Urban Economics, Commercial Activity, and Crime

Date of Award

Winter 12-22-2021

Degree Type


Degree Name

Doctor of Philosophy (PhD)




Rosenthal, Stuart S.


Commercial activity, Criminal activity, Police, Spatial patterns, Urban economics

Subject Categories

Criminology and Criminal Justice | Economics | Legal Studies | Regional Economics | Social and Behavioral Sciences


This dissertation comprises three papers that address current and relevant questions in the fields of urban economics and the economics of crime. All three papers consider neighborhood amenities to explain spatial patterns of commercial activity and equilibrium levels of crime and police patrolling.Chapter 1 estimates the value that commercial establishments place on employment density and proximity to both the city center and transit station. However, during COVID-19, costs related to disease transmission in crowded spaces and the ability to work from home might reduce the value that firms place on these neighborhood attributes. We use data on newly executed leases across 89 U.S. urban areas to document these changes. Pre-COVID, commercial rent declines 2.3 percent per mile from the city center and increases 8.4 percent if one doubles zipcode employment density. These relationships are stronger for large "transit cities" that rely heavily on subway and light rail. Post-COVID, the rent premium associated with employment density declines for all cities while the rent gradient falls only in transit cities. The premium for proximity to transit stops also decreases. Chapter 2 shows that concentrating neighborhood-level retail activity helps deter property crime and reduce equilibrium levels of local police patrols. These benefits are capitalized into higher local rent. We develop a conceptual model in which concentrating retail establishments at the street level should amplify the effect of eyes on the street by crowding shoppers into smaller areas and by making it possible to observe multiple storefronts at once. These effects will reduce property crime and the need for public and private investment in anticrime measures. Point-specific data for New York City support the model. Increasing block-level retail spatial concentration by one standard deviation reduces property crime and police stops by 8.5% and 11%, respectively, and causes retail rent to increase by at least 7.8%. Comparisons between nighttime versus daytime activity, pre-pandemic versus COVID-19 lockdown, and different measures of spatial concentration shed light on mechanisms. Chapter 3 considers the effect that perceptions of police bias have on the location of minority-owned establishments even after evidence of police bias has diminished. Customers might avoid neighborhoods where concerns about police bias are present. That will cause profits to shrink for minority entrepreneurs who cannot compensate customers with lower prices. Workers will also require a higher wage for working in neighborhoods where they feel unsafe. I examine these issues using 2010-2019 data on business activity, and police pedestrian stops in New York City. Results indicate that Census Tracts where police can be perceived as biased in 2012 have a lower minority share of small establishments in 2019. The effect is larger in predominantly minority neighborhoods. Survival models confirm that a contributing factor is the higher exit rate of minority-owned establishments in those locations.


SURFACE provides description only. Full text may be available to ProQuest subscribers. Please ask your Librarian for assistance.

This document is currently not available here.