RDP 2006-03: Australian House Prices: A Comparison of Hedonic and Repeat-sales Measures Appendix B: Data Preparation

The data used in this study were provided by Australian Property Monitors (APM) and the Real Estate Institute of Victoria (REIV). APM provided house sales data for Sydney, Melbourne and Brisbane from the March quarter 1993 to the September quarter 2005 (REIV provided data for Melbourne). The data were separated into their respective cities and all observations satisfying the following criteria were removed:

• no valid contract date (including missing observations and observations outside of the sample period);
• undisclosed price or an inconsistency in the price recorded (as noted by APM or REIV);
• outside the statistical division of each respective city;
• missing postcode;
• negative or zero sale prices;
• property types other than a cottage, house, semi-detached, terrace, townhouse or villa;
• duplicate observations, in terms of all house characteristics, the date of sale and price; and
• properties in suburbs outside of the metropolitan area (as defined by APM).

A small number of observations were also removed when comparing alternative hedonic specifications on sub-samples of the data (Table 2). Specifically, records with a land size in the top and bottom 1 percentile of the data for each city were excluded from estimation. Also, one record with 91 bedrooms was omitted for Sydney. These observations were omitted to reduce the influence of outliers and ensure plausible estimates of the implicit price relativities.