WEAI/AERE 2012 - Individual Paper Abstract
Title: Valuing Water Quality in an Urban Watershed
Author(s): Noelwah NETUSIL,Department of Economics, Reed College, 3203 SE Woodstock Blvd, Portland, OR 97202-8199, USA, 503-517-7306, 503-777-7771, netusil at reed dot edu; Michael Kincaid, Reed College; Heejun Chang, Portland State University [Photo credit: Noelwah Netusil; Burnt Bridge Creek, Vancouver, WA]
Abstract:
Urban runoff has been identified as one of the leading sources of water quality impairment of rivers, streams, lakes and estuaries in the United States (USEPA 2009). Typical pollutants found in urban runoff include sediment, pesticides, nutrients, bacteria, and heavy metals. Thermal pollution is also a common problem because urban watersheds have a high concentration of impervious surface area and urban streams often have little or no streamside vegetation (Paul and Meyer 2001).
Despite the importance of urban streams for stormwater runoff, fish and wildlife habitat, and recreation, literature on the relationship between urban stream water quality and the sale price of single-family residential properties is limited. Leggett and Bockstael (2000) estimate that increasing fecal coliform counts by 100 per 100mL decreases the sale price of waterfront properties in Anne Arundel County, Maryland by 1.5%. Poor, Pessagno, and Paul (2007) estimate a decline in sale price of 0.53% from increasing total suspended solids (mg/L) and a decline of 8.78% from increasing dissolved inorganic nitrogen (mg/L) in Maryland's St. Mary's River watershed. River, estuary and lagoon water quality is found to affect the sale price of waterfront properties in Marin County, Florida with declines in sale price ranging from 0.56% for water visibility to 2.55% for salinity (Czajkowski and Bin 2010). Our analysis combines a data set of single-family residential property sales from 2005-2007 in the Johnson Creek Watershed with five water quality parameters measured at eight monitoring sites. These parameters include lead and e-coli, which are chosen to reflect concerns about the impact of water quality on human health, pH and dissolved oxygen, which influence the health of fish and wildlife populations, and total suspended solids to capture water clarity and aesthetics.
Our first model examines if the estimated coefficients on the water quality variables vary based on a property's distance to Johnson Creek. We hypothesize that, after controlling for other important factors, distance effects will vary by creek section, for example, a property located within 1/4 mile of the creek near its headwaters may experience a different effect on sale price than a property within 1/4 mile of the creek near its confluence with the Willamette River. Water quality varies dramatically in the study area by season, so our second model compares estimated effects on sale price when water quality is modeled using an average over the year a property was sold with variables that reflect the "wet" season, defined as November-April, and the "dry" season, May-October. We hypothesize that water quality in the dry season, when water quality standards are frequently violated, will be the best modeling approach.
An increasing emphasis by federal, state and local agencies to improve urban water quality means that accurate estimates are important for benefit transfer. Our work contributes to the limited research in this field by focusing on a highly urbanized study area, expanding the number of water quality variables included in our models, modeling water quality variables to reflect the two distinct seasons in the study area, including waterfront and non-waterfront properties, and evaluating the potential for omitted variable bias from excluding land cover information.