Pricing Neighborhood Amenities: A Proxy-Based Approach

Abstract

Understanding how housing markets price neighborhood amenities is key to unpacking residential socioeconomic disparities. Yet prior approaches to amenity pricing have suffered from the key confounder of unmeasured neighborhood quality, often leading to wrong-signed estimates. In this paper, we develop a novel proxy-based method that allows us to estimate the price of housing amenities in the midst of unobserved housing quality. Our method shows, as long as one can find at least one relevant "proxy'" variable for housing quality that satisfies conditional independence, the price of amenity can be identified as the ratio of two coefficients. Using detailed migration data, we construct an innovative measure of locational desirability–Geographic PageRank–to use as the proxy variable for quality. We show that this new approach can successfully correct the “wrong-signed” problem in the amenities valuation literature when applied to a standard measure of environmental air quality. The estimated amenity prices are key inputs to evaluating returns on investment in local public goods or environmental policies, including their role in reducing housing disparities.

Keywords: Amenities, Omitted Variable Bias, PageRank

JEL Classification: R32, C39, Q51