Cities after the fall: population loss and recovery of U.S. cities
Description
TitleCities after the fall: population loss and recovery of U.S. cities
Date Created2021
Other Date2021-10 (degree)
Extent1 online resource (x, 244 pages) : illustrations
DescriptionMany U.S. cities have experienced decades of population loss between 1950 and 1980. Even today, some of them continue to hemorrhage. But neither did all cities that had lost population continue to shed it nor make recovery. Instead, since then, these affected cities have rather diverged in terms of their post-decline population trajectories (Rappaport, 2003; Glaeser & Shapiro, 2003; Voith & Wachter, 2009). Why, then, did cities that suffered population loss exhibit different, and even contrasting, pathways? What explains the divergence and recovery of cities that lose population? It is the central question that this dissertation aims to tackle in the end. In this dissertation, I look at the universe of 223 formerly large U.S. cities (that ranked the 100 largest cities in each decade from 1790 to 2010), and examines i) the overall over-time pattern of urban population loss; ii) notable variation that appears in their post-decline trajectories, and, iii) the reasons for their divergence and turnaround. Building upon a few seminal works on these topics, especially Hill et al. (2012), the current study made progress in two important ways.First, I apply a more scientific approach to typologies of shrinking cities. Unlike previous studies that classify cities by relying on a subjectively prescribed set of rules and somewhat arbitrary decision making (e.g., Turok & Mykhnenko, 2007; Hill et al., 2012), I experiment with a novel method that relies on statistical similarity: Group-based trajectory modeling (GBTM), which I apply for the first time to urban studies. Through this experiment, I am able to delineate two groups of cities: “recovering” cities and “nosedive” cities – both suffered population loss between 1950 to 1980, but distinctively different in their later pathways, i.e., one that grew after 1980 and the other that did not. Second, I investigate factors related to the recovery and turnaround of losing cities through a set of rigorous statistical analyses, along with new variables that are not tested in the literature. In doing so, I go one step further Hill et al. (2012) that only test statistical mean difference across his typologies. To be specific, I use OLS regression, OLS regression with two-stage Heckman correction, and discrete choice model with the two trajectory groups (“nosedive” and “recovering” cities) that I extracted via GBTM. With these models, I first identify the characteristics of cities that might help faster growth in general, and further examine whether those factors also help enable struggling U.S. cities to make recovery and turnaround in particular (compared to those that continue to shed population). My prime finding is the importance of a degree-intensive, high-paying job base (DIHP) on faster growth and recovery. This is something new to the field of study as they have almost always focused deeply on two variables – human capital and diversity. I also found that, unlike the mainstream view, human capital was not a factor that enables shrinking cities to recover and turn around, albeit it predicts faster growth. This may imply that building the DIHP sector might be also important, perhaps even more so.
NotePh.D.
NoteIncludes bibliographical references
Genretheses
LanguageEnglish
CollectionSchool of Graduate Studies Electronic Theses and Dissertations
Organization NameRutgers, The State University of New Jersey
RightsThe author owns the copyright to this work.
Statistical Profile