Over the last five decades, the use of medically assisted reproduction (MAR) to fulfill fertility desires has increased at an unprecedented rate. However, research has highlighted significant socio-economic disparities, with socially advantaged individuals being more likely to have MAR births. Despite this, knowledge of the mechanisms underlying these disparities remains limited. In this study, we provide novel evidence on socio-economic disparities in both MAR take-up and treatment success. We utilize population register data from Denmark covering the period between 2007 and 2017. Preliminary results indicate substantial educational gaps in both MAR take-up and treatment success. Future steps will involve examining additional social gradients and a detailed examination of various heterogeneities within our study population.

Since 2010, fertility levels have fallen unexpectedly in all Nordic countries, with some of the lowest fertility rates recorded in 2022 and 2023. Previous evidence suggests that the decline is general across different population subgroups. However, differential declines between socio-economic groups have been observed, with stronger declines among those in a weaker labour market position. This evidence points to the role of economic uncertainty as one driver of the decline. Overall, the contributing causes of these fertility declines still remain elusive, except that they are driven by fewer people entering parenthood in the first place, despite having entered cohabitation. We contribute to this literature by investigating the role of housing among young adults in explaining these changes. Increased barriers to entering homeownership and greater future uncertainty among renters can be hypothesized to have led to delayed parenthood among young adults since the early 2010s. Using harmonized Nordic individual-level register data, we assess whether the decline in first births is due to declining homeownership, changing fertility behavior within housing tenure groups, or both. Applying discrete-time event history models for the first birth rate and decomposition of this rate, we find that homeownership declined across the four countries among young adults. Preliminary results for Finland show that the first birth rate regardless of tenure status declined in the 2010s. However, the decline in first births was more pronounced among women who rented their homes than among women who lived in an owner-occupied dwelling.

The mean age at first birth has risen across Europe, especially for highly educated women, who often delay motherhood to establish a stable career, but less so for low-educated women, who are still relatively likely to start families at a young age. However, the reproductive polarization differs across welfare regimes. My paper examines how family policies shape the timing of motherhood for women of different educational levels. Using panel data for 29 European countries from the EU-SILC and the GSOEP and policy indicators from Eurostat and the OECD, I analyze how public childcare availability, parental leave length and family allowance generosity affect the probability of women with different levels of education having a child within a given age range. Employing discrete event history models, I find that public childcare provision is associated with a delayed age at first birth for low-educated women, narrowing the age gap between educational groups. Conversely, extended parental leave, which encourages women to stay home with young children, correlates with earlier family formation among low-educated women, reinforcing educational disparities. However, generous family allowances do not significantly affect motherhood timing. My results indicate that family policies that support the reconciliation of work and family life can reduce educational inequalities in the timing of motherhood, but in the absence of such support and with family policies that encourage childcare at home, low-educated women have children at an early age and do not seek to establish themselves in the labor market first.

This study examines the effect of expanding free preschool education on fertility decisions in Israel, a high-income country with exceptionally high fertility rates. We exploit the phased implementation of the Trajtenberg Reform, which extended free public preschool to children aged three and four. Prior to 2012, free education began at age five, imposing substantial childcare costs on parents. The reform was introduced in phases, starting with low socioeconomic status (SES) communities in 2012 and becoming universal by 2015. Using a natural experiment design, we analyze birth patterns before and after the reform’s implementation. We estimate logit models to assess the odds of having a second, third, and fourth child between 2008 and 2018, controlling for socioeconomic and demographic factors. Our findings indicate that fertility rates increased significantly following the reform, with the strongest effects observed among lower-middle-income groups. These results contribute to broader debates on the relationship between childcare availability and fertility. While prior research in low-fertility countries found modest effects of childcare expansions on fertility, our results suggest a stronger impact in Israel, underscoring the role of policy context and cultural norms in shaping childbearing decisions. Additional analyses examined birth spacing and subgroup differences by ethnicity and SES. We also employed advanced hazard models to better capture the timing of fertility decisions and refine our understanding of heterogeneous effects. These findings contribute to the broader debate on how early childhood education policies influence demographic patterns in high-fertility contexts.

Family transitions, such as parental separation and repartnering, are linked to lower educational achievement of involved children. These associations may vary depending on the age at which a child experiences the family transition. Two contrasting hypotheses are examined. The cumulative disadvantage hypothesis posits that early parental separation leads to accumulating negative effects over time, resulting in lower educational achievement the earlier the family transition took place. Conversely, the adaptation hypothesis suggests that younger children have more time to adjust, mitigating negative effects over time. We employ family fixed-effects models using large-scale Norwegian register and combined data to estimate within-family differences in educational achievement at age 16 based on transition timing. Our preliminary findings support the adaptation hypothesis for parental separation for young children: those who experienced separation between 1 and 10 years achieved higher educational outcomes at age 16 compared to siblings who faced it later. This suggests that younger children have more time to adjust, mitigating negative effects over time.

Parental death in childhood adversely affects educational outcomes. However, death is not only an event but also a process. The negative effects may peak around the time of death and gradually improve, or they may persist as children lose a parent’s resources. This study examines how the timing of parental death affects educational achievement during a key school transition, focusing on students’ standardized test scores in the final year of primary education (around age 12). We define timing as the interval between the test and parental death, ranging from seven years before the test to four years after the test. Using administrative data from the Netherlands (1994-2024), we apply sibling fixed effects models to analyze a sample of 1,222,946 children, including 35,823 children experienced parental death. Our finding shows that children who lost a parent before the test scored lower than those who did not. Test scores were already lower when the test was taken one year before death and were lowest when it was taken two years after death. In contrast, the negative effects of parental death did not become weaker for children who took the test three or more years after death, suggesting little recovery. We separated the analysis by causes of death (cancer, cardiovascular diseases, and others) to explore further differences. Our findings shed new light on the role of the crisis model and the resource model in understanding the impact of parental death and emphasize the need for targeted interventions to reduce inequalities among bereaved children.

Empirical research has shown a negative association between parental union dissolution or family transitions and children’s educational outcomes. However, there is a tension regarding the most appropriate measures to reflect family complexity, with recent studies calling for more comprehensive approaches. Drawing on different theoretical perspectives and previous empirical evidence, we apply dynamic sequence complexity indices to conceptualize different measures of family complexity and explore their evolution over children’s school careers. Such measures go beyond the mere number of transitions commonly applied in previous research and cover concepts such as unpredictability and theoretically informed disadvantage. We then examine the relationship between the family complexity proxies and English and math test scores using fixed-effects regression models to assess: Which family complexity measures affect English and math test scores, and at what points during children’s school careers do changes in family structure affect test scores? We use data from the UK Household Longitudinal Study (UKHLS) linked to the National Pupil Database (NPD) to analyze the family life courses of children’s parents and the educational outcomes of children. The UKHLS provides partnership histories, while the NPD offers educational attainment data for pupils in England.
Preliminary results suggest that the dynamics of family complexity during children’s schooling vary considerably depending on the different statistical measures and theoretical family perspectives employed, underscoring the need for a multifaceted approach. Our measures offer versatile tools for future research on family complexity and child development and will be tested comprehensively in relation to children’s education in the next step.

Children’s educational opportunities are shaped by their parents’ socioeconomic status (SES), even when disruptive family events such as parental separation happen. Previous research shows that high-SES families are often better able to mitigate the negative effects of separation, but the mechanisms driving this unequal separation penalty remain unclear. This study explores the role of genetic potential in shaping this relationship. We examine how separation impacts the relationship between genetic potential and educational outcomes, particularly whether this relationship differs for children from high- versus low-SES families. Two hypotheses are examined: (1) a lack of enhancement, which posits that parental separation hinders the realization of high genetic potential; and (2) a lack of compensation, which suggests that separation amplifies the influence of genetic factors, particularly for children with low genetic potential, so that genetic factors are more important for educational outcomes in separated families. We argue that the dynamics may depend on children’s SES background. A lack of enhancement may affect the educational outcomes of children with high genetic potential from low-SES families, while a lack of compensation may lead to a separation penalty for low-potential children from high-SES families, who would otherwise compensate for their genetic disadvantages. Using data from the German Twin-Life study, including polygenic indices, we investigate whether the mechanisms underlying the separation penalty vary based on children's socioeconomic background. Our findings contribute to a better understanding of the mechanisms driving the unequal separation penalty, highlighting the interplay between genetic potential, parental separation, and family SES.

In recent decades, traditional centralised models of education have undergone significant changes, with many countries gradually adopting neoliberal education policies (NLPs). These policies promote school autonomy, competition, forms of privatisation, and teacher accountability but have been criticised for potentially fostering segregation and inequality. In response, many academics have proposed the adoption of Democratic Inclusive Policies (DIPs), which advocate collaborative approaches to promoting equity and social justice in education. Despite the importance of the debate, empirical contributions on the diffusion and effects of these policies are limited. This study aims to fill this gap by providing a first description of the distribution of NLPs and DIPs across countries, while also evaluating the association between policies and educational attainment and inequalities in mathematics. Using data from the OECD-PISA 2022 survey, we apply Latent Profile Analysis to provide a classification based on 10 policy dimensions. Preliminary results show three main policy profiles: Democratic Inclusive, Traditional, and Neoliberal Autonomous. The distribution of profiles varies globally, with Democratic Inclusive policies being more prevalent in Europe and South America, while Traditional and Neoliberal Autonomous profiles being more common in the Middle East, South-East Asia, and Anglo-Saxon countries. Preliminary results from bivariate multilevel models with country fixed-effects shows that schools in the Neoliberal Autonomous profile achieve higher average scores but exhibit higher inequality compared to the other profiles. However, when the socio-economic composition of schools is taken into account, Neoliberal Autonomous schools perform comparatively worse, with better results achieved in Democratic Inclusive schools.

This study examines a widely debated issue in education policy, namely the extent to which centralisation reduces achievement inequalities between students. We consider this issue from the perspective of the standardisation of national curricula. Previous research suggests that higher levels of standardisation are associated with lower levels of inequality. However, most studies are based on a cross-sectional approach and measure curriculum standardisation as a binary variable. We extend previous evidence by conducting a longitudinal study and by constructing an index of standardization based on several items. Moreover, standardisation has been more extensively studied at the secondary school level, while there is fewer evidence on the primary school level.
We use data from the Progress in International Reading Literacy Study (PIRLS) from 2006, 2011 and 2016 on about 50 countries, representing almost 700,000 students who were tested on their reading skills in grade 4. For each country in each year, we construct a novel index of curriculum standardisation based on whether there is a curriculum for reading instruction and how detailed it is. We then estimate inequalities in reading scores between individuals and between groups (gender, parental education). Using multilevel meta-analysis, we test the role of the index of curriculum standardisation in moderating score inequalities. Unexpectedly, we found no evidence that higher levels of standardisation are associated with lower achievement inequalities in primary school. If anything, higher standardisation is associated with a small increase in overall score dispersion, but this association is entirely due to unobserved differences between countries.

To date, gender inequalities persist and vary between countries, e.g., the gender gaps in mathematics and reading competencies, indicating that school and education system characteristics differ between countries and affect gender inequalities. Educational gender inequalities lead to inequalities over the life course, e.g., health and income. Therefore, my aim is to explain how education system and school characteristics affect the gender gaps in mathematics and reading competencies of 15-year-old students.
To explain the effects of school and education system characteristics on gender inequalities, I adapt the macro-meso-micro model. At the micro level, I use theories highlighting how gender-specific expectations and stereotypes cause gendered interest and skills and therefore gender differences in mathematics and reading. Deriving a macro-meso-micro link, I explain how education system and school characteristics may affect gender-specific socialization, leading to larger or smaller gender gaps in mathematics and reading competencies.
On this basis, I perform a cross-national comparison of 78 countries participating in the Programme for International Student Assessment (PISA) 2018, combined with further researched macro data with three-level mixed-effects models. The results show that the considered school and education system characteristics differently affect boys and girls and the gender gaps in mathematics and reading. The school characteristic of school time affects the gender gaps in favor of girls, decreasing the gap in mathematics and increasing it in reading, while the education system characteristics of competition and compulsory schooling affect the gender gaps in favor of boys, increasing the gender gap in mathematics and decreasing it in reading.

Inequality in educational careers by socioeconomic background emerges through two different processes: sorting (who goes to which school) and learning (who gains skills in the enrolled school). The two processes each need their own explanation, resorting to either rational-functional arguments about efficiency, or cultural models of boundary making (e.g. Domina et al. 2017, 2019). In this paper we develop a theoretical framework that provides arguments how rational-functional and/or cultural models relate to the processes of sorting and learning. This theoretical framework leads to testable hypotheses, that will more specifically address the value of rational-functional models and cultural models for understanding inequality of educational opportunity. We test hypotheses using Dutch register data, studying the sorting into classical gymnasiums and the causal effect of attending those schools. Preliminary results indicate that cultural mechanisms of categorization and valuation offer useful explanations for the sorting process into the classical gymnasium. The learning process conditional on school enrolment follows a more rational-functional process, in which cultural mismatch in the classical gymnasiums do not harm students of disadvantaged backgrounds. More generally, these results may lead to the conclusion that tracking systems are related to inequality of educational opportunity more through the sorting process than through the learning process.

The economic downturn due to the COVID-19 pandemic has prompted speculation about workers’ participation in job-related training, with both optimistic and pessimistic scenarios. The latest release of PIAAC data covering 29 countries for 2021/22 allows for an empirical examination of the pandemic’s impact on inequalities in job-related training. The study explores how inequalities in job-related non-formal training have changed during the pandemic and examines whether and how the determinants of training participation have shifted - in particular, whether labor market allocation remains the main factor explaining the relationship between education and training (as found for PIAAC 2012/14, see Hornberg 2024, Hornberg et al. 2024), or whether worker characteristics (such as learning disposition and socio-demographics) have gained in importance. Preliminary results show that training participation declined across all education groups in many countries during the pandemic. While differences in participation between education groups persist, the magnitude of these differences, particularly of the lower training gap, has decreased in several countries. However, these reductions should not be seen as a positive development, as they are mainly due to declines in participation among intermediate-educated workers. Despite shifts in participation rates of education groups, labor market allocation remains the key factor driving the education-training relationship in most countries.

Background:
The COVID-19 pandemic exacerbated social inequalities and contributed to a labor-market-related gradient in mental health in Germany. This study investigates the relationship between employment status and mental health before, during, and after the pandemic to identify the groups most affected and assess whether these effects persisted post-pandemic.
Methods:
Using the German Socioeconomic Panel (v39.0), we applied a quasi-experimental design employing a “fake-intervention” (placebo control) approach to obtain difference-in-difference (DiD) estimates. For this purpose, we created a treatment sample (N=8,340, 2018–2022) and a control sample (N=11,869, 2014–2018), designed to mimic the treatment sample. Sequence analysis identified six typical employment trajectories (S1–S6): high income, regular income, flexible-I, flexible-II, precariously employed, and unemployed. We used difference-in-difference estimation to assess the mental health impacts of these trajectories during and post-pandemic, stratified by gender.
Results:
A significant social gradient in mental health emerged. For males, regular, flexible, and precarious employment trajectories showed a mental health decline of ~¼ standard deviation during the pandemic, while high-income and unemployed groups were unaffected. Females experienced a more generalized decline across all trajectories. Mental health showed partial recovery post-pandemic but remained below pre-pandemic levels.
Conclusions:
Middle-income and flexible employment groups were particularly vulnerable to mental health impacts during crises. Females were disproportionately affected, emphasizing the need for targeted interventions, such as childcare support and measures to reduce gender disparities.

Post-COVID-19 symptoms present significant challenges in daily life for both patients and their families and entail substantial economic costs. Social stressors are particularly relevant in the context of infectious diseases like COVID-19, as they may influence health outcomes through hormonal processes and their impact on immune function (Botzet et al., 2024). It remains doubtful that physiological mechanisms alone explain gender and socio-economic inequalities in post-COVID symptoms observed in few studies. We therefore focus on gender differences and examine whether and how multiple stressors in both work and family life are associated with post-COVID symptoms. Combining psychological and sociological perspectives with current medical insights, we hypothesize that social stressors influence the severity and development of post-COVID symptoms and may, in part, explain gender and other socio-economic disparities.
To examine the role of multiple stressors in work and family life on post-COVID symptoms, we utilize novel patient data from the National Pandemic Cohort Network (NAPKON, 2020-23, N=3598). These data include a sample from the population-based platform (POP) which follows patients after recovery from acute SARS-CoV-2 infection at intervals of approximately 12 months. We examine the relationship between multiple stressors in family and work life and the recently developed Post-COVID Syndrome Score (Bahmer et al., 2022) using stepwise linear regression models. Preliminary results reveal significant gender differences that are partly related to stress in work and family life. Overall, these results highlight that stress contributes to more severe symptoms and to the gender gap in post-COVID-19 symptom burden.

The Covid-19 pandemic transformed daily routines worldwide, yet its impact on the daily time use of adolescents and young adults remains less understood. This study utilises variation in regional Covid-19 incidence rates and transit mobility to better ascertain how the pandemic affected the time use of adolescents and young adults (N: 559; ages 15-30; female: 59.8%), including screen time usage and social interactions with family and friends. We utilize granular time use data from the Harmonised European Time Use Survey (HETUS), collected from September 2020 to September 2021, and combine it with regional Covid-19 incidence rate data and Google Mobility Report data. Using regional fixed effects models, we test whether variation in pandemic severity, measured both by Covid-19 incidence rate and by mobility relative to the pre-pandemic baseline, were associated with changes in time use, and whether these changes differed by socio-economic position (SEP). Our findings indicate that while Covid-19 incidence rates and mobility are respectively positively and negatively associated with increased screen time usage, the effects are not statistically significant – these findings are mirrored in the social interaction models. Additionally, SEP does not moderate these associations. These results might reflect the minimal restrictions in place in Finland during the study period, thereby weakening the association between the pandemic and changing time use, as well as the small sample that reduces statistical power. However, the findings might similarly support emerging literature that indicates the pandemic’s effect on time use was less substantial than initially anticipated.

Different conceptions of equality of opportunity differ in whether they view the impact of ability on life chances as just or unjust. Radical equality of opportunity views the impact of ability on life chances as unjust, while liberal equality of opportunity views it as just. We use data from the Wisconsin Longitudinal Study (WLS) to address two shortcomings of previous research. First, we use polygenic indices (PGIs) to obtain better measures of innate ability and compare them to measures of skills observed in adolescence. Second, we use family fixed effects models to estimate the causal effects of PGIs, since the genes one receives from one's parents relative to one's siblings are a random lottery. Our results show that radical inequality of opportunity is greater than liberal inequality of opportunity, but the difference is less pronounced when PGIs are used to measure innate ability rather than skills measured in adolescence. Differences between radical and liberal inequality of opportunity are larger for education, occupation, and health than for income and wealth.

The persistence of socioeconomic inequalities in educational outcomes remains a critical concern in Britain. However, prior studies adopting rigid ‘nature vs nurture’ frameworks offered limited insight into this issue. Drawing on the Millennium Cohort Study and the National Pupil Database, I examine how genetic propensities (G) and socioeconomic background (SES) interact (GxSES) to shape educational inequalities across student’s academic trajectories, from ages 7 to 16. I make two key contributions: first, I use polygenic indices (PGIs) for more accurate measures of ‘innate abilities’; second, I explore the mechanisms driving the GxSES, focusing on how different forms of family resources (economic, educational, sociocultural) moderate the relationship between PGIs and educational outcomes across the life course. Using between-family and trio designs (accounting for parental PGIs), OLS and LPM regressions show that social factors remain crucial in explaining educational inequalities, even after controlling for PGIs. At ages 7 and 12, there is partial evidence of a positive interaction between PGI and SES, suggesting that high-SES environments might enhance performance for mid- and high-PGI students. Conversely, by age 16, a negative interaction emerges, suggesting that high-SES backgrounds compensate for lower PGIs. While at earlier ages educational resources play the largest role in driving the interaction, at later stages the interaction is mostly shaped by family economic and sociocultural resources. Methodologically, these findings highlight the need to account for complex interaction effects and non-linearities in GxSES research. Theoretically, they call for a comprehensive conceptualization of social origins to unpack the mechanisms underpinning the GxSES.

This study explores the influence of the "home obesogenic environment" on childhood obesity and, in particular, the interplay between the environmental and genetic factors. To do so, we generate a latent variable for “home environment” using domains that have been established in the literature as contributing to obesity, generating what we call the “Cultural Risk Score (CRS)”. Using two U.S. datasets, the National Longitudinal Study of Adolescent to Adult Health (Add Health) and the Future of Families and Child Wellbeing Study, we employ Confirmatory Factor Analysis (CFA) and Structural Equation Modeling (SEM) to assess how environmental and genetic factors affect children's BMI and how different levels of CRS can modulate the genetic penetrance of obesity. Preliminary findings indicate that dietary factors, physical activity, household stress and sleep quality are relevant domains for the home obesogenic environment. Additionally, current findings suggest the relevant role of household environments in shaping obesity risk, showing that cultural and environmental factors not only exert independent effects but also significantly amplify genetic predispositions. Future research will refine CRS conceptualization and categorization, and incorporate additional genetic indicators to improve understanding and prediction of obesity trends.

This study applies the framework of algorithmic fairness to evaluate potential biases in polygenic prediction across demographic groups. Using data from the Wisconsin Longitudinal Study (N = 6,393), I examine whether Polygenic Indices (PGIs) for educational attainment show differential predictive accuracy across socioeconomic status and biological sex groups. The analysis employs classification metrics commonly used in machine learning—True Positive Rate (TPR) and False Positive Rate (FPR)—to assess if genetic associations with education systematically differ between advantaged and disadvantaged groups. Results show that while TPR remains stable across groups (approximately 0.65), FPR varies significantly, particularly between socioeconomic groups (0.39 for high SES vs 0.45 for low SES). This indicates that genetic endowment translates less effectively into educational attainment for disadvantaged groups, particularly among those with below-average education. These findings contribute to our understanding of gene-environment interactions in educational achievement and highlight how genetic associations inherently depend on social context. The study demonstrates the utility of applying algorithmic fairness concepts to enhance our understanding of how genetic differences manifest differently across social environments.

This study investigates the labour market integration of second generations in Italy and Spain—two Southern European countries with relatively recent, large-scale immigration and still-emerging second generations. Drawing on the notion of the “Southern European model of integration” which imply that immigrants tend to have lower unemployment risks but reduced access to highly skilled employment compared to natives, this paper examines whether and how these characteristics extend to second-generation youth, who generally face fewer language- and education-related barriers than first-generation immigrants.
Using the EU Labour Force Survey (EU-LFS) data from 2021 to 2023—now including parents’ country of birth—this study disaggregates second-generation individuals by both place of birth (country- or foreign-born) and life stage at migration (early childhood, middle childhood, adolescence). The analysis focuses on youth aged 15-29 who are not in education and on three key labour market outcomes: activity, employment, and job quality, using logistic regression models to account for composition effects and ensuring proper control for demographic variables and contextual factors.
First-generation immigrants exhibit high inactivity, while second-generation youth show varied outcomes depending on birthplace and migration timing. Those born in the destination country and children of mixed couples display participation patterns similar to non-migrants. However, foreign-born individuals arriving later struggle with integration. While first-generation immigrants face lower employment penalties but poor job quality, second-generation youth diverge: native-born youth have more difficulty entering the labour market but experience fewer job-quality penalties, whereas foreign-born second-generation youth are closer to a "double penalty" scenario.

The devaluation of immigrants’ foreign education in domestic labour markets is well-established, yet the reasons for this remain insufficiently understood. We test and contrast two of the main theoretical explanations in the literature: transferability and quality. The former argues that knowledge and skills acquired through education do not transfer perfectly across borders due to institutional differences. The latter suggests that variation in educational quality leads to differences in actual skills and knowledge at the same formal level of education, which employers recognize and devalue accordingly. A key challenge is that these explanations often overlap in their predictions, as countries with low transferability also tend to have low educational quality. Using the EU Labour Force Survey (EU-LFS) ad hoc module on the Labour Market Situation of Migrants and Their Immediate Descendants, we analyse 17,000 immigrants across thirteen European destination countries, comparing the transferability and quality of education from 167 countries of origin. We assess labour market returns to education by estimating the effect of years of education on occupational status (ISEI), using a three-way interaction between education, transferability, and quality. Our findings show that educational quality increases labour market returns but is moderated by transferability. When educational quality is high, transferability does not significantly affect returns, but for immigrants from countries with low educational quality, institutional similarity substantially improves the value of education. These results indicate a substitutionary relationship between transferability and quality, suggesting that immigrants can mitigate lower educational quality by choosing destination countries with similar institutional contexts.

This paper examines the employment and job quality penalty of second-generation migrants in Europe, and whether they reproduce the models of occupational integration observed for first-generation migrants in nine European countries.
Despite extensive research on migrant integration, limited attention has been paid to inter-generational perspectives comparing first- and second-generation migrants and their labor outcomes.

Segmented assimilation theory suggests that second-generation migrants may face ethnic penalties rather than upward mobility, depending on the context of reception. Their labor penalization may or may not mirror first-generation migrants’ occupational patterns. In Western Europe, research on first-generation migrants’ occupational integration highlights contrasting patterns. In Mediterranean countries, migrants generally face minimal employment barriers but are concentrated in low-skilled jobs. Conversely, in Continental countries migrants face high penalties in both employment and job quality.

This study explores whether first-generation occupational integration models apply to second-generation migrants. The dataset analysed is the European Union Labor Force Survey (2014 and 2021), employing linear probability and regression models with robust standard errors, as well as Heckman models to account for possible sample selection biases.
Results indicate that the occupational outcomes of second-generation migrants exhibit greater cross-country homogeneity compared to first-generation migrants. Second-generation migrants face high employment penalization but experience relatively lower penalization in terms of job quality. This pattern is consistent across both Mediterranean and Continental countries and applies to both men and women. Second generations do not replicate the same labor patterns as first-generation migrants, but they shift towards a different occupational model.

Living in the right or wrong neighborhood predicts not only school dropout rates or wages but also mental health and even suicides. While the basic link between place and inequality is well investigated, causal studies on neighborhood effects remain scarce. Using nationwide administrative data from Germany and a quasi-experimental identification approach, we propose a new strategy to investigate how local social capital shapes individual-level employment probabilities. In exploiting variation over time, within cities and between 1×1 kilometer grid cells, we provide a causal effect of gendered neighborhood employment effects on refugee women’s employment probability. Results show a direct job referral effect of full-time employed female neighbors, which is strongest for other neighborhood women from refugee countries. Analyses on locally prevalent female work norms show a positive but one-time effect only for higher employment shares of native neighbors indicating that neighbors serve as role models only before other structures are settled. In analyzing neighborhood effects by gender and ethnicity, our study reveals that presumably weak types of neighborhood networks can provide useful resources for disadvantaged social groups in the labor market. Hence, the study stresses the necessity to break down dichotomies such as gender and ethnicity not only when explaining but also when finding alternative pathways for circumventing combined hurdles of intersectionality.

In addition to examining risk factors that contribute to the onset of diseases, it is essential to investigate the factors that enable individuals to maintain good health despite exposure to various stressors throughout their life course. This research project aims to analyze the relationship between employment status and subjective health, with a particular focus on how this relationship is moderated by internal and external health resources, such as personality traits and social capital. The empirical analysis is based on data from the German Socio-Economic Panel (SOEP) spanning the years 2010 to 2019, comprising a sample of 9,310 working-age individuals (18 to 65 years). A Random Effects (RE) regression model was employed, with self-rated health (SRH) as the dependent variable. Key explanatory variables include current employment status—distinguishing between short- and long-term unemployment—alongside the Big Five personality traits and a measure of social capital. The analysis further controls for relevant covariates, including sex, age, marital status, and educational attainment. The findings reveal that employment status, personality traits, and social capital are each independently and highly significantly associated with SRH, even after adjusting for all covariates. Moreover, significant moderation effects emerge for the interactions between employment status and social capital, as well as between employment status and the personality traits of neuroticism and conscientiousness. No significant gender differences were identified. Overall, the results suggest that particularly high levels of social capital function as a protective health resource, mitigating the adverse effects of unemployment on subjective health.

In response to increasing healthy life expectancy, governments are raising retirement age to prolong working life. However, many older adults work beyond retirement due to financial need, regardless of health status. This is pertinent for South Korea because of its increasing gap between retirement and pension age, weak social security system, and traditional reliance on family support. Using logistic regression models and eight waves of data from the Korean Longitudinal Study of Aging (2006-2020), we examined the association between multimorbidity and working status amongst older adults aged 55 years and older and potential moderation by wealth and family support. Preliminary analysis shows having multimorbidity and receiving financial transfers lowers the odds of working beyond retirement, whereas higher wealth increases the odds. Wealth acts as a moderator between multimorbidity and employment status, but only for males. These findings highlight the complex role of financial resources in later-life employment, underscoring the importance of wealth and intergenerational support in decisions to continue working. Future analyses will refine health and employment measures and examine additional measures of socioeconomic resources and family context to better understand the nuances of the health and work relationship and to clarify the roles of wealth and family support.

This study investigates global trends in mental wellbeing and related mental health outcomes across birth cohorts and historical periods. According to a common narrative, we are experiencing a worsening "mental health crisis" thought to stem from rapid social and economic changes. While such claims have sparked widespread debate, little is known about actual trends in mental health, especially in a global context.

Using 23 large-scale longitudinal (n = 447,496 individuals, N = 2,170,786 observations) and repeated cross-sectional datasets (N = 998,346 observations) spanning over 30 years of data from diverse national contexts all around the world, this study applies advanced age-period-cohort (APC) modeling techniques to decompose within- and between-cohort trends in mental wellbeing.

Results reveal strikingly different patterns: while mental wellbeing has declined across recent cohorts in many high-income countries, it has actually improved in several emerging economies. Furthermore, gender differences are observed, with women experiencing more pronounced shifts—both positive and negative—across cohorts. . These divergent patterns challenge the assumption of a universal decline in mental health. Rather, national social, economical, and political conditions play a crucial role in shaping cohort trajectories.

These findings indicate that broad claims of a global mental health crisis obscure the complex, context-dependent nature of these developments. This underscores the importance of considering local social, economic, and political factors when evaluating mental health trends, and calls for nuanced, cross-national research to inform policy responses.

Prior research in the area of social support suggests that it is an important influential factor of mental health. Yet, it often remains unclear how much overlap there is between provided support, the perceived availability of support resources, and the factors that account for discrepancies. Mental health disorders tend to produce and are maintained by distorted perceptions of the social environment. Depressive symptoms, in particular, are associated with increasingly negative thought processes and evaluations of social relationships and interactions. We aim to assess the extent to which the individual attributes of mental health, loneliness, and gender are related to discrepancies in support provision.
In order to examine the research question, two school surveys will be utilized. First, we investigate dual-perspective networks of 3000 adolescents at 37 schools in Germany. The analysis is based on sociocentric network data from adolescents in the eighth grade, distinguishing between the nominations of adolescents who provide support and those who receive emotional support at the grade level. The analyses will be replicated using two Scottish school cohorts.
The preliminary findings based on the SOCIALBOND study indicate that there is a substantial amount of mismatch in individuals' perception of support relationships. Nevertheless, mental health and loneliness do not emerge as significant influential factors of discrepancies. In contrast, there is evidence for gender differences in misperceptions. Support provision is more likely to stay unnoticed when the provider is a young man, and young men are more likely to perceive support that was not intended.

The majority of Germans express concern about climate change and support climate protection, yet specific policies remain contentious, and concerns about the socio-economic impacts of the climate transition are widespread. The far-right increasingly opposes climate policy, framing climate attitudes as part of a broader ideological divide over societal openness. This study investigates whether divisions in attitudes towards climate change and the climate transition in Germany foster political polarization along the new transnational cleavage. Drawing on survey data from the second wave (2022) of the German Social Cohesion Panel (SCP), we identify five climate attitude profiles through Latent Class Analysis. Approximately 8% are climate sceptics, concerned about the socio-economic costs of climate policies. Another 22% are indifferent or undecided. The largest segment (32%) is moderately supportive of climate policy without significant socio-economic concerns. A further 17% are highly climate-conscious but share sceptics’ concerns about the transition's impacts. Finally, 22% demonstrate the highest climate awareness and support for climate policy, with minimal concern about the transition. These profiles reveal socio-economic and ideological divisions, particularly between the highly climate-conscious and sceptics, mirroring the political divide between supporters of the Green Party and the far-right AfD. However, three middle segments, comprising 70% of the population, exhibit less pronounced divisions, suggesting limited overall polarisation on climate issues. While public awareness of climate change is high, addressing socio-economic concerns about the climate transition is critical to maintaining societal cohesion.

In the face of climate change, the principles of distributive social justice have become paramount in addressing the implications of resource allocation and the unequal impacts of environmental degradation. Our study explores the relationship between distributive social justice and environmental attitudes among young people in the context of climate change. Using a natural experimental design, we examine how the 2021 European Floods influenced social justice and environmental attitudes. Our results indicate significant shifts in attitudes, particularly regarding social justice, following the flooding. We found a strong and robust relationship between social justice and environmental attitudes. A causal mediation analysis revealed that floods affected environmental attitudes indirectly through social justice attitudes beyond direct effects. Our results emphasise the importance of integrating the principles of justice in addressing climate change and suggest that young people’s perspectives on social justice play a crucial role in shaping environmental policies and responses to climate crises.

Sustainability policies have become a crucial aspect of urban planning, as cities worldwide face challenges in mitigating climate change and promoting social equity. The transport sector, as a major contributor to global emissions, provides a critical lens for examining sustainability policies in urban environments, given its direct impact on daily life. While cutting transport emissions has clear benefits, their broader socio-economic implications remain unclear. One concern is the unequal spatial distribution of sustainability policies, as they improve the quality of life in affected neighborhoods but may decrease it elsewhere. Two key questions are explored in this study: (1) Are sustainable transport policies more commonly located in wealthier neighborhoods? (2) How are such policies perceived by affected minority groups? The study utilizes 1km-by-1km grid data from 2014 to 2023 on the socio-economic composition of residents as well as rental prices for the City of Berlin and merges them with data on sustainable transport policies. Additionally, qualitative semi-structured group discussions were conducted with minority residents of one of Berlin's neighborhood, which was undergoing substantial car reduction measures at the time. Applying spatial regression analysis, findings reveal a significant correlation between wealthier neighborhoods and a higher prevalence of sustainable transport policies. Qualitative insights highlight concerns among minority residents regarding rising rents and feelings of alienation, driven by perceptions of sustainability measures as misaligned with their needs and implemented without adequate inclusion in decision-making. Our results suggest that current policies may disproportionately benefit affluent areas, raising critical questions about equity in urban policy

In 2018, climate change discourse shifted dramatically, framing it as a crisis or emergency in political debates, protests, and news. While some observers feared such rhetoric might harm democratic politics, public perceptions of climate change during this shift remain underexplored. This study analyzes Eurobarometer survey data (2002–2023, 28 countries, N=1,229,479) to examine how risk perceptions and the salience of climate change evolved, focusing on changes around 2018–2019.

Multilevel regression analyses reveal that this discursive shift coincided with marked increases in issue salience and perceived risk, though the latter was less pronounced. These changes varied across countries and were influenced by individual factors such as education, gender, age, political ideology, and trust in media. Additionally, group differences in perceptions widened in some countries during the 2018–2019 peak.

Findings suggest that intensified public discourse around climate change may heighten awareness but also polarize perceptions within and across European Union countries, highlighting challenges for fostering collective action.

Psychological factors have increasingly become important for fertility, in particular among men. Only a few studies examine the association between psychological factors and birth parity transitions. Moreover, previous studies typically measure psychological factors at the end of fertility history, i.e. when most individuals have completed their fertility trajectories. We study the prospective association between several psychological factors (leadership skills, social maturity, emotional stability) and birth parity transitions among Swedish males. We further analyze a) single partner fertility (suggesting the relevance of multipartner fertility), and b) age at first childbirth as potential mechanisms. This study makes use of high-quality Swedish register data. Information on psychological factors of males come from military conscription data, and civil registers track the fertility histories of these men up to age 39 and above. We apply linear probability and linear regression models. Findings indicate positive associations between all psychological measures and first and second birth parities. However, lower scores on all psychological factors predict higher probabilities to transition to childbirths three and four, reversing patterns from lower birth parity transitions. Further analyses suggest that multipartner fertility and age at fatherhood entry may drive these patterns.

Women's reproductive rights, such as having the ability to access legal and safe abortion, are linked to improved women's reproductive health and socioeconomic well-being. Does different access to induced abortion have an impact on infant health? In this paper, we examine the impact of access to induced abortion (IA) on infant health. We leverage variation in IA across Spanish provinces, and we combine IA registers and birth certificates between 2011 and 2018. Preliminary results show that IA access – measured as IA rate – does not impact any birth outcome. Further analyses will extend the current study by exploring population heterogeneities and investigating abortion access as proximity to a public clinic, in addition to AI rate. 

This study examines the consequences of educational assortative mating on infant health. Although the positive association between maternal education and infant health is well-established, less is known about the impact of both parents’ absolute and relative education. Yet, the homogamy premium and heterogamy penalty hypotheses suggest that couples with dissimilar educational attainment face greater stressors than their homogamous counterparts, potentially resulting in unequal infant health outcomes. Using Austrian register data, (N≈500,000 singleton births; n≈380,000 different-sex couples), we apply Diagonal Reference Models (DRM) to disentangle the effect (and effect strength) of each parent’s education level, and the impact of educational dissimilarity on detailed infant health measures. Preliminary findings indicate a socioeconomic gradient at the couple level, with improved outcomes for better-educated parents. Results further suggest that paternal education may play a stronger role in predicting infant health than previously assumed, particularly for gestational weight outcomes. Additionally, we find evidence of distinct heterogamy effects on infant health. When both partners’ educational level is accounted for, hypogamous couples are less likely to have a low-weight baby but more likely to have an excessive birth weight and large-for-gestational-age infant compared to homogamous parents. The next step in this study is to examine whether results are due to endogenous selection effects using a causal design with an instrumental variable that accounts for non-random selection into hypogamous unions. Overall, this study provides new evidence on how couples’ educational pairing affects newborn health, highlighting a potential pathway for the intergenerational transmission of health disadvantages in heterogamous couples.

Wars are known to depress fertility primarily through disruptions in the provision of civil life. Yet, it is often believed that wars lead families to shift their preferences toward having more children due to insurance or replacement effects, though these effects generally manifest in postwar fertility rebounds. We argue that fertility declines may still occur as families shift their strategies to prioritize investments in child quality over quantity. Specifically, our research examines the impact of the Spanish Civil War (1936-1939) on family fertility (child quantity) and child survival past age three (child quality). Using data from the 1991 Spanish Sociodemographic-Survey, we employ a regression discontinuity design to estimate the effects of war exposure. The analysis focuses on individuals born between 1930 and 1933, comparing them to adjacent birth cohorts to assess changes in fertility and child survival rates. Our evidence shows that war-related fertility drops are not solely driven by the disruption of civil life and hardship. Families who could still adjust their final fertility at the onset of the war had fewer, but better supported children. Our results provide new insights into how social wars reshape family investment strategies. We discuss implications for contemporary and ongoing conflicts.

Typically, poverty risk is assessed at the household level, neglecting within-couple income inequality and the role of individual characteristics in vulnerability to poverty. Using Austrian register data, we track poverty dynamics within couples over an 8-year period around first birth. During this phase, when income disparities between partners are prominent, couples may pool resources, reducing the immediate poverty risk for the lower-earning partner. However, low personal income still leaves individuals vulnerable and exposed to social risks, particularly women. By simultaneously evaluating household and individual poverty risks (with/out access to partner income), the paper enhances our understanding of economic inequality within couples and the link between parenthood and poverty.
Women with different education levels diverge in their fertility behaviour and encounter distinct motherhood penalties on income, suggesting education-specific effects of first birth on poverty risk at household and individual levels. In the context of educational assortative mating, it is crucial to consider both individual education and the couple’s combined educational profile. Focusing on first births to couples from 2011 onward, the analysis uses a quasi-experimental event study design to assess child penalties on poverty risk by education at the individual and household level, from two years before to six years after birth, controlling for period and age trends. Preliminary results show a consistently low household poverty risk for couples and similar individual poverty trajectories for men. In contrast, women face a significant increase in individual poverty risk after first birth across education levels, driven by the high prevalence of part-time work.

Over the past decades, divorce has become an increasingly common occurrence in individual life courses in most OECD countries, significantly impacting the financial situations of both spouses. Previous research, however, shows that women, on average, experience greater losses in household income than men after divorce, although results vary across countries. The institutional context shapes the necessities and opportunities for labour market inclusion during marriage through social policies. Furthermore, individual-level characteristics such as the number and age of children as well as educational attainment play an important role in shaping post-divorce trajectories. In this paper, we investigate 1) whether and to what extent women in Austria experience greater economic losses and a higher poverty risk than men after divorce; and 2) the heterogeneities behind the risk of income loss and poverty. Austria represents an underexplored yet valuable case. The high share of part-time employment of women within the country leads us to the assumption of significant differences in the poverty risk after divorce between men and women. We utilize newly accessible register data, including detailed information on around 45,000 divorces between 2015 and 2017, alongside information on labour market biographies, education and family structures before and after divorce. Fixed-effects models enable us to identify within-person changes in household income four years before and four years after divorce, resulting in a time span from 2011 to 2021. The results are important not only for research, but also to inform policy, as they reveal a systematic vulnerability of women in Austria.

Parenthood contributes substantially to wage inequality between men and women in part because it induces gender differences in care work. The Nordic countries are often portrayed as forerunners with high levels of policy support for dual-earning/dual-caregiving and comparatively low levels of parental wage equality. Yet they are not homogeneous but differ on several relevant dimensions. So far, however, systematic comparisons of parental wage inequality the Nordic countries over time are rare.

In this study, we employ a couple-level framework to understand how parenthood shapes earnings inequalities in Denmark, Finland, and Norway. We explore whether mothers’ share of couple earnings has evolved over time similarly across the countries. Specifically, we ask whether differences by couples’ educational level have decreased or increased over time and whether such time trends are similar across the three countries.

Empirically, we use high-quality administrative data to examine how earnings dynamics change following the transition into parenthood in the time frame 1987-2022. Utilizing distributed fixed-effects models, we compare different cohorts of first-time parents in diverse educational constellations. First results for Finland reveal that earnings inequalities within couples have declined somewhat over time. Mothers' share of couple earnings has improved among couples, where his and her educational level is below tertiary. As a next step, we will estimate similar models for Denmark and Norway to compare both the magnitude and the temporal development of earnings inequalities. We will also analyse drivers.

Economic inequality within couples has gained increasing attention, reflecting growing similarities in partners’ socio-economic characteristics. This trend, driven by women’s rising educational attainment and workforce participation, has reshaped family dynamics and influenced gender equality, resource accumulation, and economic disparities later in life. While research has extensively examined gender inequalities in the labour market, fewer studies have explored how these disparities evolve within couples over time. Furthermore, couples’ income patterns significantly affect life satisfaction, with financial stability and perceived fairness in earnings contributing to psychological well-being, relationship stability, and overall household welfare.
This study examines the long-term development of economic inequality within couples and its relationship with life satisfaction. Using Group-Based Multi-Trajectory Modelling (GBMTM) on German Socio-Economic Panel (SOEP) data (1984–2021), it identifies distinct earnings trajectory patterns, considering both partners’ incomes and income differences. The analysis explores how socio-economic and demographic characteristics shape these patterns and how they vary across different cohorts. Additionally, the study assesses whether earnings trajectories are linked to the subjective well-being of couple members.
By incorporating a third outcome variable—the income gap between partners— and including multiple cohorts and cohabiting couples, this research offers a novel perspective on within-couple economic inequality. Findings will contribute to understanding how relational dynamics influence broader gender disparities and household well-being, shedding light on evolving economic roles within partnerships.

Educational stratification reinforces social inequalities by separating students based on socioeconomic status (SES), ethnicity, and academic achievement. Mixing students with diverse characteristics into the same classrooms, i.e., increasing classroom diversity has been proposed as a strategy to foster cross-characteristic interactions and reduce social stratification, as suggested by Blau’s concept “cross-cutting”. Yet, the principle of homophily suggests that diverse settings do not inherently promote bridging ties. This study investigates whether multidimensional diversity, characterized by weak faultlines—where attributes of SES, ethnicity, and achievement are less aligned—enhances bridging ties in student friendship networks. Using data from the IQB Trends in Student Achievement 2018, encompassing over 30,000 ninth-grade students in Germany, we measure faultlines with the average silhouette width (ASW) and bridging ties through the proportion of actual friendships ties with peers of different subgroups based on multiple characteristics out of all potential cross-subgroup ties. Multilevel regression analyses reveal that weaker faultlines are associated with a higher proportion of bridging ties, demonstrating the positive impact of multidimensional diversity on reducing social segregation. Importantly, these effects are consistent across SES and ethnic backgrounds, suggesting that diverse classrooms benefit students broadly rather than favoring specific groups. A further two-stage regression indicate that ethnic diversity significantly shapes the faultline-bridging tie relationship, while other diversity dimensions do not show any influence. These fundings contribute to understanding how classroom diversity of multiple dimensions mitigates social stratification in schools and offers evidence-based insights for policies aimed at reducing educational inequality.

Recent research failed to provide clearcut evidence for schooling to equalize social gaps in achievement. That stands at odds with the dominant view of compensatory input stating that even unequal schooling acts towards equalization of learning because it compensates even more unequal learning inputs in the absence of schooling. We contribute to the debate by explicitly testing the previously ignored mechanism of unequal gains from schooling which could counteract the compensatory input mechanism and, hence, may explain the absence of equalisation findings in studies. The hypothesis of unequal gains posits the operation of a Matthew-effect by which intellectual gains from schooling are a function of children’s prior skills. If true, it is students from higher (lower) socioeconomic backgrounds who benefit more (less) from schooling because of their, on average, higher (lower) skill level before schooling even starts. As both mechanisms of compensatory input and unequal gains can work simultaneously, a sharper examination of the equalizing power of schooling requires accounting for achievement distributions prior to schooling. Employing data from the new-born cohort of NEPS, we first replicate earlier findings from Germany on the beneficial but non-equalising effects of school exposure on students in Grade 1. Second, after accounting for children’s achievement before schooling, we indeed find evidence for unequal gains counteracting equalisation. Once adjusting for prior achievement, school exposure does level out learning gaps by social background. Findings, although being preliminary at this stage, point to the importance of incorporating prior achievement into an examination of school equalisation effects.

Intersectionality theory has long argued that different dimensions of inequality need to be understood jointly, but school segregation by race, class or gender continues to be studied in isolation. In this study, we introduce the concept of intersectional segregation, suggest a novel way of measuring it, and apply it to the case of school segregation by race and socioeconomic status in Brazil. Using unique student-level register data, we show that between-school racial segregation varies greatly across the distribution of socioeconomic status, and is much more pronounced among economically privileged groups. Private schools play an important role in this, providing predominantly White elites the option to exit from racially integrated public schools. In doing so, we demonstrate how an intersectional understanding of segregation can help to address longstanding sociological questions about the interaction between race and class in the production of unequal outcomes.

Today’s societies are becoming increasingly diverse, yet social relationships, including the friendship networks of adolescents in schools, are often strongly segregated along demographic lines. While previous research has examined how between-school sorting and within-classroom network dynamics contribute to friendship segregation, little is known about the role of within-school sorting, i.e. the assignment of students to different classrooms. This study systematically examines class placement practices in Germany using survey data from 101 randomly selected secondary schools, collected as part of the research project Sorting Decisions and Peer Processes in School (SPINS).

Initial findings reveal substantial variation in sorting practices. Gender balancing is nearly universal, while factors such as students’ religion, parental education, and country of origin play a minor role. Students’ place of residence and former elementary school play an important role in class placements, though schools differ in whether they cluster or separate students based on these attributes. Specialized profile classes (e.g., music, sports, STEM) serve as a primary sorting criterion in nearly 40% of schools. Schools also vary in the extent to which they consider student, teacher, and parental input in their placement decisions.

Building on these initial findings, future analyses of this study will examine the structural and motivational factors shaping different sorting practices and their impact on friendship segregation among adolescents. With this study I aim to enhance our knowledge of within-school sorting practices, thereby contributing to a better understanding of the mechanisms underlying friendship segregation among adolescents.

A growing literature of research has been devoted to exploring the concept of welfare chauvinism - the desire among majority populations to exclude immigrants from the welfare system – within contemporary European societies. Much of the cross-country analyses conducted on welfare chauvinism have applied a measure provided by the European Social Survey (ESS). We argue that most previous research have not taken into account the qualitative nuances associated with the ESS variable. We therefore introduce two distinct forms of welfare chauvinism, strict and extreme welfare chauvinism. Using the European Social Survey, we show that by distinguishing between these two distinct forms of welfare chauvinism clearly nuances and adds to discussions on the prevalence of these attitudes across Europe, as well as what previous research have found to be predictors of welfare chauvinism. We find that well known factors such as education and anti-immigrant attitudes remain strong predictors of welfare chauvinism, however their impact differ considerably between the strict and extreme forms of welfare chauvinism. By investigating this distinction, we also contribute to debates on what welfare chauvinism actually is, and which of its dimensions are qualitatively distinct from general prejudices towards immigrants.

An extensive body of literature on preferences for redistribution focuses on the self-interest approach, which suggests that preferences result from rational calculations of the potential gains individuals can obtain from redistributive policies, using explanatory factors like income, labour market status or social class. However, recent research shows that welfare states consider not only individual features but also family characteristics when redistributing income. Consequently, we argue that this dimension, which the literature refers to as family-related redistribution, is relevant to preference formation from a self-interest perspective. We use a combination of survey data from the European Social Survey 2016 and simulation data on redistributive outputs from EUROMOD’s Hypothetical Household Tool to study how family characteristics and the way redistributive policies address them affect preferences for redistribution in 16 European countries. Our results indicate that, following the classic self-interest assumption, individuals in family types which benefit more from family-oriented redistribution tend to be more supportive of redistributive policies. Significant cross-country differences are found regarding how family characteristics affect preferences, and these differences are driven by the current levels of family-related redistribution in each country. These conclusions highlight the relevance of considering the family context and how it is addressed by the welfare state as a driver of political preferences.

In recent years, rising economic inequality and migration flows have intensified social science interest in social cohesion, which encompasses trust, collaboration, and relationships between civil society and institutions. Studies in sociology and social psychology suggest that social interactions play a crucial role in shaping attitudes. Particularly, individuals use their social interactions as a primary source for social comparison, evaluating both their own and others’ attributes. Recent migratory flows and labor market shifts have reshaped cultural, economic, and value domains, adding new layers to existing socioeconomic inequalities and political conflicts. These transformations are expected to impact how people trust, collaborate, or engage in conflict. This study aims to understand to what extent social interactions with different social groups impact attitudes related to social solidarity. To explore this, a vignette-based survey experiment was conducted in Germany, assessing how the cultural, socioeconomic, and value characteristics of reference groups influence attitudes toward social cohesion. The results indicate that individuals with high socioeconomic status (SES), no migration background, Christian or non-religious affiliation, politically moderate views, and self-transcendent values are seen as more trustworthy, better collaboration partners, and less prone to conflict. In contrast, individuals with low SES, a migration background, Jewish or Muslim affiliation, right-wing political orientation, and self-enhancement values are viewed as less trustworthy, poorer collaboration partners, and more prone to conflict. The study discusses what are the more salient attributes that can trigger or inhibit social cohesion and its consequences for social solidarity.

The paper investigates public opinion about social Europe by analysing attitudes related to the basic principles of European social security measures and attitudes about institutionalising a uniform European social security system. Based on the TESS survey conducted in 13 European countries, the paper explores factors responsible for the strong support for the general principles of a European social security system but more precarious support for its institutionalisation. The main findings demonstrate that value-based mechanisms are primarily responsible for translating the positive attitudes towards the general principles into positive attitudes towards a uniform European social security system. Left-leaning individuals, who identify and trust the European Union, are the supporters of the general principles who are also in favour of the realisation of a uniform European social security system. In contrast, self-interest and personal socio-economic status does not play a prominent role.

Modernization theory argues that as societies become wealthier and more educated, liberal equality of opportunity increases. This is based on the premise that new institutional structures reduce the influence of ascribed traits, such as socioeconomic background, making individuals' life outcomes increasingly associated with their innate abilities. In this study, we examine the changes in equality of opportunity across different cohorts throughout the 20th century. We use data from ELSA and HRS and leverage polygenic indices (PGIs) as proxies for innate abilities. We provide two distinct frameworks for understanding equality of opportunity over time: the liberal and the radical perspectives. Both approaches consider unfair the influence of ascribed characteristics on life chances. However, whereas the liberal view sustains that the impact of individual abilities on life outcomes is fair, the radical view sees any influence of innate abilities as unjust, as individuals have no control over the innate abilities they are born with. Our findings reveal that only liberal inequality of opportunity has decreased across cohorts, whereas radical inequality is mostly constant over time.

Western democracies changed profoundly over the last century, shifting towards post-industrial knowledge economies with democratised educational systems. According to modernisation theories this transformation has induced a meritocratic shift from ascription to individual ability in explaining occupational attainment. Yet, most existing studies do not consider the potentially changing role of genetics, an important component of both family background and individual ability. This study therefore investigates whether the roles of parental status, education, and genetics in shaping occupational attainment shifted across Finnish birth cohorts from 1940 to 1980. Based on roughly 25.000 individuals with genetic information linked to Finnish administrative registers, findings reveal a decline in the influence of parental occupational status on occupational attainment, consistent with meritocratic theories. However, genetic influences remain stable, while the importance of education in occupational attainment increases significantly across cohorts. These results highlight the complex interplay of social, genetic, and educational factors in occupational success.

Educational attainment has long been key to reducing social inequality, yet historical barriers limited women’s access to education and careers. Recent genetically-sensitive research suggests that genetic influences on educational outcomes become more pronounced as institutional constraints diminish, serving as a proxy for equal opportunities. While studies on transforming post-socialist countries indicate a growing association between genetic predisposition and educational success, evidence on gender-specific effects remains scarce. Only one U.S.-study suggests that genetic influences on education were weaker for women in earlier cohorts due to gendered norms restricting career prospects but grew stronger over time as barriers diminished (Herd et al. 2019). Whether similar dynamics exist elsewhere remains an open question.

We therefore examine how gender differences in genetic influences on educational attainment evolved in Germany before and after reunification. Using Gene-SOEP data (N=1,573), we analyse polygenic risk scores for educational attainment across birth cohorts in East and West Germany. We employ stepwise OLS, including three-way interactions, to assess whether different institutional contexts led to a convergence or divergence in genetic influences by gender. Results indicate that genetic effects on educational attainment increased over time, with gender disparities gradually eroding in both regions. However, this equalisation trend differed, with more gradual declines in the FRG. While the GDR promoted gender equality politically and ideologically, our results suggest these measures alone were insufficient to eliminate gender disparities in genetic effects before reunification. Our findings contribute to the broader discourse on the interplay between social change, institutions, and genetic influences on educational inequality.

The association between cognitive skills and income is well-documented: cognitive skill differentials are an important reason for the income disparities that we see in contemporary societies. An assumption that is commonly made is that the strength of this association has been increasing over time, following the development towards increasingly meritocratic societies. However, longitudinal evidence to this effect is surprisingly limited. Using polygenic indices (PGI) for cognitive ability created for Swedish men and women born between 1930 and 1989 (N=24 102), we here document results indicating that the association between cognitive ability and income rank has not increased, but rather decreased over the course of the 20th century. This negative trend with regards to the economic returns to cognitive skills-related genetics was the most pronounced for men. For women, the association between the PGI and income increased initially, and converged with that for men, but then follows a similar negative trend. We also provide supplementary evidence from causally robust within-family PGI models that corroborate this negative trend with regards to the economic returns to cognitive skills. Our findings serve to question the common assumption that cognitive skills have become an increasingly important predictor of income. This abstract is currently based on Swedish data only, but we aim to include comparable data from Finland in the near future. While similar in many ways, differences in labour market structure between the countries may prove important for the results.

Between 2015 and 2020, nearly two million refugees applied for asylum in Germany, with a significant proportion being of training age. Occupational integration plays a crucial role in the socio-economic inclusion of young refugees, particularly through participation in the dual vocational education and training (VET) system. At the same time, the German training market faces considerable matching problems, with unfilled training positions in certain sectors. This study examines whether young refugees are more likely to enter training occupations with vacancy problems compared to non-migrant youth and immigrant youth without a refugee background.
Using representative survey data from the BA/BIBB Refugee Migration Sample 2018 and the BA/BIBB Applicant Survey 2020 (N=7167), combined with administrative data from the Federal Agency for Employment on unfilled training positions, we analyze differences in entries to training occupations with vacancy problems. The findings indicated that refugees are significantly more likely to enter occupations with staffing shortages than non-migrant youth. Empirically, this was mainly related to unequal lower qualification levels. Additionally, an insecure residence status of refugees increased their likelihood of accepting such training positions, as this can grant access to a “training toleration” permit.
The study suggests that training occupations with vacancy problems provide young refugees with an entry point into the labor market. Further research is needed to assess the long-term stability of these training pathways and potential dropout risks.

Since the Russian invasion of Ukraine, millions of Ukrainians have sought refuge in European countries, including the Netherlands, where approximately 120,000 Ukrainian refugees reside as of February 2025. Many of them participate in the labor market, yet a substantial portion remains unemployed or employed below their education level. This study examines how different arrival infrastructures impact employment outcomes among Ukrainian refugees in the Netherlands, with a particular focus on differences between those accommodated in municipal reception centers and those staying with host families. The relatively large number of Ukrainians staying at host families in the Netherlands provides for the first time adequate data to systematically compare outcomes between persons in host families and those in municipal reception centers. Drawing on social capital theory, we hypothesize that host family arrangements provide greater access to bridging social ties, potentially improving employment prospects. To test this, we conducted a large-scale representative survey of 3,460 Ukrainian refugees in the Netherlands. We compare employment outcomes—labor force participation, job-skill match, and working hours—across accommodation types using regression models while controlling for individual and contextual factors. Our analysis also considers variations within each housing type, including accommodation conditions and levels of support. The findings have direct policy implications for optimizing refugee housing strategies to enhance labor market outcomes and reduce inequalities among refugees in different living arrangements.

The Russian invasion of Ukraine has triggered Europe’s largest refugee movement since World War II, with over a million Ukrainians seeking refuge in Germany. This study examines the complex relationship between Ukrainian refugees’ uncertainty about their stay prospects and their investments in host-country-specific human capital amidst the war’s uncertain outcome and return possibilities. Guided by literature and theories on post-migration human capital investments, we explore how human capital transferability, opportunity costs, and settlement intentions as well as war-related uncertainties shape these investments. Using a vignette experiment among Ukrainian refugees within the online high frequency survey of the IAB-BAMF-SOEP Refugee Survey, we analyze refugees’ willingness to invest in vocational training. The experiment applies a rational choice model to evaluate educational decisions, factoring in training costs – like duration and lost earnings – and benefits such as improved earnings and lower unemployment risk after training. We also consider the transferability of skills to and from the Ukrainian labor market, settlement preferences and prospects, life course factors, and the influence of war-induced uncertainties. Our findings indicate that refugees make rational educational decisions, balancing costs and benefits while factoring in the usability of qualifications over time. Individual life situations, such as age, family and career stage, also play a crucial role. Moreover, uncertainty regarding legal status, the conflict’s duration, and future outcomes significantly shape refugees’ educational strategies. These results have important implications for policies aimed at facilitating refugees’ integration and human capital development in host countries.

The refugee crisis generated by the war in Ukraine put neighbouring countries, such as Romania or Poland, in a new situation: that of receiving societies, thus having to manage significant inflows of persons fleeing the conflict area. In each country, a (constantly changing) legal framework is put in place in order to manage the access to social benefits associated with having a refugee status or temporary protection.
In this paper, we explore a specific part in the activity of service providers working with refugees in the two countries, namely the definitions of special cases or exceptions: cases of individuals who would not benefit from certain rights, such as free accommodation or food, granted to refugees from Ukraine in Romania or Poland, were it not for the special interventions of case workers. In doing so, we analyze data obtained as part of a research project focused on refugees with disabilities coming from Ukraine to Warsaw and Bucharest, through interviews with stakeholders working with refugees and, respectively, with disabled people . We argue that, in order to make up for limitations and shortages in rules and regulations, service providers use empirically grounded definitions of social categories (disabled; unable to work) that complement the formally constituted ones. Moreover, this categorization process is shaped by two distinct factors: the emotional work on behalf of case workers and the institutional infrastructure within which they operate.

In 2015, Germany first implemented a national minimum wage of 8,50€. We contribute to the literature by assessing the consequences of this reform for women. Women are more likely to work in low-wage jobs and thus to be affected by the minimum wage. Previous research supports that the reform had very small negative effects on overall employment levels. The impacts on women’s and specifically mother’s employment remain underexplored even though mothers labor force participation in Germany is low and has the largest potential for increase. We address this gap with data from the German Microcensus Panel (2012-2015) on West German mothers and a difference-in-difference approach that relies on regional wage levels to define treatment and control groups. We consider the transition probabilities between various employment states (inactive, marginal, part-time, or full-time job) of mothers separately by the age of their youngest child, their partner’s income, and regional childcare availability. Mothers who were inactive before the reform are more likely to take up a job after the reform in treated regions. This effect is driven by mothers of pre-school children who have partners with a medium income. However, for mothers who were employed before the reform, we find a tendency to decrease employment, especially for mothers with young children who live in regions with low childcare availability, or those with high income partners. Overall, our results highlight that the impact of the minimum wage reform on female employment depends on mother’s household context and local availability of child care.

In Germany, lone mothers’ economic situation is more difficult than for other family types, reflected in high dependency rates on basic income support. While their employment rates are high, earnings are often low, resulting in benefit dependency even when employed. Against this background, this paper adopts a longer-term perspective on lone mothers’ benefit and employment trajectories by studying the impact of stepping-stone jobs. We ask whether taking up a first job while remaining on benefits facilitates benefit exit in the long-term. In this context, we also look into the role of job-center counseling for lone mothers’ employment entry and benefit exit rates. From a policy perspective, results can indicate the relevance of a longer-term approach for policies for lone parents. Our analyses are based on large-scale administrative data, to which we employ a timing-of-events approach. First findings indicate strong immediate positive effects and also substantial long-term effects of taking up a first job on exiting benefit receipt even after more than two years. Job-center counseling has a substantial immediate positive effect on employment uptake, and small positive effects after three months on exiting benefit receipt. While finding a job with sufficient earnings right away is ideal, this may not always be possible, for instance when working part-time. Our findings show that in such cases, a first job nonetheless enhances opportunities for benefit exit in the long-term. Thus, while lone parents may require longer-term economic support, their employment situation is dynamic and benefit dependency is not indefinite.

Parental employment, particularly maternal employment, is strongly associated with childcare arrangements. The utilization of formal childcare is crucial for parents to combine the responsibilities of parenthood and work, potentially enabling increased maternal employment. Childcare and maternal working arrangements vary significantly across households, with disadvantaged households more likely to experience limited use of formal childcare and reduced working hours. Drawing on theories of social networks and the mobilization of social capital, this study highlights the importance of peer ties between parents of children from the same childcare cohort in explaining the socio-economic gradients in childcare arrangements. Parents who have children in the same cohort  exchange information and experience on childcare in daily life, potentially exerting peer effects within childcare centers. This study aims to investigate how peer effects within childcare centers influence childcare arrangements, focusing on changes in maternal working hours after joining the childcare centers for mothers of children aged 0–4 years.
Using longitudinal administrative data from Statistics Netherlands, we construct parental networks based on childcare cohorts in childcare centers and apply propensity score matching (PSM) analysis. Treatment and control groups are defined based on ego’s working hours and composition of childcare cohorts in terms of maternal working hours, distinguishing between cohorts with no peer effect, positive peer effects, and negative peer effects. After matching the treatment units and control units, this study estimates the average treatment effect (ATE) on the ratio of changes in weekly maternal working hours using an ordinary least squares (OLS) model.

Motherhood significantly disrupts women's careers in Germany, often leading to prolonged employment interruptions followed by part-time re-entry or permanent withdrawal from the workforce. However, the impact of motherhood differs between immigrant and native women, with immigrant women - particularly those of Turkish origin - being less likely to re-enter the labor market after childbirth. While the motherhood employment penalty affects all women, pre-existing labor market disparities exacerbate the gap between immigrant and native women.
Although research has explored factors like education and maternal role models to explain these migrant-native differentials, the role of pre-birth job characteristics - such as tenure, occupational status, and employment experience - remains underexplored. This analysis aims to fill this gap by examining how these pre-birth employment characteristics shape post-birth employment trajectories among immigrant women in Germany. Using a large-scale administrative register dataset, the analysis focuses on a subset of approximately 87,800 women, providing an in-depth examination of how prior employment experiences impact labor market outcomes after childbirth.
By identifying typical employment clusters through sequence analysis and further examining these clusters via multinomial regression, the analysis investigates the influence of pre-birth job characteristics on post-birth employment trajectories. This work addresses a significant gap in the literature, offering new insights based on robust, high-quality data and contributing to the understanding of labor market inequalities in the context of migration and gender in Germany.

Demographic shifts are reshaping the workplace, fueling intergenerational tensions: older employees are seen as more loyal to their companies, younger as less committed and more sensitive to the Impostor Phenomenon (IP). Against this background, this study examines the relationship between the IP, Perceived Discrimination (PD), and Workplace Belonging (WB). While previous research predominantly considered the IP as an individual pathology, it is understood as related to experiences of social exclusion in order to examine its economic consequences.

The analysis is motivated by bridging micro-level sentiments and macro-level burdens and links Cultural Reproduction Theory (Bourdieu & Passeron, 1990) and the concept of Social Stigma (Goffman, 1963) to explain the social stratification of IP emotions. Descriptive statistics and linear regressions including interaction terms utilizing data from a large-scale register-based survey among German adults and adolescents (n = 14,300) are used to access the link between the IP, PD, and WB.

The results show that IP intensity is lower in older age groups, while women, certain immigrant groups, and individuals from low socio-economic backgrounds are disproportionately affected, with particularly high intensity for overlapping characteristics. Moreover, the IP is positively associated with PD (based on gender, social, and ethnic origin), while negatively associated with WB (to varying degrees across occupational groups and positions). These findings demonstrate that the IP is (a) a collectively shared social phenomenon, (b) associated with exclusionary experiences, and (c) a relevant factor of employee loyalty to the workplace. Ignoring its relevance can have high social (individual) and economic (collective) costs.

Social stratification research often focuses on (binary) gender, class and ethnicity as characteristics that stratify people’s lives. Recently, researchers have started drawing attention to LGBTQ* identity as a characteristic that shapes people’s opportunities in life. LGBTQ* persons face stigma and discrimination in the labour market and Lesbian, Gay and Bisexual people earn less than heterosexual men, but often more than heterosexual women. However, previous research has likely underestimated the role of sexual and gender identity in earnings by studying a limited set of identities. We address this limitation by including extensive sexual identities that go beyond lesbian, gay and bisexual identities, as well as trans* and non-binary people.

We use data from a representative survey in Norway that is complemented with earnings data from administrative records on 19,774 individuals of which 993 have a non-heterosexual identity and 188 persons are trans* or non-binary. Sexual and gender identities that further challenge norms regarding sexuality and gender, such as, pansexual, asexual, trans* and non-binary are all related to strong earnings penalties, also once compared to gay and lesbian individuals. Our findings highlight how the privilege of cisgender heterosexuals in the labour market has been previously underestimated. Because identities available to people change over time and place, using measures that only capture a restricted number of identities complicates comparisons across contexts, especially when people with emerging identities are particularly disadvantaged in the labour market.

Many field experiments documented the prevalence of ethnic discrimination in employers’ hiring decisions across many European labour markets. Much less attention has been given to factors that could potentially mitigate such discriminatory behaviour by countering stereotypes about employability or trainability and social fit. In this study, we focus on employer selection within the Norwegian apprenticeship market. We invite all firms in the broader Oslo region to participate in a factorial survey experiment, where they will evaluate a randomly assigned set of vignettes describing fictitious applicants. We anticipate a sample size of approximately 500 employers and 3,000 applicants, with data collection scheduled for completion by February 2025. The survey experiment will be supplemented with firm- and recruiter-level characteristics obtained from register data. We investigate whether and how references from internships and vocational teachers matter in the hiring process of apprentices with and without immigrant backgrounds. We study the impact of such references on the likelihood that recruiters will proceed with applicants, how they perceive their trainability and social fit and whether this differs for applicants of Norwegian, Pakistani or Polish origin. In the context of apprentice selection, employers often operate with limited information, which can lead to a greater reliance on stereotypes. References from vocational teachers may provide insights into applicants’ learning abilities, while references from previous employers can shed light on how applicants perform in comparable workplace settings. We aim to test whether these two types of references can mitigate otherwise negative evaluations of minority applicants.

This study examines the responses of Spanish childcare centres to applications from homoparental (same-sex) and heterosexual families. Given Spain’s progressive stance on LGBTI rights, the research seeks to uncover subtle forms of discrimination within the educational sector. We conducted a field experiment involving 3,000 real applications to analyse response rates and response types, including acceptance, rejection, and neutral messages—such as requests for additional information or a call-back from the centre. Initial findings suggest minimal overt discrimination in direct acceptances or rejections. However, homoparental families are 10% less likely to receive any response and are less likely than heterosexual families to receive a neutral reply. At this stage of the research, we are using artificial intelligence to assess the warmth versus distance in the responses received by families across sexual orientations. This innovative approach aims to reveal nuanced dimensions of discrimination, offering insights into how LGBTI and migrant-origin families are treated in Spain’s educational institutions.

Discrimination remains a pressing issue with profound implications for social cohesion and equity. While extensive research has examined cross-national differences in attitudes toward marginalized groups, little attention has been paid to how regional contexts within and across countries shape these attitudes. This gap is critical because regional histories, economic conditions, and cultural dynamics often diverge significantly from national averages, challenging the notion that national borders define uniform social attitudes (Hiers et al., 2017; Dimitrova-Grajzl et al., 2016). Understanding these regional variations matters for designing more effective and context-sensitive anti-discrimination policies.
This study addresses these challenges by examining regional differences in Europe in the acceptance of social profiles characterized by ethnicity, religion, employment status, and other attributes. Using a factorial survey experiment conducted across nine European countries divided into 23 culturally and historically distinct regions, this research analyzes 16,200 respondents’ evaluations of 97,200 vignettes. Profiles are assessed on acceptance (0–10 scale) and associated stereotypes (warmth, competence, trustworthiness; Fiske et al., 2002). The use of multilevel modeling enables disentangling individual- and regional-level influences, providing a nuanced understanding of how contextual factors shape preferences.
By addressing several key limitations in prior research—including a lack of theoretical integration, limited use of multilevel analysis, and insufficient statistical power for testing cross-level interactions (Hox et al., 2012)—this study bridges social psychological theories with structural explanations of regional disparities. It challenges the focus on national-level analyses, demonstrating the importance of regions as critical units of analysis in understanding discrimination and fostering social inclusion.

Conjoint analysis is a widely applied survey experimental technique to estimate the causal effect of multiple treatment variables simultaneously on an outcome variable. However, while conjoint experiments are ideal for studying respondents’ preferences, they do not give insights into the respondents’ information-seeking behaviour leading to their final choices. In addition, although the conjoint design is believed to elicit lower social desirability (SD) bias compared to traditional survey experimental techniques, little research addresses SD bias in conjoint designs. Building on work in marketing studies, we introduce a discrete choice experiment to the field of social sciences: the click-to-reveal (CTR) conjoint. The CTR conjoint allows respondents to reveal a limited number of covered conjoint attributes, but not all. By tracking which attributes respondents reveal, the CTR conjoint assesses the importance of information in making a choice. In addition, we expect SD bias to be lower in the CTR conjoint because it captures stated preferences more indirectly, and because the choice is based on the subset of attributes that is selected by the respondent. We collected data from an online sample in the United Kingdom (Prolific, N=1,500) and present respondents with a scenario that is prone to SD bias: a simulated hiring decision testing for ethnic discrimination. We benchmark the CTR conjoint against a traditional conjoint and a traditional conjoint with information randomly covered to compare the importance weights and the parameter estimates of the attributes.

Keywords: conjoint experiment, information-search, decision-making, social desirability bias

Statistical theory makes ambiguous predictions about covariate adjustment in randomized trials. While proponents highlight (asymptotic) efficiency gains, opponents point to possible bias and a loss of precision in finite samples. The performance of adjustment is particularly controversial in case of small samples, heterogeneous treatment effects, and a large number of weakly predictive covariates. As these characteristics occur quite often in randomized trial in economics and social sciences, it is crucial for applied research to assess the finite sample performance of covariate adjustment in this data structure. In this paper, I compare the performance of machine learning covariate adjustment, namely post-double-LASSO as well as a combination of LASSO and augmented inverse probability weighting (AIPW) with unadjusted estimators. I rely on real-world data from an educational panel survey (NEPS) and conduct Monte-Carlo simulations on a wide range of different data structures, including different outcomes, sample sizes as well as homogeneous and heterogeneous treatment effects. The results confirm that machine learning covariate adjustment improves precision and power to a non-negligible extent. Over all simulation scenarios, machine learning adjustment leads to a reduction in root mean squared error of about 15-20% and is almost never worse than unadjusted analyses (and if so, negligibly). However, in sample sizes with less than 100 treated units, efficiency gains are considerably smaller, LASSO adjustment is weakly biased, and LASSO-AIPW (though unbiased) shows under-coverage of confidence intervals. In sum, machine learning covariate adjustment is a promising approach, but some open questions remain.

The transition into parenthood implies making choices about childcare that have critical
implications for parental employment trajectories, child development, and gender equality. Previous work has studied how the uptake of formal childcare is influenced by the availability of both formal and informal family care. The latter is typically measured with a few variables capturing the geographical proximity and employment status of the paternal or maternal grandparents. However, this simple approach disregards how family obligations and resources potentially depend on the characteristics and interdependencies of a larger set of kin, including grandparents, aunts, uncles, and cousins. We adopt a more multidimensional perspective that measures family care availability as a function of over thirty characteristics of parents' family networks, disaggregating their kin by age, gender, and lineage. To model the potentially complex relationships between these characteristics and formal childcare uptake, we adopt a machine learning approach in which we predict whether first-time heterosexual parents will use formal childcare services in the Netherlands. We use register data that allows us to measure the characteristics of parents, grandparents, uncles, aunts, and cousins, as well as formal care availability in terms of the location and capacity of nearby formal childcare providers. We present preliminary results comparing the predictive performance of logistic regression and XGBoost models with an increasingly complex measurement of family care availability. We discuss the promises and pitfalls of this combination of machine learning and register data for research in family sociology and demography.

The conventional Two-Way Fixed-Effects (TWFE) estimator has come under scrutiny lately. Recent literature has revealed potential shortcomings of TWFE when the treatment effects are heterogeneous. Scholars have developed new advanced dynamic Difference-in-Differences (DiD) estimators to tackle these potential shortcomings. However, confusion remains in applied research as to when the conventional TWFE is biased and what issues the novel estimators can and cannot address. In this study, we first provide an intuitive explanation of the problems of TWFE and elucidate the key features of the novel alternative DiD estimators. We then systematically demonstrate the conditions under which the conventional TWFE is inconsistent. We employ Monte Carlo simulations to assess the performance of dynamic DiD estimators under violations of key assumptions, which likely happens in applied cases. While the new dynamic DiD estimators offer notable advantages in capturing heterogeneous treatment effects, we show that the conventional TWFE performs generally well if the model specifies an event-time function. All estimators are equally sensitive to violations of the parallel trends assumption, anticipation effects or violations of time-varying exogeneity. Despite their advantages, the new dynamic DiD estimators tackle a very specific problem and they do not serve as a universal remedy for violations of the most critical assumptions. We finally derive, based on our simulations, recommendations for how and when to use TWFE and the new DiD estimators in applied research.

The present study investigates health and health inequality in East and West Germany following the reunification in 1990, utilizing this historical event as a natural experiment to explore the impact of differing health systems and cultural factors.
The analytical approach involves a comparative longitudinal analysis of GSOEP data from both German subpopulations, focusing on changes over time since 1990. Period effects on both means and variances of health satisfaction were estimated using mixed-effects location-scale models while controlling for age and cohort effects.
The findings indicate that, around 1990, West Germans enjoyed a higher average subjective health level, likely due to superior medical technology and treatment efficacy compared to the former GDR. However, health inequality was lower in East Germany due to more equal access to healthcare and socialization practices that mitigated socio-economic disparities. In the 1990s, the introduction of a two-class insurance system in East Germany led to an increase in health inequality, exacerbated by economic instability and unemployment. Since 2000, the study suggests a narrowing of the health gap between East and West Germany, driven by improvements in socio-economic conditions in East Germany. Overall, the research highlights a complex dynamic of convergence in average health levels, while health inequality trends reveal a more nuanced picture, with initial divergence in the 1990s followed by signs of convergence in the subsequent decades. This study represents a significant contribution to understanding the long-term health implications of the reunification in Germany.

Creating an inclusive school environment, in which students with disabilities are integrated as much as possible with students without disabilities, has become a commitment on the policy agenda of many countries around the world. Despite this trend toward inclusive education, however, research on the effects of being exposed to peers with disabilities in school remains limited. In this paper, we provide one of the first evidences of the noncognitive effects of exposure to peers with disabilities. Using data from the National Longitudinal Study of Adolescent Health (Add Health), we examine child disability spillovers on a range of internalizing behavioral outcomes (depression symptoms, positive outlook, self-esteem) and social inclusion outcomes (peer conflict, social acceptance) among students enrolled in middle and high school (grades 8-12) in the United States. To infer causality, we employ a quasi-experimental research design and exploit within-school and between-grade variation in school and grade composition. Our results show that exposure to a higher share of students with physical or learning disabilities is reflected in students without disabilities' increased internalizing behavioral problems and lower social inclusion. From a policy perspective, our findings suggest the need for policy makers to improve inclusion policies so that all children can benefit from a healthier and more inclusive school environment.

Given the centrality of field of study choices for later positioning in the labour market, it is essential to identify factors shaping the inequality generating process in such decisions. Yet, despite extensive research on how social inequality due to social origin and gender in the field of study choice emerge, we argue that a crucial aspect in research on horizontal differences in educational outcomes has been overlooked: individuals’ health. The increasing incidence of health problems nowadays and research findings on how (poor) health induces inequality in (vertical) educational decision-making highlights the importance of individuals’ health. We develop a theoretical model based on Morgan (2002), emphasizing the role of information and beliefs for the formation of aspirations and the level of effort students allocate towards schoolwork. Based on our model, we explicate three potential pathways on how health influences major choices: (1) beliefs about future working conditions and educational program, (2) preferences, (3) and school achievement. Leveraging data from the Nine Graders Cohort of the National Educational Panel Study (NEPS) of Germany combined with aggregated data of German Microcensus, BIBB-BAuA Job Exposure Matrix, and Student Cohort of NEPS, we employ conditional logit models to test our theoretical model. Our preliminary findings reveal that individuals with lower levels of health tend to opt for similar majors. Additionally, beliefs about job requirements and preferences such as income and flexibility preferences and career choice interests partly explain the influence of individuals’ health on the field of study choice.

Teachers are academic merit gatekeepers. Yet their potential role in reproducing inequality via assessments was overlooked or not correctly identified, being ‘an elephant in the classroom’. This article teases to what extent teacher grades and track recommendations are biased by student SES after accounting for unobserved ability, leading to overestimation in prior research. Using the German NEPS panel across elementary education, we aim to identify student ability with multiple cognitive and noncognitive composite measures and an instrumental variable design. We further assess heterogeneity along the ability distribution to test whether, according to the compensatory hypothesis, teacher bias is largest among low-performers. First, after approximating latent student ability, the residual effect of student SES on teacher assessments remains but declines by 40%, indicating substantial teacher bias overestimation in previous studies. Second, it concentrates on underperformers, suggesting high-SES parental compensatory strategies to boost teacher assessments. Thus, families and teachers might influence each other in the evaluation process. We discuss the findings’ theoretical and methodological implications for teacher bias as an educational reproduction mechanism.

This study investigates the social stratification of teachers' grading in Italy, focusing on how teachers attribute marks to students with different socioeconomic status (SES) but similar subject-specific competencies. Drawing on a dataset of 26 million student evaluations spanning ten years, the research explores systematic variations across school levels, subjects, and student ability distributions. Drawing on sociological and socio-psychological theories, a series of hypotheses are elaborated on the main and heterogeneous patterns across educational settings. Utilizing data from the National Institute for the Evaluation of the Italian School System, the study employs a novel model featuring causal mediation analysis techniques, integrated with a classroom fixed-effect models. Preliminary findings reveal a consistent association between SES and marks, with subject-specific competencies accounting for a limited portion of this relationship. SES-based grading mismatches (the direct SES effect net of competencies) is consistent across subjects (reading and mathematics) and years (2011-2021), peaking in primary and lower secondary education but diminishing in upper secondary education. The study also identifies widened gaps among low-achieving students, aligning with the compensatory advantage model, particularly in elementary school.

The impact of teachers on student outcomes has long been a central concern in research on educational inequality, yet most studies rely primarily on test scores to measure teacher effectiveness. However, cognitive skills are only one dimension of student development. This paper examines the multidimensional effects of teachers on student outcomes across three domains: (1) cognitive skills, measured through standardized test scores; (2) socioemotional skills, assessed via survey data; and (3) behavioral outcomes, including school attendance and grade progression. Using Danish administrative data covering the full population of students and teachers in Danish public schools from 2013 to 2020, we estimate teacher value-added (VA) in each of these domains.

Our findings reveal substantial variation in teacher VA across these three skill sets, indicating that teachers who excel in improving test scores do not necessarily contribute equally to socioemotional and behavioral development. Moreover, we explore teacher characteristics associated with strengths in different domains and assess the long-term consequences of these differentiated impacts. Finally, we examine heterogeneity in teacher effects by student background, addressing whether students from lower socioeconomic backgrounds benefit differently from exposure to teachers who are effective in non-cognitive domains. Our results suggest that broadening teacher evaluations beyond test scores is critical for understanding their full contribution to student development and addressing educational inequalities. These insights have implications for teacher hiring, training, and policy interventions aimed at fostering holistic student success.

In recent years, several major terror attacks linked to political Islam have shaken Europe. Leveraging social media data, this study examines how public reactions to Jihadist terrorism are expressed and shaped to explore the formation of public views of immigration and intergroup relations. The theoretical predictions derived from combining terror management theory, group threat theory, and the concept of social resilience, are empirically tested using validated Keyword-Assisted Topic Modeling (keyATM). Unlike traditional topic modeling approaches, keyATM incorporates pre-specified keywords, enhancing both the precision and interpretability of results while mitigating the impact of researcher subjectivity. I analyze over 100,000 time-stamped and geo-coded Tweets on immigration and related issues across four languages in the week following eleven major Islamist terrorist attacks in nine European cities posted by users located in the city of the attack. Consistent with the theoretical predictions, the findings reveal that both threat-related and tolerance-related topics emerge prominently across all four languages. However, deeper analyses uncover temporal shifts in the discourse: While initial reactions often reflect nationalist and exclusionary views, later stages show a rise in inclusionary and tolerance-oriented topics. These results highlight the dynamic nature of social media debates on immigration issues after dramatic events, where initial threat-driven responses often give way to more inclusive and tolerance-oriented discussions as time progresses, underscoring its role in shaping and reflecting the evolving public response to terrorism.

Awareness of past atrocities is widely seen as critical for restoring justice and building resilient democracies. Yet, confronting people with past injustice committed by their group can also lead to defensiveness. Through a survey experiment (n=2,198), we measured the effect of three different prototypical forms of information about the Holocaust on intentions to commemorate the Holocaust, intentions to counter antisemitism, and attitudes towards minoritized groups. We find that all three forms demonstrate overall effectiveness in mobilizing individuals for commemoration and against antisemitism and improving their attitudes towards minoritized groups. We find heterogeneous treatment effects by party affiliation. For far-right supporters, reading about the story of an individual victim is particularly effective. For others, formats that center the sheer extent of atrocities or focus on symbolic justice efforts have greater mobilizing potential. We repeated the survey with the same respondents three months later (Nov 2023), following the start of the Israel-Gaza war, and found that means and treatment effects are overall stable despite the change in context. The results demonstrate the power of providing people with information about past atrocities and injustice for political mobilization and prejudice reduction.

This study examines the effects of school belonging and well-being on ethnic minority adolescents’ attitudes toward the majority population in the Czech Republic, as well as the majority’s attitudes toward minority groups. Research suggests a link between well-being and attitudes toward outgroups, and a strong sense of belonging being tied to psychological well-being and social adaptation. At the same time, some studies indicate that the direction of these effects may be reversed. Using cross-lagged models, this study examines whether school belonging and well-being are positively associated with attitudes toward outgroups, or whether attitudes toward outgroups influence wellbeing. It also tests whether well-being mediates the relationship between school belonging and attitudes toward outgroups. Using data from the Czech Education Panel Survey (2023, 2024), this study analyses two subsamples: ethnic minority students (N = 1,284) and ethnic majority students (N = 13,846). The results indicate that attitudes toward outgroups are more strongly related to subsequent well-being rather than the reverse. Moreover, the findings reveal important group-specific differences. Among minority students, sense of school belonging predicted more positive attitudes toward the majority population, and this association was mediated by well-being. In contrast, while majority students’ well-being was also enhanced by a sense of school belonging, their attitudes towards minority groups were not affected by well-being. Instead, sense of school belonging exerted a strong direct effect on these attitudes. This suggests that for ethnic majority youth, feeling part of the school community may promote openness toward outgroups independently of emotional well-being.

The distinction between antisemitism and legitimate criticism of Israel has become increasingly contested, particularly following the Hamas attack on Israel on October 7, 2023. This study examines how different dimensions of antisemitism—classical, secondary, and Israel-related—manifest in contemporary Germany and whether these categories remain conceptually distinct.

Using survey data from a quota sample of 2,082 respondents in 2024, we investigate key predictors of antisemitism, including religious fundamentalism, authoritarianism, and conspiracy beliefs. Our findings indicate that the boundaries between subdimensions like classical and Israel-related antisemitism have become increasingly blurred, with Israel-related narratives often serving as a vehicle for traditional antisemitic tropes.

Political ideology also plays a crucial role: left-wing respondents tend to reject classical antisemitism but may endorse Israel-related antisemitism when framed within anti-imperialist discourse. Conversely, right-wing individuals more frequently express classical antisemitic stereotypes while simultaneously supporting Israel as part of a nationalist or anti-Muslim agenda.

The study highlights strong correlations between antisemitism and authoritarianism, as well as reinforcing previous findings on the role of conspiracy ideologies. Our results emphasize the need for a nuanced approach to differentiate between a) legitimate criticism of Israeli policies and antisemitic rhetoric and b) the different dimensions of antisemitism. Policymakers and researchers must refine existing measures to capture the evolving nature of antisemitism and ensure that efforts to combat it remain effective in the current socio-political landscape.

Using data from the German Time Use Survey 2022, this study examines how social class influences parental time investments in childcare. Parental time spent with children has increased in recent decades, with a particular increase in activities that support children's education and development. Such trends, often referred to as the "intensification of parenthood," are shaped by growing societal pressures and unequal capacities across social classes to meet these demands. We categorize parents into four social classes based on four-digit occupational codes. Based on Oesch's framework we distinguish unskilled workers, skilled workers, lower service class, and higher service class. A separate category includes persons who did not work in the past week. We employ class as an individual characteristic, but also account for the partner’s social class for individuals in couple households. Time spent with children is analyzed in two dimensions: routine care (e.g., physical care, supervision) and interactional activities (e.g., reading, playing). Our descriptive analysis shows that parents from higher social classes, especially those in the higher service class, spend more time with their children on interactional activities than parents from lower social classes. These patterns are observed for weekdays and weekends, although the differences are smaller on weekends. The multiple regression analyses further examine how social class and gender interact while controlling for other standard confounders (e.g., age of children, age, citizenship, region).

Social class is a central but contested concept in the study of social stratification. Researchers use various theoretically defined schemes to capture social positioning, which differ considerably in their understanding of the factors that define class boundaries. This ambiguity raises the question of the extent to which class reflects a tangible social structuring of occupations. Starting from the premise that class determines life chances, this article takes an inductive approach by investigating whether class boundaries can be empirically detected through patterns of social mobility.

Previous research has largely focused on intragenerational mobility, which, while informative about career trajectories, tends to cluster occupations with similar skill requirements, limiting its usefulness in assessing class boundaries. Intergenerational mobility, conversely, poses the challenge that labor market opportunities are affected by structural changes, impeding reliable comparisons between occupations of parents and children. To avoid both these issues, we propose using sibling occupation cross-classifications to identify class boundaries, leveraging the shared parental background and similar age of siblings to provide an effective assessment of social structure.

Using Swedish register data, we apply a stochastic block model (SBM) which allows us to uncover the underlying structures in a network of detailed occupations pursued by sibling-pairs. We analyze whether identified occupation groups align with established class theories or reflect alternative structuring mechanisms. By offering an empirically grounded understanding of the structure of social positioning, this article provides valuable insights into the ongoing debate on the relevance of social class as a concept.

Jobless individuals have been theoretically and empirically excluded from social class analysis, despite social class being one of sociology’s foundational concepts. Conventional frameworks often recode jobless individuals to their last main occupation or exclude them altogether, potentially neglecting crucial ways in which joblessness shapes inequalities in life chances. This paper addresses this gap by focusing on health inequalities associated with joblessness within a social class framework. Using longitudinal data from the EU-SILC, I examine France, Belgium, and Germany—countries with comparable joblessness rates but distinct labour market policies. Joblessness is defined broadly to include the unemployed, individuals fulfilling domestic tasks and other “inactives” at working age, while explicitly excluding retired individuals and students. It is theorized as a meaningful relation of exclusion from paid work that negatively impacts health through material and symbolic social penalties.
Self-rated general health serves as the outcome variable. To analyse health changes as individuals transition in and out of joblessness, I employ a fixed-effects ordered logit model, controlling for time-invariant characteristics that might confound the relationship between joblessness and health. The analysis stratifies these effects by social class. This approach tests two hypotheses: (H1) joblessness negatively affects health across all social classes, and (H2) the magnitude of this effect varies across classes. Drawing on fundamental cause theory (FCT), I argue that upper classes might be able to mitigate health penalties of joblessness due to flexible resources and power relations not fully captured by income, educational attainment or other individual-level characteristics usually analysed.

Research on right-wing populist parties (RPP) repeatedly demonstrates a higher propensity of the manual classes to vote for RPP. Further, it often concludes that, while relevant, economic factors are less important than social attitudes. We argue that this research underestimates social class effects on RPP voting since it neglects a) unobserved variables that simultaneously influence social class position and the propensity to vote for RPP and b) the interdependencies of turnout and the RPP voting. As to a), individuals self-select into occupations which form the basis of social classes, this self-selection is at least partially based on factors related to voting for RPP. Hence, social class is endogenous. As to b), voting for RPP and turnout are interdependent processes since RPP strongly mobilize previous non-voters. We analyse ESS data and combine doubly robust regressions and Heckman-regressions to estimate the effect of social class on RPP using a counter-factual approach: what would be the effects of social class on RPP voting if social class is randomly “assigned” and all classes have identical voting propensities? Compared to standard regressions we observe stronger social class effects, in particular for the manual classes. This is due to different propensities to vote, i.e. the two-stage voting process, instead of social classes’ endogeneity”, i.e. a matching correction. Hence, society in comparison with the electorate is more polarized with respect to social class than previously thought. Our findings also contribute to the debate as to whether the effect of social class on voting has diminished or vanished.

Economic integration is one among various important integration outcomes for asylum seekers, affecting individual well-being and economic development. When assigning asylum seekers to resettlement locations within a country, taking into account asylum seekers’ skills and characteristics could improve the chances of integration into the local labor markets. Researchers thus developed algorithmic procedures that match asylum seekers to resettlement locations based on asylum seekers’ characteristics. However, attempts to optimize solely for labor market integration may lead to unintended consequences for other integration dimensions. To investigate such effects, we use agent-based models to simulate different allocation mechanisms of asylum seekers to locations in Germany and their impact on integration outcomes. Concretely, we compare the current approach using the Königsteiner Schlüssel with the algorithm-based procedure GeoMatch. We study the procedures’ impacts on labor market and social integration, while taking effects on inequalities between subgroups of asylum seekers into account. Decision models and agents’ characteristics are based on the IAB-BAMF-SOEP survey of asylum seekers and refugees in Germany. This project contributes to the literature by showing to which extent algorithm-based approaches for optimizing labor market integration may impact another key integration outcome. More generally, it shows how agent-based models can be used to study unintended consequences of algorithm-based allocations of asylum seekers in dynamically changing demographic environments.

This study examines the long-term employment outcomes of immigrants who initially arrived as refugees or asylum seekers in Australia. Although the respondents in the study held different immigrant statuses upon arrival (refugee or asylum seeker), by the time of the study, all individuals had successful asylum applications and recently secured permanent residency. We refer to individuals who arrived as asylum seekers and had successful asylum applications and transitioned to permanent residency as asylees. Refugees, in contrast, refer to individuals who arrived in Australia with successful asylum applications and permanent residency. Descriptively, we first assess whether refugees and asylees differ in their labor market outcomes immediately after securing permanent residency? Do these disparities persist ten years after securing permanent residency? We also explore possible mechanisms that account for disparities between the two groups, specifically the role of welfare-state policies. Utilizing regression and propensity score weighting techniques, we find that the distinction in initial immigrant status between the two groups continues to shape their employment trajectories. We also find a surprising pattern: although refugees enjoy more welcoming policies and favored status when arriving in Australia compared to asylum seekers, they experience a relative disadvantage in the labor market both in the first year of permanent residency. However, a generous welfare state minimizes this gap ten years after permanent residency. Further, Australia’s bridging visa policy benefits asylum seekers by providing them with basic rights that has positive social and psychological effects, all of which can improve their labor market prospects.

The arrival of refugees is a major driver of the demographic changes the Netherlands has undergone in recent decades. This has increased the importance of effective policies to integrate them economically into the host country. Refugees have lower labor force participation and are often dependent on social assistance longer than other groups. This increases their risk of social exclusion, lowers their self-esteem, and depreciates their human capital. Although research has been done on refugees’ welfare dependency, the focus is largely on individual or household-level predictors. By contrast, whether and how refugee-focused policies contribute to or hinder exiting social assistance, remains under-researched, despite these policies playing a major role in the early lives of refugees in their host country.
We address this gap by analyzing how reception and integration policies affect refugees’ social assistance dependency in the Netherlands. We study how the probability of making the transition from social assistance to employment is influenced by the length of refugees’ stay in asylum seeker reception centers and number of moves between them, completion of the compulsory civic integration program, and the language level at which the integration program was completed.
We analyze register data from Statistics Netherlands of 100.000 refugees who received a residence permit between 2014 and 2022 using discrete-time duration analysis and multilevel linear probability modeling.

About 242,000 individuals were legally required to leave Germany at the end of 2023. The German government has been working to reduce this number through various political measures, including improving return processes and regularization campaigns. However, suitable and sus-tainable policies require an evidence-based understanding on how intentions to stay or to return are formed within the context of irregularity. There is a large research gap in this regard, because irregularised migrants are difficult to reach for study purposes. We tackle this challenge with innovative sampling and survey strategies to examine the factors influencing irregularised mi-grants’ intentions to stay or return. Using a mixed-methods approach, we build on existing litera-ture regarding outmigration decisions and return aspirations of regular migrants while integrating the concept of immobility.

In 2023, we collected survey data from rejected asylum seekers from West Africa, whom we sampled through an app-based Respondent Driven Sampling (RDS) as well as through a random sample drawn from the Central Register of Foreigners. The survey included a factorial survey in which respondents advised hypothetical irregularised migrants with varying characteristics on their decision to stay, return or move.

By combining quantitative findings from the survey experiments with insights gained during a comprehensive 10-month formative assessment comprising ethnographic fieldwork and qualita-tive interviews, we provide new insights into the negotiation of mobility and immobility among rejected asylum seekers. Thereby, the results contribute to a deeper understanding of the im-/mobility intentions of irregularised migrants and offer valuable perspectives for more informed policymaking.

Migrants often send remittances to their origin countries, which are crucial for both their financial well-being and that of the origin country. While most studies focus on labor migrants, less is known about the remittance behavior of refugees. Demographers expect sustained refugee-inflows in the decades ahead, hence it is important to understand how their remittance behavior differs from that of non-refugee migrants. Theoretically, refugees may be less likely to remit due to lower incomes and disrupted family ties, but they could also feel a stronger motivation to send money to support family members. This study aims to study (a) who is more likely to remit; and (b) what drives different remittance behaviors between refugees and non-refugees. Additionally, we study the difference in remittance-sending between different groups of refugees. Based on Dutch survey data from four refugee groups (Afghanistan, Iran, Iraq, and Somalia, N = 3,205) and six non-refugee groups (Turkey, Morocco, Suriname, Dutch Antilles, Poland, and China, N = 6,257), we analyze differences using mediation-approaches. Our findings show that refugees are 9.4 percentage points less likely to remit than non-refugees, with this difference explained by refugees’ lower household incomes, reduced family contact, and lower homeownership rates in their origin countries. Surprisingly, refugees are not more likely to remit despite potential motivations to support their families. Additionally, we find variation within the refugee populations, with Somali refugees being the most likely to remit, followed by Afghani, Iraqi, and Iranian refugees, highlighting important variations between and within the group of refugee migrants.

There is a research gap on loneliness in Germany among young and middle-aged adults with a migration background, especially the Afro-diasporic community, even evidence from other countries shows that this group is at higher risk of loneliness.
Based on the German Family Demography Panel Study (FReDA 2021), the prevalence of loneliness among the German Afro-diasporic community (173 respondents) is being analyzed by comparing it with natives (13.245 respondents) and other migrants (2.109 respondents). Linear logistic regression models and average marginal effects are utilized to identify potential risk factors and protective factors of loneliness.
Our descriptive findings prove that both emotional and social loneliness are common in Germany. People who have immigrated from African countries are more often emotionally lonely (36 %) than natives (26 %) or immigrants from other countries (31 %). The same pattern can be observed for social loneliness (49 vs. 36 vs. 46 %).
The multivariate results indicate that both dimensions of loneliness are related to the individuals' specific migration history. While emotional loneliness tends to affect the closest relationships, with immigrants benefiting from pre-existing intra-ethnic networks and family systems in Germany (e.g., migrants from the Maghreb region), social loneliness is associated with the effects of exclusion, discrimination, and racism in the host society. We contribute to the field by presenting findings about loneliness as a dimension of well-being for an under-researched subpopulation. In addition, besides risk factors, we were able to identify protective factors such as German citizenship as a central proxy for integration.

Immigrants often arrive with better health than the native population of their host country, a phenomenon known as the ‘healthy immigrant effect’. However, this initial health advantage declines over time, partly due to unequal access to healthcare. This study investigates disparities in healthcare access among immigrants in France, often attributed to demographic and socio-economic factors. While these determinants are often analyzed separately, I adopt an intersectional approach to examine how they interact and contribute to compounded disadvantages. Drawing on the recent Trajectoires et Origines 2 survey (N = 27,181), I apply the Multilevel Analysis of Individual Heterogeneity and Discriminatory Accuracy (MAIHDA), a recent quantitative methodological advancement that enhances the study of intersectionality beyond conventional intercategorical models. This work explores how gender, age, geographical origin and social class intersect to shape healthcare inequalities among immigrants in France. By identifying complex patterns of exclusion, this research underscores the need for frameworks that captures the multidimensional nature of healthcare access disparities. With this method, I show that it exists strong heterogeneity between groups in terms of doctors recourse. Further analysis will explore which groups face the greatest barriers to healthcare access, and will compare MAIHDA models observed outcomes to additive predictions, exposing variability from complex identity interactions.

With a shrinking population and a rising dependency ratio along with a growing ageing population, most high-income countries seek to attract well-educated and skilled immigrants who are willing to integrate and contribute to their socioeconomic development. However, it is unclear whether immigrants benefit from higher education in terms of subjective well-being (commonly known as happiness). Using the German Socioeconomic Panel and Household Income, Labour Dynamics in Australia, we explore whether higher education contributes to immigrants’ subjective well-being over time since migration, and how this experience differs between immigrant men and women. We find that in both countries, highly educated immigrants are initially less content than their less educated peers; however, as the benefits of higher education accrue over time, subjective well-being gradually improves. Highly educated immigrant women benefit the most from these outcomes, while less educated immigrant women face significant disadvantages compared to those who are highly educated. Highly educated immigrants have higher expectations, which require better life circumstances to meet them and achieve expected subjective well-being. Immigrant men, despite their high levels of education, are less happy than immigrant women. However, their subjective well-being gradually improves as compared to newly arrived cohorts. In both countries, immigrant men tend to realise their expectations over time, resulting in an increased subjective well-being regardless of their level of education. Structural and policy differences between Australia and Germany further moderate these dynamics, emphasising the importance of contextual factors in determining the extent to which education promotes subjective well-being.

This study explores multimorbidity—the coexistence of two or more chronic diseases—among immigrants and native-born populations in Norway. Drawing on data from the Norwegian Population and KUHR registers (2014–2019), we examine health patterns for 2.7 million natives, 550,000 first-generation (FG) immigrants, and 190,000 second-generation (SG) immigrants. Latent Class Analysis reveals five distinct clusters of multimorbidity, with FG and SG immigrants showing slightly lower rates (53%) than natives (57%). However, disease patterns vary: hypertension is more common among natives, while FG immigrants have higher rates of diabetes. Health outcomes also differ by region of origin and migration reasons, with Nordic immigrants generally healthier and non-Western immigrants experiencing higher multimorbidity. Gender disparities are apparent, as women are more prone to mental health conditions. These findings underscore the complexity of multimorbidity, shaped by migration and social factors, and the need to move beyond simple disease counts to understand health disparities fully.

Despite women's increasing participation in the labor market and significant expansions in formal daycare, the work-care arrangements of families with toddlers in Germany are still characterized by parental care and reduced maternal employment. Not only the limited availability of daycare spots but also widespread ideals of parental childcare and traditional views toward maternal employment contribute to the persistence of traditional work-care arrangements. Using data from the German PASS-Panel (2011, 2016, 2022), we analyze respondents' attitudes regarding the ideal age for a child to begin a) regular daycare for a few hours and b) full-time daycare outside the family as well as their attitudes regarding when a mother could work a) at least 15 hours per week and b) full-time (30+ hours). Therefore, we provide a more nuanced description of work-care attitudes than previous studies, which relied on broadly formulated items with little variance. We focus on differences across social groups and between East and West Germany. Support for part-/full-time employment is temporally lagged by a few months after the appropriate age for the child to start regular daycare and full-time daycare, respectively. Although attitudes between East and West Germany show slow convergence over time, mainly due to changes among West Germans, East Germans still think that children can attend (full-day) daycare at earlier ages and that their mothers can return to (full-time) work earlier. Personal affectedness in terms of gender and parenthood and individuals' socioeconomic background in terms of education and migration background strongly shape individuals' attitudes in both regions.

This study investigates the gendered division of contraceptive labor—unpaid tasks like researching, managing, and ensuring the use of contraception—primarily carried out by women within heterosexual couples. Using data from the German Family Demography Panel (FReDA) from 2021, we introduce the “ContraIndex” a novel measure quantifying the costs and risks associated with different contraceptive methods. Connecting contraceptive labor to the framework of the gendered division of labor, we explore how women’s relative resources and gender ideology shape their level of contraceptive labor. Preliminary findings suggest that higher-educated women chose contraception with higher contraceptive labor, while those with equal or higher employment than the male partner perform less contraceptive labor. Traditional gender ideology is associated with increased contraceptive labor for women. However, these associations are influenced by socio-demographic characteristics like age and parity. Our results suggest a complex interplay between both partners’ resources and contraceptive behavior, highlighting the need for further research on this underexplored aspect of gender inequality.

With the second demographic transition and an incomplete gender revolution, heterosexual relationships have undergone significant structural shifts that are reflected in macro-level attitudes. This study aims to contribute to our understanding of micro-level processes that may underlie or influence these trends. Applying a life course perspective and exposure- and interest-based explanations of change, we investigate how entering a relationship influences individuals’ gender, family and relationship attitudes. We want to understand whether and how partners align their beliefs over time. Specifically, we extend previous literature by considering who affects whom and under what conditions. Using nationally representative data from the German Family Panel pairfam, we provide a strong methodological approach by combining two-stage least squares (2SLS) and fixed effects (FE) regression models. Attitudes are measured with six statements covering gender ideologies regarding private and public sphere roles, the necessity of marriage and the acceptance of divorce. For the 2SLS models, we use one partner’s mother’s attitude as an instrumental variable. Findings show strong evidence of attitudinal alignment, particularly within three to five years of relationship duration. Depending on the item, attitude shifts are observed towards the more traditional, the more progressive or both partners. Often the effects towards the more progressive partner appear to be stronger, with institutional transitions such as cohabitation, marriage, and parenthood amplifying these effects. Regarding gender differences, results show no consistent patterns in who influences whom more.

Drawing on original data from a multifactorial survey experiment, we develop a theoretically informed model to examine support for gendered divisions of work among couples. We ask how respondents’ evaluations vary by division of work formulated according to contemporary gender ideology profiles. We further examine whether women and men, as well as participants with native and migrant origins (East and West German, Turkish, Dutch) vary in their evaluations of the specified gendered division of paid and unpaid work. Our findings indicate clear gender differences in participant evaluations. Women are more likely to prefer egalitarian divisions compared to men and to oppose arrangements where they are responsible for both housework and childcare. Men’s support is significantly higher for arrangements, in which women have primary responsibility both for housework and for childcare. Furthermore, there are gender differences in support for egalitarian division among West German, Turkish and Dutch (not significant), but not among East German women and men. Turkish men express the least support for egalitarian profile, while West German and Dutch are similar in their support. Participants express highest levels of support for egalitarian profiles. However, they do not support equal levels of labor force participation if this means women have to shoulder the entire “second shift.” Therefore, participants express lowest level of support for profiles, in which women work and at the same time are responsible for housework and childcare. These specific profiles, in fact, receive lower support than traditional male-breadwinner model.

Economically developed societies exhibit what is known as the gender-equality paradox: While these societies have firmly institutionalized gender equality, this often leads to seemingly contradictory developments at the individual level, particularly in the form of stronger gender-stratified economic outcomes. These paradoxical patterns are most evident in the domain of education but also extend to the gendered division of household labor and labor market participation. Such patterns challenge the intuitive assumption that the socioeconomic outcomes of women and men will converge as countries become more affluent and gender-equal.

Drawing on resource hypotheses and preference theory, our study aims to provide a broader perspective on the gender-equality paradox. We systematically analyze cross-national differences in economic and egalitarian developments and their individual-level implications for gender norms and gendered labor arrangements. In doing so, we seek to disentangle the interplay between economic wealth and gender policies to better understand their respective contributions to the gender-equality paradox.

Using four waves of ISSP data (1994–2022), which capture substantial within-country variation in institutional settings, we employ a multilevel design with country fixed effects. Preliminary results indicate that increasing contextual affluence is indeed associated with more—not less—traditional attitudes. This effect is particularly pronounced when the counteracting positive mediation effect of family policy investments is held constant and when women are excluded from the analysis.

The first years after labor market entry are an important period for young adults’ career progression. Entering the labor market during economic downturns increases the risk of unemployment, low job quality, and skill mismatches. Whether labor market outcomes improve once the economy picks up or persist over the long-term requires further investigation. This study examines the “scarring effects” of entering the labor market during economic downturns on long-term career outcomes among young adults in Europe who entered the labor market between 1993 and 2016. Using data from the European Labor Force Survey, we analyze the relationship between regional unemployment rates at the time of labor market entry and labor market outcomes 5 to 10 years later, including unemployment, temporary employment, low occupational status, and nonemployment. Preliminary findings reveal that while initial poor labor market conditions do not necessarily lead to persistent unemployment, they are significantly associated with long-term increases in temporary employment, low occupational status, and nonemployment. We find that less-educated individuals are strongly affected by initial labor market conditions while those with tertiary education experience minimal repercussions. The increased risks of unemployment and nonemployment are concentrated among less-educated women. For them, a loose labor market may imply particularly low opportunity costs of family formation and they may seek uncertainty-reduction via union formation. To better understand these interdependencies, we will take a look at spillover effects on family formation. We will also examine whether labor market institutions moderate the magnitude and persistence of adverse effects of labor market entry conditions.

This study explores how public discourse on unemployment influences the subjective well-being (SWB) of unemployed individuals, addressing two key research questions. First, we investigate whether media discussions around unemployment stigma and bad market prospects amplify or mitigate the negative impact of unemployment on SWB. Second, we examine whether these media influences vary by regional unemployment rates. Using data from 22 waves (1999–2020) of the German Socio-Economic Panel and 4,040 articles from the second largest national newspaper in Germany to assess topic salience, we apply LDA topic modeling and time-distributed fixed-effects models. Our findings reveal that a salient media focus on the everyday life challenges faced by unemployed individuals and bad labor market conditions intensifies the SWB decline among the unemployed. Contrary to the social norm hypothesis, these negative conditional effects on SWB persist regardless of the extent of regional unemployment. Taken together, the findings emphasize that societal attitudes reflected in media discourses can reinforce unemployment’s psychological costs, highlighting the need to consider media influence in employment policies for the unemployed.

Recipients of basic income support (BIS) for jobseekers in Germany often have low levels of qualification and either very little employment experience or long histories of cycling between unstable employment and unemployment. Specific groups among them may be barred from better quality employment due to discrimination, language barriers, or missing or not recognized educational degrees. In this context, we study the determinants of exposure to physical and psychosocial strain at work as one indicator of job quality. We use data from the Sample of Integrated Welfare Benefit Biographies (SIG) for people receiving BIS between 2005 and 2020. We employ a three-process hazard model, with entering employment and leaving benefit receipt without a job specified as competing risks. As a third process, we specify a probit model for the new job having high exposure to occupational strain. Our findings indicate that some vulnerable groups - older benefit recipients and those with low education - appear to be generally disadvantaged, both with respect to job entries as well as the quality of the jobs taken up, as measured by exposure to physical and psychosocial strain. Other groups such as men without German citizenship take up a job quickly but also suffer from high occupational exposure. As high job exposure to physical and psychosocial strain can negatively affect health and employability in the long term, improving benefit recipients‘ options for better quality jobs and occupational mobility is important.

In response to increasing difficulties faced by companies in filling vacant training positions, Germany introduced a minimum apprenticeship wage of €515 at the beginning of 2020. From a theoretical point of view, the effect of the introduction of a MAW is ambiguous: higher appren-ticeship costs may reduce firms' incentives to offer training positions, while increased wages could make apprenticeships more attractive to young adults, potentially increasing the number of applicants. It remains unclear which effect predominates. This paper provides the first empirical analysis of the MAW's impact on offered apprenticeship positions and actual take-ups in Ger-many. Using a difference-in-difference design with data from the IAB Establishment Panel and the Establishment History Data, we find negative effects on both offered as well as filled places. This finding is robust to a wide range of robustness and sensitivity analyses, including placebo tests for diverging trends before treatment or effects on other groups of employees, inclusion of control variables covering affectedness by Covid-19, post-treatment selection and attrition as well as different model specifications. Consequently, these results suggest that the negative im-pact on the incentive to offer trainings dominates possible gains in attractiveness in terms of filled places. The impact on educational outcomes such as drop-out from dual education are left to future research.