Who works longer hours in smart cities?

This paper investigates how human capital concentration in cities is associated with working hours across different worker groups, an important but understudied dimension of urban agglomeration effects. Using microdata from the American Community Survey covering 240 metropolitan statistical areas in 2018, the study finds significant heterogeneous effects: A one percentage point increase in college graduate share is associated with a 0.043% increase in working hours for college graduates but a 0.023% decrease for non-college workers. The effects vary between employment types: College-educated paid workers work 0.054% more hours, while the self-employed work 0.071% fewer hours in cities with higher human capital stocks. Through a two-step two-stage least squares approach, the study reveals that these effects operate primarily through income changes rather than non-income channels. Alternative measures of human capital stock and various robustness checks confirm the main findings. These heterogeneous labor supply responses suggest that the welfare impact of place-based development initiatives depends not only on productivity gains but also on workers’ capacity to capture these benefits through skill development, highlighting the importance of complementing talent attraction policies with workforce development programs.