Saturday, September 1, 2012

Water Quality Index Aggregation and Cost Benefit Analysis Abstract: The water quality index (WQI) has emerged as a central way to convey water quality information to policy makers and the general public and is regularly used in US EPA regulatory impact analysis. It is a compound indicator that aggregates information from several water quality parameters. Several recent studies have criticized the aggregation function of the EPA WQI, arguing that it suffers from “eclipsing” and other problems. Although past papers have compared various aggregation functions in the WQI (usually looking at correlation), this is the first paper to examine these functions in the context of benefit-cost analysis. Using data from the 2003 EPA CAFO (Concentrated Animal Feeding Operation) rule, the present paper examines four aggregation functions and their impact on estimated benefits. Results indicate that the aggregation method can have a profound effect on benefits, with total benefit estimates varying from $82 million to $504 million dollars. Furthermore, a sensitivity analysis does not find convincing evidence to substitute the current aggregation function, although several changes to the underlying WQI methodology may be warranted.
From the CAFO data, it is clear that estimated benefits are quite sensitive to the subindex aggregation function. Over the four different functions, benefits range from $82 million to $504 million. The geometric mean, which is used in EPA RIAs, sits near the middle of that range at $287 million. Although these monetized benefits need to be added to other monetized benefits, such as the value of reduced nitrification of private wells ($30.9 – $45.7 million), reduced public water treatment costs ($1.1 - $1.7 million), and reduced livestock mortality ($5.3 million), they represent the lion’s share of monetized benefits. Since the total social costs of the rule were estimated to be $335 million (US EPA, 2003b), the choice of the aggregation function could move the rule from positive monetized net benefits to negative. Policy recommendations from the benefit-cost analysis could vary drastically depending on the aggregation function used.

The sensitivity analysis did not support a switch from the geometric mean. With the geometric WQI, the importance of individual parameters to estimated benefits is a good reflection of the weights provided by a panel of hydrology experts. Also, the geometric mean does not inflate high valued indicators or eclipse the most impaired indicator as much as the other aggregation functions. The harmonic and minimum functions were found to be extremely sensitive to the most impaired variable, while the arithmetic mean was subject to eclipsing and is dependent on the level of water quality.

Although this paper does not support a move from the geometric mean, the results highlight another pressing issue: an updating of the WQI weights and subindex curves. The weights had a relatively strong influence on estimated benefits in the sensitivity analysis. Since the weights presently used by the EPA are based on a survey from the 1970’s (McClelland, 1974); an update may be in order. The biology, ecology, and limnology underlying water quality analysis have all improved in the last 40 years and expert opinion has likely evolved as well. Furthermore, most state and national water quality criteria have become more refined to different uses and there are now additional criteria for different pollutants.

Some of the criticism of the current WQI may be assuaged by developing new weights and subindex curves. A regional approach to the subindex curves, popularized by Cude (2001), represents a promising future path. That approach has already been used in the Construction and Development (2009) and Florida Numeric Nutrients (2010) rules, and has been met with widespread approval.

by Patrick J. Walsh and William Wheeler
National Center for Environmental Economics (NCEE)
Working Paper Number: 2012-05; Document Date: 07/06/2012
Subject Areas: Water Pollution; Economic Damages/Benefits; Benefit-Cost Analysis
Keywords: water quality; valuation; cost-benefit analysis

No comments:

Post a Comment