-institutionalised participation, it is the other way around. There are four countries where institutionalised participation is almost the same across age groups, but there are also sizeable differences in non-institutionalised participation in favour of young people. These are Belgium, Estonia, Finland and Spain. Britain shows a pattern of relatively small differences in non-institutionalised participation but very strong differences in institutionalised participation. Older British people are not very far from young British people in non-institutionalised participation. However, they are much more active in institutionalised forms of participation. Lithuania is the most unequal in terms of both measures. Older people in Lithuania are much more active than young people in institutionalised forms of participation and much more inactive in non-institutionalised participation. In other words, older and young people in Lithuania differ completely in their patterns of political participation.

Figure 6: Scatter plot of mean age ratios institutionalised and non-institutionalised forms of participation Fig6Diamonds=Democracies since before 1945, rectangles=democracies since after 1945 and before 1961, circles=democracies since the 1970s, crosses= democracies since 1989. For country acronyms, see Table 2.

To sum up the findings of this essay, older people’s political participation should be considered within a broader picture of mass political participation, since political participation is a multi-dimensional phenomenon. Most importantly, institutionalised participation through formalised routines of representative democracies is used more by older people than young people in many, but not all democracies. Exceptions include Belgium, Poland and Spain. Non-institutionalised forms of participation are characterised by a loose commitment outside of the formal channels of representative democracies. These are favoured more by young people, but in some countries like Iceland, the age gap is practically negligible. Across all modes of participation, there are a lot of differences between countries. Sometimes, there is a relationship between the levels of participation in a country and the size of the age gap. For example, in countries with higher turnout, the age gap in favour of older people is weaker. This pattern is often stratified by the age of a country’s democracy with more recently democratised countries showing lower levels of participation overall. Overall, 13% of older Europeans do not participate in politics at all and only 2% limit themselves to non-institutionalised forms of participation. This implies that 85% are firmly embedded in the electoral process.

Technical Appendix

The data are mostly taken from the Eurostat database and were provided by the team of the European Social Survey.

Estimates of participation rates and support for certain political stances are my own, based on waves 5 to 7 of the European Survey. They apply to the years 2010-14. The estimates are weighted by design weight to reflect differences in sampling techniques.

Note that the estimates are based on reported behaviour. Since citizens know that democratic participation is socially desirable, some respondents indicate their having complied with this expectation while actually not having done so in reality. This is due to conscious lying and more importantly to our inner subconscious urge to be consistent with our own image of ourselves and our behaviour.

Table 6: Full table of estimates on contacting a public official

Country Name

Country Code

Overall

Level among older people (60+)

Level among middle-aged (30-59 yrs)

Level among young (18-29 yrs)

Ratio old by young

Lithuania

LT

8

7

11

3

2.78

Czech Republic

CZ

13

14

14

6

2.19

France

FR

14

14

17

6

2.17

Slovakia

SK

9

10

10

5

2.16

Italy

IT

16

16

19

8

2.06

Britain

GB

17

19

18

9

2.06

Switzerland

CH

16

15

18

8

1.83

Austria

AT

19

20

20

12

1.75

Finland

FI

20

18

24

11

1.71

Sweden

SE

18

18

20

11

1.63

Denmark

DK

19

18

23

11

1.62

Cyprus

CY

17

15

21

9

1.59

Netherlands

NL

17

16

18

10

1.58

Norway

NO

22

20

27

13

1.55

Ireland

IE

17

19

18

13

1.51

Germany

DE

17

16

18

11

1.47

Bulgaria

BG

5

5

5

4

1.38

Greece

GR

9

7

11

5

1.38

Iceland

IS

28

26

31

19

1.33

Hungary

HU

8

7

10

6

1.25

Belgium

BE

15

12

17

12

1.00

Slovenia

SI

10

9

12

9

0.92

Portugal

PT

8

7

9

7

0.90

Spain

ES

15

9

18

11

0.86

Estonia

EE

14

10

17

13

0.81

Poland

PL

8

5

10

8

0.65

Croatia

HR

7

4

9

7

0.54

Mean

14

13

17

9

1.51

Minimum

5

4

5

3

0.54

Maximum

28

26

31

19

2.78

 

Figure 8: Full table of estimates on signing a petition

Country Name

Country Code

Overall

Level among older people (60+)

Level among middle-aged (30-59 yrs)

Level among young (18-29 yrs)

Ratio old by young

Britain

GB

34

30

37

29

1.03

Switzerland

CH

34

31

37

31

1.00

Sweden

SE

42

41

43

42

0.98

Germany

DE

35

29

39

31

0.95

Cyprus

CY

9

7

11

8

0.88

Bulgaria

BG

8

6

8

7

0.87

Iceland

IS

57

46

63

56

0.82

Czech Republic

CZ

18

13

20

17

0.81

France

FR

33

25

37

31

0.81

Austria

AT

29

25

31

31

0.79

Netherlands

NL

26

20

29

27

0.76

Hungary

HU

4

3

4

3

0.76

Belgium

BE

22

16

25

23

0.69

Ireland

IE

22

17

23

25

0.66

Norway

NO

35

27

37

42

0.64

Italy

IT

23

18

24

28

0.64

Croatia

HR

24

17

26

26

0.63

Slovakia

SK

21

15

22

25

0.62

Spain

ES

31

19

37

31

0.62

Denmark

DK

29

19

32

36

0.54

Lithuania

LT

9

5

11

12

0.44

Estonia

EE

9

5

11

12

0.43

Greece

GR

5

3

6

7

0.41

Slovenia

SI

9

5

11

13

0.39

Finland

FI

29

15

36

39

0.39

Poland

PL

11

6

12

16

0.38

Portugal

PT

8

4

10

12

0.29

Mean

23

17

25

24

0.67

Minimum

4

3

4

3

0.29

Maximum

57

46

63

56

1.03

 

Table 7: Full table of estimates on taking part in a demonstration

Country
Name

Country
Code

Overall

Level among older people (60+)

Level among middle-aged (30-59 yrs)

Level among young (18-29 yrs)

Ratio old by young

Iceland

IS

18

16

19

16

1.00

Croatia

HR

8

6

9

7

0.86

Lithuania

LT

3

2

3

3

0.80

Ireland

IE

11

9

10

14

0.70

Britain

GB

4

3

4

4

0.69

Hungary

HU

3

2

4

4

0.66

Sweden

SE

8

7

7

11

0.65

Norway

NO

10

7

11

11

0.64

Italy

IT

16

14

16

22

0.64

Czech Republic

CZ

6

4

6

6

0.63

France

FR

14

10

16

16

0.59

Belgium

BE

6

3

8

6

0.56

Slovakia

SK

3

2

3

4

0.53

Bulgaria

BG

4

3

5

5

0.53

Spain

ES

22

13

25

28

0.48

Estonia

EE

3

2

3

5

0.46

Poland

PL

2

2

2

4

0.45

Portugal

PT

5

3

6

7

0.43

Germany

DE

9

6

9

14

0.41

Cyprus

CY

5

2

5

6

0.40

Austria

AT

7

4

8

11

0.38

Slovenia

SI

3

1

4

4

0.36

Switzerland

CH

5

2

5

7

0.36

Netherlands

NL

3

1

3

4

0.34

Denmark

DK

6

3

6

11

0.25

Greece

GR

10

4

11

17

0.24

Finland

FI

2

1

2

4

0.16

Mean

7

5

8

9

0.53

Minimum

2

1

2

3

0.16

Maximum

22

16

25

28

1.00

 

REFERENCES

Franklin, Mark N. 2004. Voter Turnout and the Dynamics of Electoral Competition in Established Democracies. Cambridge: Cambridge University Press

Goerres, Achim. 2007. „Why are Older People more Likely to Vote? The Impact of Ageing on Electoral Turnout in Europe.“ British Journal of Politics and International Relations 9 (1):90-121.

———. 2009. The Political Participation of Older People in Europe: The Greying of Our Democracies. Basingstoke: Palgrave Macmillan.

Veröffentlicht unter Ageing Democracies? | Verschlagwortet mit , , , , | Schreib einen Kommentar

Können Wähler mit Migrationshintergrund die Bundestagswahl 2017 entscheiden? Aussiedler aus der Sowjetunion können eine schwarz-gelbe Koalition ermöglichen oder verhindern

Von Achim Goerres

Dieser Beitrag arbeitet eine einfache Schätzung auf, nach der aufgrund der jetzigen politischen Großwetterlage, der Größe der Gruppe der Deutschen mit sowjetischen Hintergrund (Russlanddeutsche, Aussiedler) und ihrer politischer Präferenzen die Stimmen  dieser Gruppe eine schwarz-gelbe Mehrheit ermöglichen könnten. Die türkeistämmigen Deutschen sind eine zu kleine Gruppe und haben die „falsche“ politische Färbung (eher links) zur jetzigen Großwetterlage, um einen Unterschied zu machen.

 

Eine der häufigsten Fragen, die wir vor der Bundestagswahl 2017 in unserer Forschungsgruppe (Dennis Spies, Sabrina Mayer und ich) zur ersten Deutschen Migrantenwahlstudie gehört haben, war, ob Wählerinnen und Wähler mit Migrationshintergrund die Wahl entscheiden können.

Die Antwort ist einfach: ja, klar.

Aber nicht, weil sie einen Migrationshintergrund haben, sondern weil letztendlich jede einzelne Stimme das Gesamtergebnis entscheiden kann; dass das passiert, ist sehr unwahrscheinlich aber möglich.

Da diese Antwort zwar korrekt aber unbefriedigend ist, wollen wir die Frage etwas umformulieren: können Wähler mit Migrationshintergrund aufgrund der Größe ihrer Gruppe und aufgrund der Tatsache, dass sie in Summe anders wählen als die autochthonen Deutsche, das Wahlergebnis grundsätzlich verändern?

Hier wollen wir uns auf zwei Dinge beschränken: wer in den Bundestag kommt und die Mehrheiten, die für eine Koalition notwendig sind.

Nun brauchen wir ein paar Zahlen.

61,5 Millionen Menschen sind bei der Bundestagswahl wahlberechtigt.

Darunter sind 6,3 Millionen Deutsche, die selbst nach Deutschland migriert sind oder mindestens einen Elternteil haben, der nach Deutschland migriert ist.

Gehen wir von einer Wahlbeteiligung unter den autochthonen Deutschen von 73 % aus (beim letzten Mal war es 71,5 %, ich glaube, dass die Zeiten polarisierender geworden sind und deswegen die Wahlbeteiligung wieder etwas höher sein wird).

Dann gehen (61,5-6,3)*0,73 Mio. = 55,2 Millionen autochthone Menschen zur Wahl. Nota bene: Migranten der dritten Generation zähle ich zu den Autochthonen.

In der Gruppe der Wähler mit Migrationshintergrund gehen wir von einer Wahlbeteiligung von 67,0 % aus, weil in dieser Gruppe die formelle Bildung anders verteilt ist als in der autochthonen Gruppe. Das bedeutet, dass 6,3*0,67=4,2 Mio. Migrantenwähler zur Wahl gehen.

Bei insgesamt 59,4 Millionen tatsächlichen Wählerinnen und Wählern reichen 2,97 Millionen Stimmen, um eine Partei in den Bundestag kommen zu lassen. Alle 4,2 Mio. Wähler mit Migrationshintergrund zusammen könnten also locker eine Partei in den Bundestag hieven. Es würde sogar reichen, wenn 70,7 % (2,97 von 4,2) aller tatsächlichen Migrantenwähler für eine Partei stimmten, um diese in den Bundestag zu bringen.

Das Problem ist nur: diese Gruppe ist extrem heterogen und keinesfalls geeint in dem Bemühen, EINE Partei gemeinsam zu wählen. Schauen wir uns deswegen die beiden größten Gruppen an, die türkeistämmigen Deutschen und die Deutschen mit einem sowjetischen Hintergrund (Aussiedler, Russlanddeutsche).

 

Die politische Großwetterlage am Freitag vor der Bundestagswahl

Bevor wir das tun, brauchen wir noch eine Schätzung für das Gesamtergebnis bei der Bundestagswahl, um ein wahrscheinliches Szenario vor Augen zu haben. Dafür nehmen wir die Mittelwerte der sieben großen Umfragehäuser von wahlrecht.de vom 21.9.2017. Dass bei den Befragten auch Migranten dabei sind und es Überhangmandate im Bundestag gibt, vernachlässigen wir.

 

Zweitstimmen ohne Migranten

Sitzanteile ohne Migranten

CDU/CSU

36.4%

38.0%

SPD

22.0%

23.0%

GRÜNE

7.8%

8.1%

FDP

9.5%

9.9%

DIE LINKE

9.6%

10.0%

AfD

10.4%

10.9%

Sonstige

4.4%

0.0%

Koalitionen nach Zweitstimmen

Koalitionen nach Sitzanteilen

GroKo

58.4%

61.0%

Jamaica (schwarz-gelb-grün)

53.6%

56.1%

Schwarz-Gelb

45.9%

48.0%

Schwarz-Grün

44.1%

46.2%

Ampel (rot-gelb-grün)

39.3%

41.1%

Rot-rot-grün

39.4%

41.2%

 

Man sieht, dass alle kleinen genannten Parteien weit weg von der 5-Prozenthürde sind. Dass Wähler mit Migrationshintegrund in dieser Hinsicht einen Unterschied ausmachen, ist also unwahrscheinlich.

Interessanter wird es bei den möglichen Koalitionen. GroKo und Jamaica sind safe, egal was die Wähler mit Migrationshintergrund machen. Sie werden 61,0 bzw. 56,1 % der Sitze auf sich vereinen und machbar sein. Am unteren Ende sind Ampel und Rot-rot-grün sehr weit davon entfernt, eine Mehrheit zu haben.

Es bleiben also – Stand heute – nur zwei „wackelige“ Szenarien, bei denen Migrantenwähler den Unterschied machen können:

  1. Ob eine Schwarz-gelbe Mehrheit möglich ist
  2. Ob eine Schwarz-grüne Mehrheit möglich ist

 

Die Einflussmöglichkeiten der beiden größten Gruppen bei der Bundestagswahl 2017

Türkeistämmige Deutsche

Es gibt etwa 700.000 Wählerinnen und Wähler, die selbst aus der Türkei migriert sind oder mindestens einen Elternteil mit diesem Merkmal haben. Bei einer geschätzten Wahlbeteiligung von 67 % gehen nur 469.000 türkeistämmige Wählerinnen und Wähler zur Wahl. Das sind 0,8 % aller abgegebenen Stimmen. Wenn alle diese Stimmen an eine Partei gingen, würde diese Partei nicht im Parlament vertreten sein. Diese Gruppe kann somit für keine der wackeligen Szenarien Stand heute der Königsmacher sein. Dass sie, nach allem, was wir bisher wissen, auch eher links stimmen, spielt aufgrund der politischen Großwetterlage und ihrer geringen numerischen Größe keine Rolle.

 

Deutsche mit sowjetischem Hintergrund

Von diesen gibt es etwa 2,0 Millionen Wahlberechtigte. Bei einer geschätzten Wahlbeteiligung von 67 % gehen 1,34 Mio. Wählerinnen und Wähler mit diesem Hintergrund zur Wahl. Das sind 2,3 % aller Stimmen. Auch für sie gilt: selbst wenn sie alle völlig geeint stimmten, würden sie es nicht schaffen, eine  Partei allein in den Bundestag zu bringen.

Anders als bei den Türkeistämmigen aber könnten sie das Zünglein an der Waage für schwarz-gelb und vielleicht sogar für schwarz-grün sein. Wenn sie alle für CDU/CSU/FDP stimmten, wären sie die Königsmacher für schwarz-gelb und brächten 2,4 % der Sitze. Wenn sie alle für schwarz-grün stimmten, brächten sie die Koalition auf 48,6 % der Sitze, also immerhin ziemlich nahe an die benötigte Mehrheit.

Wenden wir uns nun der Frage zu, wie wahrscheinlich es ist, dass die Deutschen mit sowjetischem Hintergrund für CDU/CSU, FDP oder B‘90/Grün stimmen. Nach allem, was wir bisher wissen, stimmen die Aussiedler deutlich häufiger für konservative Parteien und auch die FDP als die Autochthonen.

Zusammengefasst kann es also sehr gut sein, dass die Aussiedler und ihre Nachkommen bei der Bundestagswahl 2017 die Mehrheit für eine schwarz-gelbe Mehrheit oder sogar eine schwarz-grüne bringen, weil sie tendenziell konservativer wählen und viele Stimmberechtigte in ihren Reihen haben. Die türkische Gruppe ist zu klein und gegeben die politischen Verhältnisse im Moment in Deutschland „in der falschen Farbe gefärbt“ (nämlich tendenziell links), um einen Unterschied zu machen.

Alle diese Schätzungen basieren auf zahlreichen Annahmen, die falsch sein können. Es lohnt sich also trotzdem noch wählen zu gehen.

 

Quellen:

Bundeswahlleiter, https://www.bundeswahlleiter.de/info/presse/mitteilungen/bundestagswahl-2017/01_17_wahlberechtigte.html

Mikrozensus, Berechnungen für IMGES (Immigrant German Election Study)

Wahlrecht.de, 21.9.2017, für die Zusammenfassung der Umfrageergebnisse

Veröffentlicht unter Allgemein | Verschlagwortet mit , , , , , | Schreib einen Kommentar

Wählerinnen und Wähler nehmen TV-Duell zur Bundestagswahl 2017 zum Anlass, sich über Parteipositionen zu informieren

Von Sebastian Krause.

Die Analyse von Google Trends-Daten zeigt einen Höhepunkt des Suchaufkommens im Anschluss an das „TV-Duell“ zur Bundestagswahl 2017. Mithilfe der Suchmaschine wird insbesondere nach den Wahlprogrammen von FDP und AfD gesucht.

In meinen Analysen zur Bundestagswahl 2013 hatte ich gezeigt, dass Internetnutzungsdaten und insbesondere Google-Suchanfragen eine geeignete Datengrundlage sind, um Vorhersagen für die Ergebnisse der Wahl zu berechnen. Während für diese Analysen Suchanfragen in Form absoluter Zahlen verfügbar waren, stehen diese für die Wahl 2017 noch nicht zur Verfügung. Davon unberührt bietet der Online-Dienst Google Trends bereits die Möglichkeit, die aktuelle Popularität einzelner Suchbegriffe im Zeitverlauf zu analysieren, auch wenn die Daten lediglich in Relation zum totalen Suchaufkommen präsentiert werden.

Die Berechnungen zur Wahl 2013 hatten ergeben, dass der Zusammenhang zwischen den parteibezogenen Suchanfragen und dem späteren Stimmenanteil am größten ausfällt, wenn die Parteinamen mit Begriffen verknüpft werden, die einen Bezug zur Wahl erkennen lassen (z. B. im Falle der SPD: „SPD Wahlprogramm“). Auf diese Weise erhöht man die Wahrscheinlichkeit nur diejenigen Suchanfragen zu erhalten, die Wählerinnen und Wähler hinsichtlich ihrer Wahlentscheidung tätigen.

Abbildung 1: Suchaufkommen der letzten 30 Tage. Eigene Darstellung, Quelle: Google Trends

Abbildung 1: Suchaufkommen der letzten 30 Tage. Eigene Darstellung, Quelle: Google Trends

Bei Google Trends kann das Suchaufkommen für fünf verschiedene Suchbegriffe gleichzeitig abgerufen werden, um das relative Suchvolumen zu vergleichen. Die Suchanfragen für einzelne Suchbegriffe werden so ins Verhältnis zum Suchvolumen desjenigen Suchbegriffs gesetzt, nach dem am häufigsten gesucht wurde. Da der nächste Bundestag gemäß der Sonntagsfrage aus voraussichtlich sechs Fraktionen bestehen wird, musste eine Partei bei den Analysen vernachlässigt werden. Bündnis 90 / die Grünen bildet in den derzeitigen Umfragen das Schlusslicht der Parteien, die es voraussichtlich über die Fünf-Prozent-Hürde schaffen werden. Zudem ist das Suchaufkommen für diese Partei aufgrund unterschiedlicher Schreibweisen des Pateinamens schwerer zu erfassen [1]. Abgerufen werden daher die Suchvolumina der Begriffe „SPD Wahlprogramm“, „CDU Wahlprogramm“ (durch eine Verknüpfung mit der CSU erhält man deutlich weniger Anfragen), „FDP Wahlprogramm“, „Linke Wahlprogramm“ (mit „Die Linke Wahlprogramm“ erhält man deutlich weniger Anfragen) und „AfD Wahlprogramm“. Zusätzlich wurden nur die Suchanfragen gefiltert, die aus Deutschland stammen.

In Abbildung 1 wird der Suchverlauf für die letzten 30 Tage dargestellt. Zunächst ist zu erkennen, dass ein Anstieg der Suchanfragen ab Anfang August 2017 zu verzeichnen ist – ein Zeichen dafür, dass die verwendeten Suchbegriffe tatsächlich mit dem Interesse an der Bundestagswahl zusammenhängen. Bemerkenswert ist ein deutlicher Ausschlag der Suchanfragen für alle Parteien am 3. September. An diesem Tag fand das „TV-Duell“ zwischen Angela Merkel und Herausforderer Martin Schulz statt, das etwa 16 Mio. Menschen am Fernseher verfolgten [2]. Verkürzt man den betrachteten Zeitraum auf die vergangenen sieben Tage (Abbildung 2), erhält man auf der Zeitachse die Uhrzeit des Suchaufkommens. So beginnt der Anstieg bereits ab etwa 16 Uhr, um im Anschluss des Duells gegen 22 Uhr seinen Höhepunkt zu erreichen. Interessanterweise steigen hierbei nicht nur die Suchanfragen nach den Parteien der Duellanten, sondern auch die nach allen anderen Parteien. Insbesondere das Suchaufkommen für die AfD erreicht hier seinen bisherigen Höhepunkt.

Zieht man das Verhältnis der parteispezifischen Suchanfragen heran, so werden derzeit für AfD und FDP die meisten Suchanfragen getätigt. Dieser Befund spiegelt sich auch in den von Google Trends bereitgestellten durchschnittlichen Suchanfragen wider, die im Übrigen auch das Suchinteresse an SPD und CDU in etwa gleich auf sehen. Dies kann als Hinweis darauf interpretiert werden, dass die Zustimmung zu den genannten Parteien in derzeitigen Umfragen unterschätzt wird.

Abbildung 2: Suchaufkommen der letzten 7 Tage. Eigene Darstellung, Quelle: Google Trends

Abbildung 2: Suchaufkommen der letzten 7 Tage. Eigene Darstellung, Quelle: Google Trends


[1] Probehalber wurden die Suchanfragen für „Linke Wahlprogramm“ durch Suchanfragen für „Grüne Wahlprogramm“ ersetzt. Auch bei Google-Trends ist das Suchaufkommen für Bündnis 90 / die Grünen am geringsten und liegt unter den Anfragen für die Linke.

[2] Quelle: http://www.spiegel.de/kultur/tv/tv-duell-mit-merkel-und-schulz-hat-16-11-millionen-zuschauer-a-1165996.html

Veröffentlicht unter Allgemein | Verschlagwortet mit , , | Schreib einen Kommentar