Gender Differences in Returns to Skills

Mathias Fjællegaard Jensen

Student thesis: Master thesis


Women, men, and their differences in the labour market are widely studied, e.g. in terms of segregation, participation, pay gaps, and discrimination. In this literature, workers are normally categorised into mutually exclusive occupations, so that job characteristics and their variation across occupations can be imported from databases such as O*NET and DOT. However, recently available data from online job vacancies enable analyses that move beyond across-occupation variation to also include within-occupation variation in terms of skills: in which occupations, but also, in which firms do workers employ certain task-specific skills? Such a question can be answered by the use of job vacancy data alone. More interestingly, I also test how the employment of skills and their returns depend on the gender of the worker by exploiting a novel combination of Danish job vacancy data and Danish individual-level register data. I am the first to operationalise this combination of data. I use the novel combination of vacancy and register data to show that, in aggregate, similar skills are required by women and men. Some gender differences are observed, but they are small relative to the high degree of gender segregation in the labour market. Furthermore, variation in skills cannot be fully accounted for by occupations and other sets of covariates. Even with a very extensive set of control variables, regression analyses indicate that women face lower returns to character, writing/language, customer service, management, financial, and computer skills when compared to men. This result is driven by workers in the private sector. Whereas these results are merely correlational, they serve to warrant future casual studies of the impact of task-specific skills on wages, and thus, om the gender pay gap. An empirical analysis of gender differences in the deployment of and returns to skills does not only provide further insights into pay differentials between women and men. Equally important, my analysis of skills opens up a discussion of the difference between the female worker and female work, as well as between the male worker and male work. This represents a move away from the essentialist approach taken by many economists where no such distinction is made. The distinction becomes vital as the skills framework breaks down mutually exclusive occupational categories, which are inherently gendered. Occupational gender segregation is costly, not only by causing gender pay inequalities, but also because of the tokenisation of workers whose occupational choice and gender identity do not align. However, reconceptualising work through the skills framework facilitates a transgression of the boundaries between occupations, and thus, between female and male work. Female work becomes multi-dimensional and so does its male counterpart; it becomes clear that the perceived reality of the gendered worker is only superficially constructed. This realisation allows a move beyond segregation; a move beyond an occupational division of workers to a multi-dimensional approach to work. Conceptualising work as skills, rather than simply occupational categories, therefore both enrich quantitative studies of pay inequalities, and it potentially allows workers to forsake identities of work based on inherently gendered occupational categorie

EducationsMsc in Business Administration and Philosophy, (Graduate Programme) Final Thesis
Publication date2018
Number of pages91
SupervisorsFane Naja Groes