188.8.131.52.1. Hedonic: Air
Urban Air Pollution Damage Functions: Theory and Measurement (EE0206) by Crocker is a 1971 property value study. It covers 1,288 residential property transactions in Chicago from 1964 through 1967. In addition to valuation data, data on suspended particulates and sulfur dioxide were available for each of the 48 months by location as were data on the structural, site, purchaser, sale and neighborhood characteristics of each home. The investigator found roughly a $450 value diminution for each 10% increase in pollution. It was also found that land values were more sensitive to pollution than land plus improvements values. “The hypothesis that residential property values decline at a decreasing rate with respect to increasing air pollution dosages” could not be rejected.
The 1979 study Methods Development for Assessing Air Pollution Control Benefits Volume II, Experiments in Valuing Non-market Goods: Benefit Measures of Air Pollution Control in the South Coast Air Basin of Southern California (EE-0271B) by Brookshire, d’Arge, Schulze and Thayer is an early and ambitious attempt to value clean air. It uses two methods in parallel: hedonic analysis and contingent valuation. For the hedonic analysis, the investigators identify six neighborhood pairs. Within each pair, the neighborhoods were similar geographically, demographically and in terms of housing but one was more polluted than the other. Hedonic regressions were run to determine the housing price differential associated with clean air. Both techniques found a willingness to pay or value for cleaner air. Very crudely, households were willing to pay $135 per month in the hedonic study and $26 per month (1978$) for 30% cleaner air. EE-0067 is an overview of the subject document.
A sequel to the Los Angeles study was done in 1985. It is Methods Development for Environmental Control Benefits Assessment, Vol IV, Measuring the Benefits of Air Quality Improvements in the San Francisco Bay Area (EE-0278D) by Loehman, Boldt and Chaikin. This study reports the results of using two alternative methods of measuring benefits of air quality changes in the San Francisco Bay Area; both methods are based on the hypothesis that benefits are measured by willingness to pay. In one method, property values were used to infer willingness to pay for air quality. In the other method, willingness to pay was obtained from a contingent value survey.
Pollution was shown to have a statistically significant negative effect on property values. Willingness to pay obtained from the survey was shown to be correlated with visibility and health effects (as defined by the PSI index). Both methods resulted in an estimate of average benefits per household of about $80 annually for a 30 percent improvement in air quality; this represents an annual total benefit of about $130 million for a 30 percent improvement for the Bay Area (using 1977-1978 air quality conditions as a base and 1978 socioeconomic conditions and property values).
In 1985, Horowitz produced Inferring Willingness to Pay for Housing Amenities from Residential Property Values (EE-0005). He built a hedonic model of house prices in Baltimore. He begins by noting that the housing market does not necessarily assign a house to the potential buyer valuing it most highly for three reasons all related to uncertainty. First, house prices “are established through sequential bidding, so sellers do not necessarily accept the highest bids that might be made .... Second, ignorance ... may lead potential buyers to make bids that are either higher or lower than the ones they would make if they had complete information. Finally, the distribution of bids for a house is truncated at the seller’s asking price.” The paper presents an alternative hedonic model that addresses these considerations.
In a test against a traditional model that ignores uncertainty, the alternative model does much better at predicting observed values. Among its arguments are three that describe air quality. These turn out in most regressions to have the “wrong” or unexpected sign and are insignificant.
Improving Accuracy and Reducing Costs of Environmental Benefit Assessments Vol VI Improving the Accuracy of Hedonic Price Methods: Econometric Analysis of Existing Data Sets (EE0325) by Graves, Murdoch, Thayer and Waldman (1985) is a hedonic property value study of the benefits of air pollution control. It is distinguished from numerous other such studies by the attempt to produce robust econometric results. That is, the investigators are mindful of the criticism occasionally leveled at statistical exercises that the result reflects researcher bias and could be otherwise if another researcher of different bias had performed the study. To address the robustness problem, the investigators address questions of variable selection, measurement error, functional form and estimation technique.
The NSF/EPA-funded study from 1997, “Improving Air Quality Benefit Estimates from Hedonic Methods” by Thayer, Murdoch and Beron will conduct a hedonic property value investigation using a single, pooled cross-section, time-series data set. The primary focus will be on the hedonic price of air in the South Coast Air Basin of California for the period 1980-1996. A set of case studies will become the benchmark for future benefits transfer work.