The four individual papers in this thesis all explore some aspect of the relationship between productivity and the density of economic activity.

The first paper (co-authored with Martin Andersson, and Johan Klaesson) establishes the general relationship between regional density and average labor productivity; a relationship that is particularly strong for workers in interactive professions. In the paper, we also caution that much of the observed differences are not causal effects of density, but driven by sorting of actors to

dense environments.

Paper number two (co-authored with Martin Andersson, and Johan Klaesson) addresses the attenuation of density externalities with space. Using data on the neighborhood-level, and information on first- and second-order neighboring areas, we conclude that the neighborhood effects are stronger for highly educated workers, and that the attenuation of the effect is sharp.

In the third paper, I estimate an individual-level wage equation to assess appropriate levels of aggregation when analyzing density externalities. I conclude that failure to use data on the neighborhood level will severely understate the  benefits of working in the central parts of modern cities.

The fourth paper departs from the conclusions of the previous chapters, and asks whether firms position themselves to benefit from density externalities. Judging by job switching patterns, the attenuation of density externalities are a real issue for the metropolitan workforce. Employees, especially those in interactive professions, tend to move short distances between employers, consistent with clustering to take advantage of significant but sharply attenuating human capital externalities.