Measuring cities' internal demographic change as the movement of emergent boundaries

Abstract

Do neighborhoods “spread”?

American cities are often highly segregated, divided into large regions of blocks

with similar ethnic and racial composition with sharp boundaries between them.

While most neighborhood research uses fixed boundaries, mostly Census tracts,

I argue that boundaries are not predetermined, but emerge endogenously and can

move over time. In this dissertation, I ask to what extent, over the last decade,

has population change occured by these boundaries moving in space, and to what

extent by the overall internal mix of these regions changing? Neighborhood change

that occurs by within-cluster change would be large-scale, gradual, and diffuse,

whereas change by boundary movements would be sharp, occur to specific blocks,

and perhaps continue to segregate residents’ use of space.

In Chapter 1, I develop a Bayesian algorithm, the Space-Time Chinese Restaurant

Process (Space-Time CRP), to identify these spatial clusters of households’ race

and ethnicity from block-level Census data, with boundaries that can move over

time. I examine the changes in those ethnoracial clusters in Philadelphia, PA from

2000 to 2010, and decompose Philadelphia’s ethnoracial change to find a previously

unmeasured dynamic: more-White and Asian clusters are growing spatially, even

as all clusters are internally becoming more Black and more Hispanic.

In Chapter 2, I replicate the analysis for the 100 largest U.S. central cities. In most

of the cities, the overall trend was one of internal changes, led by increasing proportions

Hispanic and Asian. However, I find that population change by boundary

movement was prevalent in denser cities, and usually by Whiter clusters expanding

into Blacker clusters. While all cities experienced gradual increases in non-White populations within clusters, the gentrification of dense cities occured by the sharper

change of Whiter clusters expanding.

In Chapter 3, I measure the change in number and type of businesses as a boundary

moves. Businesses are a compelling symbol and perhaps agent of neighborhood

change. I find that business change is typically a lagging indicator around gentrifying

boundaries, with businesses changing only after the demographic change has

occured. In neighborhoods experiencing increasing non-White populations, businesses

change in advance of the boundary’s approach.