Quality & Time-on-the-Market in Residential Markets

See Table 3, page 26, Haurin et al. 2006. Haurin’s method for calculation in the Appendix, Part 3. Creation of the Atypicality Measure: An Example of a Hedonic House Price; is as follows: Haurin, et.al, states: “The atypicality measure (A) for the i-th property in the j-th jurisdiction is defined as:

where the h _kij are the characteristics of a property, \( {\overline h_{kj}} \)is the mean value of the k-_th characteristic in the j-_th suburb, and the p _k are the implicit prices of the property’s characteristics. Thus, the atypicality variable measures the dollar value of the absolute value of the deviations from the submarket mean of a property’s characteristics.”

We also estimate a probit model to test whether the property quality has a significant influence on the probability of a sale. We do not include the results in the current paper, but they are available from the authors upon request. According to Taylor (1999), the quality affects the possibility of a sale depending on the interaction between the option-value effect and the signal-dampening effect under various information structures. The probability of being sold is regressed on variables described by X and under market conditions described by M as: \( Pi = Xi\alpha X + Mi\alpha M \). Where, X & M are the set of variables described in the text for models 1 and 2. The rationale for this model is to examine whether quality impacts the probability of a sale occurring. The results support this hypothesis. High quality properties are more likely to sell. Whereas, lower quality properties, properties with more time on the market and properties that are overpriced are associated with a lower probability of a sale. These results are consistent with Taylor’s argument that properties with more time on the market and lower quality are less likely to obtain a match.

Haurin’s et al. (2006) atypicality estimate for their full sample is 66.83 (reported in thousands) compared to our estimate of 61.96 (reported in thousands). See footnote 1 above.