108 Chapter 5 Variable Measure N Mean SD Source Socioeconomic University (Dummy) Respondent reports having higher-level degree (Degree = 1) ESS (2016) Income feeling Respondent’s subjective financial satisfaction ESS (2016) Unemployment (Dummy) Ever unemployed and seeking work for a period more than three months (Yes=1) ESS (2016) Job insecurity Respondent’s self-reported likelihood of becoming unemployed and looking for work within the next 12 months for at least four consecutive weeks on a 4-point scale (Very unlikely=1; Very likely=4) ESS (2016) Union membership Respondent reports (ever) being a member of a trade union or similar organisation ESS (2016) Sociocultural Ideological position Self-reported placement on a 11-point scale (1=left; 10=right) ESS (2016) Religiosity Self-reported religiosity on an 11-point scale ESS (2016) Contextual Log of GDP per capita GDP per capita, constant prices & OECD base year – 2010 OECD (2017c) Unemployment rate The share of the labour force that is without work but available for and seeking employment World Bank (2017) Social spending as a % of GDP Aggregated social expenditures as a percentage of GDP OECD (2017d) 5.5 Empirical Analysis Table 5.2 displays the multilevel ordered logit models for the dependent variable ‘the government should take steps to reduce differences in incomes’ indicating general attitudes towards redistribution. It presents six different mixed effects models, I first introduce a model using only micro-level factors and build up to including several contextual-level factors, testing the two different immigration indicators both separately and together, as well as testing an alternative control for welfare state generosity. The coefficients represent the log-odds of being in a higher category of the dependent variable relative to all lower categories. For example, a positive and significant coefficient indicates that as the independent variable
RkJQdWJsaXNoZXIy MjY0ODMw