This article was authored by Nicholas Parrillo, Professor of Law at Yale Law School, and Lee Liberman Otis, Senior Vice President and Director of the Federalist Society's Faculty Division.

This article first appeared in the Regulatory Review's series on "Five Recommendations for Improving Administrative Government" that focuses on the ACUS Recommendations adopted at the December 2017 Plenary Session. Reposted with permission. The original may be found here, and the Regulatory Review's entire series may be found here.

For individuals and firms who live under federal regulation, actual regulations are just the beginning of the story. Despite being voluminous and complex, regulations leave many matters open to interpretation, and they leave agency officials with discretion over implementation and enforcement. Individuals and firms want to know how the agency reads the ambiguous provisions and how it will use the discretion left it. And officials want individuals and firms to have that knowledge to facilitate voluntary compliance. So officials provide the public with lots of “guidance”: advisories, interpretive letters, enforcement manuals, fact sheets, FAQs, and more.

Although ubiquitous, guidance is also controversial. It can be issued without the process that the Administrative Procedure Act (APA) mandates for regulations—most prominently notice and comment—because it qualifies for the APA’s exemptions for interpretive rules and policy statements. Thus, guidance can be produced and altered much faster, and in higher volume, than regulations.

What justifies this disparity, in the familiar telling, is that guidance, unlike a regulation, is not binding. It is only a suggestion of what the agency is inclined to do, not something the agency will follow in an automatic, ironclad manner. If an individual or firm wants to do something different, the agency is supposed to give fair consideration to that alternative approach. If officials treat guidance with this kind of flexibility, it does not seem so bad for the agency to be unconstrained in issuing guidance to begin with.

Yet individuals and firms subject to regulation have long complained that, although guidance is officially nonbinding, it is coercive in real life. If true, this complaint reveals a giant loophole in the APA: Agencies can issue de facto regulations with little constraint simply by calling them “guidance.”

The classic articulation of this fear was penned a quarter-century ago by former Administrative Conference of the United States (ACUS) chairman Robert Anthony, who said that the “capital problem” was that agencies will frequently issue guidance documents with the improper “intent or effect of imposing a practical binding norm upon the regulated or benefited public.” Anthony did little to elaborate on “effect” and instead focused largely on “intent,” for he believed agencies’ abuses were very often deliberate. Relatedly, the ACUS recommendation arising from Anthony’s report spoke exclusively in terms of intent, denouncing the issuance of documents other than regulations “that are intended to impose binding substantive standards.”

Twenty-five years later, ACUS sponsored a new study of guidance to reexamine the subject in light of current administrative practice, and it has recently issued a new recommendation based on that study.

The study draws on interviews with 135 individuals across agencies, industries, and NGOs, finding that, consistent with Anthony’s worries, regulated parties often (though not always) have no practical choice but to follow a guidance document. But the study also finds that Anthony was mistaken to view this phenomenon primarily in terms of the agency’s “intent” to produce this outcome. Rather, the problem of coercive guidance is far more often one of institution-level behavior that nobody fully intends. Coercive guidance therefore calls primarily for an institutional-reform response.

Certain structural features of modern regulation impose a strong incentive on regulated parties to figure out what the agency wants and to do just that, regardless of the format in which the agency’s wants are expressed (regulation or guidance).

For example, some regulatory schemes have pre-approval requirements, in which a regulated party must win the agency’s affirmative assent to obtain some legal advantage, such as a license or benefit. Think of the U.S. Food and Drug Administration’s requirement for pre-market approval of new drugs, or Medicare’s reimbursements to health providers. If the regulated party stands to suffer major losses from a denial or delay, it will do whatever it can to win the agency’s approval in the first go-around, and fast. That means following guidance.

Other regulatory schemes create a strong incentive for a regulated party to maintain a positive image and relationship with the agency—most commonly, if the party is subject to repeated evaluations by the agency and if the law is so complicated that the party will inevitably fail to comply with some requirements some of the time. For example, banks work to win the trust of the agencies that supervise them, knowing these relationships may need to be “cashed in” someday.

Also, most guidance is addressed to firms, and the firm is a “they,” not an “it.” The employees at a firm who confront agency guidance increasingly work in compliance departments, and they typically possess a disposition and socialization that may give them an overriding motivation to stay in the agency’s good graces.

Finally, following guidance may be the best way to avoid the initiation of an enforcement proceeding, which firms often do not want to risk, regardless of how it turns out. Firms may seek to avoid enforcement altogether because markets may take it as a signal that the firm cannot be trusted by its counter-parties (as in accounting) or because the statutorily authorized sanctions are so draconian that many firms dare not refuse whatever “deal” the enforcers offer them and thus never get to a neutral adjudicator (as with Medicare exclusion).

These structural features, present in various combinations in many although far from all regulatory schemes, often make it difficult for regulated parties to act out of line with agency guidance. This pressure occurs regardless of whether agency officials “intend” the documents to be coercive—it is hard-wired into the regulatory scheme.

These pressures are diminished if agency officials are sufficiently open to considering alternative approaches to those delineated in a guidance document and the regulated entities know it. But agency officials are under institutional pressures that often lead them to be inflexible.

Agencies are frequently under active and legitimate stakeholder pressure to be inflexible (that is, to maintain consistency). When regulated parties are competitors, an agency that departs from guidance in the case of one firm may put other firms at a competitive disadvantage. Individualized flexibility on guidance can also draw accusations of favoritism from NGOs, the media, and Congress.

These concerns can be mitigated if the agency gives public rationales for guidance departures and makes the rationales applicable to similarly situated parties going forward. That approach provides flexibility, but in a principled, even-handed, and transparent manner. But that is easier said than done. Reason-giving requires officials to spend scarce time and money working up the explanation, usually obtaining approval from a higher-level official, which creates a bottleneck.

There are also organizational obstacles to flexibility. The distinction between binding and nonbinding rules is not intuitive; training helps, but can be costly. In addition, some offices tend not to develop the kind of cooperative relationships with regulated parties that foster flexibility; this is particularly likely for enforcement offices, which are distinct from program offices. Finally, higher-level officials authorized to override their subordinates’ adherence to guidance are often loath to do so, because it feels like pulling the rug out from under the people who work for them.

The new ACUS recommendation—which covers policy statements and may also be helpful for interpretive rules—sets forth a series of measures to foster flexibility in the face of these challenges. Some of the measures are low-cost and should be adopted universally, like making sure that policy statements include a disclaimer of binding status on the public, and ensuring that, if some agency employees are bound to follow a policy statement, other higher-level employees are authorized to depart from it.

But other measures are not low cost. They will take up resources and managerial initiative that might be spent elsewhere. These measures include: publishing reasons for departures from a policy statement; training and monitoring frontline personnel to ensure they treat departure requests appropriately; fostering dialogue about departures through intermediaries like ombudspersons and associations; assigning departure authority to agency offices that have give-and-take relationships with regulated parties; and assigning appeals of departure denials to officials who do not regularly work with the people whose decisions might be overridden. As result, priorities about which measures to use will have to be set.

The recent ACUS recommendation says that agencies should invest more in promoting flexible use of a policy statement when regulated parties are otherwise under strong pressure to comply with the statement or the statement has a significant impact on private interests. On the other hand, agencies should invest less in flexibility when legitimate concerns about the need for consistency dominate in the minds of officials and stakeholders.

We have focused thus far on how guidance affects regulated parties. But it also affects third parties. For example, guidance may harm the beneficiaries of regulation, like citizens breathing polluted air, since some guidance may be practically favorable to regulated parties compared to the status quo. The agency might say it thinks a safety measure, previously viewed as necessary, actually is not, or that some activities that violate the law will be a “low priority” for enforcement. Thus the agency may effectively deregulate without APA procedures for amending or rescinding regulations, thereby blocking regulatory beneficiaries from participating in those decisions. Or guidance may harm other third parties like consumers by promoting installation of a costly but not legally required pollution control device.

The remedy for this kind of problem could, theoretically, be the same as the remedy for guidance that unduly coerces regulated parties: The agency should treat the guidance flexibly on a case-by-case basis, generally adhering to the path outlined in the guidance—but not always. The idea is that an approach that allows or encourages some parties to harm others, unless done by notice and comment, should not give ironclad protection to those who may do harm.

But this may not work very well in the real world. When an agency offers “safe harbors” that are not actually 100 percent safe, it puts regulated parties in a difficult position. Flexibility also may not do much to protect regulatory beneficiaries and interested third parties. It can help in situations where these parties, or NGOs representing them, can and do participate in individual cases and argue against applying the guidance. This may happen if the cases are big and salient, and if the NGOs are well-resourced and sophisticated—as seems to be the case for some U.S. Environmental Protection Agency permitting proceedings. But often these conditions do not hold.

This brings us to the final part of the new recommendation, on getting members of the public to participate when an agency first adopts a policy statement. Public participation can be achieved through soliciting comments or through less formal means like workshops and targeted outreach to stakeholders. Such interaction may be quite valuable to regulated parties, and it may be the only time regulatory beneficiaries and interested third parties can practically provide input on the statement.

The recommendation does not adopt any set formula for when to invite participation or how. Instead, it lays out factors for the agency to consider. On the one hand, participation can provide officials with better information and increased stakeholder acceptance of the policy statement (On the other hand, participation can backfire if the agency lacks the resources or inclination to be responsive to the views it hears.). Participation can also slow the adoption of a policy statement, potentially prolonging the limbo stage in which the statement is in published “draft” form, and might cause stakeholder confusion about whether to follow the statement. In the end, the quest to pursue a highly participatory process for crafting guidance might well eat up resources that could be used to formulate guidance on other subjects where stakeholders are in the dark.

We do not expect the measures laid out in the recommendation to serve as a panacea for the various challenges that agency guidance presents. But we do hope that they provide agency officials and members of the public some useful ways to think about these conundrums and a broader toolkit of options to help address them.