Authors: Emily Molfino*, U.S. Census Bureau, Shawn Bucholtz, Housing and Urban Development, Jed Kolko, Indeed
Topics: Applied Geography, Quantitative Methods, Urban Geography
Keywords: urban, suburban, rural, machine learning, american housing survey, small area estimation
Session Type: Paper
Presentation File: No File Uploaded
Definitions of urban and rural, and to a lesser extent, suburban, are abundant in government, academic literature, and data-driven journalism. Equally abundant are debates about what is urban, suburban, or rural. Absent from most of this discussion is evidence about how people perceive their environments. In 2017, the U.S. Department of Housing and Urban Development added a simple question to the 2017 American Housing Survey (AHS) asking respondents to classify their neighborhood as urban, suburban, or rural. Recent analysis of the data reveal that more than half of Americans describe their neighborhood as suburban. This paper describes how the AHS data was used to create the Urbanization Perceptions Small Area Index (UPSAI). To create the UPSAI, we first applied machine learning techniques to the AHS neighborhood description question to build a model that predicts how out-of-sample households would describe their neighborhood (urban, suburban, or rural), given characteristics of both the neighborhood and the household. We then applied the model to the American Community Survey (ACS) household-level microdata, thereby creating a predicted likelihood a household would describe their neighborhood as urban, suburban, and rural. Finally, we aggregated the household predictions to small areas (Census tracts), permitting us to classify each Census tract as urban, suburban, or rural. These later two steps are commonly referred to as small area estimation. Our approach provides an illustrative example of the use of existing federal data to create innovative new data products of substantial interest to researchers and policy makers alike.
To access contact information login