Estimating Preferences for Neighborhood Amenities Under Imperfect Information

We introduce a generalized neighborhood choice model that allows for heterogeneity in knowledge of local amenities. In reality, people often make decisions about where to live without complete information, which can influence their choices and distort the estimated value of amenities. To mitigate this bias, we construct a latent quality index using panel data from a neighborhood choice program that provided information on rents and same-school networks to graduating students. Our analysis shows that individuals tend to switch to neighborhoods with larger networks and lower rents, and these effects persist even after graduation, influencing their actual residential choices. Our marginal willingness-to-pay estimates indicate that living in a neighborhood with a larger network is worth an additional $123 per month, and not accounting for endogeneity could overestimate this amount by as much as 70%.