Knektid.com

The Hedonic Price Model in Housing Valuation

December 14, 2008

In an article appearing in the Southern Economic Journal in 2002, Ted Gayer, James Hamilton, and Kip Viscusi used the hedonic price model to examine housing price fluxuations that resulted from environmental information and associated perceived cancer risk. Specifically, the authors used housing price changes which occurred “after the release of a regulatory agency’s environmental risk information to estimate the value people place on cancer risk reduction” (Gayer, Hamilton, & Viscusi, 2002, p. 1).

This study used a sample of 16,928 houses which were sold more than once in the “Greater Grand Rapids area from 1988 to 1993″ (p. 11). The decision to limit the sample to only dwellings with multiple sales records was made in an effort to “focus on changes in risk over time” and “estimate whether price effects vary over time as risk beliefs change”. “Environmental risk information” was primarily concentrated on data releases which pertained to nearby Superfund sites. Prior knowledge regarding Superfund sites among residents was accounted for in the study and used as a baseline to measure changing risk perceptions.

The authors were able to correlate newspaper coverage of positive environmental risk assessments with a rise in home values in the range of $56 to $87 per dwelling in response to each article published. These figures implied a “statistical value per case of cancer of $4.3 million to $8.3 million”, which was consistent with labor study analyses of the “value of a statistical life”. Newspaper coverage of the decreased cancer risk, it was therefore determined, had contributed to the residents’ positive changes in perception.

The authors noted that their results differed from previous studies in which residents had “either alarmist reactions or no reaction at all to risk information” (p. 22). Specifically, two related studies published in 1980 and 1985, respectively, were deemed to have results contrary to those of this study. Nevertheless, the authors concluded that residents do exhibit a tendency to respond to “expert risk information” provided by the EPA, which results in a corresponding change in housing prices.

In another study of housing prices which used the hedonic price model, David Brasington sought to explore “which measures of public school quality the housing market values” (Brasington, 1999, p. 1). In particular, the author attempted to determine which of 37 traditional measures of school quality had the greatest influence on housing prices. In doing so, Brasington hypothesized, it may be possible to convey the particular school data that should be used in future housing price assessments.

The author used “housing transactions from the major metropolitan areas of the state of Ohio” and “a variety of school quality measures” to conduct his study. These quality measures included, but were not limited to: proficiency test passage rates, expenditures per pupil, teacher/student ratio, teacher salary, student attendance rates, value-added measures, graduation rate, teacher experience levels, and teacher education levels.

Brasington determined that the housing market is most significantly affected by those measures of school quality which are “readily available to homebuyers”, including “test passage rates, expenditures per pupil, and a low teacher/pupil ratio” (p. 15). Average teacher salary and the student attendance rate were found to have a marginal effect on housing prices, while value-added measures, teacher education levels, teacher experience levels, and graduation rate had either little bearing on housing prices or were determine to be unreliable.

The author’s findings concerning housing values were consistent with similar studies in which proficiency test passage rates were deemed to have the most pronounced affect on housing prices. This study refuted other findings, however, that identified value-added measures to be a primary determinant of housing price variation. The author attributed this distinction to the fact that this study used a large sample size and significantly more measures of school quality than previous studies.

Stuart Gabriel also used the hedonic price model in his 1987 study which sought to quantify the economic effects of racial integration on housing prices. The study defined the racial attribute “in terms of percentage of minorities that live in the immediate neighborhood” and attempted to determine an economic bias in response to this ratio (Gabriel, 1987).

The basis of this study’s information was “census tract data for the city of Oakland, California, for 1977″ (p. 4). This geographic location was chosen primarily because of its significant minority population and its role in previous studies of racial discrimination. The area included 73 census tracts and obtained its “housing information, including sales prices and structural characteristics” from the Society of Real Estate Appraisers.

Results from this study indicate a pronounced negative economic effect following recent neighborhood racial integration. Additionally, the study’s author determined a “continued household willingness to pay for diminished minority neighborhood presence” (p. 10). These results are consistent with previous studies, but indicated that those previous analyses may have actually understated the economic effects of racial integration. In light of these findings, the author suggests modifying public policy to reflect the “limited efficacy” of “open housing laws and the like”.

Works Cited

Brasington, D. M. (1999). Which Measures of School Quality Does the Housing Market Value? Journal of Real Estate Research , 395-413.

Gabriel, S. A. (1987). Economic Effects of Racial Integration: An Analysis of Hedonic Housing Prices and the Willingness to Pay. American Real Estate and Urban Economics Association Journal , 268-279.

Gayer, T., Hamilton, J. T., & Viscusi, W. K. (2002). The Market Value of Reducing Cancer Risk: Hedonic Housing Prices with Changing Information. Southern Economic Journal , 266-289.

Filed under: Urban Economics