Date of Award

Spring 2021

Degree Type




Director of Thesis

Conor Harrison, PhD

Second Reader

Zhenlong Li, PhD


The purpose of this thesis is to describe and analyze the South Carolinian eviction crisis from the perspective of radical geography. South Carolina was chosen for the severity of its crisis and the lack of research at a sub-state level. Court records of eviction filings from 2019 were geocoded and tested for spatial clustering, which was clearly visible. Plaintiff names were used to identify the most frequent filers and distinguish landlords by type. At the census tract level, eviction filing counts were compared with neighborhood characteristics using negative binomial regression, and most were found to be significant in South Carolina. To better capture spatial variation in how eviction filings may be best explained, the paper introduces Geographically Weighted Regression (GWR) to the field of eviction research. This novel approach is shown to be useful at identifying the interactions between eviction and localized housing markets, although it was not established as statistically stronger that linear regression. Finally, this report urges a reorientation of eviction research towards the application of its findings.

First Page


Last Page