Estimating Preferences for Neighborhood Amenities Under Imperfect Information

This paper presents a new framework to estimate preferences for neighborhoods in the presence of individual imperfect information about every amenity in each neighborhood. We estimate the model with data from a new neighborhood choice program that provided information about market rents and same-school network, and collected neighborhood rankings for the same individual before and after receiving information. We find that switchers - who change rankings after the information intervention - increase network shares by 1.46 percentage points and decrease rents by $430. This variation from the panel data of individual rankings is critical to produce a latent quality index that addresses biases arising from imperfect information. Estimates from the neighborhood sorting model reveal a strong negative marginal utility of rents, and a positive marginal willingness to pay of $123 per month to live in a neighborhood with a larger network. Finally, information also influenced residential choices after graduation.