This project was realized during my internship in the Department of Sociology at Trinity College Dublin in summer 2023 (M1). I worked under the supervision of Jan Skopek, the Head of the Department, on the intergenerational transmission of skills.
One of the main results of sociology is the intergenerational transmission of inequalities. It was often analyzed by the analysis of some correlation between parents and their children: income, occupation, or educational path. However, a way that wasn’t contemplated was the intergenerational transmission of skills. For example, does a parent with high verbal skills have children with high verbal skills?
Then, I made a literature review to analyze this phenomenon. Firstly, I established a definition; a skill is an ability that a person acquired through practice. To complete this definition, I compared some typologies: cognitive (verbal, mathematical, reasoning…), social, and manual… to make a definition adapted to the contexts. The second point was to find a way to measure skills. There were direct methods: declaration or psychometric measure. Searchers often used psychometrics methods for children. Also, when direct measures weren’t available for parents, some searchers used indirect measurement strategies: level of education or skills linked to their occupation. Therefore, I saw there was a positive correlation between children's and parents’ cognitive and social skills but a negative one between the manual skills of parents and the cognitive ones of children. Also, I saw a negative correlation between the size of the family and the children's outcome and gendered effects: coding speed was influenced by the mother when verbal fluency was more linked with the own-gendered parent.
Thus, I could develop a theoretical framework. I compared two theories. One was the selection theory, the other was the theory of secondary socialization. The selection theory is the idea that parents with more capital can help their children to be selected for the highest social groups. I took the Bourdieusian theory of cultural capital: the way to act, the cultural objects and the diplomas are capital that people can mobilize to reach the prevailing social class. Therefore, the intergenerational transmission of skills is explained by the selection of school systems and cultural institutions. Another way to approach this is to consider the secondary socialization of the parents. It means people continue to incorporate after childhood and early adulthood skills and attitudes that can influence the education of their children. Moreover, I highlighted that we have to consider that parenting skills and financial resources have a role.
With this information, I decided to compare the effects of selection theory and the theory of secondary socialization. First, I created the database. My tutor gave me access to the German National Educational Panel Study (NEPS). I used it to have data about the verbal and mathematical skills of children of various ages between 3 and 9 years old. However, I didn’t have a direct measure of parents’ skills, but I knew the ISCO code of each parent of a child. Thus, I used the skills' database of O*NET, which links skills and occupations. It refers, for each job, to the importance of each skill and a job and their level of requirement by skill. There were a great number of skills, so I regrouped it using a principal factor analysis to reduce the dimensionality and I obtained four categories: Non-STEM Cognitive and decision skills, manual skills, STEM skills, and managerial skills. Therefore, for each parent of each child, I had the importance and the level's requirements and importance of these four skills in their occupation. The idea was to compare if the correlation was stronger with the importance of the skills or their level's requirements. Indeed, in the first case, the theory of secondary socialization is probably dominating, whereas it’s the selection theory in the other case. To estimate the correlation, I ran linear regression of the children's outcome on the parental skills, sometimes controlled by the number of years of education or socioeconomic status.
The results show that, generally, managerial skills aren’t statistically significant. The correlation is globally positive between cognitive and decision skills and children's outcomes. It’s the opposite for manual skills. The correlation of the level requirements is higher for Non-STEM Cognitive, Decision, and Manual Skills than the correlation with the importance, and it’s the reverse for STEM skills. With age, we see that the highest correlation is at 5 and 6 years old. When controlled by education or socioeconomic status, coefficients become non-significant or are divided by 2. The STEM competence of the mother is more impactful for mathematical outcomes, whereas the non-STEM Cognitive Skills of the father are more impactful for vocabulary outcomes. Then, we conclude that selection theory is dominant in explaining the transmission of Non-STEM Cognitive skills, whereas the theory of secondary socialization has to be considered for mathematical skills. However, I highlight some limits: we should reproduce it with other data and analyze other skills, or measure parental skills in other ways, by using, for example, the skills acquired during the educational path.
Details available on GitHub: https://github.com/jmeynent/TrinityResearch
Comments