Committee Co-Sponsors 9th Fall Meeting Panel: Rules versus Guidances

Ken Weinstein

Ken Weinstein is a partner in Latham & Watkins' Washington, D.C. office and is a member of its Litigation and Environmental Departments.

At the 9th Section Fall Meeting in St. Louis on October 5, 2001, our Committee co-sponsored a well-received panel which thoughtfully explored the continuing debate over EPA's extensive use of "guidances" to set forth agency policy. The panel considered the circumstances in which EPA uses guidances (also variously known as policy statements, nonlegislative rules, or interpretive rules) rather than rules to effectuate policy, whether such uses are "good" or "bad," and the criteria used by the courts to define the dividing line. The eminent panelists included Professor Ronald Levin of the Washington University School of Law, past chair of the ABA Section on Administrative Law and Agency Practice; Jeffrey Holmstead, EPA assistant administrator for Air and Radiation; and James Conrad, counsel for the American Chemistry Council. I was the moderator.

Ron Levin discussed the characteristics that the courts have used to distinguish "rules" from "guidances." Jeff Holmstead shared the key factors that influence EPA's thought process in deciding whether a policy should be published as a rule or merely as a guidance. Jamie Conrad synthesized the principles that industry often looks at in separating "good" guidance from "bad" guidance. The following summary is extracted from their papers.

Pros and Cons. Professor Levin observed that the basic quest is to distinguish between true rules, which have the force of law and must be published in accordance with the APA, and guidances, which do not. Because the procedural requirements imposed on the regular rulemaking process have grown more elaborate over time, agencies like EPA have responded by relying on informal documents, such as guidances, to articulate many of their policy decisions. Critics of this widespread use of guidances argue that it undercuts the benefits to the public that the rulemaking process is supposed to secure. Rulemaking provides an opportunity for the agency to learn from external perspectives through the comment process, induces them to give thorough consideration to new policies, subjects them to the discipline of having to respond in writing to criticisms of proposed actions, and is conducive to political accountability. Despite the absence of all these safeguards, the critics argue, guidances (which theoretically are nonbinding) acquire de facto "practical binding effect," because regulated parties are loath to challenge them.

Defenders of agencies' use of guidances argue that the creation and publication of these documents benefit the public, which can thereby learn the agencies' thinking and find out what they need to do to avoid enforcement difficulties. Guidances also promote regularity of administrative action, because they enable an agency's central leadership to coordinate and oversee the work of personnel in the field. Many documents that serve these functions simply would not be generated if they had to be subjected to the cumbersome rulemaking process.

Openmindedness. An agency is required to follow its regulations, but a guidance is not binding in that sense. The agency has the option of not following an outstanding guidance (although a decision to depart from it would call for an explanation). Thus, if a private person makes a strong challenge to the guidance, the agency must defend its position substantively - it cannot simply say that it has already enunciated its policy. Flagstaff Broadcasting Found. v. FCC, 979 F.2d 1566 (D.C. Cir. 1992); Bechtel v. FCC, 957 F.2d 873 (D.C. Cir. 1992).

While this distinction between regulations and guidances is relatively clear (at least in concept), it is not the usual focus of legal controversy. Normally, regulated persons try to avoid getting embroiled in enforcement proceedings. Thus, a more common context in which the regulation-guidance distinction matters is the pre-enforcement stage, when private persons seek legal intervention on the basis of how they expect or fear the rule would be applied if an enforcement proceeding were to ensue.

The label that an agency applies is not controlling, because courts do not always accept agencies' labels at face value. If a court decides that an agency has written or used a guidance in a binding fashion, it may recharacterize the rule as a regulation and hold it void for lack of notice and comment, or it may admonish the agency not to apply it in a binding way. McLouth Steel Prods. Corp. v. Thomas, 838 F.2d 1317 (D.C. Cir. 1988).

The question thus becomes: In what ways may an agency make use of a guidance - including taking actions that constrain private parties in some degree - without being held to have given the document "practical binding effect" (a phrase highlighted in the writings of Professor Robert Anthony)? It is widely agreed that an agency can give mandatory instructions to low-level staff, thus forcing private persons who disagree with the instructions to raise their grievances at a higher level in the agency's hierarchy. ACUS Recommendation 92-2, III. Moreover, an agency can look to its policy statements as precedents, although it may not treat them as entirely dispositive of significant challenges raised against them. Panhandle Producers & Royalty Owners Ass'n v. Economic Regulatory Admin., 847 F.2d 1168 (5th Cir. 1988).

If the agency is entirely unreceptive to the private persons challenge at every level of review, however, or if the opportunities for challenge that it leaves open are too constricted, it is running larger legal risks. In practice, the courts have had difficulty administering this doctrine. The outcome of a given case often turns on somewhat superficial evidence as to whether the agency intends to be bound by the guidance - such as whether the document uses imperative words like "must" or "shall." In other instances, the court engages in a wide-ranging factual inquiry to decide whether the agency has, in fact, remained openminded in its use of the document. U.S. Telephone Ass'n v. FCC, 28 F.3d 1232 (D.C. Cir. 1994).

Finality. Under traditional doctrine, it is easier to get immediate (pre-enforcement) review on the merits of a regulation than of a guidance. A rule must be "final" to be amenable to review. (That requirement is sometimes seen as derived from the APA, 5 U.S.C. § 704, and sometimes as a component of the case-law doctrine of "ripeness.") To be final, an action must "mark the 'consummation' of the agency's decisionmaking process [and] not be of a merely tentative or interlocutory nature." Bennett v. Spear, 520 U.S. 154, 177-78 (1997). A guidance or other informal statement may well be a work in progress, and thus not meet that test. See, e.g., Ohio Forestry Ass'n, Inc. v. Sierra Club, 523 U.S. 726 (1998); Florida Power & Light Co. v. EPA, 145 F.3d 1414 (D.C. Cir. 1998) (statements in preamble to proposed rule were not "final").
In practice, however, the "binding" or constraining manner in which a guidance is implemented may induce a court to treat it as final in this sense. The court may feel that the agency's use of the statement is imposing real-world burdens on the public that overshadow the theoretical possibility that the agency might change its mind about the policy. See Whitman v. American Trucking Ass'ns, Inc., 121 S.Ct. 903 (2001); Appalachian Power Co. v. EPA, 208 F.3d 1015, 1021 (D.C. Cir. 2000).

Scope of Review. The Supreme Court has recently held in United States v. Mead Corp. that issues of law presented by informal documents such as interpretive rules and policy statements are subject to a different - and reputedly more intrusive - standard of review than are issues of law presented by regulations. Regulations are reviewed under Chevron U.S.A., Inc. v. NRDC, 467 U.S. 837 (1984): The courts are expected to enforce the "clear intent" of Congress; but if the statute is found to be ambiguous on the "precise issue" presented, the court must uphold the agency's interpretation if it is "reasonable" or "permissible."

Mead concludes that Chevron should apply only to agency actions as to which Congress expected the agency would speak with the force of law. Thus, interpretive rules and policy statements should not be reviewed under Chevron, but instead under Skidmore v. Swift & Co., 323 U.S. 134 (1944). Under Skidmore, the extent of judicial deference depends on the extent to which the agency's interpretation has "power to persuade," in light of such factors as the timing and consistency of the agency's position and the nature of its expertise.

Good Versus Bad Guidance. After Professor Levin established the conceptual framework for distinguishing between rules and guidances, Jamie Conrad discussed the factors that make the use of guidances appropriate or inappropriate. The following summarizes some of his remarks.

Mr. Conrad observed that trade associations and other representatives of regulated entities frequently decry federal agencies' use of guidance documents and the like in lieu of notice and comment rulemaking. Congress has denounced such "back-door regulation." "Non-Binding Legal Effect of Agency Guidance Documents," H.R. Rep. No. 1009, 106th Cong., 2d Sess. 1 (Oct. 26, 2000). Even public interest groups and individuals will sue over "de facto" or "spurious" rules when it suits their purposes. (E.g., Warder v. Shalala, 149 F.3d 73 (1st Cir. 1998) (Medicare beneficiaries challenge HCFA ruling regarding classification of orthotic management devices), cert. denied, 526 U.S. 1064 (1999); Animal Legal Defense Fund v. Quigg, 932 F.2d 920 (Fed. Cir. 1991) (animal group challenges PTO notice announcing patentability of living nonhuman organisms); Community Nutrition Inst. v. Young, 818 F.2d 943 (D.C. Cir. 1987) (NGO challenges FDA guideline limiting enforcement to cases where aflatoxin levels in corn exceed "action level" of 20 ppb).)

Whenever we collect examples of agency use of guidance documents, however, we find as many cases where industry groups support the use of guidance as where they oppose it.

In thinking about these issues, it is helpful to remember that people are mainly concerned about two things: The opportunity to have input before a document is issued - the essence of notice and comment rulemaking; and the opportunity to seek judicial review of the document afterwards.

Regulation or Guidance? In cases where an agency has the power to issue a regulation, when is it appropriate to proceed by guidance? Put another way, how can you tell if a given guidance document is a "good" one, or a "bad" one that is likely to generate criticism and maybe lawsuits?

Does the guidance bind the public? All concerned, it seems, agree that regulations - which have the force of law - may bind the public, and that guidance cannot. (See, e.g., Anthony, "Interpretive Rules, Policy Statements, Guidances, Manuals and the Like - Should Federal Agencies Use Them to Bind the Public?," 41 Duke L. J. 1311, 1315 (1992) ("[N]onlegislative rules like policy statements, guidances, manuals and memoranda should not be used to bind the public") (emphasis in the original). Accord Strauss, "The Rulemaking Continuum," id. at 1465 (Where "the citizen is bound but the government is not . . . seems the very antithesis of the 'rule of law'"); Levin, "Nonlegislative Rules and the Administrative Open Mind," id. at 1498. Actually, there is some dispute about whether interpretive rules can bind the public, but we will not analyze that dispute here.)

Even EPA has stated unambiguously that it "does not intend its policy statements and guidance documents to be binding and they have no binding legal effect on the public." Letter from Lisa K. Friedman to Rep. David McIntosh (Mar. 2, 1999), at 3, reprinted in H.R. Rep. No. 1009, supra, at 14. The same EPA document makes reference to "binding guidelines," but it seems to refer to documents that would have been regulations but for some exception from notice and comment requirements. Id. n.1.

Therefore, a guidance document should not read like a rule (in striking down EPA's "Periodic Monitoring Guidance," the D.C. Circuit noted that "the entire guidance, from beginning to end . . . reads like a ukase. It commands, it requires, it orders, it dictates." Appalachian Power Co. v. EPA, 208 F.3d at 1023), and ought to make clear that it does not impose binding requirements on the public. What exactly "binding" means in the context of guidances can be the subject of some disagreement. At a minimum, however, it means that the agency takes the guidance to be dispositive of an issue; that a private party is precluded from raising an objection to it before the agency. See Anthony, supra, at 1312 n.1; Levin at 1498.

The more difficult sense of "binding" is the "practical effect" concept. In 1997, OSHA created a "Cooperative Compliance Program," directed at workplaces with high injury rates. Companies opting into the program would establish a health and safety program in cooperation with OSHA, and in exchange would be dropped from OSHA's regular inspection program. High-injury workplaces not opting in the program would be the subject of heightened inspections.

The U.S. Chamber of Commerce challenged the program as being a de facto rule. The D.C. Circuit agreed and struck it down for lack of notice and comment. Chamber of Commerce v. U.S. Dep't of Labor, 174 F.3d 206 (D.C. Cir. 1999). OSHA argued that the CCP was a procedural rule because it did not require anyone to do anything and thus was not "binding." (Notice and comment is not required for rules of agency organization, procedure, or practice. 5 U.S.C. § 553(b)(3)(A).) The court disagreed, finding that the "practical" effect of the program was to "place the burden of inspection" on employers who did not opt in. OSHA also argued that its rule was only a policy statement, but the court again disagreed, stating that the rule did not announce the agency's "tentative intentions for the future," but rather "inform[ed] employers of a decision already made." The American Chemistry Council did not join this lawsuit.

At some level, any document announcing an agency's intentions will have some practical effect of coercing regulated entities' behavior, even if those intentions are tentative or subject to challenge before the agency. Query what would have happened if OSHA had said employers who did not opt in would "likely" be inspected.

Does the guidance bind the agency? Even if a guidance purports to leave discretion for the agency, a court may hold it to be a de facto rule if, in practice, agency staff follow it rigidly. E.g., McLouth Steel Prods. Corp. v. Thomas, 838 F.2d at 1321 (agency staff would not depart from model). Moreover, courts and academics alike have assailed the standard EPA disclaimer parodied above as being "boilerplate" and "a charade." Appalachian Power, 173 F.3d at 1023 (quoting Strauss, supra, at 1485).

For example, in May 1993, EPA Administrator Browner issued a memorandum announcing her draft "Combustion Strategy" for reducing hazardous waste incineration. Among other things, the memo declared that incinerators, boilers, and industrial furnaces burning hazardous waste, though already subject to RCRA regulation, would be required to meet tighter standards taken from the Clean Air Act New Source Performance Standards for new municipal solid waste incinerators. EPA Press Release, "EPA Administrator Browner Announces New Hazardous Waste Reduction and Combustion Strategy" (May 18, 1993). Several groups representing hazardous waste combustors sued, arguing that the memo was an illegal attempt to amend the existing RCRA rules. Cement Kiln Recycling Coalition v. Browner, No. 93-1413 (D.C. Cir. filed May 1993). Shortly after the lawsuit was filed, EPA issued a memo "clarif[ying]" that the NSPS numbers were "targets" to be considered in conducting site-specific risk assessments at individual facilities. Memorandum from Richard J. Guimond to Regional Administrators (Aug. 11, 1993).

There is significant disagreement about how good or bad it is for vast bureaucracies to be able to enforce consistency among their myriad offices. (The whole thrust of Strauss, supra, is that on balance society is better off if bureaucracies can act consistently.) However, just as a guidance must leave regulated entities free to challenge it before the agency, so a guidance must leave agency staff - at some level in the hierarchy - free to depart from it, and they must do so where regulated entities present compelling reasons to do so, to avoid a successful legal challenge.

Is there a history of rulemaking on this subject? The essence of the "regulation by guidance" objection is that an agency is circumventing or evading both notice and comment and judicial review. Sometimes, the facts are pretty compelling. For example, the 1990 Clean Air Act Amendments required EPA to issue rules for operating permit programs, including monitoring requirements. 42 U.S.C. § 7661a(b)(2). EPA issued proposed and final rules on the topic, which included reference to "periodic . . . monitoring." 40 C.F.R. § 70.6(a)(3)(B). The preamble to the final rule also stated that if any existing federal air standards lacked sufficient monitoring to satisfy this regulation, "EPA will issue a rulemaking to revise such requirement." 57 Fed. Reg. 32278 (July 21, 1992). Instead, six years later, EPA issued a "Periodic Monitoring Guidance" that implicity found the monitoring of numerous federal standards to be inadequate. In cases where an agency has already begun a rulemaking process, therefore, it would behoove the agency to finish that course, rather than opt for guidance.

Does the issue involve controversial legal calls? Faced with politically sensitive issues of law with vocal proponents on both sides, agencies are often tempted to craft compromise positions in guidances that are frequently "draft" or "interim." A prominent recent example is EPA's recent Title 6 guidance. "Draft Title VI Guidance for EPA Assistance Recipients Administering Environmental Permitting Programs" (June 16, 2000). This guidance has generated heated opposition from both sides of the issue and has satisfied no one. The heightened sensitivity of the issue only feeds suspicion, on both sides, that the agency is attempting to circumvent its legal obligations through guidance. Moreover, questions that primarily involve legal issues - as opposed to factual ones - are the easiest for an agency to address. The comments they attract are more in the nature of legal briefs than the voluminous, fact-laden filings of typical rulemakings. Such an administrative record is thus easier to manage and review. Defending a decision in such a case is also more straightforward. (This point is made persuasively in Levin, supra, at 1506-07.)

Is the subject matter highly technical and case-specific? EPA regulations already vie with those of the IRS and SEC as being among the most complex and detailed in the CFR. And yet, those regulations frequently do not contain sufficient detail to enable a determination of their applicability to a particular set of facts. Also, where a regulation addresses a very heterogeneous subject matter - for example, the concept of "major modification" at the heart of the Detroit Edison case study - it makes sense for the regulation to be expressed in simple, general terms, and for its applicability to be articulated, and developed case-by-case, through guidance. The following are examples where EPA has properly used guidance for highly technical or fact-specific subjects: EPA's guidance for implementing the High Production Volume Chemical Challenge program, EPA's Soil Screening Guidance under the Superfund program, and the IRIS database.

Is the subject, or the agency's understanding of the subject, evolving rapidly or likely to do so? A familiar criticism of the regulatory process is that it has "ossified," slowed to a snail's pace by the continual accretion of procedural requirements imposed by Congress, courts and administrations. In theory, guidance documents can be issued more quickly since they are not subject to these impediments. Particularly in areas where the subject, or the agency's understanding of it, is changing rapidly, it makes sense for the agency to proceed by guidance. Thus, as events and agency knowledge progress, the agency can more easily and quickly update the material.

Regulation by information. Many of the "guidances" that irk regulated entities are published by agencies on subjects where it is questionable, at best, whether the agency actually possesses the power to issue a regulation. Instead, the agency is simply relying on its statutory authorities to collect and disseminate information. (For example, the Clean Air Act authorizes EPA to "collect and disseminate . . . basic data on chemical, physical, and biological effects of varying air quality and other information pertaining to air pollution and the prevention and control thereof." 42 U.S.C. § 7403(b)(6).) These agencies have realized, in the words of former EPA General Counsel Jonathan Cannon, that "[i]nformation . . . can be a supplement, sometimes even an alternative, to regulation. When broadly available, information can change behavior." 15 Environmental Forum 36 (July/August 1998).

Moreover, the advent of the Internet and the ubiquity of computers have qualitatively changed the nature and extent of governmental information dissemination. Virtually all new documents released publicly by federal agencies, and many historic documents, are now available on their websites. Agencies are also actively developing information "products," projects that combine databases and/or models to produce multidimensional characterizations of particular companies, facilities, or regions. For example, EPA's Risk Screening Environmental Indicators model ranks sources of toxic chemical releases by a methodology that combines hazard and exposure data. (See http://www.epa.gov/opptintr/env_ind/index.html.) In creating such products, agencies make the same sorts of policy and scientific choices that they make in issuing regulations.

This use of information has many actual and potential benefits, including speed, flexibility, efficiency, and public right-to-know. On the other hand, if the government's disclosures are inaccurate or misleading, they can inflict unwarranted harm on private parties and disserve the public and the government. Interests in business confidentiality and security may also be jeopardized.