Country effects on outcomes for individuals are often analysed using multilevel (hierarchical) models applied to harmonized multi-country data sets such as ESS, EU-SILC, EVS, ISSP, and SHARE. We point out problems with the assessment of country effects that appear not to be widely appreciated, and develop our arguments using Monte Carlo simulation analysis of multilevel linear and logit models. With large sample sizes of individuals within each country but only a small number of countries, analysts can reliably estimate individual-level effects but estimates of parameters summarizing country effects are likely to be unreliable. Multilevel modelling methods are no panacea.
Logistic regression estimates do not behave like linear regression estimates in one important respect: They are affected by omitted variables, even when these variables are unrelated to the independent variables in the model. This fact has important implications that have gone largely unnoticed by sociologists. Importantly, we cannot straightforwardly interpret log-odds ratios or odds ratios as effect measures, because they also reflect the degree of unobserved heterogeneity in the model. In addition, we cannot compare log-odds ratios or odds ratios for similar models across groups, samples, or time points, or across models with different independent variables in a sample. This article discusses these problems and possible ways of overcoming them.
This study attempts to further our understanding of the contextual sources of anti-immigrant sentiments by simultaneously examining the impact of immigrant group size, negative immigration-related news reports and their interaction on natives' perceived group threat. We test our theoretical assumptions using repeated cross-sectional survey data from Spain during the time period 1996–2007, enriched with regional statistics on immigrant group size and information from a longitudinal content analysis of newspaper reports. Drawing on multilevel regression models, our findings show that a greater number of negative immigration-related news reports increases perceived group threat over and above the influence of immigrant group size. Additionally, our findings indicate that the impact of negative immigration-related news reports on perceived group threat is amplified (weakened) in regions with a smaller (larger) immigrant group size. Collectively, these results testify to the importance of immigrant group size and negative immigration-related news reports as key contextual sources of natives' perceived group threat.
This article discusses the determinants and the development of public concern for the state of the natural environment. First, we review some theoretical approaches that try to explain individual as well as cross-national differences in environmental attitudes. Particularly, we discuss Inglehart's theory of post-materialism. Dunlap and Mertig's globalization explanation, and the prosperity hypothesis. Second, we test these hypotheses by applying multilevel analysis to the International Social Survey Programme (ISSP) data from the years 1993 and 2000. The results support, above all, the prosperity hypothesis. Individuals with higher relative income within countries display higher levels of environmental concern than their compatriots, and additionally, more concern is reported in wealthier countries than in poorer nations. The results indicate that environmental concern is also closely associated with post-materialistic attitudes and various sociodemographic variables. Comparing the environmental concern measured in the ISSP in 1993 with that in 2000 shows that environmental concern has more or less stabilized since the early 1990s in the countries under scrutiny.
Many surveys of respondents from multiple countries or subnational regions have now been fielded on multiple occasions. Social scientists are regularly using multilevel models to analyse the data generated by such surveys, investigating variation across both space and time. We show, however, that such models are usually specified erroneously. They typically omit one or more relevant random effects, thereby ignoring important clustering in the data, which leads to downward biases in the standard errors. These biases occur even if the fixed effects are specified correctly; if the fixed effects are incorrect, erroneous specification of the random effects worsens biases in the coefficients. We illustrate these problems using Monte Carlo simulations and two empirical examples. Our recommendation to researchers fitting multilevel models to comparative longitudinal survey data is to include random effects at all potentially relevant levels, thereby avoiding any mismatch between the random and fixed parts of their models.
This study focuses on ethnic competition as a contextual explanation of cross-national differences in anti-immigrant prejudice. It contributes to the existing literature by refining the concept of ethnic competition into a socio-economic and a cultural aspect, which is reflected in two different measures of outgroup size. To improve cross-national comparability, the outgroup size measure is based on foreign country of birth instead of citizenship. Moreover, as outgroup size does not only measure competition, but also contact opportunities and familiarity with immigration, intergroup contact theory is taken into account and a non-linear relationship between outgroup size and perceived ethnic threat is tested. This study employs multi-level linear regression and uses the first round data set of the European Social Survey. The main conclusions of this analysis are that economic and social competition between groups might play a lesser role in the explanation of cross-national differences in anti-immigrant attitudes than often assumed, and that it might be rather lacking familiarity and fear of conflict over values and culture that drive the relationship between outgroup size and anti-immigrant attitudes.
A number of studies have found a negative relationship between ethnic diversity and social capital and assumed from this a harmful effect of diversity on social cohesion. This article suggests that social cohesion must be treated as a multifaceted concept and any analysis into the relationship between diversity and social capital needs to be complemented by an analysis of diversity's effect on 'relations between ethnic groups'. Our results show that while increasing diversity does have a negative impact on social capital, it simultaneously improves perceptions of, and relations between, ethnic groups. Furthermore, we find that forming 'bridging' ties in diverse environments plays a significant role in the positive relationship between diversity and tolerance, and that the presence of 'bridging' ties can also reduce the negative impact of diversity on social capital. However, while our results show that diversity has both positive and negative effects on social cohesion, we find that it is disadvantage which has the most detrimental impact, undermining both social capital and interethnic relations. We also find evidence that using a more sensitive measure of diversity (creating an area typology based on the proportional size, number, and type of ethnic groups in an area) reveals that living in different structures of diversity may lead to different social cohesion outcomes.
We present a new theory of erotic capital as a fourth personal asset, an important addition to economic, cultural, and social capital. Erotic capital has six, or possibly seven, distinct elements, one of which has been characterized as 'emotional labour'. Erotic capital is increasingly important in the sexualized culture of affluent modern societies. Erotic capital is not only a major asset in mating and marriage markets, but can also be important in labour markets, the media, politics, advertising, sports, the arts, and in everyday social interaction. Women generally have more erotic capital than men because they work harder at it. Given the large imbalance between men and women in sexual interest over the life course, women are well placed to exploit their erotic capital. A central feature of patriarchy has been the construction of 'moral' ideologies that inhibit women from exploiting their erotic capital to achieve economic and social benefits. Feminist theory has been unable to extricate itself from this patriarchal perspective and reinforces 'moral' prohibitions on women's sexual, social, and economic activities and women's exploitation of their erotic capital.
Two issues have been especially contentious in debates over religious change in Europe: the unity or diversity of the trends observed across the continent, and the significance of the large subpopulation that is neither religious nor completely unreligious. This article addresses these problems. An analysis of the first wave of the European Social Survey (ESS) shows that each generation in every country surveyed is less religious than the last. Although there are some minor differences in the speed of the decline (the most religious countries are changing more quickly than the least religious), the magnitude of the fall in religiosity during the last century has been remarkably constant across the continent. Despite these shifts in the prevalence of conventional Christian belief, practice and self-identification, residual involvement is considerable. Many people are neither regular churchgoers nor self-consciously non-religious. The term 'fuzzy fidelity' describes this casual loyalty to tradition. Religion usually plays only a minor role in the lives of such people. Religious change in European countries follows a common trajectory whereby fuzzy fidelity rises and then falls over a very extended period. The starting points are different across the continent, but the forces at work may be much the same.
Divergent findings on trends in inequalities in educational attainment associated with individuals' social origins have led to much discussion of how far these reflect real differences by place and time or, rather, differences in research procedures. But in this latter regard, one issue has received relatively little attention: i.e. that of the conceptualization and measurement of social origins. We propose decomposing social origins into parental class, parental status, and parental education. Following this approach, we analyse data from three British birth cohort studies. We show that these three components of social origins have independent and distinctive effects on educational attainment, and ones that persist or change in differing ways across the cohorts. We also make some assessment of their combined effects. We consider the methodological implications of our findings, in particular for analyses of trends in educational inequalities, and, further, how they might result from other, independently established, changes in social stratification in Britain over the historical period covered.
This analysis of variations in the level of generalized social trust (defined here as the belief that others will not deliberately or knowingly do us harm, if they can avoid it, and will look after our interests, if this is possible) in 60 nations of the world shows that trust is an integral part of a tight syndrome of social, political and economic conditions. High trust countries are characterized by ethnic homogeneity, Protestant religious traditions, good government, wealth (gross domestic product per capita), and income equality. This combination is most marked in the high trust Nordic countries but the same general pattern is found in the remaining 55 countries, albeit in a weaker form. Rural societies have comparatively low levels of generalized trust but large-scale urban societies do not. Cause and effect relations are impossible to specify exactly but ethnic homogeneity and Protestant traditions seem to have a direct impact on trust, and an indirect one through their consequences for good government, wealth and income equality. The importance of ethnic homogeneity also suggests that the difference between particularized and generalized trust may be one of degree rather than kind.
"Research has repeatedly shown that educational opportunities are distributed unevenly in all countries. Therefore, the question is not whether family background and educational outcomes are related but to what degree they are related. This latter question then invites a comparative perspective. That is, does social inequality in education differ across time and countries? If yes, which institutional characteristics can explain differences in educational inequality? Educational inequality is conceptualized as the association between individuals' and their parents' highest educational level attained. Intergenerational educational mobility processes are analysed for 20 industrialized nations by means of log-linear and log-multiplicative models. The results show that the degree of educational mobility has remained stable across the second half of the 20th century in virtually all countries. However, nations differ widely in the extent to which parents' education influences their children's educational attainment. The degree of educational inequality is associated with the institutional structure of national education systems. Rigid systems with dead-end educational pathways appear to be a hindrance to the equalization of educational opportunities, especially if the sorting of students occurs early in the educational career. This association is not mediated by other institutional characteristics included in this analysis that do not exert notable influences on educational mobility." Die Untersuchung enthält quantitative Daten. Forschungsmethode: empirisch-quantitativ; empirisch; Sekundäranalyse; Querschnitt; Längsschnitt. Die Untersuchung bezieht sich auf den Zeitraum 1994 bis 1998. (author's abstract, IAB-Doku).
Educational aspirations are generally based on past academic achievement and families' endowment with the resources needed to reach targeted educational levels. However, although they perform worse at school and hold lower social status, previous research observes that some ethnic minorities tend to express higher educational ambitions than natives. This study discusses and tests possible reasons for this striking finding using German data from the Young Immigrants in the German and Israeli Educational Systems project, which includes families from Turkey and the former Soviet Union. The results reveal that Turkish students hold higher aspirations than their native counterparts, whereas no aspiration gap was found between natives and adolescents from the former Soviet Union. While German students' aspiration patterns can mainly be ascribed to status attainment motivation, Turkish students' high educational ambitions seem to be stimulated by a desire of status upward mobility.
In this study, we set out to explain anti-Muslim attitudes in the Netherlands. Although the presence and immigration of Muslims have become widely discussed, there is little systematic evidence about the determinants underlying anti-Muslim attitudes. Using data from the Social and Cultural Developments in the Netherlands (SOCON) survey (2005, 2006), containing a more detailed measurement of anti-Muslim attitudes, we tested two contradictory mechanisms, derived from ethnic competition theory and intergroup contact theory. Results from hierarchical structural equation modelling indicate that the relative outgroup size induces both intergroup friendship contact as well as perceptions of ethnic threat. However, only the latter turned out to affect anti-Muslim attitudes directly. Moreover, our findings revealed that contact with colleagues belonging to ethnic minority groups reduces negative attitudes towards Muslims and mediates the effect of individual-level determinants on anti-Muslim attitudes. The complementary nature of both ethnic competition theory and intergroup contact theory is illustrated by negative correlation between both mediating mechanisms, as well as the support for a curvilinear relationship between outgroup size and perceived ethnic threat.
In Europe, on average, three times as many adult children occasionally help their parents with the housekeeping than do provide regular physical care. This is not surprising, considering the great differences between these two types of support. Care follows needs, whereas help tends to be given sporadically when one has the opportunity. In the familial welfare states in Southern Europe, where little professional support is available, provision of care by children is more likely—whereas parents in the north are more likely to receive help in the household or in dealing with the authorities. Logistic multi-level models enable these differences to be traced back to the availability of social and health services in the individual countries. There is a 'crowding in' of the help children give their parents, but a 'crowding out' of physical care. Overall, the results based on the Survey of Health, Ageing and Retirement data thus support the specialization hypothesis: professional providers take over the medically demanding and regular physical care, whereas the family is more likely to provide the less demanding, spontaneous help. Everyone does what they do best. The overall care of older people thus tends to be assured both quantitatively and qualitatively by well-developed service systems.
"Using data for seven European countries we analyse trends among women in class differences in educational attainment over the first two-thirds of the 20th century. We also compare educational attainment between men and women; we ask whether class differences among the two sexes are similar or not; and whether trends in class differences over birth cohorts have differed between men and women. We find that, as expected, over the 20th century, inequalities between men and women in their educational attainment declined markedly. More importantly, changes in class inequalities in educational attainment have been similar for both men and women, although, in some countries, women displayed greater inequality at the start of the 20th century and have shown a somewhat greater rate of increase in equality. Patterns of class inequality were also largely similar for both sexes, though in some countries daughters of farmers and the petty-bourgeoisie did relatively better than their brothers. While some of these results reinforce what has long been believed, our central finding of a decline in class inequality in educational attainment for both men and women contradicts the 'persistent inequality' in education that earlier scholars claimed existed." Die Untersuchung enthält quantitative Daten. Forschungsmethode: empirisch; empirisch-quantitativ; Aktenanalyse; Querschnitt; Längsschnitt. Die Untersuchung bezieht sich auf den Zeitraum 1900 bis 1970. (author's abstract, IAB-Doku).
Rent-based accounts of inequality argue that institutionalized barriers to the access to labour market positions create artificial restrictions on the supply of labour and, in turn, generate wages for workers in protected positions in excess of the wages they would receive in a competitive labour market. In this article, we extend this argument to the comparative context, and elaborate a rent-based explanation of between-occupation wage inequality in Germany and the United Kingdom. We test it with new and unique data on four institutionalized sources of closure (educational credentialing, licensure, unionization, and apprenticeships), matched to newly constructed measures of occupational skills and to national labour force survey data. We show that in both countries, between-occupation wage inequality is substantial, and much of it can be traced to variations across occupations in closure and to the positive association between closure and wages. We also show that the prevalence and the payoff to each of the four closure institutions differ across the two countries: Specifically, vocational credentialing and unionization have a particularly high payoff in Germany, while tertiary credentialing and licensure have a particularly high payoff in the United Kingdom. These results have important implications for understanding between-occupation wage inequality and cross-national differences in aggregate levels of wage inequality.
China's rural-to-urban migration has affected 12.6 million school-age rural children who have migrated with their parents and another 22 million who have been left behind by their migrant parents. Not enough is known, either theoretically or empirically, about the causal impact of migration on the well-being of this large number of Chinese children affected by migration. Propensity score matching methods are applied to estimate the effects of migration in children 10-15 years old from a 2010 national survey (N = 2,417). Children's migration has significant positive effects on their objective well-being but no negative effects on their subjective well-being. There is little difference between the left-behind and non-migrant children across multiple life domains. The Rosenbaum bounds tests indicate that the causal effects of child migration are sensitive to hidden bias for certain outcomes, but not for others.
Are more equal societies 'better' societies? This article addresses the question as to whether and why income inequality lowers the degree of Europeans' subjective well-being. While in broad international comparisons typically no clear-cut link between income inequality and (un)happiness exists, we can demonstrate that Europeans are somewhat less happy in more unequal places. We further discuss and empirically test three explanations as to why Europeans are inequality-averse, namely (dis)trust, status anxiety, and perceived conflicts. Each of these three potential mediators is hypothesized to be shaped by the extent of a nation's income inequality, and in turn to result in lower subjective well-being. A multilevel mediation analysis with data from the European Quality of Life Survey 2007 for 30 countries reveals that distrust and status anxiety are important mediators of inequality aversion, whereas perceived conflict is not. We can further show that trust is the crucial mediator among affluent societies, whereas status anxiety is crucial among the less affluent societies. The results are discussed with reference to the Spirit Level theory developed by Richard Wilkinson and Kate Pickett.
The 'income inequality hypothesis' holds that beyond a certain level of gross domestic product (GDP) per capita, the association between absolute income, health and mortality weakens, and the distribution of income across a society becomes more important as a determinant of a range of outcomes including average mental and physical health. Recent reviews suggest that the empirical reality of the income inequality hypothesis is now established but fierce debate remains about what explains the association. In this article we describe three hypotheses that have emerged in the literature to explain the association—the social capital, the status anxiety and the neo-materialist hypotheses before operationalizing each, and testing their ability to explain the relationship between income inequality measured using the GINI coefficient and mental well-being measured using the WHO5 scale. We use multi-level models and data from the European Quality of Life Survey which contains information from 30 countries and over 35,000 individuals. Results give most support to the status anxiety and social capital hypotheses and almost no support to the neo-materialist hypothesis. Measures representing the social capital hypothesis reduce the coefficient measuring income inequality by 55% in high GDP countries and render it insignificant as well as providing the best fitting model as measured by AIC and BIC value. However, variables representing the status anxiety hypothesis reduce the income inequality coefficient by more in lower GDP countries.