Research

Working with business partners in distant countries often requires operating outside standardworking hours, potentially harming workers' mental well-being—especially for women balanc-ing professional and family responsibilities. This hypothesis is investigated using matchedemployer-employee data from Denmark merged with information on workers' use of prescrip-tion antidepressants (AD) and firms' import and export transactions. The analysis exploitsvariation within job spells and controls for both unobservable firm-year heterogeneity and forthe average differential in annual AD use among men and women. The results indicate thata decline in the business hour overlap between a Danish firm and its foreign trading partnersleads to an increased AD use among women relative to men. This effect is economicallymeaningful, largest for college-educated workers under 45 and more pronounced for singlemothers. In addition to the gender effect, workers in occupations requiring establishingrelationships with others are more strongly affected than those in other occupations.

Misperceptions about the relative wages offered in different jobs may distort job search, especially at labor market entry. We survey early career job seekers in Denmark and link responses to administrative data. Our survey i) elicits beliefs about typical wages in three relevant jobs, ii) contains a randomized information treatment revealing actual typical wages, and iii) elicits planned search behavior and beliefs about own potential wages. Comparing beliefs about typical wages to administrative data reveals large relative misperceptions. In 80% of cases, the perceived wage gap between two jobs differs from the truth by more than 50%. Using the information treatment, we show that these misperceptions causally affect search intentions and realized post-survey wages in administrative data. Interpreting estimates through a discrete choice model, we find that 9.5% of job applications are misallocated due to wage misperceptions, resulting in the same welfare loss as a 0.8% reduction in all wages.

How does job flexibility affect hours worked for workers and their partners, and what are the broader effects on family wellbeing and fertility? We apply BERTopic modeling to the full text of 2 million Danish job advertisements (2012–2024) from jobnet.dk, extracting approximately 300 job characteristics from 40 million sentences. Linking these text-based measures to administrative data on employment, wages, and family outcomes, we exploit within-firm variation driven by M&A events and CEO transitions to estimate the causal effects of family-friendly working conditions on labour supply, fertility, and job separations — including spillovers to partners.