Three essays on urban location choice and urban growth

This dissertation encompasses three papers. My first paper contributes to the larger literature on the effect of individual-level characteristics on urban location choice by examining whether young people aged 25--34 with a bachelor's degree or higher are more likely to live in central cities in 2011 than in 1990. In 1990 37% of 25--34 year olds (Baby Boomers) living within a metropolitan statistical area (MSA) lived in a central city. By 2011 the percentage of young people (Millennials) living in a MSA that lived in a central city had declined to 33%. However, when 25--34 year olds are segmented by education it is clear that this decline was driven by young people with less than a bachelor's degree. Conditional on living in a MSA the percentage of young people with a bachelor's or advanced degree that lived in a central city was approximately 36% in both 1990 and 2011. When I control for individual-level characteristics I find that the effect of education on the probability of living in a central city remains similar in both generations. I estimate that having a bachelor's degree increases the probability that a 25--34 year old will locate in a central city by 8.3% in 1990 and 8.2% in 2011. The increases in the probability of living in a central city from having a master's degree or a doctorate in 2011 are also similar in magnitude to their counterparts in 1990. This is evidence that to the extent education plays a role in the larger population of high human capital 25--34 year olds in cities it is due to a composition effect rather than cities becoming more attractive to educated people at the margin. While educated young people are not more attracted to cities across generations there have been some intertemporal regional changes. I also analyze individual cities in each region to demonstrate that the regional changes obscure city level heterogeneity. I find that in Cleveland, Chicago, New York and Portland the effect of a bachelor's degree on living in the central city of those MSAs increased from 1990 to 2011. In the Houston MSA the effect of a bachelor's decreased and in the Los Angeles and Atlanta MSAs the effect of a master's decreased.;In my second paper I use 2011 IPUMS data to estimate the effect of education on living in a central city for various age groups, with a focus on the 25--34 year old age group. Consistent with other studies I find that the effect of education on living in a central city declines with age but that this decline is not monotonic. For example, relative to a high school graduate a bachelor's degree increases the probability of living in a central city for 25--34 year olds by 8%, has no significant effect on 35--44 year olds, and increases the probability by 4% for people over age 64. When I separate the 25--34 year old age group into sub-populations several interesting results emerge. First, the effect that education has on living in a central city varies by metropolitan statistical area (MSA). In MSAs that contain cities that experienced a relatively large increase in their population of 25--34 year olds from 2005--2011 the effect of a bachelor's or advanced degree is positive. In MSAs that contain cities in which that age group grew more slowly or declined the effect of a bachelor's or advanced degree is not statistically significant. This means that cities that experienced a larger increase in their population of 25--34 year olds from 2005--2011 were more attractive to the educated members of that age group. Second, the positive effect that a bachelor's degree has on living in a central city can largely be attributed to white 25--34 year olds. I estimate that a bachelor's degree increases the probability that a white 25--34 year old will locate in a central city by 11% compared to that of a high school graduate, while a bachelor's degree has no effect on the probability that a black 25--34 year old will locate in a central city. This difference is robust to specific MSAs. There are also differences by gender; relative to high school graduates 25--34 year old males with a bachelor's or master's degree are more likely to locate in a central city than similar females.;My final paper examines the effect of state government spending on city population growth. State government spending as a percentage of gross state product (GSP) has been increasing over the last 40 years. In 1970 state government spending as a percentage of GSP averaged 13.8% across all 50 states. By 2000 it had risen to an average of 16.9% and by 2012 it had further increased to an average of 19.1%. As state government spending increases relative to GSP it crowds out private investment, decreasing employment opportunities in other parts of the state while simultaneously increasing them in the state's capital where most of the government offices are located. As state spending increases and resources become increasingly concentrated in capital cities the demand for labor will increase in MSAs that contain capital cities relative to other MSAs in the state. This demand increase for labor will affect the population distribution of a state. Using data from IPUMS I find that conditional on being a capital city an increase in state spending increases a city's population, though the effect is imprecisely measured. Additional data at both the MSA and year level will allow me to more precisely measure the effect and examine whether it changes over time. When I sub-sample the data I find that during the three decades from 1980--2010 state government spending negatively impacted the population growth of non-capital cities.