Valuing water quality improvements in the United States using meta-analysis: Is the glass half-full or half-empty for national policy analysis?

Abstract

The literature estimating the economic value for water quality changes has grown considerably over the last 30 years, resulting in an expanded pool of information potentially available to support national and regional policy analysis. Using 131 willingness to pay estimates from 18 studies that use a similar definition of water quality, we performed a meta-regression analysis and found mixed results. We find that WTP varies in systematic and expected ways with respect to factors such as the size of the water quality changes, average household income, and use/nonuse characteristics of respondents. As a whole, we conclude that our meta-regression results provide a reasonable basis for estimating expected WTP values for defined changes in water quality. However, despite a large number of existing economic valuation studies, relatively few could be meaningfully combined through meta-analysis due to heterogeneity in the commodities being valued in the original studies. Based on these findings, we provide recommendations for future research, including suggestions regarding more standardized approaches for defining water quality and reporting information in valuation studies.

Introduction

Protecting and improving water quality in the U.S. has been a focal point of national environmental policy for over 30 years, particularly since the passage of the Clean Water Act (CWA) in 1972. The overall goal of the CWA is to “… restore and maintain the chemical, physical, and biological integrity of the Nation's waters.” Although much has been accomplished toward achieving this goal, policy makers continue to face the challenge of setting priorities and implementing effective and efficient clean water programs. Evaluating the social welfare implications of environmental policy options – now often required by executive order or statute – has proven to be particularly challenging. However, recent efforts by federal agencies and research organizations to synthesize and integrate the various natural, physical, and social sciences used in such evaluations are beginning to show tangible progress (see, for example, NRC/NAS, 2005).

In light of these challenges, this paper provides an assessment of the existing literature on the economic value (“benefits”) of improvements in surface water quality. In particular, the purpose of this research is to determine the usefulness of the literature for supporting environmental benefit assessments and policymaking at the national level. Our approach involves two basic questions: First, can we estimate a regression function(s) that represents societal preferences for water quality contained in a collection of existing (primary) studies, which can then be used for benefit transfer? Second, in what ways can the literature be augmented or adapted in the future to best address the needs of national policy analysis?

We find that the existing body of research provides a potentially rich and diverse source of secondary data for evaluating national water policy; however, this diversity presents both opportunities and obstacles for benefit transfer. In practice, relatively few existing studies have been actively used in benefit transfers to evaluate the benefits of national water quality policies. One reason is that few of these studies are national in scope, and extrapolating values beyond the spatial boundaries of a study area inevitably adds to uncertainty.

Another specific challenge in using the existing literature for national analysis is that, across valuation studies, water quality is defined in a number of different ways.¹ In many cases it is difficult to match these definitions with those used to predict policy-related water quality changes. For example, two alternative paradigms provide a useful illustration of the potential for heterogeneity across the literature. One approach involves estimating the willingness to pay for reductions in one or more specific pollutants (e.g., Bockstael et al., 1987), which can be viewed as inputs in an ecosystem production function. The other approach involves estimating willingness to pay for enhancements to ecosystem services (e.g., suitable for swimming (Carson and Mitchell, 1993) or fish become safe to eat (Viscusi et al., 2004)), which can be viewed as outputs of an ecosystem production function.

To quantitatively evaluate results from this literature, we use meta-analysis. This approach allows us to synthesize information from selected studies in a systematic way and to test hypotheses regarding the determinants of these estimates. It also allows us to explore a format for predicting values, which is a critical component of any benefit transfer method. Although it has evolved primarily in health sciences research, meta-analysis is increasingly being applied in economics (Stanley, 2001), including nonmarket valuation research (Smith and Pattanayak, 2002). Recently, Johnston et al., 2003, Johnston et al., 2005 also conducted meta-analyses involving value estimates for surface water quality improvements. Although similar in several respects to our analysis, the two meta-analyses focus on water quality changes that specifically involve improvements in fish habitat and Johnston et al. (2003) focuses mainly on nonuse value estimates.

For our analysis, we identified and reviewed over 300 publications related to water quality valuation; however, we selected values from a subset of these studies that were sufficiently comparable for inclusion in a meta-analysis. To maintain comparability, we limited the meta-analysis to stated preference studies conducted in the U.S. and to studies that described water quality in terms that could be converted to a common 10-point scale.

Our investigation provides both good and bad news for policy analysis and benefit transfer. On one hand, we find that WTP varies in systematic and expected ways with respect to factors such as the size of the water quality changes, average household income, and use/nonuse characteristics of respondents. These results show promise for benefit transfer because they suggest that it may be possible to reliably predict WTP for appropriately selected changes in water quality. On the other hand, despite the large number of existing economic valuation studies, relatively few could be meaningfully combined through meta-analysis. The main obstacle to including more valuation results is the heterogeneity in water quality descriptions across studies, making it difficult to convert findings to a common scale for analysis. As a result, information from a large number of studies could not be synthesized for use in policy analysis. In addition, although the meta-analysis results provide a reasonable basis for predicting WTP for broad water quality changes, they are more limited for explaining how the characteristics of the affected population and the spatial variation in water quality changes affect WTP.

Based on our review and analysis, we provide recommendations for future research that may enhance the utility of the literature for national policy analysis. One of the main challenges is developing definitions and characterizations of water quality changes that are both policy-relevant and easily communicated and replicated across studies. We argue that the “designated use” classifications, which are widely used to establish water quality standards across the country, provide a particularly useful framework for standardization.