This post was originally published on Legal Planet and was co-authored with Professor Farber’s colleague Eric Biber. Reprinted with permission. Click to read the first, second, and third posts in this series.
This is the final installment in our series of posts about the causation issue under NEPA. In our previous post, we laid out NEPA’s purposes and why analogies to tort law can misfire because that area of law has very different purposes. Today, building on our recent working paper, we explain the functional approach to causation that we believe courts should apply.
We begin with a very basic point. Any effects to be analyzed must be environmental. Effects that do not result from environmental changes, directly or indirectly, need not be analyzed. This first point was made years ago in a Supreme Court case (Metropolitan Edison Co. v. People Against Nuclear Energy) and should be taken as axiomatic.
Second, any effects to be analyzed must be reasonably foreseeable — this is in the text of the 2023 NEPA amendments, which basically codified past practice. As a result, effects that are overly speculative, beyond the technical capabilities of the agency to analyze, too complex to assess, or are otherwise intractable for an agency to assess need not be documented in an environmental impact statement. For instance, long causal chains or attempts at predicting the strategic behavior of individual actors in the future are not necessarily out of bounds but may well not be tractable, an issue where CEQ can provide expert guidance.
Foreseeability may be much greater at the aggregate level. For instance, when oil is produced, it is very likely that it will eventually be burned and release carbon dioxide, contributing to climate change. But it could be far less certain where it will be burned and what local population may be impacted by pollution there.
Subsequent permitting decisions by other agencies are relevant to the extent they can make analysis less tractable by making future outcomes less predictable and thus less foreseeable. This will be particularly true where those effects are indirect ones that are site-specific. For instance, it may not be tractable to assess how a particular permitting decision in the future for an individual, yet-to-be-proposed project will proceed, what conditions that permitting process will impose, and what effects might be result. However, aggregate impacts analysis that involves subsequent permitting decisions may well be tractable, because that analysis will not necessarily depend on the behavior of individual agencies or private parties in the future.
Third, the analysis requirement under NEPA is to be followed by agencies consistent with other statutory requirements. Most importantly, where an agency has no discretion to consider some or all environmental impacts, analysis of those relevant environmental impacts of that decision would serve no purpose under NEPA and would be inconsistent with those other statutory requirements. The 2023 amendments to NEPA codify this analysis in section 102(2)(C), which qualifies requirements for an environmental impact statement when “compliance would be inconsistent with other statutory requirements.”
Similarly, section 106(a)(4) excuses compliance entirely when the agency “does not have authority to take environmental factors into consideration in determining whether to take the proposed action.” Notice that this exclusion is keyed to whether the agency can consider environmental effects in making its own decision about an action, not to whether another agency can independently regulate the effects.
As a rule of thumb, we would suggest that where an agency has discretion to consider the economic costs or benefits of some effect of its actions, it presumptively has discretion to consider the corresponding environmental effects. After all, as the Supreme Court said in Michigan v. EPA, “One would not say that it is even rational, never mind ‘appropriate,’” to ignore the costs of a decision, with the term costs encompassing “harms to human health or the environment.” Thus, if the Surface Transportation Board could consider the economic benefits of producing and selling oil, it should be able to consider the corresponding environmental costs. Otherwise, its final decision would be hopelessly skewed.
Some aspects of the Seven Counties case are easily analyzed under our approach. It is highly foreseeable that the project could result in a large expansion in oil production in the basin, and even without knowing the exact locations, some aggregate impacts are readily foreseeable. A more difficult problem, however, is posed by the downstream effect on communities near refineries. While it is obvious that such communities could be impacted, analyzing the impacts could be intractable given the difficulty of predicting precise locations and the potential for intervention by air pollution regulators.
This functional approach is consistent with Supreme Court precedent, based on the text and purposes of NEPA, and provides workable guidelines for agencies to determine what kinds of effects to examine when conducting environmental reviews. It is the approach the Court should follow when deciding Seven Counties and when giving guidance to lower courts and agencies about how to apply NEPA.