Data Science for Housing Affordability

The Urban Data Science Workshop (CP290E) brought students together to collaborate on developing skills in data science and apply them to a common theme of housing affordability.  We developed web scraping tools to collect data from online Craigslist rental listings and shared room listings, Airbnb listings, and Redfin sales transactions, for the entire United States.  From the past few months we have now accumulated over 5 million rental listings, enabling us to analyze detailed spatial and temporal patterns in rents.  We have developed interactive national maps of these data to enable others to explore them and use them to inform their own planning and analysis.  We compared these rental listings to HUD Fair Market Rent to evaluate how more current data might impact the determination of FMR values.  We analyzed the relationships between Airbnb listings and rents within metro areas and neighborhoods to evaluate whether higher densities of Airbnb listings were correlated with higher rents in the broader market.  We analyzed accessibility and other neighborhood variables and developed predictive models of rents to better understand how local context influences rents.  Come join us for a final review to see how we leveraged data science to better inform our understanding of housing affordability, and to provide feedback on our plans to turn this into a publicly accessible platform.

RSVP here!