Courts Have a Role to Play Reining in Federal Agencies

The Trump administration recently issued two executive orders designed to roll back existing regulations, and Congress is currently considering numerous bills that would reform the regulatory state in various ways. A major question in comprehensive regulatory reform is the role of the courts: To what extent should judges review agencies’ analysis to ensure that the agencies have evidence that they are solving a real problem and that they understand the benefits and costs?

When creating new regulations, executive branch agencies must conduct a regulatory impact analysis that defines the problem a regulation is intended to solve, considers a range of possible alternatives, and examines the benefits and costs of the key alternatives. Currently, agencies often produce low-quality regulatory impact analyses.

Regulatory reformers argue that expanded judicial review of agency analysis could produce more effective and less costly regulation. Opponents express skepticism that generalist judges can competently review agency economic analysis, fearing that that judges would either substitute their policy views for those of the agency or defer excessively if they agree with the agency.

Under existing law, federal judges have sometimes reviewed regulatory impact analyses, but they have done so inconsistently and have shown a high degree of variation in the rigor with which they assess the underlying analysis. We recently examined 37 cases in which federal appeals courts considered the agency’s regulatory impact analysis or other equivalent economic analysis in deciding whether to uphold or strike down the regulation. We found that courts are fully capable of reviewing agency regulatory analysis, and agencies have responded by improving their analysis.

Courts have competently considered all four of the major elements that a complete regulatory impact analysis ought to include: analysis of the problem, development of alternatives, and estimates of both the benefits and costs of each alternative. They have set aside regulations when agencies failed to do some basic homework.

In one case (Chamber of Commerce v. SEC (2005)), the agency completely ignored a less burdensome alternative that was actually proposed by two of the agency’s own commissioners. The court sent the rule back to the agency, directing it to at least consider this alternative option. In another case (Center for Biological Diversity v. NHTSA (2008)), the agency that sets fuel economy standards for cars made the unfounded assertion that the social benefit of reducing carbon emissions was zero. Again, the court struck down the rule and asked the agency to perform additional analysis.

Our study finds that the quality of the agency’s analysis improves noticeably after a court sent a rule back to an agency. We also found no evidence that the court system is biased against regulation. In 57 percent of the cases where courts examined an agency’s economic analysis, the court upheld the regulation. In 56 percent of the cases where courts struck down a regulation, the court’s decision implied that flaws in the analysis led the agency to regulate too little.

Unfortunately, courts currently review agency regulatory analysis inconsistently. Unless a statute requires a different standard, review occurs under the Administrative Procedure Act’s “arbitrary and capricious” standard. What this rather vague verbal formulation means in practice depends on the judge applying it: Some judges interpret this standard to require a careful look at the evidence the agency offers in support of the regulation. Other judges think it’s sufficient that the agency gave a reason (however inadequate or far-fetched) for its decisions.

A clear statutory standard could promote consistency and rein in any potential judicial tendencies toward either excessive activism or excessive deference. 

We believe an effective standard would consist of three provisions: First, it would outline the topics a sound regulatory analysis must cover. Second, it would require agencies to use the best available scientific and economic evidence. Finally, a court could only overturn a regulation due to faulty analysis if the error in the analysis makes a material difference to the regulatory decision. This prevents regulations from being thrown out because of trivial “gotcha” details.

A key function of courts is to assess which set of evidence is more persuasive when facts are in dispute. The record shows that courts are quite capable of evaluating the evidence agencies provide that underlies major regulatory decisions—including economic analysis. There’s no need to fear judicial review of agency regulatory analysis, as long as Congress provides a clearer standard to guide that review.