Subjective social class has a bad name , but predicts life chances well

In this paper, I analyze the association between individuals’ perceptions of income inequality and their subjective social status. Schneider (2019) posits that self-perception of social status is affected by income inequality through two mechanisms. First, a higher level of income inequality means that incomes are shifting apart, and in the presence of upward comparison, a discrepancy between the individual’s income and the reference income is higher. In other words, a higher level of income inequality may increase the feeling of relative deprivation and therefore may lower subjective social status. Second, income inequality may increase the frequency of social comparison and may increase the salience of people with high income, which results in stronger effects of upward social comparison. I use four waves (1992, 1999, 2009, 2019) of the “Social Inequality” module of the ISSP that cover 28 countries and almost 70,000 individuals. I use a measure of subjective inequality perception that is similar to the conventional Gini coefficient. It is based on questions that ask respondents to estimate the earnings of people working in different occupations. I use respondents’ earning estimations for five occupations (doctor, chairman of a large national corporation, cabinet minister in the national government, shop assistant, unskilled worker in a factory) that are available in all four waves. The main outcome variable is subjective social status measured on a 10-point scale. I show that the higher the level of perceived income inequality is, the lower the individual’s perception of her social standing, even if objective income inequality and preferences for the legitimate level of income inequality are controlled for. The effect size is relatively large: a one standard deviation change in perceived inequality is associated with a 10.8% standard deviation change in subjective social status. The results are robust to the measure of perceived inequality and the choice of the outcome variable. It is worth noting that these results are based on estimates that control for income and other objective measures of social status (e.g., education, labor market status, occupation) of the respondents. Whereas empirical evidence shows that social status and inequality perception are correlated (Knell & Stix, 2020; Kuhn, 2011), the estimated negative association between perceived inequality and subjective social status captures other mechanisms beyond economic self-interest. These mechanisms might be the increased frequency and importance of social comparison or the increased feeling of relative deprivation (Cheung & Lucas, 2016; Schneider, 2019; Sommet et al., 2019). The analysis also provides evidence that the estimated association is weaker for individuals with higher income, higher education, and countries without postcommunist history. Overall, the results suggest that not only do objective inequality and perception of fairness have consequences regarding subjective social status but also the perceived level of income inequality itself.


Subjective social class has a bad name, but predicts life chances well
Daniel Oesch and Nathalie Vigna (University of Lausanne)

Research question
Over the last decades, the study of subjective class has been eclipsed by research on objective class. The recurrent mismatch between subjective and objective class has led to the common wisdom that self-reported class is a poor measure of people's life chances. This paper questions this common wisdom. We examine whether subjective class account for less variance than objective class in life chances as measured by income and wealth. We thus compare the predictive power of different class measures.

Hypotheses
Several elements suggest that subjective class could be a good measure of people' social standing. The hierarchy of occupations seems clearly visible to laypeople. This is suggested by the strong crossnational correlation of Treiman's prestige scales -scales based on laypeople rank-ordering sets of occupations. Moreover, individuals have much more information than do survey researchers, and they have not only their own material situation in mind, but also their household's -and life chances primarily depend on the economic means of the household. We then test the following hypotheses: H 1: subjective class accounts for more variance in household income and wealth than does objective class. H 2: the advantage of subjective class over objective class is greater in predicting household income than personal income.

Data, countries and methods
Our analyses are based on 55 country surveys from ISSP 2009 and 2019. We present results aggregated by six groups of countries: continental West Europe; Eanglish-speaking countries; Scandinavia; Suthern Europe; Easter Europe; Africa, Asia and South America. Our dependent variables are income and wealth. Our two key predictors are 5-category measures of subjective class and 5-category measures of objective class (we also run robustness with several class measures such as ESEC9 or Oesch5). We compare the variance in income and wealth that is accounted for by different class indicators. In practice, we estimate linear regressions on household income, personal income and wealth for each country, and compare the adjusted R2 for either subjective or objective class.

Main results
We show that subjective class systematically accounts for more variance in household income than do the three indicators of objective class. Particularly notable are the Continental Western European countries where the sole indicator of subjective class accounts for 23% of variance in household income, as compared with 13 to 15% for objective class measures. The disparity between objective and subjective class in predicting wealth is even larger than for income, with the strongest contrast in the English-speaking countries where subjective class accounts for 16% of variance as compared to 5% for objective class. But subjective class also accounts for twice as much variance in wealth as objective class in the Scandinavian, Continental and Southern European countries. The advantage of subjective over objective class is specific to the household level. At the personal level, subjective and objective class account for similar amounts of variance in income. But then subjective class only requires one single question and is thus much easier to measure.

Abstract
In this paper, I analyze the association between individuals' perceptions of income inequality and their subjective social status. Schneider (2019) posits that self-perception of social status is affected by income inequality through two mechanisms. First, a higher level of income inequality means that incomes are shifting apart, and in the presence of upward comparison, a discrepancy between the individual's income and the reference income is higher. In other words, a higher level of income inequality may increase the feeling of relative deprivation and therefore may lower subjective social status. Second, income inequality may increase the frequency of social comparison and may increase the salience of people with high income, which results in stronger effects of upward social comparison.
I use four waves (1992,1999,2009,2019) of the "Social Inequality" module of the ISSP that cover 28 countries and almost 70,000 individuals. I use a measure of subjective inequality perception that is similar to the conventional Gini coefficient. It is based on questions that ask respondents to estimate the earnings of people working in different occupations. I use respondents' earning estimations for five occupations (doctor, chairman of a large national corporation, cabinet minister in the national government, shop assistant, unskilled worker in a factory) that are available in all four waves. The main outcome variable is subjective social status measured on a 10-point scale.
I show that the higher the level of perceived income inequality is, the lower the individual's perception of her social standing, even if objective income inequality and preferences for the legitimate level of income inequality are controlled for. The effect size is relatively large: a one standard deviation change in perceived inequality is associated with a 10.8% standard deviation change in subjective social status. The results are robust to the measure of perceived inequality and the choice of the outcome variable. It is worth noting that these results are based on estimates that control for income and other objective measures of social status (e.g., education, labor market status, occupation) of the respondents. Whereas empirical evidence shows that social status and inequality perception are correlated (Knell & Stix, 2020;Kuhn, 2011), the estimated negative association between perceived inequality and subjective social status captures other mechanisms beyond economic self-interest. These mechanisms might be the increased frequency and importance of social comparison or the increased feeling of relative deprivation (Cheung & Lucas, 2016;Schneider, 2019;Sommet et al., 2019). The analysis also provides evidence that the estimated association is weaker for individuals with higher income, higher education, and countries without postcommunist history. Overall, the results suggest that not only do objective inequality and perception of fairness have consequences regarding subjective social status but also the perceived level of income inequality itself.
Keywords: subjective social status, inequality perception, income inequality JEL codes: D31, D63, I31, J31 The moderating effect of subjective socioeconomic status on the association between economic comfort and happiness

Atefeh Bagherianziarat
Introduction: The level of happiness and the effect of income on happiness are found to be paradoxical. Some studies confirmed that high-income society has a lower income effect than low-income society. However, a huge body of studies also showed a strong cyclical relationship between different indicators of mental well-being including happiness and both GDP per capita and household income. To explain this paradox some socio-psychological factors have been suggested to be considered as the moderating/mediating mechanisms such as perceived fairness, deservingness, promising future, and subjective socioeconomic status (SSS). The latter index, which refers to a person's conception about their socioeconomic standing in reference to others, has attracted a huge attention in studies on the relationship between inequality and mental well-being and mental health. However, to the best of our understanding, the relevant studies have considered SSS as a mechanism that mediates the association between income and the components of wellbeing and also little studies have explicitly focused on the moderator effect of SSS on income-happiness link.

Method:
Exploring the mediator effect of SSS on the link between income and happiness, In our study, we first reviewed some arguments related to social comparison theories elaborating the socio-psychological mechanisms through which the investigated association can be assumed. Accordingly, we utilized ISSP data, module 2019, Czech modification, to test the study's main hypothesis which assumes that the higher levels of SSS decreases the positive association between individuals' economic condition and their happiness. We applied a linear regression model on the data while controlling for age, gender, education level and marital status of respondents.

Results:
The results supported the study's hypothesis and at the same time challenged some previous studies that confirmed the effect of SSS can be a stronger predictor of happiness rather than actual economic condition. The main effect of economic comfort on happiness was significantly stronger than that of SSS on happiness, although both links were statistically significant even after entering the controlling variables into the model. This result is contradictory with some previous studies showing that perception of social status affects mental-wellbeing components above the actual economic condition. Nevertheless, consistent with social comparison theories and our main hypothesis, the moderation interaction effect of SSS and economic comfort on happiness was negative and significant, illustrating that having higher levels of economic comfort will positively affect everyone the same. It deepens to some significant extends to the degree people consider themselves higher in socioeconomic status.
Conclusion: Accordingly, we argue that both community and personal-level efforts regarding happiness should consider the factors that led people to get involved in evaluating themselves compared to others which both directly and indirectly impact their happiness. However, this effect has been shown not to be the same in every society. The influence of social and cultural contexts are required to be considered and investigated in other studies.

Family Origin versus Individual Effort in the Creation of Human Capital: China in Comparative Perspective
Tony Tam (tony.tam@cuhk.edu.hk ) and Peng Wang Department of Sociology, The Chinese University of Hong Kong This paper examines the relative role of SES origin and individual effort in the determination of educational achievement, with a special emphasis on understanding the case of China from a comparative perspective.
Making use of three optional retrospective behavioral items in the 2009 ISSP, this paper develops an operational measure for the concept of a student's self-discipline at age 14/15/16. This measure of self-discipline (SD15) provides a novel operationalization of a student's level of effort relevant to educational achievement. Conceptually, the measure reflects personal investments that take the form of (a) behavioral consistency and dependability (compliance costs) or (b) forgoing behavioral options (opportunity costs). This investment-based measure of a student's effort can be added to, and thus significantly extend, a standard model of inequality of educational opportunity (IEO). For China, the optional items include three for the student phase and three for the current phase. After incorporating pairwise correlated errors between adolescent and current indicators of self-discipline, a structural equation model (SEM) shows that the indicators of student self-discipline have statistically significant but small correlations with similar indicators of adult respondents.
Previous studies have indicated that individuals with different hukou origins were facing different mobility opportunity structures (Chan and Zhang 1999;Wu and Treiman 2004;2007). During the reform era, there can be two distinct worlds of social mobility for people of rural and urban hukou origin. It is necessary to rule out the confounding effect of hukou barrier in gauging the intergenerational association in China's society. Thus we added an analysis in which we focus on people of urban hukou origin.
To examine China's extended model of IEO from a comparative perspective, we conducted cross-national analyses to compare China to two sets of societies in the 2009 wave of the ISSP, specifically, post-Communist societies in East Europe and Confucian miracle growth economies in East Asia. The percentile ranking score is used to provide a relatively margin-free measure of IEO for cross-national comparison.
The preliminary results ( Fig. 1 and 2) show that the effect of family origin on a child's educational attainment is quite high in China, whereas the impact of effort is much lower than that of family origin (just a quarter of the latter). These findings challenged the prevalent idea that China is a socially open society with merits as important factors in getting ahead (Blau and Ruan 1990; Xie 2016; Xie and Zhang 2019), while echoing the view that intergenerational association is strong in China. However, when we focus on the sample of urban hukou origin, although the effect size of effort did not change much, the association between family origin and destination dropped significantly. In this case, the ratio of effort to family origin in affecting social outcomes turned out to be not significantly different from 1.

Institute of Sociology of the Czech Academy of Sciences
Relations between the level of corruption and perception of income fairness:

Multilevel analysis of ISSP countries
Over the last decades, we have witnessed important changes in the society, such as economic, political or social changes as well as an unprecedented increase in economic inequality and its impact on the society as a whole. While income inequality is a crucial and at the same time well-researched issue in understanding the values and behaviour of the society, its perceived fairness is much less analysed issue, especially at a macro level -across countries and regions.
As previous research suggests, at the individual level, income fairness is associated among others with age, economic status, gender, religious or political views. At the macro level, the research is rather scarce with the exception of effects of Gini coefficient or GDP on income fairness. However, an important gap in research of perceived income fairness is related to a political situation in a country, especially to the quality of governance or corruption levels. This research wants to fill this gap. In general, high levels of corruption can damage the public sense of fairness regarding income levels. If there is a widespread corruption in a society, people can perceive the income distribution as unfair, because the allocation of resources is not based on merit, but rather on the basis of connections or bribery. High corruption in a country might be therefore associated with low perceived income fairness.
However, perceived income fairness varies greatly not only within countries but also across countries and cultures. This proposal looks specifically into the issue of perceived income fairness in countries taking part in the ISSP and aims to explain the relationships between corruption and the perception of fairness. The connections are analysed using the ISSP 2019 data, specifically on the "Social inequality" module. Moreover, the World Bank "Control of Corruption" is used for corruption measure at the country level. In addition to pooled data analysis of all the countries surveyed, the paper also looks at different countries and regions to explore different dynamics of connections between income fairness and level of corruption. This research employs multilevel modelling in order to analyse the individual and country effects in income fairness perceptions at the same time and looks into different micro and macro variables influencing income fairness perception at the individual level in different countries and regions.
Findings suggest that corruption at the country level is indeed a strong predictor of perceived income fairness at individual level, meaning that not only economic situation in a country and at individual level is connected to perception of income fairness, but also general political situation and governance, suggesting the gravity of policy implications. However, the strength of this relation varies significantly across regions. All over the world, income inequality remains a central concern for societies as higher levels may have negative effects on the legitimacy of the political system, social cohesion, and democracy in general. The effects of income inequality, and more particularly, the politicization of income inequality depends on the perceptions of actual levels as well as accepted or tolerated levels, a questions that still remains unsettled.

Keywords
The relationship between objective, perceived and tolerated income inequality has been extensively studied in recent decades. Using the ISSP 2019 module, we build on such research and analyze the perceptions of inequality and the levels of tolerance to income differences in 27 countries. We explore their variation between and within countries by educational level and perceptions of social mobility, as well people's own status in society. In addition, we explore how tolerated levels of income inequality are related to objective indicators of inequality such as the Gini Index. In order to understand whether the redistributive role of the State influences tolerance to inequality, we analyze how the levels of tolerated inequality are related to the changes in the Gini Index between measures pre and post transfers and taxes. Finally, we hypothesize that countries that have a greater the level of income inequality post taxes and transfers, there is also a greater the tolerance of inequality due to a resignation or acceptance of the status quo.

A Few People Make all the Difference -an International Comparison of "Fair" Pay Differentials
Abstract Social inequality has always been an important topic of public debate in almost all societies in the world, but how much do people actually disagree and who is that does? This paper uses the 1987, 1992, 1999, 2009, and 2019 waves of International Social Survey Programme data to compare attitudes toward "fair" pay ratios in Germany, Italy, Hungary, Norway, Great Britain, the USA, and Russia. Although respondents generally underestimate the actual size of current earnings gaps, in all countries an overwhelming egalitarian majority agrees that "income differences are too large". As well, since the ISSP has also asked respondents how much different occupations "should earn", one can compare the fair pay ratios.
In all countries, for all years examined, fair pay ratios are (a) remarkably small and (b) remarkably similar for roughly 75 to 80% of the population. Cross-country differences in average attitudes do not occur due to higher general levels of support for income inequality but are rather concentrated in the 'inegalitarian few'. Our analysis of cross-country differences in attitudes toward inequality, therefore, concentrates on how the inegalitarian few differ from the egalitarian many, to differing degrees in different countries, in order to better explain the political tensions underlying differences in redistribution in market economies.

Should we care of pay ratios?
Renzo Carriero

Department of Cultures, Politics and Society -University of Torino
Collegio Carlo Alberto -Torino Measures of attitudes toward income inequality based on actual and just pay ratios are a distinctive feature of ISSP modules on social inequality. However, they have been heavily criticized by Pedersen and Mutz (2019) who found such measures to be prone to anchoring and ratio biases. In this contribution, drawing on the latest edition of the "Social Inequality V" module (2019), I show that pay ratios can still be useful to compute a measure of individual sensitivity to income inequality, i.e. the extent to which estimated inequality (perceived pay ratio) should be reduced to match ideal inequality (just pay ratio). Such sensitivity measure significantly and substantially predicts attitudes toward income inequality based on a composite index of nine standard (word-based) items. Specifically, it raises R squared by 33% (or +6 points) after controlling for social class, education, party voted and demographic variables (age, sex, country). However, partial associations between the composite index, on the one hand, and class, education and party voted, on the other hand, are hardly affected. This suggests that the ratio of estimated to ideal inequality captures a peculiar feature of attitudes toward income inequality which is not correlated with classic predictors such as class and political orientation. Further research is needed to understand the sources of this individual characteristic, but contrary to Pedersen and Mutz's conclusions, pay ratios should not be easily dismissed.
Overcoming the structuralist/individualist dichotomy: However, this categorization is derived from factor analytic techniques, which were repeatedly found to underestimate the number of factors underlying data, especially when highly correlated.
Moreover, despite attitudes toward inequality being an heterogenous set of judgments, perceptions and beliefs, scholars usually studied them in isolation 1 , neglecting their interconnections. Hence, this paper seeks to answer a two-folded research question, by adopting a complex system perspective originated from network psychometrics. Firstly, the alleged two-dimensionality of the inequality beliefs battery will be investigated trough an innovative technique called Exploratory Graph Analysis 2 . Secondly the interplays between these items and other judgments and beliefs toward inequality will be studied through the lens of the Causal Attitude Network model 3 . Here an attitude is conceptualized as a network in which nodes represent attitude elements that are connected by direct causal interactions.
The following hypotheses will be tested.
H1: The inequality beliefs battery will show more than two dimensions in the majority of the ISSP countries.
H2: The items composing the individualist beliefs dimension will be more central in the attitude networks of countries characterized by high GINI.
H3: A Network Comparison test 4 will show structural differences between attitude networks of countries with high (versus low) GINI.
To test our hypotheses we will use data from the 2019 ISSP Module of Social Inequality. For H1, variables Q1a to Q1j and all 29 countries of the sample will be analysed. H2 and H3 will investigate a broader set of variables tapping perceptions, beliefs, and judgments, comparing the attitude networks of three countries with low, medium, and high GINI.
Our preliminary results are summarized in Figure 1 (see below). This plot aggregates the results of the EGA performed on each country in the sample. Remarkably, inequality beliefs are clustered in more than two components in all but three countries. EGA seems to indicate that besides the classical distinction between individualists and structuralists beliefs, a political dimension is present.
Our contribution represents the first application of EGA to the inequality beliefs battery, as well as the first comprehensive and comparative study of the miscellaneous elements composing the attitude toward inequality construct.  Figure 1: Results of EGA performed on the inequality beliefs country in each ISSP country

Dynamics of Low-Effort Responding across ISSP Social Inequality Module
Public opinion surveys can serve as barometers measuring the dynamics of prevailing attitudes in societies. Participation in surveys may be seen as especially useful in democratic societies, where the governing elites and their contenders presumably monitor the demand side of the electoral equation. However, survey estimates can be biased due to a number of potential sources of errors. Analysts use various weighting strategies to correct for biases caused during the design and collection stage of the process (e.g., coverage or non-response errors). As opposed to mere absence of responses from particular groups, however, it is more difficult to make adjustments for low quality of responses. One source of low quality of responses is careless and insufficient (low-effort) responding, meaning lack of engagement with the questions.
We explore a proposition that one of the reasons for contemporary rates of low-effort responding in Eastern European countries could be that, after the initial surge of optimism about the nature of democracy and expectations about elite responsiveness in the early post-communist era, the surveys stopped being seen as a potentially useful novelty and, as troubles tied to transition settled in, people became disillusioned and perhaps even cynical about the value of carefully considering their answers. We predict that there will be an overall trend of increase in low-effort responding over time. We also predict that the quality of responses measured via indicators of low-effort responding will be initially similar among both the Western and the Eastern European countries, and the increase in the incidence of low-effort responding will be greater in Eastern European countries.
In the empirical part of the paper, we analyse 5 waves of ISSP on social inequality, covering a period from 1987 to 2019. Using multiple measures of low-effort responding, we look at shares of low-effort responders across particular batteries of questions on particular topics and shares of low-effort responders across whole surveys. We also consider mode of survey administration, reasoning that respondents may be motivated to respond differently based on the immediate situation during which they consider their answers. We then compare these over the years, controlling for a number of individual and country-level measures, and describe the patterns showing sources of differences in low-effort responding.
Preliminary findings show an overall pattern of increase in incidence of low-effort responding. Finally, we juxtapose and discuss the patterns of low-effort responding in the Eastern European countries compared to the Western Europe countries and patterns of low-effort responding based modes of administration.
Keywords: cross-cultural surveys, data quality and low-effort responding

Assessing the Quality of the Occupation Indicators in the ISSP Social Inequality Modules IV and V Using a Double Indicator MTMM Design
Harry B.G. Ganzeboom, Department of Sociology, VU University Amsterdam Tamira E. Sno, Faculty of Social Sciences, Anton de Kom University of Suriname

Summary
The ISSP 2019 Social Inequality V module as well its 2009 predecessor contains a ten-category crude measure of occupations as an optional question, next to its ISCO-coded detailed measures of occupations. This repeats a design implemented in the ISSP 1987 for a small (5)  As argued elsewhere (Ganzeboom, 2005;De Vries & Ganzeboom, 2008), the crude measure of occupation may be as effective a measurement of occupational status as a detailed indicator. Moreover, including the crude indicator next to the detailed indicator in a multiple-indicator measurement model allows one to examine the measurement quality of each indicator. In this paper, we examine the quality of the occupation indicators in the ISSP for all countries and rounds that included a double measurement one way or the other. We estimate validity and reliability coefficients in an intergenerational status attainment model that connects father's and mother's occupations with respondent's education, first occupation, current occupation and personal income. Validity and reliability are separated using the Saris & Andrews (1991) reparametrization of the classical Multi-Trait Multi-Method model.
The results show (tentatively): • Crude and detailed measures have about the same quality with respect to systematic measurement error (invalidity) -validity is almost perfect for both. • However, the detailed measure suffers from a higher level of random measurement error (unreliability) than the crude measure. • More importantly, both measures suffer from total measurement error, which can be only diagnosed and corrected when they are simultaneously used in a latent variable model. These latent variable models give a substantially corrected account of status attainment and social reproduction patterns in the ISSP countries.