June 19, 2007

Environmental Regulation

EPA Data-Crunching Tools Need Regular Checkups

National Research Council calls for evaluation of agency models used for regulating

Cheryl Hogue

EPA needs to evaluate, on a regular basis, the computational models it uses in making regulatory decisions, the National Research Council says in a report released on June 14.

The agency’s mathematical and computer models run the gamut from predicting how a pollutant will move through groundwater to estimating the costs and benefits of a proposed regulation. Results of modeling, for example, help regulators decide whether to allow a new pesticide to go to market or set a cleanup level for a particular pollutant at a Superfund site. Over the years, EPA has increased its use of models.

The NRC report says regular, rigorous evaluation of models may be costly. But, it says, "such investments are essential if environmental regulatory modeling is to meet challenges now and in the future."

These reviews may invite challenges to models’ results, either in court or under the Information Quality Act, "because the agency is documenting features of its models that need to be improved," the report says. "However, an improved model is less likely to generate erroneous results that could lead to additional challenges, and it better serves the public interest."

In addition, if EPA has a formal evaluation process for its models, lobbying groups and federal courts may give deference to regulatory decisions that, in part, hinge on results from modeling, the report says.

The report also recommends that EPA, as much as possible, use models that are not proprietary. EPA stirred up controversy in 2005 when it used a proprietary computer model to evaluate various legislative options to cut emissions of sulfur dioxide, nitrogen oxide, and mercury from coal-fired power plants. The Congressional Research Service found the agency’s analysis to be faulty. EPA’s analysis favored the Bush Administration’s less stringent proposal over other legislation, CRS said (C&EN, Dec. 12, 2005, page 8).

EPA "should only adopt proprietary models when a clear and well-documented case has been made that the advantages of using such models outweigh the costs in lower credibility and transparency that accompanies reliance on proprietary models," the report says. "Furthermore, proprietary models should be subject to rigorous quality requirements and to peer review."

The NRC report, "Models in Environmental Regulatory Decision Making," is available online at books.nap.edu/catalog.php?record_id=11972.

Chemical & Engineering News
ISSN 0009-2347
Copyright © 2007 American Chemical Society