A new working paper by UBS Foundation Professor David Dorn and co-authors finds that self-reported skills on LinkedIn reveal more about earnings and gender gaps than traditional measures of human capital like education or work experience.

This article by Tessa Rauber was originally published by the UZH Department of Economics on 19.5.2025. Edited for layout purposes by the UBS Center.

Recent research examines the skills people list on their LinkedIn profiles and examines their relationships to education, work experience, income, and gender differences in the workplace.

Traditional economic theories of the labor market suggest that people can improve their job prospects by going to school or learning on the job. However, just looking at years of schooling or experience may provide a limited picture of people’s human capital. The authors analyze the self-reported skill sets of nearly 9 million U.S. college graduates from LinkedIn profiles, and group these skills into three types: general skills, occupation-specific skills, and managerial skills.

More skills mean higher pay

If skills are informative about human capital, then they should relate meaningfully to educational investments and work experience. Indeed, recent college graduates on average report fewer skills, and lower shares of occupation-specific and managerial skills, than their more experienced co-workers. These patterns align with the idea that young workers primarily gain general skills through education, while older workers additionally acquire specific and managerial skills through work experience and post-university training. Workers with degrees in STEM disciplines (science, technology, engineering, and mathematics), health, and law, and those with more advanced degrees, also report higher shares of occupation-specific skills.

The paper also finds a clear tendency for workers who report more skills to hold higher-paying jobs. In particular, occupation-specific and managerial skills are strongly associated with higher job-based earnings. While one might worry that self-reported skills provide only an imprecise measure of human capital, the authors show that these skills explain a significantly

Fewer skills, lower pay for mothers

There are striking gender differences in skills. While young women and men report similar numbers of skills early in their careers, the subsequent growth in workers’ skill sets is slower for women, especially between their late twenties to early forties. This slower skill growth is largely explained by time taken off for motherhood: Women with children on average work fewer hours in the labor market than comparable men, and thus accumulate new skills more slowly. They also differ from men by having smaller fractions of highly valued occupation-specific and managerial skills in their skill profiles. These gender differences in skills account for a substantial portion of the gender earnings gap.

Overall, even though LinkedIn skills are self-reported, they provide a remarkably rich and meaningful picture of workers' abilities, and this skill information is thus widely used by recruiters who look for suitable job candidates.