Selected effects of urban mobility

Abstract

[EMBARGOED UNTIL 5/1/2024] This dissertation presents four essays characterizing urban mobility through cellphone location data. Chapter 1 surveys the recent literature using location data to understand urban interactions and mobility. Chapter 2 defines home and work locations within the St. Louis MSA, and uses these definitions to characterize changes in workforce mobility during and after the COVID-19 pandemic. Chapter 3 focuses on workplace interactions by demographic group. Finally, chapter 4 illustrates shortterm changes in crime reporting during the COVID-19 pandemic. Chapter 1: Recent articles leveraged cellphone location data to research locationbased policy questions. Initially, economists focused on mobility and consumption patterns such as distance travelled for commuting and non-commuting trips. Subsequent analysis expanded to the effects of social distancing policies in response to COVID-19, segregation, and the benefits of in-person interactions. Chapter 2: Using high-frequency smartphone location data between 2019 and 2021, this paper describes the likelihood of working on-site during and after the COVID-19 induced shutdown in the St. Louis MSA conditional on residential block group demographics. While the probability of working on-site significantly declines during the shutdown, there is a partial resumption of on-site work once the strictest social distancing measures are lifted. Those living in lower income and less educated areas are more likely to continue working on-site during and after the pandemic. Conversely, race has no sizeable effect on the probability of working on-site during the COVID-19 shutdown. Chapter 3: This paper depicts the characteristics of on-site workplace interactions using anonymous cellphone location data for a representative sample of the St. Louis MSA during the months of April and June between 2020 and 2021. Employees with less than a bachelor's degree tend to have their workplaces most evenly dispersed throughout the city, while those with a bachelor's degree or higher tend to be the most spatially concentrated. Among a sample of employees at the largest firms in the St. Louis MSA by observed 2021 employment, more educated employees tend to work in larger firms and are more likely on average to interact with a higher educated person at work. Chapter 4: Following the spread of COVID-19 to the US, human mobility fell significantly in major metropolitan areas, leading to a compositional change of criminal reporting. We use the timing of the rise in social distancing and crime reporting data to model the impact of reduced personal interactions outside the home on personal and property crime reports. While models vary in their statistical significance, most models are consistent with about a 15 percent decline in property and personal crime reports in response to doubling the percentage of those staying fully at home.