Friday, June 10, 2011

Using Frontier Models to Mitigate Omitted Variable Bias in Hedonic Pricing Models: A Case Study for Air Quality in Bogotá, Colombia

Abstract: Hedonic pricing models use property value differentials to value changes in environmental quality. If unmeasured quality attributes of residential properties are correlated with an environmental quality measure of interest, conventional methods for estimating implicit prices will be biased. Because many unmeasured quality measures tend to be asymmetrically distributed across properties, it may be possible to mitigate this bias by estimating a heteroskedastic frontier regression model. This approach is demonstrated for a hedonic price function that values air quality in Bogotá, Colombia.

Coefficient estimates for the second (heteroskedastic) frontier model (HFM) for structural characteristics are of expected signs. Apartments located in neighborhoods with higher socioeconomic stratum rent for more. Apartments located at higher elevations rent for more. Apartments far away from main roads rent for more. The estimated coefficient for distance to drainage ditches is negative, but not statistically significant. Drainage ditches in Bogotá are generally unattractive, and may be viewed as a disamenity. Apartments near metropolitan (large) parks rent for more than apartments located far away. Zonal (small local) parks did not have a significant impact on rents. Apartments located closer to the central business districts (as measured by commute time) have higher rents. Apartments located in areas with higher crime rates have lower rents. Of particular interest is the estimated coefficient on PM10 (particulate pollution) was -0.0908. This was negative and significant at all conventional levels, implying that higher PM10 concentrations are a disamenity.

Some effort was made to determine which neighborhood characteristics were most important for explaining variation in the variance of the asymmetric error component. In particular, we searched for neighborhood characteristics that were correlated with air quality. We found a strong negative correlation between elevation and environmental quality (-0.5447). When elevation and its interaction with PM10 are included as variables that explain the variance of the asymmetric error component, these variables were not significant in the HFM specification.

The coefficient for Crime was -0.0039, Elevation 0.0604, the log of distance from a drainage ditch -0.0018, the log of distance from the main road 0.0232 and the log of commuting time was -0.0643.

by Fernando Carriazo, Richard Ready and James Shortle

Universidad de los Andes–Facultad de Economía–Cede
Document 2011-1; 2011

No comments:

Post a Comment