INTRODUCTION

The Environmental Protection Agency (EPA) issued an advanced notice of proposed rulemaking (ANPRM) regarding increasing consistency and transparency in considering costs and benefits in the rulemaking process. Benefit-cost analysis (BCA) is a critical component of good rulemaking. Yet it is only valuable if done well.

EPA is not alone among agencies that struggle to perform consistent and thorough BCA. The agency deserves credit for recognizing that it can improve, and for asking the public to provide ideas.

EPA faces unique challenges in that many of its authorizing statutes differ on how or whether, and in what form, BCA can inform decisions. To make things simpler and to provide the analyses necessary to make good regulatory policy, EPA should utilize BCA to the maximum extent possible under the law. In situations where the law is vague about its application, EPA should utilize BCA since it can improve decisions and limit arbitrary and capricious decisions. Even in situations where BCA is not allowed to be considered in a final regulatory determination, such as National Ambient Air Quality Standards, EPA should still produce an analysis and publish a summary in the Federal Register to inform the public.

EPA asked a number of specific questions in the ANPRM. These questions are addressed in sequence below.

RESPONSE TO QUESTIONS

A: The Nature of Potential Concerns Regarding Perceived Inconsistency and Lack of Transparency

EPA requested more information about the nature and extent of the concerns relating to possible inconsistency and lack of transparency in considering costs and benefits in the rulemaking process. In particular, EPA asked for “specific examples with context and specify relevant statutory provisions,” and sought input on whether greater consistency or transparency could improve outcomes for regulated entities, states, tribes, and localities, and the public.

One area of inconsistency has been regarding the use of co-benefits, or benefits derived from non-targeted pollutants. Co-benefits can be used to inflate benefits estimates to the point of overwhelming direct costs imposed by a rule. One such example is the Mercury and Air Toxics Standards for Power Plants rule issued in 2012 (77 Fed. Reg 9304). EPA estimated a range of benefits of $36-$90 billion, but reduction of mercury accounted for just $4-$6 million of that total. Virtually all the rest of the benefits came from reducing particulate matter. Costs of the rule were $9.6 billion. But the estimate only looked at direct costs to complying entities. The imbalance caused by estimating a broad range of estimates while narrowly considering costs is problematic because EPA has an entire other regulatory scheme aimed at particulate matter.

Not only was EPA drastically overstating the benefits of the mercury rule to justify the costs, but it was also likely double counting the benefits associated with a pollutant addressed in other rules. For instance, say two emissions-related rules (implemented concurrently) claim to yield $X and $Y (respectively) in co-benefits due to reductions in particulate matter. How can EPA be sure of the causal linkage between whatever reductions emanate from either rule? Are such reductions a product of both rules working in concert? These are issues that EPA ought to more fully address in its analysis of such rules.

Another example of inconsistent methodology occurred in the analysis of the Clean Power Plan (80 Fed. Reg. 61661), where once again a minority of the calculated benefits came from the targeted pollutant. Here, EPA used global benefits (rather than just those to the United States, contrary to guidance from the Office of Management Budget) to inflate total benefit numbers to justify direct costs of $11.9 billion. EPA seems to have accounted for this in the proposed rule that would repeal the Clean Power Plan. Although, in that proposal the analysis includes seemingly competing methodologies and limited clarity on which approach EPA uses to justify that action. Hopefully the final version of that rulemaking will include a more direct approach.

B: Potential Approaches for Increasing Consistency and Transparency in Considering Costs and Benefits in the Rulemaking Process

1: What would increased consistency look like?

Increased consistency would allow the public to know that a consistent method to BCA had been applied from rule to rule. It would also allow the public to know that EPA had applied a consistent methodology to what types of costs and benefits had been considered.

On a rule to rule basis, EPA should – to the extent possible under applicable statutes – standardize the breadth of benefits and costs it will use to analyze rules. This will ensure that all rules are considering the same types of costs and benefits. The benefit of this approach is that it will minimize the arbitrary inclusion or avoidance of including certain types of costs or benefits.

EPA should apply a consistent methodology. Within each rule, costs and benefits should be viewed through the same scope. As the examples included in discussion of section A illustrate, some previous rules have included a broad range of benefits while narrowly analyzing costs. EPA ought to establish parameters on the scope and definition of various forms of costs and benefits. An agency-wide memo providing such parameters (in the spirit of OMB Circular A-4) could help guide both the agency and interested stakeholders with a clear idea of what to expect in a given rulemaking.

2: What would improved transparency look like?

A critical component of transparency is accessibility. Too often, information that could help policymakers and the public develop knowledgeable decisions hides in plain sight in the Federal Register. For some rules the costs and benefits are buried in narrative explanations. In others, they are presented on clearer tables. In some cases, it is a mix between the two. The variance in presentation can cause confusion among readers. Transparency of BCA could be improved by standardizing how this information is presented in preambles of rules published in the Federal Register. The public would be best served by a clearly titled “Benefit-Cost Analysis” section within the regulatory analysis section of preambles. This section would lead with a clear table showing the benefits and costs on a total and annualized basis. An example is included below:

BENEFIT-COST ANALYSIS

Analysis Component	Value
Total Benefits	$XXXXXXXX
Total Costs	$XXXXXXXX
Duration	XX Years
Annualized Benefits (discount rate)	$XXXXXXXX
Annualized Costs (discount rate)	$XXXXXXXX

Following this table, EPA should include a narrative discussion of how it arrived at these values, including assumptions and discount rates, as it often does now. This table would make discerning this information much more accessible to the public. Other agencies include such sections on a far more consistent basis. Some particularly helpful examples include rules from the Federal Aviation Administration (e.g. here and here) and the Centers for Medicaid and Medicare (in “Accounting Statements” such as here and here). EPA should also consider including similar tables breaking down the relevant topline values for other required analyses, such as those required by the Regulatory Flexibility Act and Paperwork Reduction Act.

3: To what extent would requiring a systematic retrospective review element in new regulations help to provide ongoing consistency and transparency in how regulatory decision making will adapt over time to new information?

A 2007 Government Accountability Office report found that every administration since President Carter has called for retrospective review of regulations. Since then, Presidents Obama and Trump sought to continue the practice through their own executive orders on regulatory reform. The need is obvious. Absent such review, agencies and the public are unable to tell if regulations are effective. Failure to review existing regulations also leads to reams of outdated, unnecessary regulatory requirements that impose compliance costs on regulated entities and require taxpayer funded resources to enforce.

Unfortunately, the practice has not been meaningfully institutionalized by agencies. According to the GAO report, typical problems agencies have encountered is a lack of resources, documentation, and impetus, just to name a few. EPA has the opportunity to emphasize the critical nature of ex post review in this rulemaking. Incorporating regulatory language establishing retrospective review would help address many of the problems identified in the GAO report that have rendered previous efforts disappointingly inadequate.

With retrospective review, EPA can take what actually happened and use that to better determine a regulatory program’s efficacy. The American Action Forum has compiled a sample of retrospective studies here that covers a wide array of regulatory issues, including EPA’s portfolio. A regular, institutionalized review schedule (such as the 10-year cycle employed by the Federal Trade Commission) could help EPA become a model agency in terms of sharpening its research techniques and addressing outdated, overly burdensome regulations.

CONCLUSION

EPA should utilize BCA to the maximum extent possible under the law. Importantly, EPA should strengthen its BCA to make it more consistent and transparent. The scope of benefits and costs and costs should be consistent from rule to rule. Transparency can be improved by clearly identifying the topline benefits and costs in a table published with the rule in the Federal Register. EPA should also systematically require retrospective reviews as part of it BCA procedures to ensure targets are being achieved and to identify potential regulatory improvements.

EPA deserves credit for soliciting input from the public on its use of BCA. Other federal agencies would benefit from a similar effort. We appreciate the opportunity to comment and the consideration of our input. Should EPA require additional information, please contact us at 202-559-6420.