Housing Prices in Tucson, Arizona

The hedonic approach is used to analyze selling prices for housing units in greater Tucson, Arizona, during 1998. Using over 6,400 Multiple Listing Service transactions, ordinary least-squares regression is used to estimate implicit prices for the amount of living space, age of the unit, and other housing attributes in this rapidly growing southwestern city. Price premiums are estimated for nine different real estate districts, which realtors believe are different housing submarkets. Regression estimates are also given for the difference between original list price and selling price, thereby indicating how sellers and buyers differentially evaluated housing attributes. Nearly 3,700 transactions (from an overlapping data set) were geocoded, and these cases are examined in somewhat more detail. Spatial heterogeneity in selling prices is addressed but spatial dependency is not. The statistics of the hedonic price functions compare favorably with those for other U.S. cities.