Preferences for redistribution in Europe

Olivera, Javier Article Preferences for redistribution in Europe IZA Journal of European Labor Studies Provided in Cooperation with: IZA – Institute of Labor Economics Suggested Citation: Olivera, Javier (2015) : Preferences for redistribution in Europe, IZA Journal of European Labor Studies, ISSN 2193-9012, Springer, Heidelberg, Vol. 4, Iss. 14, pp. 1-18, https://doi.org/10.1186/s40174-015-0037-y This Version is available at: https://hdl.handle.net/10419/127458 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. http://creativecommons.org/licenses/by/4.0/ ORIGINAL ARTICLE Open Access Preferences for redistribution in Europe Javier Olivera Correspondence: [email protected] Institute for Research on Socio-Economic Inequality (IRSEI), Research Unit INSIDE, University of Luxembourg, Route de Diekirch, L-7220 Walferdange, Luxembourg Abstract This paper uses pseudo panel techniques and a fixed effects estimator to analyse the determinants of preferences for redistribution in 34 European countries over the period 2002–2012. The data is drawn from the six available waves of the European Social Survey. The main result is that changes in income inequality positively affect changes in preferences for redistribution over time. Though this result is predicted by standard political economy models, it has found little previous empirical support. This study shows that, at least in Europe, growing income inequality leads to more individual support for redistribution. The empirical results hold after performing a variety of robustness checks regarding the construction of pseudo panels, the use of lags and different measures of income inequality. JEL codes: D31; D63; D72; H20 Keywords: Redistribution; Income Inequality; Preferences for redistribution; Pseudo panels 1 Introduction A common topic of interest for economists and other social scientists is the formation of preferences over how much income redistribution must be implemented, if any. As pointed by Alesina and Giuliano (2011), this is the most important dividing line between left and right political views concerning economic issues. Through political voting, these preferences can play a significant role in the final level of redistribution accomplished by the government. Early models of voting (Meltzer and Richard 1981) show that the median voter is decisive in pushing for redistribution when the median income is placed left of the mean income, i.e., when the income is unequally distributed. Although this model is insightful, there are missing mechanisms that, if accounted for, will produce different results. For example, individuals belonging to the lower part of the income distribution may have the expectation of upward mobility so that they will prefer less redistribution (Piketty 1995; Benabou and Ok 2001). Alesina and Angeletos (2005) show that societies where individual effort is believed to be the main source of income formation will prefer less taxes and redistribution. The contrary holds for societies that believe luck is important to create income, so they will prefer more redistribution. Furthermore, Karabarbounis (2011) finds empirical support for the ‘one dollar, one vote’equilibrium, which means that richer groups of individuals are able to put forward their agenda on less taxes and redistribution through their economic and political influence. © 2015 Olivera. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http:// creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. Olivera IZA Journal of European Labor Studies (2015) 4:14 DOI 10.1186/s40174-015-0037-y There are a number of studies analysing the determinants of preferences for redistribution, mainly exploiting the cross-country variation 1 . Although all these works are important in the literature of preferences for redistribution, they do not directly address the determinants of changes in these preferences. One of the reasons of this deficiency is the scarcity of adequate data for this purpose, i.e., panel data surveys that include questions on redistributive preferences. The aim of this paper is to study the determinants of preferences for redistribution in Europe taking into account variation over time and country. Within this framework, particular attention is paid to the effects of income inequality on these preferences. In this way, this study attempts to discern whether growing income inequality, as is widely observed, has an effect on the formation of preferences for redistribution. It is important to mention that this study does not deal with the realisation of these preferences, i.e., this study does not analyse whether the actual degree of redistribution observed in a country corresponds to the realisation of the preferences for redistribution of its citizens. For such analysis, one would need to use a longer period of observations in order to account for political and economic cycles (like the study by Georgiadis and Manning (2012) for the UK). The present paper differs from the existing empirical literature in several respects. It uses a harmonised dataset composed of the total six waves of the European Social Survey (ESS) carried out between 2002 and 2012, which comprises a total of 34 countries and 235,842 individuals with non-missing information. This data is collapsed to construct synthetic panels based on birth year cohorts, sex and country in order to use pseudo panel techniques (Deaton 1985) and study the changes in inequality and preferences more fully. For this purpose, a fixed effects estimator is used. This strategy allows for overcoming the data limitations and assess the role of changing inequality on redistributive preferences. Furthermore, the analysis considers that individuals are not only influenced by the level of income inequality—as measured, for example, by the Gini coefficient or the top 1% income share—but also by the degree of redistribution which is already taking place in the country. In all these cases, the analyses use comparable and harmonised country level variables that vary over time. The results indicate that variations in income inequality over time affect preferences for redistribution. These findings are robust to different measures of income inequality and specifications with different sizes and numbers of synthetic panels, and therefore the results provide evidence that preferences for redistribution are not immobile and that their evolution is influenced by changing income inequality. The positive effect of the Gini index computed with gross incomes (before tax and transfers) is particularly relevant for preferences for redistribution because this index is less influenced by the contemporaneous tax and transfer system. In addition, it is found that the level of actual redistribution operates in the opposite direction of income inequality, which helps to explain why some welfarist-oriented countries such as Denmark, Norway and Sweden exhibit a lower preference for redistribution. This can be interpreted as individuals living in economies where substantial redistribution already exists and do not want more redistribution. It is important to bear in mind that all these findings must be interpreted as short-time responses given the limited length of time of the data. The paper is organised as follows. The next section briefly discuss the relevant literature. The third section presents the data. The fourth section presents the empirical Olivera IZA Journal of European Labor Studies (2015) 4:14 Page 2 of 18 strategy. Section 5 presents and discusses the econometric results, and section 6 concludes. 2 The study of preferences for redistribution The main prediction of the median voter theorem (Meltzer and Richard 1981) is that the level of income inequality positively affects the size of income redistribution in the country. This result has led to the emergence of an important body of empirical studies trying to test its validity. This literature can be roughly subdivided into two branches: one branch uses measures of income inequality and redistribution (mostly the Gini coefficient and the ratio of median to mean income) at the country or state level. The other branch uses individual preferences for redistribution. In the first group of studies, the effect of inequality on redistribution has not received significant empirical support. Examples are Rodriguez (1999), Persson and Tabellini (1994), Perotti (1996), Moene and Wallerstein (2001, 2003), Lind (2005) and Shelton (2007). Exceptions are Milanovic (2000, 2010) and Karabarbounis (2011). The studies of the second branch attempt to uncover the determinants of individual preferences for redistribution, and some of these evaluate the effects of income inequality on redistributive preferences (see a summary in Table 1). Examples of studies assessing the role of economic inequality on redistributive preferences are Pittau et al. (2013), Kerr (2014), Tóth and Keller (2011), Yamamura (2012) and Jaeger (2013). The results of the effect of inequality on the preferences for redistribution are mixed, although a majority of them find a positive effect. The analysis is mainly based on cross-country differences so that the problems of unobserved effects and reverse causality come with caution when interpreting the results. However, Kerr (2014) uses an IV model to detect a positive effect of inequality on the demand for redistributions across American states, and Jaeger (2013) uses a pseudo panel approach. This last study, however, has three problems that affect the correct estimation and interpretation of the effect of income inequality on preferences for redistribution. The first one is the use of synthetic panels that are constructed on the base of variables (social class position) that are not immobile over time and not observable for all individuals 2 ,whichisacondition to build proper pseudo panels (Verbeek 2008). The second problem is the use of Gini indexes from seven different data sources for different countries, and even for the same country observed in different years 3 . This mix of sources and years severely limits the comparability of income inequality within and across countries and over time. Furthermore, the year of some Gini indexes does not correspond with the year of the ESS wave. The use of household income, without any adjustment, is also problematic as the ESS does not have a uniform income question across waves 4 . Other approaches employed to understand how redistributive preferences are shaped pay particular attention to the formation of beliefs about income position, informational limitations on inequality levels and the influence of reference groups. These studies mostly use experiments as the empirical strategy to deal with the demanding set of required variables. Relevant examples are Kuziemko et al. (2013) and Cruces et al. (2013). Furthermore, a recent effort aimed at integrating the many findings and approaches in the formation of redistributive preferences is presented in Schokkaert and Truyts (2014) in the form of a model that considers the possibility of assessing Olivera IZA Journal of European Labor Studies (2015) 4:14 Page 3 of 18 income differences caused by ability as unjust. This model is built on the premise that income differences caused by luck are seen as illegitimate, while those arising from effort are legitimate, and considers a utility function composed by a self-interested and a social justice part. It is shown that the desired degree of redistribution of an individual depends on the relative importance of luck, effort and ability assigned to explain income differences in an environment of informational bias originating from the reference group. This model also helps to understand recent experimental research finding that better-off individuals are more prone than poorer individuals to recognise other’s effort when making allocations (Barr et al. 2015). 3 Data and variables 3.1 The data The data is drawn from the complete set of available bi-annual rounds (six rounds) of the European Social Survey (ESS) from 2002 to 2012. This survey is designed to Table 1 Literature on preferences for redistribution Study Dataset Region Modelling Effect of inequality Pittau et al. (2013) ESS 2002-2008 23 EU countries Logit multilevel + GSS 2000-2006 US states - Kerr (2014) GSS 1972 -2000 US (states) OLS + or insignificant ISSP 1987, 92, 99 Many countries IV OLS WVS 1990, 95, 00 Tóth and Keller (2011) Eurobarometer 1999 EU-27 OLS Multilevel + Yamamura (2012) JGSS 2000-2008 Japan Ordered Probit + for high-income earners, otherwise insignificant Jaeger (2013) ESS 2002-2008 31 EU countries FE Pseudo Panels insignificant Luttmer and Singhal (2011) ESS 2002-2006 32 EU countries OLS Not studied Guillaud (2013) ISSP 2006 33 countries Ordered Logit Not studied Alesina and Giuliano (2011) GSS 1972-2004 US OLS Not studied WVS 1981, 90, 95, 99 Many others Alesina and FuchsSchundeln (2007) Panel GSOEP 1997-2002 Germany Probit Not studied Georgiadis and Manning (2012) BSAS UK OLS Not studied Alesina and La Ferrara (2005) GSS 1978-91 US Ordered Probit Not studied Corneo and Grüner (2002) ISSP 1992 12 developed countries Logit Not studied Fong (2001) Gallup Poll Social Audit Survey 1998 US Ordered Probit Not studied Acronyms: ESS: European Social Survey GSOEP: German Socio Economic Panel ISSP: International Social Survey Program GSS: General Social Survey WVS: World Values Survey JGSS: Japanese General Social Survey BSAS: British Social attitudes Survey Olivera IZA Journal of European Labor Studies (2015) 4:14 Page 4 of 18 measure attitudes, beliefs, values and behaviour patterns of individuals in Europe. There is a core set of questions implemented in each wave and additional modules in specific waves. Income inequality measures are drawn from the Standardized World Income Inequality Database (SWIID version 4.0) because this data—although not without its problems—provides the broadest coverage across countries over time, allowing for the attainment of the largest number of country-year points, whereas data on income inequality from Eurostat covers fewer observations 5 . The SWIID provides Gini indexes computed with incomes both before and after taxes and transfers, and the top 1% of share. While the Gini index captures inequality along the total distribution of incomes, the top income share is useful to capture income concentration at the very top of the income distribution. These measures cover different forms of income inequality and therefore make this study more complete. Given the recent revival of the importance of top incomes (Atkinson et al. 2011), the share of the top 1% of the income distribution is considered as an additional measure of inequality in the analysis. Pittau et al. (2013) also usedatafromapreviousversionofSWIIDtoanalyse preferences for redistribution, although they do not use the variation of inequality indices over time. The other macro variables used in the present analysis are the real and PPP-adjusted GDP per capita from the World Bank’s World Development Indicators and public social protection expenditures (as percentage of GDP) from Eurostat. A total of 235,842 individuals have complete information for the variables of interest. This selection comprises a total of 34 countries and represents 152 country-year points. The included countries are the EU-28 (except Malta) plus Norway, Iceland, Russia, Switzerland, Turkey, Ukraine and Israel. Not all countries have observations in each wave (see Table 2). There are 15 countries with observations in all 6 waves: Belgium, Denmark, Finland, Germany, Hungary, Ireland, the Netherlands, Norway, Poland, Portugal, Switzerland, Slovenia, Spain, Sweden and the United Kingdom. Furthermore, 3 countries have information in 5 waves: Czech Republic, Estonia and Slovakia. Austria, Bulgaria, Cyprus, France, Greece, the Russian Federation and Ukraine have observations in 4 waves. Israel, Italy and Lithuania are observed in 3 waves. Latvia and Romania are only observed in one period. 3.2 The dependent variable The key question measuring individual preferences for redistribution is repeated in each wave, which is: “To what extent do you agree or disagree with the statement: the government should take measures to reduce differences in income levels”. The individual must choose one of five responses, which are rescaled in the following way: strongly agree (5); agree (4); neither agrees nor disagree (3); disagree (2) and strongly disagree (1). Therefore, the higher this number, the more it favours redistribution. The average score of preference for redistribution is increasing across waves, though there are important differences among countries and years. For example, for those countries with observations in 2002 and 2012, the score increased in 14 countries and decreased in 3. The measures of inequality also show variation among countries and years. Between 2002 and 2012, the Gini of net incomes increased in 11 countries and decreased in 6. The average increase is 5.7% and the average decrease is −5.8%. For the same period, the Gini of market incomes increased in 10 countries and decreased in 7, with 6.9% Olivera IZA Journal of European Labor Studies (2015) 4:14 Page 5 of 18 being the average increase and −5.8% the average decrease. In the case of the top 1% of income share, this has increased in 12 countries and declined in 5 countries. The data shows a great deal of variability across countries and over time in redistributive preferences. The mean score for the variable measuring preferences for redistribution (from 1 to 5) in the complete 2002–2012 period is 3.84. The countries with the highest and lowest scores are Greece with 4.35 and Denmark with 3.03, respectively. Confirming some regional differences, the Mediterranean countries are placed well Table 2 Composition of sample Total 2012 2010 2008 2006 2004 2002 Year 4 x x x x Austria 6 x x x x x x Belgium 4 x x x x Bulgaria 2 x x Croatia 4 x x x x Cyprus 5 x x x x x Czech Rep 6 x x x x x x Denmark 5 x x x x x Estonia 6 x x x x x x Finland 4 x x x x France 6 x x x x x x Germany 4 x x x x Greece 6 x x x x x x Hungary 2 x x Iceland 6 x x x x x x Ireland 3 x x x Israel 3 x x x Italy 1 x Latvia 3 x x x Lithuania 2 x x Luxembourg 6 x x x x x x Netherlands 6 x x x x x x Norway 6 x x x x x x Poland 6 x x x x x x Portugal 1 x Romania 4 x x x x Russian Fed 5 x x x x x Slovakia 6 x x x x x x Slovenia 6 x x x x x x Spain 6 x x x x x x Sweden 6 x x x x x x Switzerland 2 x x Turkey 4 x x x x Ukraine 6xxxxxxUK 152 Total Olivera IZA Journal of European Labor Studies (2015) 4:14 Page 6 of 18 above the Nordic countries (see Figure 1). The relation between preferences for redistribution and income inequality is positive when attention is paid to cross-country differences, which is reported in the left-hand panel of Figure 2. At first glance, it is surprising that traditional pro-welfare states like the Nordic countries have simultaneously low levels of inequality and lower preferences for redistribution. However, it is possible that individuals who in general are in favour of income redistribution are less willing to favour more redistribution if the scale of redistribution already taking place is high enough. An indication of this can be observed in the right hand panel of Figure 2. The share of public social protection expenditures to GDP may be interpreted as a rough measure of the size of redistribution implemented in the country. The figure suggests some tendency for preferences for redistribution to be lower where the size of redistribution to GDP is higher. 3.3 The control variables The variables used in the regression analysis are the standard individual controls employed in the literature of redistributive preferences and include sex, age, marital status (living with partner), education level in the form ISCED dummies, belonging to a minority ethnic group in the country (ethnic), and the religious position of the individual regardless of any particular religion (religious) in a scale from 1 (not at all) to 10 (very religious). The ESS has no uniform question on household income, but an income proxy that is asked in every wave is included. This is “which of the descriptions on this card comes closest to how you feel about your household’s income nowadays?” with four possible scales: living comfortably on present income (1), coping on present income (2), difficult on present income (3) and very difficult on present income (4). Of course, this question may refer to satisfaction with income, so one should be cautious in interpreting the estimates of this variable. Another group of control variables refers to labour conditions of the individual and includes the dummy variables retired and unemployed. Finally, left-right political scale denotes the self-placement of the individual in the political spectrum from 0 (left) to 10 (right). The descriptive statistics of the variables are reported in Table 3. 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Greece Turkey Portugal Hungary Bulgaria Lithuania Slovenia Latvia Israel Ukraine Romania Cyprus Italy Russian Fed Spain Croatia France Iceland Poland Finland Estonia Slovakia Ireland Belgium Austria Sweden Switzerland Norway Germany Czech Rep Luxembourg UK Netherlands Denmark in favour of redistribution Figure 1 Preferences for redistribution by country, 2002-2012. Olivera IZA Journal of European Labor Studies (2015) 4:14 Page 7 of 18 4 Empirical strategy In the empirical literature of preferences for redistribution, it is a common practice to use the multi-scale variable of preferences for redistribution and estimate with OLS. Examples of this are Georgiadis and Manning (2012), Kerr (2014), Alesina and Giuliano (2011) and Luttmer and Singhal (2011). All of them argue that the use of alternative modelling approaches such as the ordered logit model do not change the results. Differently, Pittau et al. (2013) recode the original 5-scale question on preferences for redistribution into 1/0 and apply a logistic regression with multi-level modelling. Guillaud (2013) use an ordered logit and Alesina and La Ferrara (2005) use ordered probit and probit models. Even though the specification models of some of these studies have controlled for country and time effects with dummy variables, one cannot fully assure that changes in income inequality and redistribution over time have the same effects over the preferences for redistribution. Panel data can help to study the effects of income inequality over time because the reaction of the same unit of analysis to changing inequality can be followed across time. The application of a fixed effects model will allow for controlling for time-invariant observed and unobserved effects. This is an essential distinction with respect to pooled OLS models (such as Kerr 2014; Luttmer and Singhal 2011; Alesina and Giuliano 2011; and Alesina and La Ferrara 2005) because the differences in the preferences for redistribution may vary irrespective of the differences in income inequality across countries. In that case, the difference in the preference for redistribution will be more related to specific and persistent factors of the country that shape the preferences of their citizens. For example, Karabarbounis (2011) cite legal origins, political institutions, persistent cultural characteristics, ethnic fragmentation, prospects of upward mobility, and social beliefs about fairness. Country differences in culture (Berigan and Irwin 2011) and national identity (Shayo 2009) are also part of those specific factors that can affect the demand for redistribution. In a panel data structure with i=1,…Nindividuals followed across t = 1,…,Tperiods, it is common to use the following specification: yit ¼δtþαiþβXit þγZit þμit ð1Þ The dependent variable y it measures the individual preference for redistribution in year t. The vector X it contains inequality measures that are the same for individuals of the same country and year. Z it denotes individual and time specific socio-demographic variables. The term α i is the year-invariant individual unobserved effect; δ t is a common 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 20.0 25.0 30.0 35.0 40.0 45.0 50.0 in favour of redistribution Gini of net income correl = 0.43 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.0 10.0 15.0 20.0 25.0 30.0 in favour of redistribution Social p rotection ex p enditures ( GDP) correl = -0.30 Figure 2 Preferences for redistribution and income inequality, 2002-2012. Olivera IZA Journal of European Labor Studies (2015) 4:14 Page 8 of 18 Table 7 FE estimators with different sizes and numbers of synthetic individuals Variable Baseline: Cohort size ≥30 Cohort size ≥50 No minimum cohort size Cohort size ≥30 Cohort size ≥30 & 10 birth year cohorts & 10 birth year cohorts & 10 birth year cohorts & 7 birth year cohorts & 14 birth year cohorts coeff std error adj R 2 n coeff std error adj R 2 n coeff std error adj R 2 n coeff std error adj R 2 n coeff std error adj R 2 n lagged one year: gini net 0.0097*** (0.0036) 0.180 2657 0.0140*** (0.0041) 0.216 2211 0.0083** (0.0036) 0.141 3037 0.0090** (0.0037) 0.208 1963 0.0117*** (0.0037) 0.180 3372 gini market 0.0060** (0.0023) 0.180 2657 0.0065*** (0.0025) 0.213 2211 0.0074*** (0.0022) 0.144 3037 0.0071*** (0.0025) 0.211 1963 0.0066*** (0.0023) 0.180 3372 top 1% income share 0.0082** (0.0037) 0.179 2657 0.0097** (0.0041) 0.213 2211 0.0037 (0.0044) 0.139 3037 0.0065* (0.0039) 0.206 1963 0.0083** (0.0036) 0.178 3372 social protection expend, % GDP −0.0111** (0.0047) 0.161 2348 −0.0148*** (0.0049) 0.204 1957 −0.0082* (0.0045) 0.135 2677 −0.0103* (0.0054) 0.187 1736 −0.0111** (0.0044) 0.165 2979 lagged two years: gini net 0.0071** (0.0033) 0.183 2657 0.0108*** (0.0036) 0.218 2211 0.0052 (0.0032) 0.143 3037 0.0058* (0.0033) 0.209 1963 0.0078** (0.0033) 0.182 3372 gini market 0.0033* (0.0018) 0.182 2657 0.0035* (0.0019) 0.215 2211 0.0035** (0.0017) 0.143 3037 0.0036* (0.0019) 0.210 1963 0.0032* (0.0018) 0.181 3372 top 1% income share 0.0196*** (0.0052) 0.191 2657 0.0207*** (0.0056) 0.225 2211 0.0124** (0.0061) 0.145 3037 0.0172*** (0.0058) 0.216 1963 0.0192*** (0.0049) 0.188 3372 social protection expend, % GDP −0.0049 (0.0045) 0.157 2329 −0.0074 (0.0047) 0.197 1942 −0.0021 (0.0046) 0.135 2657 −0.0037 (0.0050) 0.183 1722 −0.0044 (0.0041) 0.161 2954 1 year lag x2 year lag: gini net 0.0088** (0.0035) 0.181 2657 0.0131*** (0.0040) 0.217 2211 0.0070** (0.0035) 0.142 3037 0.0077** (0.0035) 0.208 1963 0.0102*** (0.0035) 0.181 3372 gini market 0.0047** (0.0021) 0.181 2657 0.0051** (0.0022) 0.214 2211 0.0055*** (0.0020) 0.143 3037 0.0053** (0.0023) 0.210 1963 0.0049** (0.0021) 0.180 3372 top 1% income share 0.0137*** (0.0046) 0.184 2657 0.0151*** (0.0051) 0.218 2211 0.0074 (0.0055) 0.142 3037 0.0116** (0.0050) 0.210 1963 0.0135*** (0.0045) 0.183 3372 social protection expend, % GDP −0.0091* (0.0050) 0.159 2348 −0.0125** (0.0052) 0.201 1957 −0.0056 (0.0049) 0.134 2677 −0.0080 (0.0057) 0.185 1736 −0.0087* (0.0045) 0.163 2979 ***p < 0.01, **p < 0.05, *p < 0.1. Standard errors are robust and clustered by country and cohort. Each row corresponds to a different regression and only reports the corresponding income inequality coefficient. Each regression includes the same covariates as in Table 5, and year dummies. In the first panel, inequality measures are lagged one year with respect to the ESS wave as in previous specifications. In the middle panel, inequality measures are lagged two years. In the bottom panel, the inequality measures are the averages of the one and two year lags. Olivera IZA Journal of European Labor Studies (2015) 4:14 Page 15 of 18 6 Conclusions This paper has shown that income inequality of the country matters for preferences for redistribution when one focusses on changes over time. These results arise from fixed effects estimators applied to pseudo panels for the period 2002– 2012 in 34 European countries. The findings are robust to different measures of income inequality and specifications with different sizes and numbers of synthetic panels. It is shown that increases in pre-tax and transfer income inequality over time raise the demand for redistribution, which is in line with the predictions of early political economy models (Meltzer and Richard 1981) that have not found much empirical support. Income inequality measured with disposable income and the top 1% income share also positively affects the demand for redistribution. Looking at the evolution of the top 1% or further concentration shares of income have become increasingly important in the recent empirical literature about inequality, and therefore this study contributes to the understanding of the effects of income concentration on the demand for redistribution. Another important result is that the actual level of redistribution implemented in the country, measured with public expenditures in social protection expenditures, reduces the demand for redistribution. This helps to explain why some Nordic countries exhibit a rather low support for redistribution. In sum, at least in Europe and being mindful of the short length of the period of analysis, one can observe that increasing income inequality leads to more individual support for redistribution. 7 Endnotes 1 Georgiadis and Manning (2012); Pittau et al. (2013); Kerr (2014), Alesina and Giuliano (2011); Alesina and La Ferrara (2005); Alesina and Fuchs-Schundeln (2007); Luttmer and Singhal (2011); Guillaud (2013); Corneo and Grüner (2002); Fong (2001); Yamamura (2012). 2 The social class position is based on ISCO88 occupational codes, which is only observed for 74% of the ESS 2002–2012 sample. 3 Jaeger (2013) (see Appendix’s Table 1) uses the following databases to gather information for Gini indexes for the years 2002–2010: World Inequality Database, Wikipedia, http://www.nationmaster.com, http://www.indexmundi.com, Eurostat, World Bank Databank and CIA World Factbook. In addition, 22 country-year points of a total of 125 are allocated Gini indexes corresponding to another year. 4 The ESS includes a question that indicates which range of total household income the individual belongs to. However, there are two problems in using this question across all waves. There are 12 ranges in waves 2002–2006, and 10 in waves 2008–2012. Furthermore, there is a high percentage of individuals that do not answer the income question of the survey (23% of the full sample). 5 The SWIID dataset is built with the United Nations University’s World Income Inequality Database (WIID), the Luxembourg Income Study dataset (LIS), the World Top Incomes Database and other country specific data on incomes. This employs a custom missing-data algorithm to standardised the WIID data by using the LIS data as the standard. For more details see Solt (2009). 6 The oldest birth cohort is 1920–1926, and the youngest is 1983–1989. Olivera IZA Journal of European Labor Studies (2015) 4:14 Page 16 of 18 7 For more about the asymptotic properties and conditions of pseudo panel estimators, see Verbeek (2008), Verbeek and Vella (2005), Collado (1997) and Moffit (1993). 8 GDP per capita is not included in the regression as this is collinear with the social protection expenditures-GDP ratio. 9 In both datasets, the minimum size of the cells is 30 respondents. The average cell size of the dataset of 7 and 14 birth cohorts is 113 and 62, respectively. Competing interests The IZA Journal of European Labor Studies is committed to the IZA Guiding Principles of Research Integrity. Theauthordeclaresthathehasobservedtheseprinciples. Acknowledgements I would like to thank the comments and suggestions made by Brian Nolan, Koen Decancq, an anonymous referee and conferences and seminar participants at the University of Antwerp, EPCS meeting at the University of Cambridge, University of Luxembourg, ZEW-Mannheim and Irish Economic Association meeting. The author alone is responsible for the analysis carried on in this study. Responsible editor: Sara de la Rica Received: 9 March 2015 Accepted: 30 March 2015 References Alesina A, Angeletos GM (2005) Fairness and redistribution. Am Econ Rev 95(4):960–80 Alesina A, Fuchs-Schundeln N (2007) Good Bye Lenin (or not?). The Effect of Communism on People’s Preferences. Am Econ Rev 97(4):1507–1528 Alesina A, Giuliano P (2011) Preferences For Redistribution. In Handbook Of Social Economics. Benhabib J, Bisin A, Jackson MO (Eds). North Holland. 93-132 Alesina A, La Ferrara E (2005) Preferences for redistribution in the land of opportunities. J Public Econ 89:897–931 Alesina A, Di Tella R, MacCulloch R (2004) Inequality and Happiness: Are Europeans and Americans Different? J Public Econ 88:2009–2042 Atkinson AB, Piketty T, Saez E (2011) Top Incomes in the Long Run of History. J Econ Lit 49(1):3–71 Barr A, Burns J, Miller L, Shaw I (2015) Economic status and acknowledgement of earned entitlement. Journal of Economic Behavior and Organization. doi: 10.1016/j.jebo.2015.02.012 Benabou R, Ok EA (2001) Social mobility and the demand for redistribution: the POUM hypothesis. Q J Econ 116:447–487 Berigan N, Irwin K (2011) Culture, Cooperation, and the General Welfare”. Soc Psychol Q 74(4):341–360 Collado D (1997) Estimating Dynamic Models from Time Series of Independent Cross Sections. J Econ 82:37–62 Corneo G, Grüner HP (2002) Individual preferences for political redistribution. J Public Econ 83:83–107 Cruces G, Perez-Truglia R, Tetaz M (2013) Biased perceptions of income distribution and preferences for redistribution: evidence from a survey experiment. J Public Econ 98:100–112 Deaton A (1985) Panel Data from Times Series of Cross-Sections. J Econ 30:109–126 Fong C (2001) Social preferences, self-interest, and the demand for redistribution. J Public Econ 82(2):225–246 Georgiadis A, Manning A (2012) Spend it like Beckham? Inequality and redistribution in the UK, 1983–2004. Public Choice 151:537–563 Guillaud E (2013) Preferences for redistribution: an empirical analysis over 33 countries. J Econ Inequal 11(1):57–78 Jaeger MM (2013) The effect of macroeconomic and social conditions on the demand for redistribution a pseudo panel approach. Journal of European Social Policy 23:149–163 Karabarbounis L (2011) One Dollar, One Vote. Econ J 121:621–649 Kerr W (2014) Income Inequality and Social Preferences for Redistribution and Compensation Differentials. J Monet Econ 66:62–78 Kuziemko I, Norton M, Saez E, Stantcheva S (2013) How Elastic are Preferences for Redistribution? Evidence from Randomized Survey Experiments. NBER Working Paper 18865 Lind JT (2005) Why is there so little redistribution? Nordic Journal of Political Economy 31:111–125 Luttmer EFP, Singhal M (2011) Culture, Context and the Taste for Redistribution. American Economic Journal: Economic Policy 3(1):157–179 Meltzer AH, Richard SF (1981) A rational theory of the size of the government. J Polit Econ 89(5):914–27 Milanovic B (2000) The median voter hypothesis, income inequality and income redistribution: an empirical test with the required data. Eur J Polit Econ 16(3):367–410 Milanovic B (2010) Four critiques of the redistribution hypothesis: An assessment. Eur J Polit Econ 26:147–154 Moene KO, Wallerstein M (2001) Inequality, social insurance and redistribution. Am Polit Sci Rev 95:859–874 Moene KO, Wallerstein M (2003) Earnings, inequality and welfare spending: a disaggregated analysis. World Politics 55:485–516 Moffit R (1993) Identification and Estimation of Dynamic Models with a Time Series of Repeated Cross-Sections. J Econ 59:99–124 Perotti R (1996) Democracy, income distribution and growth: what the data says. J Econ Growth 1:149–187 Persson T, Tabellini G (1994) Is inequality harmful for growth?: theory and evidence. Am Econ Rev 84:600–621 Piketty T (1995) Social mobility and redistributive politics. Q J Econ 110(3):551–84 Olivera IZA Journal of European Labor Studies (2015) 4:14 Page 17 of 18 Pittau MG, Massari R, Zelli R (2013) Hierarchical modelling of disparities in preferences for redistribution. Oxf Bull Econ Stat 75(4):556–584 Rodriguez FC (1999) Does Distributional Skewness Lead to Redistribution? Evidence from the United States. Economics & Politics 11(2):171–199 Schokkaert E, Truyts T (2014) Preferences for Redistribution and Social Structure. KU Leuven Discussion Paper Series DPS14.01 Shayo M (2009) A Model of Social Identity with an Application to Political Economy: Nation, Class and Redistribution. American Political Science Review 103:147–174 Shelton CA (2007) The size and composition of government expenditure. J Public Econ 91:2230–2260 Solt F (2009) Standardizing the World Income Inequality Database. Soc Sci Q 90(2):231–242, SWIID Version 3.1, December 2011 Tóth, IG, Keller T (2011) Income distributions, inequality perceptions and redistributive claims in European societies. Gini Discussion Paper Number 7 Verbeek M (2008) Pseudo Panels and Repeated Cross-Sections. In: Matyas L, Sevestre P (eds) The Econometrics of Panel Data: Fundamentals and Recent Developments in Theory and Practice. Springer, Berlin Verbeek M, Vella F (2005) Estimating Dynamic Models from Repeated Cross-Sections. J Econ 127:83–102 Yamamura E (2012) Social capital, household income, and preferences for income redistribution. Eur J Polit Econ 28:498–511 Submit your manuscript to a journal and benefi t from: 7 Convenient online submission 7 Rigorous peer review 7 Immediate publication on acceptance 7 Open access: articles freely available online 7 High visibility within the fi eld 7 Retaining the copyright to your article Submit your next manuscript at 7 springeropen.com Olivera IZA Journal of European Labor Studies (2015) 4:14 Page 18 of 18