Social Mobility and Demand for Redistribution. A Comparative Analysis

Social Mobility and Demand for Redistribution. A Comparative Analysis Antonio M. Jaime-Castillo1Ildefonso Marqu´es-Perales2 1Universidad de M´alaga 2Universidad de Sevilla Junio, 2016 XII Congreso Espa˜nol de Sociolog´ıa Federaci´on Espa˜nola de Sociolog´ıa Jaime-Castillo & Marqu´es-Perales Social Mobility and Demand for Redistribution (1 / 21) Contents 1Theoretical approach 2Data and methods Data The model 3Findings Jaime-Castillo & Marqu´es-Perales Social Mobility and Demand for Redistribution (2 / 21) Introduction Introduction The literature on preferences for redistribution has paid little attention to the effect of social mobility on the demand for redistribution, which is in contrast with the literature on class-voting. Some works have addressed this issue but no systematic test of the hypotheses connecting social mobility and preferences has been done. The fact that the majority of people in advanced societies experience social mobility, at some extent, posit the question about the way in which these experiences may affect to preferences for redistributive policies. We use the diagonal reference model (Sobel, 1985), which has been used to analyze the impact of social mobility on voting, attitudes towards ethnic minorities, fertility and cultural consumption. Jaime-Castillo & Marqu´es-Perales Social Mobility and Demand for Redistribution (3 / 21) Theoretical approach Social mobility and preferences According to the acculturation hypothesis, attitudes toward redistribution will be closer to the typical preferences of the destination class than to the preferences of the origin class (Heath, Jowell and Curtice, 1985). The self-interest approach predicts that individuals have incentives to adapt their preferences to their current socio-economic status. Mobile individuals receive pressures to adapt their attitudes to those of their destination class to maintain psychological well-being. The influence of the destination class becomes stronger as individuals get older (De Graaf, Niewbeerta and Heath, 1995). Jaime-Castillo & Marqu´es-Perales Social Mobility and Demand for Redistribution (4 / 21) Theoretical approach Social mobility and preferences According to the socialization hypothesis, attitudes towards redistribution policies should be closer to the origin class. Basic norms and values are acquired relatively soon and tend to endure across the life course (Glenn, 1980). Most of the socialization process takes place primarily early in life (Tolsma, 2009). Since core political ideas are acquired relatively early (Easton y Dennis, 1967), family was considered the main source of political attitudes (Lane, 1959). Jaime-Castillo & Marqu´es-Perales Social Mobility and Demand for Redistribution (5 / 21) Theoretical approach Social mobility and preferences The status attainment hypothesis focuses on the individual aspiration to keep the more prestigious identity (Lipset and Bendix, 1959; Bourdieu, 2002). Downward mobile individuals tend to be influenced by their origin class while upward mobile individuals tend to adapt their views to their destination class. In the context of attitudes to redistribution, individuals who experience upward mobility are expected to be more willing to reduce their taste for redistribution while individuals who experience downward mobility will be less inclined to embrace support for redistribution. Jaime-Castillo & Marqu´es-Perales Social Mobility and Demand for Redistribution (6 / 21) Data and methods Data Variables The dependent variable is support for redistribution (dichotomized from ordinal scale) The key explanatory variables are origin and destination social classes, following the EGP class schema (Erikson and Goldthorpe, 1992): I+II. Service class IIIab. Routine non-manual workers IVab. Petty bourgeoisie V+VI. Supervisors and skilled manual workers VIIab. Unskilled manual workers For the origin class, we use the highest of the status of father and of mother when the respondent was 14. Jaime-Castillo & Marqu´es-Perales Social Mobility and Demand for Redistribution (7 / 21) Data and methods Data Variables Additional controls: Gender Age group Working status Data: European Social Survey (2002-2014) Sample: The final dataset contains information on 25 different countries and about 217,000 individuals aged 18 years and older. The average effective sample size for each country is 8,677. Jaime-Castillo & Marqu´es-Perales Social Mobility and Demand for Redistribution (8 / 21) Data and methods The model Diagonal reference model The diagonal reference model can be written as: Yijk = log pijk 1−pijk =wµj+ (1 −w)µk+Xβpxip where µhdenotes the mean for core members in social class h(h=jor k) and wis estimated weight of destination. By definition the esimtated weight of origin equals (1−w). Jaime-Castillo & Marqu´es-Perales Social Mobility and Demand for Redistribution (9 / 21) Findings Social mobility and class polarization Using the probabilities of supporting redistribution for each destination class, we can compute the degree of polarization between classes in preferences for redistribution through the kappa index (Hout, Brooks and Manza, 1995), proposed within the literature on class-voting: κ=v u u t 1 J J X j=1 (pj−¯p)2 The kappa index reflects the degree of polarization in preferences for redistribution and, thus, it is useful as a measure of the distribution of preferences between classes. Jaime-Castillo & Marqu´es-Perales Social Mobility and Demand for Redistribution (16 / 21) Findings Class polarization (kappa index) A. Actual data B. Constant mobility C1. Constant preferences C2. Constant preferences D. No mobility AT 0.067 0.066 0.016 0.110 0.089 BE 0.062 0.062 0.016 0.109 0.085 BG 0.055 0.054 0.019 0.107 0.071 CH 0.078 0.080 0.017 0.104 0.098 CZ 0.104 0.109 0.017 0.104 0.140 DE 0.075 0.077 0.016 0.107 0.100 DK 0.055 0.058 0.016 0.105 0.077 EE 0.070 0.075 0.018 0.104 0.097 ES 0.031 0.031 0.017 0.108 0.054 FI 0.072 0.076 0.016 0.103 0.107 FR 0.056 0.057 0.017 0.107 0.085 GB 0.075 0.077 0.017 0.105 0.102 HU 0.036 0.036 0.017 0.104 0.063 IE 0.046 0.046 0.017 0.109 0.076 LT 0.035 0.036 0.017 0.103 0.061 LU 0.081 0.077 0.018 0.111 0.121 NL 0.089 0.091 0.016 0.106 0.106 NO 0.056 0.058 0.017 0.103 0.077 PL 0.077 0.078 0.016 0.105 0.135 PT 0.016 0.016 0.016 0.109 0.042 RU 0.064 0.071 0.018 0.105 0.098 SE 0.077 0.081 0.017 0.104 0.118 SI 0.062 0.063 0.017 0.105 0.108 SK 0.070 0.073 0.017 0.103 0.110 UA 0.033 0.035 0.019 0.107 0.052 Jaime-Castillo & Marqu´es-Perales Social Mobility and Demand for Redistribution (17 / 21) Findings Origin weights and upward mobility AT BE BG CH CZ DE DK EE ES FI FR GB HU IE LT LU NL NO PL PT RU SE SI SK UA .2 .3 .4 .5 .6 .7 Origin weights .15 .2 .25 .3 .35 .4 Upward mobility Jaime-Castillo & Marqu´es-Perales Social Mobility and Demand for Redistribution (18 / 21) Findings Origin weights and class polarization AT BE BG CH CZ DE DK EE ES FI FR GB HU IE LT LU NL NO PL PT RU SE SI SK UA .2 .3 .4 .5 .6 .7 Origin weights .02 .04 .06 .08 .1 Class polarization Jaime-Castillo & Marqu´es-Perales Social Mobility and Demand for Redistribution (19 / 21) Conclusions Conclusions Main findings Social origin matters to a little extent to explain preferences for redistribution, as newcomers tend to adopt the preferences of the destination class. Only limited evidence supporting the acculturation hypothesis and no support for the status maximization hypothesis. Cross-national differences The effect of social origin varies largely between countries. High rates of upward social mobility sharply reduce the effect of social origin on preferences for redistribution. Jaime-Castillo & Marqu´es-Perales Social Mobility and Demand for Redistribution (20 / 21) Thank you. Comments are welcome!! Jaime-Castillo & Marqu´es-Perales Social Mobility and Demand for Redistribution (21 / 21)