Urban environmental evaluation using an affiliated private value auction model

In this study, we evaluate a subsidy program that reduces urban fire risk. Using apartment auction data from Osaka, Japan, we employ econometric models of first-price sealed-bid auctions to assess the program’s benefits. The data include the attributes of the auctioned apartments, reserve prices, and number of observed bids as well as the highest, second-highest, and third-highest bids. In our analysis, we use the affiliated private value (APV) model to infer the bidder’s valuation. We model affiliation using the Archimedean copula and parametric marginals. To estimate the parameters of the copulas and parametric marginals, we use moment conditions that are satisfied by observed and equilibrium bids when our model specification is correct. We apply the method of simulated moments to the moment conditions. Overidentification and goodness-of-fit tests reject the independent private value (IPV) models, although some APV models are not rejected. Our hypothetical policy evaluation suggests that the subsidy program has a reasonable effect. We note that IPV models under- or overestimate the benefit of the policy when the reduction in fire risk is not only non-marginal but also marginal.