carbon sinks or sequestering, biodiversity, and hydrological regulation is still highly speculative. Inclusion of such estimates in the national accounts is questionable today and might be postponed until data and methodologies in this area are improved.
More research is needed on the effect of stock changes on the value of these service flows because the relationship is complex and current information may be inaccurate. For example, a reduction in standing volume of timber may change water outflows from a forest, increase habitat for some animals and decrease habitat for others, and increase some kinds of recreational services while decreasing others. Storage and diversion of waterways for irrigation purposes may likewise provide habitat for some aquatic species and destroy it for others, and increase some recreational uses but eliminate others.
Many of the same issues arise in accounting for the market-related functions of renewable resources and subsoil assets. Much work already exists on valuation of forests and timber, but further research on valuation is necessary to determine the accuracy of the Hotelling approach. The major challenge in estimating both asset values and service flows lies in determining appropriate values for nonmarket aspects, which are particularly important for forests. Recommendations for forests are in Chapter 4 (see particularly recommendations 4.5, 4.8, and 4.9).
Developing improved accounts for environmental assets such as air and water quality or nonmarket services of natural-resource and environmental assets is an important goal of augmented accounting. Accomplishing this goal involves both measurement of the costs of pollution abatement and estimates of the value of the market and nonmarket services provided by these assets. One important initial step undertaken by BEA was the construction of a set of estimates of pollution abatement and control activities. This effort has unfortunately been discontinued because of budget cuts imposed on BEA. These estimates are an important aspect of any economic assessment of the environment.
The development of accounts for changes in air and water quantity was postponed to Phase III of the IEESA effort, as was accounting for uncultivated biological resources such as wild fish and undeveloped land. Though ambient environmental quality represents an important dimension of current consumption and from a conceptual point of view belongs within an expanded set of environmental accounts, data needed to implement this approach are currently unavailable except in a small number of cases.
Data on air and water pollution illustrate the difficulties. Although