Panel: Using Big Data to Identify, Build, and Preserve Affordable Housing
(Housing, Community Development, and Urban Policy)

Friday, November 8, 2019: 10:15 AM-11:45 AM
I.M Pei Tower: Terrace Level, Columbine (Sheraton Denver Downtown)

*Names in bold indicate Presenter

Organizer:  Christian Redfearn, University of Southern California
Panel Chair:  Brian Y. An, University of Tennessee
Discussants:  Jennifer Newcomer, Gary Community Investments, Shift Research Lab and Robert Collinson, University of Notre Dame

This panel investigates the real causes and effects of the housing affordability crisis at the resident and property levels. All across the United States, renters are threatened with cost burdens, eviction, expiring affordability, and landlord harassment. Until recently, researchers have not been able to measure these challenges. The four analyses in this panel unpack exciting new "Big Data" sets with millions of records apiece, allowing researchers to identify the concentration of ownership, the propensity of tenant rights violations, the risk of eviction, and the variation of cost burdens across the city, neighborhood, and income distributions. In order to harness the full power of these data, the researchers employ sophisticated statistical methods, from machine learning to cluster analysis to Bayesian hierarchical modeling. Their presentations will illuminate not only critical levers for policymakers to address the needs of local residents but also methodological advances that can extend across sectors to extract the full value of Big Data for the broad future of public policy.


Using Big Data and Social Media to Understand Neighborhood Conditions
Constantine E. Kontokosta, New York University, Lance Freeman, Columbia University and Yuan Lai, Massachusetts Institute of Technology



Using Machine Learning to Target Assistance: Identifying Tenants at Risk of Landlord Harassment
Rebecca Johnson1, Teng Ye2, Samantha Fu3, Jerica Copeny4, Bridgit Donnelly5, Joe Walsh6 and Rayid Ghani6, (1)Princeton University, (2)University of Michigan, (3)London School of Economics and Political Science, (4)Evansville Public Library, (5)City of New York, (6)University of Chicago



A Data-Driven Approach to Implementing Strategies to Preserve Naturally Occurring Affordable Housing
Brian Y. An, University of Tennessee, Andrew Jakabovics, Enterprise Community Partners, Anthony Orlando, California State Polytechnic University, Pomona and Seva Rodnyansky, University of California, Berkeley



Estimating the National Prevalence of Eviction Using Millions of Public Court Records
Ashley Gromis1, Ian Fellows2, James Hendrickson1, Lavar Edmonds1, Lillian Leung1, Adam Porton1 and Matthew Desmond1, (1)Princeton University, (2)Fellow Statistics