How does working from home affect different types of places?

In this article, we set out the relationships between the behavioral and spatial responses to working from home. The analytical framework centres explicitly on the choice of commuting frequency as the key decision‐making variable that endogenously reshapes the relationships between other spatial and nonspatial variables as a result of the work‐from‐home revolution. We find that optimal commuting frequency is positively related to the opportunity costs of less‐than‐continuous face‐to‐face interaction and inversely related to commuting costs. As well as a “ donut effect” with growth in the suburbs and hinterlands around cities, our results also identify a “shadow effect” in smaller cities. The reason is that, somewhat counterintuitively, commuting frequency optimisation magnifies the benefits of working from home in larger cities because of a greater decrease in the burden of commuting. Our results imply enhanced productivity of larger cities over smaller cities, suggesting that the economic divergence between large cities and left‐behind places is likely to persist.