2001 V29 1: pp. 55–84 REAL ESTATE ECONOMICS A Metropolitan Transaction-Based Commercial Price Index: A Time-Varying Parameter Approach Henry J. Munneke ∗ and Barrett A. Slade ∗∗ This study examines the usefulness of time-varying parameter techniques for constructing reliable transaction-based commercial price indices for metropoli- tan areas. Time-varying parameter techniques allow the implicit prices of differ- ing quality characteristics to vary intertemporally, overcoming the potential bias imposed by holding implicit prices fixed and simply interpreting time dummy variables as in a conventional hedonic approach. This paper empirically in- vestigates three time-varying parameter methods (Chained, Laspeyres, and Paasche) and considers the potential for sample selection bias. Precision mea- sures are constructed to examine the reliability of the respective indices. Price changes for heterogeneous goods often result from changes in quality characteristics; therefore, estimating the intertemporal pure price change re- quires that the differing quality characteristics be held constant. Since the early study of land pricing by Haas (1922) and the often cited study of automobile prices by Court (1939), numerous studies have attempted to capture intertempo- ral pure price changes by regressing quality-related variables and time dummy variables on price. 1 This traditional hedonic time dummy variable approach assumes the implicit prices on the quality variables are constant over the time period, allowing the parameters on the time dummy variables to capture the pure price change. However, if the implicit prices are not structurally stable, the resulting index will not reflect the pure price change. This study seeks to overcome this form of potential bias by employing varying parameter techniques: a chained hedonic price index, a Laspeyres hedonic price index, and a Paasche hedonic price index. These procedures allow the implicit prices to vary over time but also capture the pure price change. The chained ∗ Terry College of Business, University of Georgia, Athens, GA 30602-6255 or hmunneke@terry.uga.edu. ∗∗ Marriott School, Brigham Young University, Provo, Utah 84602 or bslade@byu.edu. 1 For an interesting discussion of Haas and Court and the early history of hedonic price analysis, see Colwell and Dilmore (1999) and Goodman (1998).