Authors: Alicia Sabatino*, UMBC, Dillon Mahmoudi, UMBC
Topics: Economic Geography, Urban Geography, Geographic Information Science and Systems
Keywords: Market Value Analysis, urban planning, algorithmic governance, capitalism, racial justice
Session Type: Virtual Paper
Start / End Time: 8:00 AM / 9:15 AM
Room: Virtual 32
Presentation File: No File Uploaded
Classification algorithms designed to guide “community development” often entrench the racist spatial organization of people in cities around the United States. The historical redlining classification of neighborhoods, notorious for reproducing a pattern of wealthy white neighborhoods and impoverished Black neighborhoods, has been examined through various social science lenses. Yet, this research fails to examine how processes of spatial ordering are recreated through new classification algorithms. One such algorithm, developed by The Reinvestment Fund (TRF) provides data to client municipalities to direct investment strategies which “strengthen communities”. The algorithm “objectively” ranks neighborhoods according to the health of their housing market and associated factors, ignoring the racist and patriarchal algorithms which previously ordered urban space. We argue the TRF’s algorithm is used to rationalize the racist and patriarchal ordering of the city. We examine 9 cities that have both Home Owners Loan Corporation classifications and TRF classifications alongside TRF’s classification of Baltimore over a 12 year period to demonstrate: (1) how redlining is reproduced with algorithms, and (2) how TRF reproduces spatial segregation and prejudice. We consider these outcomes in relation to the persistence of segregation by race/ethnicity, median household income, and city investment. Following recent literature on computational praxis for social justice, we argue spatial algorithms, presented as “objective,” are susceptible to reproducing hierarchies of difference. Computational praxis toward social justice must excavate the historical production of hierarchies of difference to combat the entrenchment of technopolitical power structures, necessitating social science researchers to produce explicitly anti-racist and anti-patriarchal algorithms.