Unequal participation: Why workers don t vote (anymore) and why it matters Political and Economic Inequality: Concepts, Causes and Consequences Armin Schäfer Zürich, 28.1.2016
The increase of income inequality (23 countries) 120 115 110 105 100 1980 1990 2000 2010 Market income Household income Data: SWIID. 1980=100.
Inequality Feedback Loop Income inequality Unequal responsivenes s Unequal participation Unequal representation
Outline I. Income inequality and unequal participation (cross section) II. Unequal turnout in Germany III. Unequal representation IV. Opinion differences between social classes V. Unequal Responsiveness?
Income inequality and self-reported voter turnout (87 elections, 23 countries, ~1980-2010) 1.9 Reported turnout.8.7.6 20 24 28 32 36 40 Income inequality (Gini) Note: This graph shows the predicted probability to vote calculated from a logistic regression based on 87 elections in 23 countries over three decades. It controls for age, age², gender, unemployment, union membership, and several country level variables.
Income inequality and voter turnout (23 countries) 1.9.8.7.6 20 24 28 32 36 40 Inequality (Gini index) Education high medium low Note: This graph shows the predicted probability to vote calculated from a logistic regression based on 87 elections in 23 countries over three decades. It controls for age, age², gender, unemployment, union membership, and several country level variables.
Income inequality and voter turnout (23 countries) 1.9.8.7.6 20 24 28 32 36 40 Inequality (Gini index) Income quintiles highest 2. 3. 4. lowest Note: This graph shows the predicted probability to vote calculated from a logistic regression based on 87 elections in 23 countries over three decades. It controls for age, age², gender, unemployment, union membership, and several country level variables.
UNEQUAL TURNOUT IN GERMANY
Unequal turnout in German general elections 1 Income 1 Education 1 Subjective class.9.9.9.8.8.8.7.7.7.6.6.6 1983 1987 1990 1994 1998 2002 2005 2009 2013 1983 1987 1990 1994 1998 2002 2005 2009 2013 1983 1987 1990 1994 1998 2002 2005 2009 2013 low medium high Source: ALLBUS 1980-2014. The graph shows predicted probabilities based on a logistic regression that includes age, age², gender, political interest, and a trend variable as controls.
Social Class and turnout in German general elections 1983 1987 1990 1994 1998 2002 Social class Worker Technician Non-manual routine Lower service Higher service Self-employed Employers 2005 2009.6.7.8.9 1 Reported turnout Source: ALLBUS 1980-2014. Social class is based on occupations and follows the Goldthorpe- Erikson scheme.
UNEQUAL (DESCRIPTIVE) REPRESENTATION
Social Class and Party Membership in Germany (2000-2010) Semi- and Unskilled Workers Routine Non-Manual Skilled Workers Lower-grade Professionals Self-Employed Higher-grade Professionals 0.02.04.06.08 Source: ALLBUS 1980-2012.
Workers in Parliament (Bundestag) Civil Servants (Beamte) Political Organization Freelancers Salaried Employees Self-Employed Other Public Sector Employees Missing Workers 0 50 100 150 200 Source: Kintz 2010: 495-496.
OPINION DIFFERENCES BETWEEN SOCIAL CLASSES
Range of opinion differences Data: Elsässer-Hense-Schäfer 2015. Each vertical line stands for one survey item (N=255). The box includes 50 percent of all cases and the vertical line inside the box is the median. Self-employed Civil servants Higher ranking white collar Routine white collar Skilled korkers 0 10 20 30 40 50 60 Percentage point differences to unskilled workers
Opinion differences within different groups Unskilled workers vs. self-employed Low vs. high education Women vs. men East vs. west German 0 10 20 30 40 50 60 Percentage point difference Data: Elsässer-Hense-Schäfer 2015.. Each vertical line stands for one survey item (N=255). The box includes 50 percent of all cases and the vertical line inside the box is the median
UNEQUAL RESPONSIVENESS?
How the Bundestag decides 1 Selfemployed 1 Unskilled workers.8.8.6.4.2 Probability of policy change.6.4.2 0 0.1.2.3.4.5.6.7.8.9 1 Share of self-employed who support policy change 0 0.1.2.3.4.5.6.7.8.9 1 Share of workers who support policy change Data: Elsässer-Hense-Schäfer 2015. The graph shows the predicted probability of a policy change depending on the share of a social class who supports this change. The graph is based on a logistic regression.
How the Bundestag decides 1 Civil servants 1 Skilled workers.8.8.6.4.2 Probability of policy change.6.4.2 0 0.1.2.3.4.5.6.7.8.9 1 Share of civil servants who support policy change 0 0.1.2.3.4.5.6.7.8.9 1 Share of workers who support policy change Data: Elsässer-Hense-Schäfer 2015. The graph shows the predicted probability of a policy change depending on the share of a social class who supports this change. The graph is based on a logistic regression.
Summary With rising social inequality, voter turnout declines. Lower turnout means unequal turnout. Turnout rates have dramatically declined for members of the working class. At the same time, workers are neither present in parties nor in Parliament. Political preferences between social classes differ systematically and more strongly so than educational groups of women and men. Political decisions do not reflect the preferences of workers but those of civil servants and the self-employed.
Unequal participation: Why workers don t vote (anymore) and why it matters Political and Economic Inequality: Concepts, Causes and Consequences Armin Schäfer Zürich, 28.1.2016
Opinion differences in different policy fields Foreign Policy Social & Cultural Issues Environment & Energy Labor Market & Social Policy Economy & Finance Migration 0 5 10 15 20 25 Opinion differences among social groups (median) Education Region Gender Social class Data: Elsässer-Hense-Schäfer 2015.
Turnout differences within the working class 1983 1987 1990 1994 1998 Income low medium high 2002 2005 2009.5.6.7.8.9 Reported turnout Source: ALLBUS 1980-2014.
Social Class and turnout in German regional elections, 1978-2011 1.8.6.4.2 0 1978 1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2011 higher civil servants lower civil servants higher employees medium-level employees lower-level employees high-skilled workers skilled workers unskilled worker Source: Based on 110 Länder elections, various surveys. The graph shows predicted probabilities that result from a multinomial logistic regression with controls for age, cohort, gender and religion.
Workers choice between Social Democracy and abstention, 1978-2011.6.5.4.3.2.1 0 SPD Abstention 1978 1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2011 Source: Based on 110 Länder elections, various surveys. The graph shows predicted probabilities that result from a multinomial logistic regression with controls for age, cohort, gender and religion.