How informative are average effects? Hedonic regression and amenity capitalization in complex urban housing markets

Variations in real estate prices within an urban area are commonly exploited to value local amenities. Be they public or private goods, local amenities should be capitalized into property prices. This paper, however, demonstrates that standard hedonic models used to recover implicit prices are highly sensitive to sample choice and to model specification. In the case of Los Angeles, complex local housing markets produce attribute prices that vary spatially and temporally, violating the common assumption they are fixed. This misspecification yields estimated average effects of the value of access to light rail stations that vary widely across seemingly innocuous choices regarding samples and model specification. The paper proposes an alternative, more flexible approach, which yields a finding of no capitalization of light rail access into surrounding home prices.