Working Paper Series
The GK SOCLIFE working paper series is published by the Research Training Group SOCLIFE, an interdisciplinary research training group, which is at the crossroads between social and economic sciences, statistics and law. The Research Training Group SOCLIFE develops under the aegis of the Faculty of Economic and Social Sciences of the University of Cologne. The Working Papers are meant to share work in progress before formal publication. The aim of the series is to promote conceptual, methodological and substantial discussion between the members of SOCLIFE and the larger scholar community.
Submissions. Papers are invited from all the members of Research Training Group SOCLIFE, and also from other interested scholars. All papers are subject to review by one member of the Editorial Board. Papers can be submitted either in English or in German. Authors interested in including their work in the GK SOCLIFE series may contact or send their papers to Prof. Dr. Hans-Jürgen Andreß, the managing editor of the series.
The Social Investment Welfare Expenditure data set (SIWE): a new methodology for measuring the progress of social investment in EU welfare state budgets
The social investment perspective has become a reference framework for comparative welfare state analysis, and a powerful idea influencing the European social dimension since the Lisbon Strategy. A number of empirical studies in the field have focused on the budgetary side of welfare state change, tracking the dynamics of “new” social investment versus “old” social protection spending. Still, many data limitations (e.g. scarce country/years coverage) and the prevailing use of rough spending-over-the-GDP indicators have hindered the progress of our empirical knowledge over social investment in Europe. This working paper presents a new data set and methodology for the comparative analysis of welfare state budgets from the perspective of social investment. Based on various Eurostat data sources, the Social Investment Welfare Expenditure data set (SIWE) includes social spending data finely disaggregated into welfare functions for 29 countries (EU-28 less Croatia, plus Norway and Switzerland), and covers years from 1995 to 2013. Building on previous contributions, I develop a new methodology for measuring “budgetary welfare effort” (BWE), that is, the effort effectively put by governments on selected welfare programmes, net of the interferences due to economic and demographic oscillations. I also construct two composite BWE indices that allow to directly compare the whole social investment and social protection dimensions of welfare state budgets, in a way more accurate than what done so far. This provides researchers with a fresh tool for empirical analyses of the dynamics, causes and consequences of welfare state change from the perspective of social investment. The SIWE data set can be requested from the author’s web page.
The fixed effects approach as an alternative to multilevel analysis for cross-national analyses
Multilevel models that combine individual and contextual factors are increasingly popular in comparative social science research; however, their application in country-comparative studies is often associated with several problems. First of all, most data-sets utilized for multilevel modeling include only a small number (N<30) of macro-level units, and therefore, the estimated models have a small number of degrees of freedom on the country level. If models are correctly specified paying regard to the small, level-2 N, only a few macro-level indicators can be controlled for. Furthermore, the introduction of random slopes and cross-level interaction effects is then hardly possible. Consequently, (1) these models are likely to suffer from omitted variable bias regarding the country-level estimators, and (2) the advantages of multilevel modeling cannot be fully exploited.
The fixed effects approach is a valuable alternative to the application of conventional multilevel methods in country-comparative analyses. This method is also applicable with a small number of countries and avoids the country-level omitted variable bias through controlling for country-level heterogeneity. Following common practice in panel regression analyses, the moderator effect of macro-level characteristics can be estimated also in fixed effects models by means of cross-level interaction effects. Despite the advantages of the fixed effects approach, it is rarely used for the analysis of cross-national data.
In this paper, I compare the fixed effects approach with conventional multilevel regression models and give practical examples using data of the International Social Survey Programme (ISSP) from 2006. As it turns out, the results of both approaches regarding the effect of cross-level interactions are similar. Thus, fixed effects models can be used either as an alternative to multilevel regression models or to assess the robustness of multilevel results.
Hawal Shamon and Hans-Jürgen Andreß
Daten über die personelle Einkommensverteilung: Was man beachten sollte
Abstract: Income inequality data are widely used in the empirical literature. Sometimes it seems to be the case that the conceptual heterogeneity of these data are disregarded by the papers' authors. In the extreme case, estimations in cross-sectional and/or longitudinal analyses reflect rather this conceptual heterogeneity of the used income inequality data rather than true empirical differences. Using the World Income Inequality Database (WIID V2.0c) of the United Nations University, this paper aims at clarifying the different conceptual issues that should be thought of in the use of income inequality data. In addition, it recommends different research questions for which income inequality data can be adequately used.
Mara Boehle and Christof Wolf
Understanding time as socio-historical context: Analyzing social change within the framework of multilevel analysis
Abstract: From a methodological and sociological perspective, analyzing social change is best done by using repeated cross-sectional data and by including individual level variables, time, and time-dependent macro variables. Furthermore, interest often focuses on whether the effects of explanatory variables change over time, e.g. the impact of education on one’s social position. We argue that multilevel models with time as context best meet these requirements. However, the methods for applying a multilevel approach to study time-dependent social processes are relatively poorly developed and therefore rarely used. Instead, most social change related analyses of repeated cross-sectional data, such as APC models or changing effect models, include time as an individual level variable and thereby neglect the contextual characteristics of time and changes at the macro level. In contrast, pooled time series analysis does not take account of individuals and micro level effects. Thus, the aim of this paper is to overcome these shortcomings by adopting a new approach to social change studies. We describe the application of multilevel analysis with years as contextual units on a step-by-step basis and emphasize the substantive advantages of each model for the study of social change.
A new color in the picture: The impact of educational fields on fertility in Western Germany
Abstract The extensive research on the impact of educational attainment on fertility behavior has been expanded by a new dimension. According to these recent findings, not only the level, but also the field of education has to be taken into account. The field of education determines a great deal about labor market options and influences opportunities to combine emplozment and family life. The question this paper aims to answer is: How does the educational field influence the transition to parenthood of women and men in Western Germany? The German Socio Economic Panel (1984-2010) provides the data. Discrete time event history models are applied to examine the impact of the field of education on the transition to parenthood, looking at the time after graduation until a first child is born. Educational fields are grouped according to their most salient characteristic with regard to the share of women, the occupational specificity, the share of public-sector employment, and the share of partßtime employment among people educated in the field. The models take the educational level into account and control for marital status, episodes of educational enrollement, and migration background. The results show that educational fields matter for the transition to a first borth only for women. For men, the results do not show a significant impact of educational fields on the transition rates to parenthood. However, they point at the importance of educational level for the probability of men to become fathers. High transition rates are found among women educated in both female-dominated and male-dominated fields. The finding of low transition rates among women aducated in public-sector fields comes as a surprise, since, given the high workplace security in the public sector, they were expected to be among the women with high transition rates.
Pei-Chun Ko and Karsten Hank (corresponding author)
Grandparents caring for grandchildren in China and Korea: Findings from CHARLS and KLoSA
Objectives: To provide an overview of the prevalence and profiles of grandparents caring for their grandchildren in two East Asian countries, China and South Korea, sharing a similar cultural background but very different contemporary institutional, demographic and socioßeconomic circumstances.
Methods: We apply logistic models to analyze pilot data from the China Health and Retirement Longitudinal Study (CHARLS) and data from the Korean Longitudinal Study of Aging (KLoSA, wave 2). Our analytic sample comprises 772 Chinese respondents and 4,958 Korean respondents aged 45 to 79.
Results: The proportions of grandparents providing care for grandchildren differ considerably between China (58%) and South Korea (6%). Stil,, the determinants of gradparents' involvement in childcare (e.g. age, geographic proximity) are fairly similar in both countries. However, financial support from adult children to grandparents is found to be significant in China only, whereas Korean grandparents exhibit a greater propensity to care for their (employed) daughters' children than for their sons'.
Discussion: Our analysis suggests that patriliniar considerations may begin to lose some of their importance in shaping downward functional solidarity between generations in East Asia and that instead (grand-)children's actual needs, particularly such as those related to maternal employment, receive more attention.
Paper on request
Living green but poor? Investigating inequality in household energy costs among income groups
Abstract: What are the social consequences of climate mitigation policies? International agreements and activities of international organizations have led to an increase in national policies to mitigate climate change. The question of who has to pay the costs of intensive public programs for mitigating greenhouse gas emissions, however, is mainly addressed on a national level. So far, research has widely neglected whether the costs of climate change mitigation policies are unequally distributed among socio-economic groups. As a consequence, the impact of climate policies on individuals’ life situations is largely unknown, even though the design of, for example, a general tax on energy suggests a highly regressive effect on national patterns of housing utility costs among low-income households. Taking into account that lower income groups of Western societies are also characterized by a significantly smaller carbon footprint than those of higher income groups, the implementation of such policy instruments in the residential housing sector creates a challenge for social justice. This conflict between effectiveness of climate mitigation policies and unintended social consequences, such as increasing fuel costs, is analyzed in this paper by applying a before-after design for three country cases: Austria, Denmark, and the United Kingdom. Depending on their design, climate mitigation policies can foster patterns of inequality in household energy costs by stimulating consumption behavior and the promotion of more efficient appliances, support for small-scale renewable energies, and/or increasing taxes and contributions included in the housing utility costs. The reason is, for example, that the poor cannot cope with increased costs as effectively as the rich can. Depending on the welfare or social housing regime, this effect can also be compensated for by high levels in housing allowances or social transfers. The analysis is based on longitudinal data of housing utility costs from the EU-SILC study using multivariate panel-regression models for the period of 2005 to 2008.
The development of public demand for redistribution. A pseudo-panel model for decomposing within- and between-effects
Abstract: Public demand for redistribution has been a subject to research for the last couple of decades. While research has identified quite robust individual-level effects, we are still lacking solid evidence on the effects of country-level variables. In this paper I investigate the influence of macro-economic conditions on public demand for redistribution. I use data from the European Values Study (1990-2009). I argue that most cross-sectional research suffers from omitted variable bias. I propose a pseudo-panel model that makes use of longitudinal variation to draw better causal inference. The proposed model makes use of the hybrid-approach and allows disentangling within- and between-unit effects. The results show that within- and between-unit effects can be substantially different. I argue that within-unit effects should generally be better estimates than between-unit effects. Therefore, the analysis casts general doubt on the validity of cross-sectional analysis of country-level effects. My results suggest that increasing social spending leads to less demand for redistribution after a certain level of redistribution is reached (saturation effect). This result contradicts many results published in previous cross-sectional research. I also find a significant positive within-effect of unemployment rates on public demand for redistribution. I find a negative within-effect of economic inequality, which casts doubt on the median voter hypothesis. Finally, I find that increasing economic wealth has a positive but diminishing effect on public demand for redistribution.
Employment biography and income in later life: the impact of pension systems in European welfare states
Abstract: In this paper I analyse the impact of the design of national pension systems on the relation of individuals' work histories and their pension income. Therefore, data on employment histories of individuals in 13 European countries from the third wave of the Survey of Health, Ageing and Retirement in Europe (SHARELIFE) is combined with macro-data describing national pension systems. To examine how the relation of the work history and the pension income is mediated by national pension systems, country fixed effects regression models with cross-level interaction effects are estimated. The results show that a deviation from the normal biography of continuous standard employment is negatively related to the later pension income of men and women. However, especially for women, the strength of this relation differs among countries and is mediated by the pension system: redistributive elements as well as the integration of private pension schemes compensate for the negative effect of non-standard employment histories.
Paper on request:
Environmental concern for climate change - A welfare state perspective
Abstract: Which societies are willing to take steps towards the mitigation of climate change? Major theories explain variations in the degree of environmental concern of countries according to factors of economic wealth and postmaterialist value orientations. This article investigates the means by which economic wealth is distributed within countries between socio-economic groups. It is argued that countries of equal economic strength but different institutional social policy arrangements also exhibit variation in terms of individual and contextual characteristics, which support individual concern for climate change. By addressing the dual nature of climate concern as a long-term value vs. short-term interest based orientation, an argument is forwarded for differences in welfare regimes’ support for socio-economic positions holding an interest in climate policies and the emergence of postmaterialist value orientations.
Relying on survey data from the 2008 Eurobarometer, this issue is examined using multilevel regression analysis for EU-27 countries. The empirical analysis demonstrates country variation in terms of concern for climate change, which is explained by institutional differences in the welfare regime. This effect, however, is mainly compositional, where differences in the number of individuals holding an interest in climate policies or postmaterialist value orientations account for the variation in climate concern across welfare regimes.
Paper on request:
Gerechtigkeitsbewertungen des eigenen Einkommens im gesellschaftlichen Kontext – Spielt der soziale Vergleich und die Einkommensungleichheit eine Rolle?
In this paper I analyze the relevance of the social context for reflective income justice evaluations. On the basis of "A New Theory of Distributive Justice" (Jasso 1980) it is assumed that social context comprises first and foremost the influence of the actual income inequality within a social group (in my case country) on the justice evaluations of the individuals. Secondly, social context includes the impact of the relevant average income on justice evaluations, and thus implicitly on justice judgments. This follows Berger et al. (1972) “Status-Value Theory”, according to which an individual's desired vision of a just reward is derived from the comparison with relevant generalized others. A multilevel analysis on the basis of the ISSP 1999 shows that justice evaluations are determined by both, income inequality and a person's relevant average income. Individuals evaluate their income on average the more unfair, the more unequal the incomes are distributed in their country. And individuals evaluate their income the more unfair, the less they earn compared to people with the same income relevant characteristics.
Paper on request
The experience of social mobility and the formation of attitudes Ttoward redistribution
Abstract: The study aims to explain the genesis of attitudes toward redistribution by applying the theory of causal attribution to the phenomenon of subjective social mobility. The paper asks two questions: (1.) Are attitudes toward income redistribution affected by the subjective experience of social mobility, and, (2.) how are these effects moderated by cultural contexts? These questions refer to a potential long-term feedback process between a welfare state's success in providing equal opportunities and individual attitudes toward welfare state actions. The hypotheses are tested with a multilevel design based on 21 countries and three time-points using international survey data from the ISSP (International Social Survey Programme). The results suggest that the experience of downward mobility leads to an increase in the support for redistribution. Furthermore, a negative effect of upward mobility could be identified for the population of men. These effects are independent of the indirect mobility effects that are due to changes in vertical positions within the society. The analysis demonstrates that the individual effect of being upwardly mobile is moderated by the cultural context. In particular, the analysis suggests that the negative effect of upward mobility is amplified in strongly individualistic countries, while it is weakened in collectivistic countries.
The influence of immigrant status on health. Exploring the subjective health status of first and second generation Turkish immigrants in Germany
Abstract: Using data from the Generations and Gender Survey (GGS), the health status of Turkish immigrants in Germany is observed in this paper. The GGS includes approximately 10,000 German natives (surveyed in 2005) and 4,000 Turkish immigrants (surveyed in 2006) living in Germany. Logistic regression models are estimated to compare the health of first and second generation Turkish immigrants to that of German natives. Differences in health are clear and, contrary to the expectations derived from existing literature, Turkish immigrants do not seem to be in worse health than the native German population when different variables are taken into account. Especially when socio-economic status and coping resources are considered, migrant status has no significant effect on the subjective health status. Furthermore, Turkish immigrants are, to some extent healthier than their German counterparts when variations between East and West Germany are taken into account. Additionally, separate models for Turkish immigrants and German natives are estimated, and it can be shown that Turkish immigrants who have lived in Germany for a shorter period of time (as it is the case in East Germany) have a health advantage.
Karsten Hank & Michael Wagner
Parenthood, partnership status, and well-being in later life: Evidence from SHARE
Abstract: Using data from the first two waves of the Survey of Health, Ageing and Retirement in Europe (SHARE), we address the question of whether and how parenthood and partnership status are associated with various dimensions of elders’ well-being, which we define by the individual’s economic situation, psychological distress, and social connectedness. The results of our multivariate analysis suggest that childless individuals do not fare worse than parents in terms of their economic, psychological, or social well-being. Rather, unmarried and childless individuals are less likely to report financial difficulties than parents do. Childless elders and parents do not differ with regard to symptoms of depression, and neither does having a partnership, per se, contribute to greater psychological well-being: only those reporting satisfaction with the extent of reciprocity in their relationship exhibit lower odds of depression than the currently married. We observe no statistically significant association between parenthood (partnership, respectively) and individuals’ propensity to participate in social activities. These findings are fairly universal; that is, they hold for both men and women, they are stable across various cohorts, and they do not vary systematically across countries.
Social spending generosity and income inequality: A dynamic panel approach
Abstract: This paper explores whether more generous social spending polices in fact lead to less income inequality, or if redistributive outcomes are offset by behavioral disincentive effects. To account for the inherent endogeneity of social policies with regard to inequality levels, I apply the System GMM estimator and use the presumably random incidence of certain diseases as instruments for social spending levels. The regression results suggest that more social spending effectively reduces inequality levels. The result is robust with respect to the instrument count and different data restrictions. Looking at the structure of benefits, particularly unemployment benefits and public pensions are responsible for the inequality reducing impact. More targeted benefits, however, do not significantly reduce income inequality. Rather, their positive effect on pre-government income inequality hints at substantial disinctive effects.
Dennis Spies and Simon Franzmann
A two-dimensional approach to the political opportunity structure of Extreme Right Parties in Western Europe
Abstract: Previous studies on the electoral fortunes of Extreme Right Parties (ERPs) have pointed to the importance of variables of party competition for the success – or failure – of ERPs. These studies vary greatly when it comes to describing the political opportunity structure of the Extreme Right. Apart from their methodological differences, existing studies differ especially with regard to the assumed underlying dimension of party competition. In this article, we test the impact of three frequently discussed variables in the political opportunity structure of ERPs (mainstream party convergence, position of the established right and party system polarization) on the vote share of ERPs in Western Europe. In addition to examining previous studies in this field, we focus on the interplay between the economic and the cultural dimensions as part of the political opportunity structure. We show that a decrease in polarization with regard to economic questions is accompanied by a growing salience of ERPs’ core issues, leading in the end to an increase in ERPs’ vote share.
Why the working-class turned Right
Abstract: While the overrepresentation of working-class members among the electorates of extreme right parties (ERPs) in Western Europe is well documented, previous studies have usually explained this pattern as a result of this voter group’s changing political preferences. In contrast to these studies, this article argues that it is not the changing political preferences of the working class that lead them to vote for ERPs, but changes in the supply side of party competition that have caused the re-orientation of these voters toward the extreme right.
Differentiating between an economic and a cultural dimension of party competition, it is shown that both the policy options offered by parties to voters as the salience of the two issue-dimensions have changed dramatically over the last three decades. While the salience of economic issues as well as of party system polarization among these issues have declined in most Western European countries, the very opposite trend can be identified for non-economic issues, including the core issues of ERPs (immigration and law and order).
These changes on the supply side of party competition cause working-class voters to base their vote decisions solely on their authoritarian, non-economic preferences and not — as in the past — on their left-wing economic demands. The theoretical assumptions are tested empirically with data from the Eurobarometer Trend File for the period from 1980 to 2002. In contexts where the economic dimension is more polarized than the cultural dimension, the positive impact of being a member of the working class on the vote decision for an ERP is significantly reduced.