Research

How does artificial intelligence (AI) affect the actual task content of jobs, and how do workers adjust to AI? We combine novel patent-based measures of AI exposure with survey data on tasks and administrative data on worker careers to answer these important questions. While robots replaced routine tasks, AI shifts the content of jobs from non-routine analytical tasks to `high-level' routine tasks. This task shift occurs mainly within detailed occupations and is stronger among low- and medium-skilled workers. AI exposure increases job mobility with workers moving away from exposed to non-exposed industries resulting in small negative employment and earnings effects overall. 

We study the heterogeneous earnings effects of industry-level exports, using a

generalized random forest algorithm. Our approach allows the effects on workers’

earnings to differ across a large range of worker, firm, and job characteristics as

well as their interactions. We find that workers with highly specialized skills and

workers employed in a large firms experience substantial earnings gains due to

exports, with these two factors strongly interacting. Our method identifies a share

of workers who incur sizable earnings losses due to exports, in contrast to more

conventional methods to estimate heterogeneous effects. Our results suggest that

specific human capital is more relevant than general human capital (e.g., formal

education) in explaining who benefits and who loses from increased exports.

We examine preferences for gender diversity in high-profile jobs, using stated-choice experiments with more than 9,000 highly educated individuals in Germany. Across three distinct samples covering university students, Ph.D. students, and university professors, we uncover a substantial willingness-to-pay for gender diversity at the workplace of up to 5% of earnings on average. Women have a much higher willingness-to-pay for gender diversity compared to men across all samples. Our findings carry insights for why organizations with a high share of men in top positions may find it difficult to attract and retain top-talent women.

What explains relative wage losses of import-exposed workers? I study the link between manufacturing import exposure, relative losses in firm wage premiums or firms rents, and wage inequality, using German micro data. Import exposure induces worker mobility to the service sector where firm wage premiums on average are lower. More skilled workers, compared to less skilled workers, are better able to transition to the highest-paying firms within narrowly defined service industries and are better able to transition to service industries where firm wage premiums are high on average. As a result, the degree of worker-firm sorting or assortative matching increases. Relative losses in firm wage premiums explain 50% (70%) of the resulting rise in wage inequality between (within) skill groups.