Date of Award

August 2020

Degree Type


Degree Name

Doctor of Philosophy



First Advisor

Scott D Drewianka

Committee Members

Scott J Adams, John S Heywood, Rebecca Neumann, Kevin E Thom


Human Capital, Individual-level Task Data, Skill Differences, Wage Determinants, Wage Gaps


A key interest in labor economics is to understand quality differences between workers and why technology helped some types of labor, while hurting others. Conventional methods rely on formal qualifications such as education or experience to measure skill differences between workers. These are crude measures, however, as they assume that workers with comparable formal qualifications perform similar activities at work and thus earn similar wages. To provide remedy, this dissertation extends the task approach to labor markets, popularized by Autor, Levy & Murnance (2003), by utilizing information on tasks performed at work. This strand of the literature utilizes information on job-related activities to enhance our understanding of the concept of `skill' and how it translates into wage differences.

The first chapter uses novel survey data from Germany which provide self-reported information on job-related activities by individuals, thus task requirements at the worker level. Commonly used data such as the Occupational Information Network (O*Net) database in the US are based on occupational analysts and therefore provide external assessment on the job-specific task requirements at the occupation-level. Comparing this task data with Expert-based data provided by the German Federal Employment Agency, similar in spirit to the O*Net database, my findings document substantial heterogeneity in task assignments at the individual level. This variation in job-related activities is predictive of wage differences between and within occupations and robust to a series of alternative model specifications. Importantly, various statistical tests favor individual-level information on tasks over occupational measures due to greater explanatory power on wages. The superior statistical performance of Survey data is related to intra-occupational efficiency gains workers earn as a result of task specialization within occupations. Suggestive evidence indicates this enhanced degree of task specialization may become even more important if greater weight is given to the time allocation of job-related activities. Overall, the results suggest incomplete information on the part of Expert data and recommend worker-level information in studies on job tasks.

The second chapter applies the detailed information on individual tasks to explore the wage gap between native and foreign workers. In this study, I decompose wage differences along the wage distribution, adopting a statistical tool called `Recentered Influence Function' (RIF). This way I estimate unique wage responses resulting from a change in job activities by nativity and at different points of the distribution. According to this distributional analysis, variation in interactive tasks has been a key contributor to the rising native-foreign wage gap, suggesting that native and foreign workers perform distinct activities at work. Importantly, variation in task assignments is most pronounced among high-wage earners, explaining up to 25% of wage differences, and can also be found among workers with similar formal qualifications. Previous research has documented how natives utilize their comparative advantage in interactive tasks by choosing occupations intensive in communication-heavy activities. However, my research is the first to demonstrate that this specialization pattern can likewise be found within occupations and as this trend has become more meaningful in recent years it reinforced already existing wage disparities. These idiosyncratic differences can explain small migration-induced wage effects despite assimilation in formal qualifications. My research thus has important implications for the integration of immigrant workers and offers a novel source of imperfect substitutability between native and foreign workers, which is at the core of small migration-induced wage effects usually found in the literature.