Do sales prices overstate underlying house prices in market downturns? Evidence from the Canadian house price crash of 1991

Canadian Journal of Administrative Sciences, Dec 2002 by Marion Steele, Richard Goy

10 The Appraisal Institute of Canada (AIC) recognizes this definition of value. The Appraisal Foundation (2002) also gives the definition of the International Valuation Standards 2000: "Market value is the estimated amount for which a property should exchange on the date of valuation between a willing buyer and a willing seller in an armslength transaction after proper marketing wherein the parties had each acted knowledgeably, prudently and without compulsion" (2002, p. 47).

11 Goetzmann (1996) refers to this as the "unconditional price" and the price of houses actually sold as the "conditional" price.

12 Of course, this means we assume that the stochastic term in the reservation price equation is not iid. In our assumptions we part company with Gatzlaff and Haurin (1998). They assume that the means of the stochastic terms in both the reservation price and the offer price equations are zero, and so implicitly define the true price differently than we do. Our specification is more consistent with Fu's (1996) theory than it is with Gatzlaff and Haurin's.

13 Specifically, records without at least one private flush toilet (i.e. with no flush or chemical toilet inside dwelling, with a "chemical or other type" toilet, or with toilet shared with another household) were deleted. Other exclusions are records without at least one private bathroom with installed shower or bathtub (i.e. with no such bathroom or with bathroom shared with another household); with no piped hot water; with fuel for piped hot water bottled gas (including propane), wood, or other; with principal heating equipment heating stove or cook stove; with principal heating fuel propane, wood, or other.

14 Strictly, these dummies in terms of HIFE categories are as follows: one bathroom (with installed baths or shower), two toilets; two or more bathrooms, two toilets; one bathroom, three or more toilets; two or more bathrooms, three or more toilets. In the text, we refer to the number of

toilets minus the number of bathrooms as either the number of half bathrooms or the number of washrooms.

15 The R^sup 2^ rises dramatically, to 0.693, when CMA dummy variables are included. A rationale for inclusion is the differences in land values among CMAs.

16 The estimated depreciation rate in the regression (not reported here) in which CMA dummies are included is closer, but still below, the U.S. estimates. The DiPasquale and Somerville (1995) regression does not have SMSA dummies, but it does have three regional dummies.

17 Regressions, using the sample untrimmed of observations revealed by their DFFITS to be outliers, show greater drops in 1991 in most cases than the regressions reported here.

18 This was calculated using Stan Hamilton's hedonic price estimates for Vancouver.

References

Abraham, J.M. & Shauman, W.S. (1991). New evidence on house prices from Freddie Mac repeat sales. AREUEA Journal, 19 (3), 333-352.

Appraisal Foundation (2002). Uniform standards of professional appraisal practice. Washington, DC.