Suchergebnisse

To what extent social inequalities in socioeconomic standing, SES, are multidimensional is still unresolved. This paper uses the methodology of Karlson and Birkelund (2023) to study how sibling correlations, as a measure of family background influence in intergenerational transmissions, overlap across SES dimensions. Four dimensions of SES are considered: education, occupation, income (earnings and disposable income), and wealth. The findings reveal three clusters represent the overlap pattern of sibling correlations across these dimensions: (1) education/occupation, (2) income, and (3) wealth. This pattern remains consistent across various subsamples based on sex. The data analysis shows that a one-dimension sibling factor, assuming a single latent SES factor, fails to adequately capture the observed sibling correlations. Still, the one-dimension factor contributes modestly to substantial portions to the sibling correlations in all outcomes. Furthermore, the idea of a single latent family factor is also refuted when including additional outcomes such as crime, cognitive ability, and physical capacity (health) in a younger male sample born between 1958 and 1968. In conclusion, while a unidimensional sibling factor exists, it is trumped by multidimensionality.

This paper estimates the size of the effect of parental union dissolution on offspring union dissolution, as a share of the sum of all social origin factors. A large literature has documented a positive correlation between divorce and separation among parents and their children. Parental union dissolution is one out of several aspects of social origin that is associated with union dissolution. It is difficult to contrast the relative impact of parental union dissolution to other aspects of social origin because many apects of social origin are unobservable. Swedish administrative data for the 1960 to 1965 birth cohorts that cover the individuals' life events until 2018 is used to estimate sibling correlations in divorce and childbearing union dissolution, adjusting for parental union dissolution. The variance in union dissolution attributable to factors shared by siblings ranged from 6% to 13%. Parental union dissolution and factors associated with parental union dissolution explained between 15% and 28% of this variance. Sister correlations are greater than brother correlations, and sibling correlations of childbearing union dissolution are higher than sibling correlations of divorce. It is pertinent to estimate the total effect of all social background factors on offspring union dissolution. It is likewise of interest to specify the share of all social background effects that are constituted by factors caused by or correlated specifically with parental union dissolution. Sibling correlations can be a useful tool for quantifying these relationships.

We introduce a novel approach for decomposing sibling correlations in socioeconomic attainment into group-specific components. These groups are defined at the family level, based on factors such as parental income, marital status, race, or the gender composition of siblings. Unlike existing studies that compare sibling correlations across different groups, our method accounts for differences in average outcomes between groups in the overall outcome distribution. This approach results in a group-specific measure of persistence in the overall outcome distribution, contrasting with the 'within-group' measure commonly used in the literature. Additionally, we propose a straightforward 'counterfactual' framework to identify which groups have a greater impact on trends or country differences in sibling correlations. We present three examples that demonstrate the utility of our method.

Abstract
We introduce a novel approach for decomposing sibling correlations in socioeconomic attainment into group-specific components. These groups are defined at the family level, based on factors such as parental income, marital status, race, or the gender composition of siblings. Unlike existing studies that compare sibling correlations across different groups, our method accounts for differences in average outcomes between groups in the overall outcome distribution. This approach results in a group-specific measure of persistence in the overall outcome distribution, contrasting with the 'within-group' measure commonly used in the literature. Additionally, we propose a straightforward 'counterfactual' framework to identify which groups have a greater impact on trends or country differences in sibling correlations. We present three examples that demonstrate the utility of our method.